Fabrication, Potentiometric Characterization, and Application of Screen-Printed RuO2 pH Electrodes for Water Quality Testing

Screen-printed sensing electrodes attract much attention for water pollution monitoring due to their small size, physical and chemical durability, and low cost. This paper presents the fabrication and broad potentiometric characterization of RuO2 pH sensing electrodes deposited by screen printing on alumina substrates and sintered in the 800–900 °C temperature range. All the fabricated electrodes showed close to Nernstian sensitivity, good linearity, fast response, small drift, low hysteresis, and low cross-sensitivity toward various interfering cations and anions. Furthermore, decreasing the sintering temperature led to better adhesion of the RuO2 layer and a negligible response to interfering ions. The measurements in real-life samples from different water sources showed that the fabricated electrodes are on par with conventional glass electrodes with a maximum deviation of 0.11 pH units, thus indicating their potential for application in water quality monitoring.


Introduction
The pollution of rivers and lakes creates a great danger for the environment and human health. Pollution can originate from industrial wastewaters, municipal sewage, and substances used in agriculture (fertilizers, pesticides, and manure). At present, the quality of water is determined against numerous parameters: electrical conductivity, turbidity, dissolved oxygen, toxic inorganic and organic substances, etc. [1]. However, the determination of all the parameters is time-consuming, and sometimes an express test is needed.
A change in pH can be a simple and fast signal of a pollution appearance in water [2,3]. Thus, the availability of cheap, accurate, and stable pH sensors is of high importance. pH sensors, developed in the last decades, vary in detection principles, sensing materials, and fabrication methods [3][4][5][6][7][8][9]. One of the most effective and inexpensive techniques of pH detection is the potentiometric method where the pH of the sample is determined by measuring the potential difference between a pH-sensitive electrode and a reference electrode. The value of the pH is calculated from the Nernst equation [10].
Among the sensing materials investigated for pH detection, metal oxide-based electrodes are well-known for their potential to overcome such drawbacks of glass electrode as fragility (and related risk of contamination with dangerous shattered glass) and poor stability in harsh conditions (strong alkaline or acidic solutions, high pressure, and high temperature) [11,12]. from the point of view of the firing process, does not negatively impact the performance of the pH electrode.

Preparation of RuO 2 Paste
Ruthenium oxide paste for screen printing was prepared by mixing anhydrous RuO 2 powder (density: 6.95 g/cm 3 , Sigma Aldrich, St. Louis, MO, USA) with ethyl cellulose (analytical grade purity) and terpineol (anhydrous, Fluka Analytical, Switzerland) in an agate mortar. Mixing was carried out for 20 min to achieve optimal consistency of the paste.

Fabrication of RuO 2 Electrodes for Potentiometric Sensors
Standard alumina (Al 2 O 3 , 96%) plates were chosen as the substrates for the pH electrodes due to their compatibility with thick films and high tolerance toward various environmental conditions. First, Ag/Pd thick-film paste (9695, Electro-Science Laboratories, King of Prussia, PA, USA) was screen-printed on the substrates, dried at 120 • C for 15 min, and fired at 860 • C for 30 min. Freshly prepared RuO 2 paste was then screen printed on the substrates in such a way that the RuO 2 layer slightly overlapped the Ag/Pd conducting layer. After drying at 120 • C for 15 min, electrodes were sintered at 800 • C (RuO 2 -800), 850 • C (RuO 2 -850), or 900 • C (RuO 2 -900) for one hour. Three RuO 2 electrodes were prepared for each sintering temperature. Next, electrical contact was attached to an open end of the conducting layer by soldering a copper wire. To avoid any contact between the conducting layer and the electrolyte, the electrical contact and the conducting layer were covered with noncorrosive polydimethylsiloxane coating (DOWSIL™ 3140 RTV Coating, Dow Chemical Company, Midland, MI, USA), and the sensitive area was left uncovered. Finally, the silicone resin cover was hardened at room temperature for 48 h. The schematic representation of the various stages of fabrication of RuO 2 electrodes is presented in Figure 1.
using a glass electrode. Additionally, cross-sensitivity of the fabricated RuO2 electrodes to the ionic contaminants that can be found in environmental water due to overfertilization of agricultural lands was evaluated. Another important objective was to show that lowering the sintering temperature, which enables reduction of the environmental and economic impact from the point of view of the firing process, does not negatively impact the performance of the pH electrode.

Preparation of RuO2 Paste
Ruthenium oxide paste for screen printing was prepared by mixing anhydrous RuO2 powder (density: 6.95 g/cm 3 , Sigma Aldrich, St. Louis, MO, USA) with ethyl cellulose (analytical grade purity) and terpineol (anhydrous, Fluka Analytical, Switzerland) in an agate mortar. Mixing was carried out for 20 min to achieve optimal consistency of the paste.

Fabrication of RuO2 Electrodes for Potentiometric Sensors
Standard alumina (Al2O3, 96%) plates were chosen as the substrates for the pH electrodes due to their compatibility with thick films and high tolerance toward various environmental conditions. First, Ag/Pd thick-film paste (9695, Electro-Science Laboratories, King of Prussia, PA, USA) was screen-printed on the substrates, dried at 120 °C for 15 min, and fired at 860 °C for 30 min. Freshly prepared RuO2 paste was then screen printed on the substrates in such a way that the RuO2 layer slightly overlapped the Ag/Pd conducting layer. After drying at 120 °C for 15 min, electrodes were sintered at 800 °C (RuO2-800), 850 °C (RuO2-850), or 900 °C (RuO2-900) for one hour. Three RuO2 electrodes were prepared for each sintering temperature. Next, electrical contact was attached to an open end of the conducting layer by soldering a copper wire. To avoid any contact between the conducting layer and the electrolyte, the electrical contact and the conducting layer were covered with noncorrosive polydimethylsiloxane coating (DOWSIL™ 3140 RTV Coating, Dow Chemical Company, Midland, MI, USA), and the sensitive area was left uncovered. Finally, the silicone resin cover was hardened at room temperature for 48 h. The schematic representation of the various stages of fabrication of RuO2 electrodes is presented in Figure 1.

Microstructural Studies
Scanning electron microscopy (SEM) and energy dispersive spectroscopy (EDS) (Nova Nano SEM 200 with EDAX Genesis EDS system, FEI, Hillsboro, OR, USA) were

Microstructural Studies
Scanning electron microscopy (SEM) and energy dispersive spectroscopy (EDS) (Nova Nano SEM 200 with EDAX Genesis EDS system, FEI, Hillsboro, OR, USA) were used to examine the microstructure and elemental composition of the fabricated electrodes.

Electrochemical Studies
Electrode potential was measured by standard potentiometric technique. The fabricated RuO 2 electrodes were used as pH-sensitive electrodes; meanwhile, an ion-selective glass electrode (ISE, Ag|AgCl|KCl, Hydromet, Poland) was used as a reference electrode. To minimize signal loss, electrodes were attached to a unity gain buffer amplifier (LMC6044, Texas Instruments, Dallas, TX, USA) which was further connected to a voltage input module (9205, National Instruments, Austin, TX, USA). LABVIEW software (National Instruments, Austin, TX, USA) was utilized to record all measured data.

pH Measurements
Potentiometric determination of pH relies on selective identification of H + ions present in the investigated solution [39]. The standard potentiometric setup consists of an electrochemical cell and a measuring device (potentiometer, voltmeter, multimeter, etc.). The electrochemical cell consists of a sensing electrode, sensitive to pH change, and a reference electrode (usually silver chloride electrode).
The electrical characteristic of an electrochemical cell is electromotive force (Emf ). The Emf of the cell is determined as the difference in electrode potentials (E) of the two half-reactions proceeding at the sensing and reference electrodes. Usually, the reference electrode is grounded, its potential is considered equal to zero, and the Emf of the cell is equal to the potential of the sensing electrode.
The half-reaction taking place on the sensing electrode is quantitatively explained by the Nernst equation: where E 0 is standard potential, V; R is the universal gas constant, 8.314 J/K·mol; T is temperature, K; n is the number of the electrodes participating in the redox reaction; F is the Faraday constant, 96,485 C/mol; and [Red] and [Ox] are the activities of reduced and oxidized forms of the electrode material, respectively, mol/L. The standard potential is a measure of the individual potential of the reversible electrode (in equilibrium) in the standard state (concentration 1 mol/L, pressure 1 atmosphere and temperature 25 • C). For the RuO 2 electrode, the mechanism of pH-sensing can be explained by a simplified equation [11]: The Nernst equation for this process takes the following form: The value of 58.3 mV is called electrode sensitivity or theoretical Nernst response at T = 22 • C. At the given temperature, the sensitivity value should be the same for all the pH-sensitive electrodes when n = 1. However, in practice deviation from the theoretical response is observed (see Table S1) [33,34,40].
As a conditioning protocol, all electrodes were immersed in distilled water for 24 h prior to their first measurement in order to hydrate pH-sensitive surfaces.
The sensitivity of the fabricated electrodes was determined by measuring the Emf of the electrochemical cell (the potential difference between the reference electrode and the fabricated pH-sensitive electrode) as a function of pH. For that, electrodes were submerged into buffer solutions of pH range from 1 to 14. Buffers were purchased from Chempur (PiekaryŚląskie, Poland) and used as received. The pH of the solution was monitored with a combined glass electrode (ELMETRON, Zabrze, Poland).
The Emf was recorded for 5 min with data points being collected every 10 s. The Emf at the specific pH was determined as the average value of the last 10 data points. Electrode sensitivity, E 0 , and linearity of the response were determined by plotting the electrode potential as a function of pH and calculating the equation describing this dependency using the least-squares approach. Electrode sensitivity was calculated as the slope of the linear equation, E 0 was calculated as the potential at pH = 0 by extrapolating the data, and the linearity of the response of the electrode to pH change was calculated as correlation coefficient (R 2 ).

Response Time, Drift Rate, and Hysteresis
The response time was determined as the time needed for the electrode potential to reach 90% of the stable value.
To measure the drift of electrode response in time, electrodes were left overnight in distilled water, and the drift rate (in mV/h) was calculated using the slope of the line-of-best-fit approach.
To study the hysteresis, the memory effect of an electrode, the fabricated electrodes were exposed to a series of pH buffers. First, the electrodes were exposed to a pH change from acidic to basic (pH change 1.1 → 4.1 → 7.0 → 10.0 → 13.4), and then, the same pH changes were carried out in the opposite direction. Electrode response was recorded for 3 min after submerging the electrode into a new buffer solution. Electrodes were washed with distilled water and dried with a pressure gun after each measurement.

Cross-Sensitivity
The interference of ions with the performance of the fabricated RuO 2 electrodes was evaluated by measuring the Emf of the cell and determining the sensitivity of the electrodes in the presence of KCl, KNO 3 , NH 4 NO 3 , and (NH 4 ) 3 PO 4 . For that, the abovementioned salts were added to buffer solutions to reach the concentration of 0.01 M. Electrodes were immersed into the samples for 5 min, and their Emf response was recorded.

Measurements of Real-Life Samples
The fabricated electrodes were used to measure the pH values of different types of water samples: distilled and tap water, mineral water and water from a river and two lakes. Samples were stored at 4 • C prior to any measurement.
Commercially available "Wysowianka" still water was used as mineral water. River water samples were collected from the Vistula River (Kraków, Poland).
Lake water samples were collected from Zakrzówek Lake (Lake Z, Kraków, Poland, Figure 2a) from the surface and the depth of 6 m and the surface of a mountain lake-Lake M (Lake Morskie Oko in Tatra Mountains, Poland, Figure 2b).

Microstructure of RuO2 Electrodes
Thermal treatment temperature was previously shown to affect the properties of metal oxide solid-state electrodes due to change in relative density and/or crystallinity

Microstructure of RuO 2 Electrodes
Thermal treatment temperature was previously shown to affect the properties of metal oxide solid-state electrodes due to change in relative density and/or crystallinity with temperature. The sensing layer morphology significantly impacts its sensing performance [17,21,41,42]. For the fabricated RuO 2 electrodes, there was only a slight change in surface morphology and microstructure of the layers sintered at different temperatures in the range of 800-900 • C. The EDS analysis confirmed the presence of both ruthenium and oxygen in the fabricated pH-sensitive layers. SEM images of fractured cross-sections of the electrodes screen printed on Al 2 O 3 substrates ( Figure 3) indicated that both the RuO 2 -800 and the RuO 2 -900 electrodes were characterized by small porosity and uniform, fine-grained microstructure with grain sizes of 0.5-2 µm. However, for RuO 2 -800 electrodes (Figure 3a), improved adhesion to the substrate was observed, probably due to their lower porosity related to densification proceeding with a higher contribution of the amorphous phase at grain boundaries. Higher content of pores in the sensing layer can entail scattering of the charge carriers, leading to reduced carrier mobility and decreased sensitivity [41].

Microstructure of RuO2 Electrodes
Thermal treatment temperature was previously shown to affect the properties metal oxide solid-state electrodes due to change in relative density and/or crystallin with temperature. The sensing layer morphology significantly impacts its sensing p formance [17,21,41,42]. For the fabricated RuO2 electrodes, there was only a slight chan in surface morphology and microstructure of the layers sintered at different tempe tures in the range of 800-900 °C. The EDS analysis confirmed the presence of both thenium and oxygen in the fabricated pH-sensitive layers. SEM images of fractur cross-sections of the electrodes screen printed on Al2O3 substrates ( Figure 3) indicat that both the RuO2-800 and the RuO2-900 electrodes were characterized by small poros and uniform, fine-grained microstructure with grain sizes of 0.5-2 µm. However, RuO2-800 electrodes (Figure 3a), improved adhesion to the substrate was observe probably due to their lower porosity related to densification proceeding with a high contribution of the amorphous phase at grain boundaries. Higher content of pores in t sensing layer can entail scattering of the charge carriers, leading to reduced carrier m bility and decreased sensitivity [41].

Sensitivity of the Fabricated Electrodes
Sensitivity is the key characteristic of an electrode that allows determining if the electrode is working properly. The pH sensitivity of the fabricated RuO 2 electrode was determined by exposing the electrodes to buffer solutions of different pH and calculating the slope of the Nernst equation for the electrodes based on measured Emf. The results are presented in Figure 4 and Table 1. It can be seen that the sensitivity is close to the theoretical response with good linearity (R 2 = 0.994 − 0.996) for all the electrodes. Furthermore, the sensitivity values are slightly decreasing with an increase in sintering temperature. Therefore, decreasing the sintering temperature of the RuO 2 pH electrodes does not negatively impact their sensitivity and can be implemented to reduce the power consumption that is needed to achieve higher firing temperatures in the furnace. are presented in Figure 4 and Table 1. It can be seen that the sensitivity is close to the theoretical response with good linearity (R 2 = 0.994 − 0.996) for all the electrodes. Furthermore, the sensitivity values are slightly decreasing with an increase in sintering temperature. Therefore, decreasing the sintering temperature of the RuO2 pH electrodes does not negatively impact their sensitivity and can be implemented to reduce the power consumption that is needed to achieve higher firing temperatures in the furnace.  The sensitivity of the pH electrodes under investigation became stable almost immediately. Moreover, the response is faster than that of the RuO2 electrodes fabricated from the commercial paste containing a glass addition [43].

Long-Term Stability and Repeatability
It was previously shown that solid-state electrodes might require some time to reach stable sensitivity values [17,29]. However, no significant change in sensitivity for all the fabricated electrodes was noticed over two weeks of usage ( Figure 5, Table S2). The daily variation in temperature was taken into account, and the observed sensitivity values were normalized according to expected theoretical sensitivity at a given temperature.
The sensitivity values of RuO2-800, RuO2-850, and RuO2-900 electrodes remained close to the theoretical value and did not significantly change over time. Furthermore, the decrease in sensitivity with sintering temperature, observed during the first usage of the electrodes, did not occur during further measurements.  The sensitivity of the pH electrodes under investigation became stable almost immediately. Moreover, the response is faster than that of the RuO 2 electrodes fabricated from the commercial paste containing a glass addition [43].

Long-Term Stability and Repeatability
It was previously shown that solid-state electrodes might require some time to reach stable sensitivity values [17,29]. However, no significant change in sensitivity for all the fabricated electrodes was noticed over two weeks of usage ( Figure 5, Table S2). The daily variation in temperature was taken into account, and the observed sensitivity values were normalized according to expected theoretical sensitivity at a given temperature. The sensitivity values of RuO 2 -800, RuO 2 -850, and RuO 2 -900 electrodes remained close to the theoretical value and did not significantly change over time. Furthermore, the decrease in sensitivity with sintering temperature, observed during the first usage of the electrodes, did not occur during further measurements.
The porosity and surface oxidation states of the RuO 2 layer can be impacted due to the ageing of an electrode [17]. Nevertheless, in our study, ageing did not impact the pH sensing properties of the electrodes even though they were only conditioned once before the first measurement. The pH sensitivity of RuO 2 -850 and RuO 2 -900 electrodes fabricated six months earlier and stored in the air was only reduced by 2.7 and 6.5 mV/pH, respectively. For general water quality monitoring, the pH is not expected to change very rapidly and frequently. However, the fabricated electrodes were repeatedly subjected to fast changes in pH range 1-13.4. The stability in sensitivity over days thereby exhibits the endurance of RuO 2 to multiple measurements in electrolytes with varying pH across a wide range. Therefore, fabricated electrodes exhibit not only long-term stability in operation but also have a long shelf life. Moreover, the loss in sensitivity over 6 months was lower at a lower sintering temperature. The porosity and surface oxidation states of the RuO2 layer can be impacted due to the ageing of an electrode [17]. Nevertheless, in our study, ageing did not impact the pH sensing properties of the electrodes even though they were only conditioned once before the first measurement. The pH sensitivity of RuO2-850 and RuO2-900 electrodes fabricated six months earlier and stored in the air was only reduced by 2.7 and 6.5 mV/pH, respectively. For general water quality monitoring, the pH is not expected to change very rapidly and frequently. However, the fabricated electrodes were repeatedly subjected to fast changes in pH range 1-13.4. The stability in sensitivity over days thereby exhibits the endurance of RuO2 to multiple measurements in electrolytes with varying pH across a wide range. Therefore, fabricated electrodes exhibit not only long-term stability in operation but also have a long shelf life. Moreover, the loss in sensitivity over 6 months was lower at a lower sintering temperature.

Drift and Response Time
For continuous measurements that can last for hours, an important characteristic of an electrode is the drift rate. Drift rate is used to evaluate if the readings of an electrode maintain the same over a long period of observation. The values of the drift rate for the electrodes are presented in Table 1. These values remained small and did not exceed 1 mV/h. Both the potential drift of an electrode and the response time are closely related to the mechanisms that govern them. Based on the material composition, drift of the Emf during pH measurement in RuO2 electrodes can be caused due to slow hydration of the pH-sensitive layer or slow H + ions diffusion [17]. Other critical aspects of the RuO2 electrode that determine the drift of potential are homogeneity, porosity and thickness of the pH-sensitive layer, and composition and structure of the pH-sensitive material [17]. For a given temperature and electrode composition, the drift rate is expected to be small as it does not significantly depend on the interface between the pH-sensitive surface and the electrolyte, while pores at grain boundaries trapping H2 govern the H + ions transport through this layer [17]. Therefore, the low drift rate related to fast surface hydration and H + diffusion experienced by the fabricated RuO2 electrodes could be attributed to their single-phase composition without additives and the uniform microstructure.
Response time is the characteristic that tells how long one needs to wait before the electrode reaches stable potential readings. Response time plays an important role when a fast change in pH must be detected. All the fabricated electrodes showed a response time of 2-36 s ( Table 1).
The disadvantage of pure RuO2 could be lesser adhesion to the substrate [17]. To overcome it, glass can be added to the screen-printing paste [33,35]. However, the addition of glass has been reported to increase the response time due to reduced porosity of

Drift and Response Time
For continuous measurements that can last for hours, an important characteristic of an electrode is the drift rate. Drift rate is used to evaluate if the readings of an electrode maintain the same over a long period of observation. The values of the drift rate for the electrodes are presented in Table 1. These values remained small and did not exceed 1 mV/h. Both the potential drift of an electrode and the response time are closely related to the mechanisms that govern them. Based on the material composition, drift of the Emf during pH measurement in RuO 2 electrodes can be caused due to slow hydration of the pH-sensitive layer or slow H + ions diffusion [17]. Other critical aspects of the RuO 2 electrode that determine the drift of potential are homogeneity, porosity and thickness of the pH-sensitive layer, and composition and structure of the pH-sensitive material [17]. For a given temperature and electrode composition, the drift rate is expected to be small as it does not significantly depend on the interface between the pH-sensitive surface and the electrolyte, while pores at grain boundaries trapping H 2 govern the H + ions transport through this layer [17]. Therefore, the low drift rate related to fast surface hydration and H + diffusion experienced by the fabricated RuO 2 electrodes could be attributed to their single-phase composition without additives and the uniform microstructure.
Response time is the characteristic that tells how long one needs to wait before the electrode reaches stable potential readings. Response time plays an important role when a fast change in pH must be detected. All the fabricated electrodes showed a response time of 2-36 s ( Table 1).
The disadvantage of pure RuO 2 could be lesser adhesion to the substrate [17]. To overcome it, glass can be added to the screen-printing paste [33,35]. However, the addition of glass has been reported to increase the response time due to reduced porosity of the printed layer and diffusion rate of protons [6]. The choice of the best RuO 2 -based composition is therefore strongly dependent on the requirements of the user and the area of the sensor application. In our research, we have similarly experienced that the layer made of commercial RuO 2 paste containing glass particles adheres better to the alumina substrate than a layer of pure RuO 2 . However, the slightly weaker mechanical strength and integrity of the latter do not impact its sensitivity to pH as shown in Section 3.2.
The response at the pH < 7.0 is almost instantaneous (2 s). However, as the pH increases, the response time also increases, and at pH 13, the maximum response time of 52 s is observed. The negative effect of pH increase on the response time can be attributed to the low concentration of H + ions at higher pH values [6]. Furthermore, compared to the OH − ions that are bigger and diffuse more slowly, H + ions are small, which makes their diffusion in the RuO 2 layer faster and therefore facilitates relevant ion exchange [6].

Hysteresis
Another important characteristic of an electrode is hysteresis. Differences in the Emf values at the same pH occur when the measurement is repeated multiple times. The analysis of this phenomenon, also known as the memory effect, allows for estimating if the previous measurement affects the consequent. To evaluate the hysteresis, the fabricated electrodes were exposed to a series of pH buffer changes. The RuO 2 electrodes exhibited a small hysteresis effect ( Figure 6). All the fabricated electrodes showed hysteresis values not exceeding 21 mV when the pH changing started from the acidic region (1.1 → 13.4 → 1.1). For the opposite direction of pH change (13.4 → 1.1 → 13.4), the hysteresis values did not exceed 25 mV. These results show that all electrodes have a proper response to pH changes. The lowest hysteresis was observed for the RuO 2 -850 electrodes.

Hysteresis
Another important characteristic of an electrode is hysteresis. Differences in the Emf values at the same pH occur when the measurement is repeated multiple times. The analysis of this phenomenon, also known as the memory effect, allows for estimating if the previous measurement affects the consequent. To evaluate the hysteresis, the fabricated electrodes were exposed to a series of pH buffer changes. The RuO2 electrodes exhibited a small hysteresis effect ( Figure 6). All the fabricated electrodes showed hysteresis values not exceeding 21 mV when the pH changing started from the acidic region (1.1 → 13.4 → 1.1). For the opposite direction of pH change (13.4 → 1.1 → 13.4), the hysteresis values did not exceed 25 mV. These results show that all electrodes have a proper response to pH changes. The lowest hysteresis was observed for the RuO2-850 electrodes.
The pH measurement loop and crystalline properties of the pH-sensitive material are two factors that mainly influence the hysteresis effect in metal oxide-based pH electrodes [6]. The hysteresis effect in this study was higher for all electrodes when pH change started from the basic to acidic region (13.4 → 1.1 → 13.4) because OH − ions H + ions diffuse slower than H + ions in the RuO2 layer. This finding is in agreement with previous studies [6,20,23]. However, it can be seen that similar to the response time, the hysteresis effect was also lower for higher sintering temperature, which may be attributed to higher porosity.

Cross-Sensitivity
One of the main limitations of the usage of metal oxide pH electrodes is the sensitivity of the electrode to interfering ions. The interferences caused by various ions were investigated by measuring the sensitivity in the presence of K + and NH4 + cations and Cl -, NO3 − , and PO4 3− anions. The observed pH sensitivities in the presence of the interfering ions for all three types of electrodes are listed in Table 2. Furthermore, Table S1 presents The pH measurement loop and crystalline properties of the pH-sensitive material are two factors that mainly influence the hysteresis effect in metal oxide-based pH electrodes [6]. The hysteresis effect in this study was higher for all electrodes when pH change started from the basic to acidic region (13.4 → 1.1 → 13.4) because OH − ions H + ions diffuse slower than H + ions in the RuO 2 layer. This finding is in agreement with previous studies [6,20,23]. However, it can be seen that similar to the response time, the hysteresis effect was also lower for higher sintering temperature, which may be attributed to higher porosity.

Cross-Sensitivity
One of the main limitations of the usage of metal oxide pH electrodes is the sensitivity of the electrode to interfering ions. The interferences caused by various ions were investigated by measuring the sensitivity in the presence of K + and NH 4 + cations and Cl − , NO 3 − , and PO 4 3− anions. The observed pH sensitivities in the presence of the interfering ions for all three types of electrodes are listed in Table 2. Furthermore, Table S1 presents the comparison of electrode properties and sensitivity characteristics of RuO 2 -based pH sensors reported previously by other authors with those obtained in this work. The Emf versus pH plots for the RuO 2 -800 electrode in the presence of interfering ions are presented in Figure 7. For RuO 2 -850 and RuO 2 -900 electrodes, similar behavior was observed. All the fabricated electrodes exhibited close to the Nernstian response and did not show any significant deviation from the theoretical sensitivity in distilled water when no salt was added ( Table 2).
In general, the effect of the interfering salts was small, especially for RuO 2 -800 sensing electrodes fired at the optimal temperature of 800 • C. However, in the case of RuO 2 -850 and RuO 2 -900 electrodes, a decrease in sensitivity was observed. The maximum impact observed was a reduction of 5.1 mV/pH for RuO 2 -850 in the presence of 0.01 mol/L of KCl. K + ion is smaller in size as compared to NH 4 + , and therefore, it has higher mobility. Due to its faster diffusion into the double layer at the sensing surface of the RuO 2 electrode, it influences the sensitivity of the electrode more than the bigger and slower NH 4 + cation. The most distinct deviation in the E 0 potential observed for (NH 4 ) 3 PO 4 (for RuO 2 -800 and RuO 2 -900 electrodes) implies that the phosphate changes the contribution of Ru 4+ and Ru 3+ ions governing the redox reaction in the sensing layer. Such shifts in the E 0 values were attributed by Lonsdale et al. [21] to the changes in the Ru 4+ /Ru 3+ ratio caused by oxidizing/reducing agents. composition of the sensing layer based on pure RuO2 used to fabricate the sensors. In the study by Labrador et al. [38], where commercial RuO2 paste was used, interference from halides, carbonates, and sulfates were attributed to the presence of lead (II) oxide (PbO), which causes the formation of insoluble products with the anions. Commercial RuO2 paste contains lead borosilicates because it aids in modifying material properties such as temperature coefficient of expansion of the glass phase [44]. Pocrifka et al. [26] reported a pH electrode developed by mixing RuO2 with 30 mol.% TiO2 that was insensitive toward cations (Li + , Na + , and Ca 2+ ) without a significant improvement in the pH sensitivity when compared to the undoped RuO2 electrode. In this study, it was stated that a pure RuO2 sensing electrode is capable of being uninfluenced by various interfering ions without the addition of lead borosilicate-based glass phase or TiO2. This indicates that the addition of PbO to RuO2 only assists in improving the thick-film properties of the pH electrode, such as sheet resistivity, adhesion, low temperature coefficient of resistance, and low noise indices [44]. Similarly, the advantage of adding TiO2 is less related to improvement in sensitivity and selectivity, and it has more to do with lowering the cost [26] and avoiding corrosion by oxidation of RuO2 to RuO4 or RuO4 2− [45]. Therefore, the additives to pure RuO2 may have an adverse, favorable, or neutral impact on the cross-sensitivity of the pH electrode depending on the properties of the additive, and the decision of their use rests upon the requirements and priorities of the researchers and manufacturers.
All the compounds used in the cross-sensitivity experiment performed in this study are also agricultural fertilizers used for plant growth, and the combined presence of nitrogen, phosphorus, and potassium in the composition gives such fertilizers the common name N-P-K fertilizers. Albeit their great contribution to tackling the ever-growing global demand for food, N-P-K fertilizers are also notorious for nutrient pollution through groundwater contamination and surface run-off from farms to in-land waters. In order to be absorbed by the plant's root hair cells, fertilizers must be soluble in water, and the water solubility of KNO3, KCl, and NH4NO3 is 100%, whereas for (NH4)3PO4, it is 35% The lack of interference from various ions in the solution may be due to the undoped composition of the sensing layer based on pure RuO 2 used to fabricate the sensors. In the study by Labrador et al. [38], where commercial RuO 2 paste was used, interference from halides, carbonates, and sulfates were attributed to the presence of lead (II) oxide (PbO), which causes the formation of insoluble products with the anions. Commercial RuO 2 paste contains lead borosilicates because it aids in modifying material properties such as temperature coefficient of expansion of the glass phase [44]. Pocrifka et al. [26] reported a pH electrode developed by mixing RuO 2 with 30 mol.% TiO 2 that was insensitive toward cations (Li + , Na + , and Ca 2+ ) without a significant improvement in the pH sensitivity when compared to the undoped RuO 2 electrode.
In this study, it was stated that a pure RuO 2 sensing electrode is capable of being uninfluenced by various interfering ions without the addition of lead borosilicate-based glass phase or TiO 2 . This indicates that the addition of PbO to RuO 2 only assists in improving the thick-film properties of the pH electrode, such as sheet resistivity, adhesion, low temperature coefficient of resistance, and low noise indices [44]. Similarly, the advantage of adding TiO 2 is less related to improvement in sensitivity and selectivity, and it has more to do with lowering the cost [26] and avoiding corrosion by oxidation of RuO 2 to RuO 4 or RuO 4 2− [45]. Therefore, the additives to pure RuO 2 may have an adverse, favorable, or neutral impact on the cross-sensitivity of the pH electrode depending on the properties of the additive, and the decision of their use rests upon the requirements and priorities of the researchers and manufacturers.
All the compounds used in the cross-sensitivity experiment performed in this study are also agricultural fertilizers used for plant growth, and the combined presence of nitrogen, phosphorus, and potassium in the composition gives such fertilizers the common name N-P-K fertilizers. Albeit their great contribution to tackling the ever-growing global demand for food, N-P-K fertilizers are also notorious for nutrient pollution through groundwater contamination and surface run-off from farms to in-land waters. In order to be absorbed by the plant's root hair cells, fertilizers must be soluble in water, and the water solubility of KNO 3 , KCl, and NH 4 NO 3 is 100%, whereas for (NH 4 ) 3 PO 4 , it is 35% [46]. Therefore, the low cross-sensitivity of fabricated RuO 2 pH electrodes makes them suitable for measurement in water that has been contaminated by fertilizers and is rich in nutrients. Applicability of such pH sensing without interference from ions such as K + , NO 3 -, NH 4+ , and PO 4 3− ranges from water quality monitoring in agriculture to municipal and industrial wastewater treatment.

Performance of pH Electrodes in Real-Life Water Samples
The fabricated electrodes were tested in several water samples to evaluate their applicability to real-life measurements. The results obtained for distilled, mineral, tap, river, and lake water are presented in Table 3. The pH values measured with the fabricated electrodes were similar to those measured with a conventional glass electrode with the maximum discrepancy of 0.11 pH units observed for Lake M water samples. Furthermore, all the fabricated electrodes showed good uniformity of the measured pH value, with the standard deviation not exceeding 0.13 pH units. The highest difference between the measurement by glass electrode and the fabricated RuO 2 electrodes was only 0.11 units for surface water of Lake Z and Lake M measured by RuO 2 -900 electrode. The RuO 2 -850 electrode exhibited the closest to the glass electrode pH value and a very small variation (±0.01 pH units) in parallel readings for all water samples except Lake M water.
The industrially manufactured and commercially available mineral water tested in this study contains cations such as Na + , Ca + , Mg + , and K + , and anions such as HCO 3 − , SO 4 2− , Cl − , and F − . In 2017, an analysis of municipal tap water in Krakow, Poland, revealed the presence of calcium, magnesium, and nitrates in it [47]. The quality of water in the Vistula River is quite impacted by urbanization, and a high content of contaminants such as NH 4 + , K + , Ca 2+ , Mg 2+ , Zn 2+ , Mn 2+ , Fe 2+ , Fe 3+ , etc., can be found in it [48]. Until 1991, Lake Z was an upper Jurassic limestone quarry, and a 2003 study by Galas [49] reported high conductivity and chloride content of water due to infiltration from the nearby Vistula River. Lake M was reported to have a cadmium concentration of about 0.13 mg/dm 3 due to anthropogenic activities [50].
The obtained results of the measurements of the water samples highlight the effectiveness of the fabricated RuO 2 pH electrodes in the presence of various chemical species and elements. Therefore, reported electrodes are capable of functioning accurately in environmental, industrial, and municipal water quality monitoring.

Conclusions
Thick-film pH-sensitive electrodes based on pure RuO 2 without the addition of glass or other metal oxides were successfully prepared by screen printing and subsequent sintering at different temperatures. Excellent sensing characteristics were attained owing to the single-phase composition of the electrodes, their high conductivity, and good electrochemical and catalytic properties of RuO 2 . The advantageous properties comprised the Nernstian behavior in a broad pH range of 1-13 with the sensitivity of 56.1-61.8 mV/pH, fast response of 2 s for pH ≤ 7, low hysteresis, good long-term stability, and high resistance to cross-sensitivity from various interfering cations and anions. The analysis of the influence of sintering temperature of RuO 2 electrodes on the performance of potentiometric pH electrodes showed small differences for the samples sintered at 800, 850, and 900 • C. However, a sintering temperature of 800 • C should be preferred not only from the economic and ecological point of view but also for the better adhesion of the RuO 2 layer to the alumina substrate and its lowest cross-sensitivity. Furthermore, the excellent applicability of the fabricated electrodes for the pH measurements of real-life water samples (tap, mineral, river, and lake water) was proved in this work. The water samples used in the study exposed the RuO 2 pH-sensitive layers to a variety of chemical elements, and the precise pH measurement in such conditions showed their ability to function correctly in different environments. In most water samples, the electrodes sintered at 850 • C showed the best precision and repeatability. The developed screen-printed potentiometric electrodes seem to be promising candidates for online monitoring systems of water quality due to their excellent pH sensing performance and potential for miniaturization and wireless measurement data transfer.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/s21165399/s1: Table S1. Summary of RuO 2 -based pH electrode characteristics reported previously and in this work, Table S2. Change of the sensitivity of the fabricated electrodes with time.