Miniaturized water quality monitoring pH and conductivity sensors

Abstract

The development of cost-effective, durable and sensitive on-line water quality monitoring sensors that can be installed across water distribution networks has attracted attention to increase the frequency of monitoring and hence reduce the risk of accidental or deliberate contaminations or improve routine control of water quality. This paper presents microsensors fabricated and implemented in a serpentine channel interface for the measurement of pH and conductivity, two important water quality parameters. The performance of these sensors was tested in both still (static) and flowing (dynamic) water. The serpentine channel interface, numerically optimized, provides a constant flow and a constant outlet pressure condition for dynamic experiments. Tests conducted for evaluating the effect of the exposure time on the sensor performance show no change in the sensor response even after one month. Finally, the pH and conductivity sensors were compared against the common commercial sensors used for evaluating water quality. The results show that the pH and conductivity sensors are as precise as the commercial sensors in both static and dynamic conditions. However, the cost of the sensors presented here is significantly lower than that of the commercial sensors.

Introduction

Maintaining proper drinking water quality is of utmost importance to prevent any outbreak. Chemical quality of drinking water is usually monitored by measuring turbidity, free chlorine content, pH, conductivity and temperature [1], [2]. Online water quality monitoring or early warning systems are becoming a necessity for large water distribution system (WDS) [3] to prevent accidental [4] or deliberate contamination. Thus, the implementation of nearly real time water quality monitoring sensors across WDS has attracted much attention. The sensors developed in the laboratories have been tested in the steady state conditions and for laminar flow at atmospheric pressure. Installing these sensors for online water quality monitoring in a real WDS requires certain conditions, e.g., reduction in pressure and flow rate. Available flow control devices [5], [6] are complicated and expensive, and the implementation of online water quality sensors highly depends on their costs. One of the objectives of this study is to develop a simple interface which can reduce the WDS pressure to the atmospheric pressure and maintain the laminar flow even when the change in the flow rate and pressure occurs in water mains. The other objective of this study is to fabricate cost effective, low maintenance, durable and online pH and conductivity sensors, that can be implemented in the interface for continuous online monitoring of drinking water. The recommended range for pH of drinking water is 6.5–8.5 [7], [8], [9], [10]. A change in the response time or the response of the sensor may indicate biofilm development [11] or leaching and nitrification [12] respectively. In addition, pH of water has an impact on the degree of corrosion of metal pipes [13], disinfection efficiency [14] and the speciation of disinfection by-products. Conductivity of the water, on the other hand, changes with inorganic dissolved solids [15]. Conductivity indicates the amount of ionic salts in water and can be an important indication of ionic salt contamination. The conductivity of drinking water is usually between 0.05 and 0.5 mS/cm. However, there is no guideline since conductivity can vary based on the amount of non-toxic salts present.

There are different pH sensors developed based on the potentiometric method [16], the amperometric method [10], or the change of polymer (e.g., hydrogel) properties. Potentiometric pH sensors usually consist of a reference electrode, and a working electrode having a pH-sensitive polymer layer deposited on it. Several polymers have been considered appropriate for fabricating potentiometric pH sensors as outlined in [16], [17], [18], [19], [20], [21]. Because of the formation of the metal oxide layer of potentiometric pH sensors, they may eventually become inaccurate which makes them less favorable for continuous online monitoring applications. In amperometric sensors, a constant and/or varying potential is applied at an electrode's surface and the resulting current is measured with a three electrode system [22]. The amperometric pH sensors are usually developed using layers of gold, polypyrole and gold. Despite the accuracy of these sensors, they have not been tested under continuous flow of large volume water samples.

Real time, miniaturized, integrated pH sensors have been fabricated using polymers like hydrogel. There are several hydrogel-based sensors [23], [24], [25], [26], [27] developed for the measurement of different properties of the fluid of interest. For instance, transducers made of a bending plate [28], [29], [30], [31], [32], [33] or microcantilevers [36], [37], [38] have been developed to relate the change in the hydrogel properties (due to the change in pH) to an electrical signal or to optical signal [34], [35]. The electrical properties, including conductivity and capacitance of the hydrogel that changes during the swelling/deswelling process can also be used as indicators for pH change. Relevant to our study are the two sensors developed by Sheppard et al. [39] and Gill et al. [40] for the measurement of pH. The sensors prepared by Sheppard et al. [39] use hydrogel copolymers prepared from 2-hydroxyethyle methacrylate (HEMA) and N,N-dimethyelaminoethlmethacrylate (DMAEMA) by redox initiated free radical solution polymerization [39]. Tetraethylene diacrylate (TEDA) was used as crosslinker. The hydrogel was deposited on the interdigted electrode. The response of the sensor is linear for the pH range of 6.5 to 8. Gill et al. [40] developed a hydrogel-based pH sensor for which the hydrogel was prepared from poly aniline emerald salt (PANI ES) doped with poly pyrrole (pPy) and Poly vinyl butaryl (PVB). The sensor showed good sensitivity to the changes of pH in the range of 2 to 8.

Conductivity sensors are mainly two types: electrode (contacting sensors) and toroidal (inductive sensors). Electrode sensors may have two, three, or four electrodes. Ramos et al. [41] and Li et al. [42] developed a four-terminal water quality monitoring conductivity (electrode) sensor. The range of the conductivity sensing unit developed by Ramos et al. [41] is between 0.05 and 0.5 S/cm, which is higher than the usual conductivity of drinking water. The sensor developed by Li et al. [42] can be used for fresh water. The main advantages associated with this sensor are its wide measurement range, low cost, sensitivity to contact resistance, and linear response, which can be extended by varying the excitation-voltage level of the cell. In the toroidal conductivity sensor, on the other hand, the input and output circuitry does not come in contact with the sampled water. This reduces the probability of fouling. However, electrical interference and signal loss can occur for toroidal conductivity sensors. In this study, interdigitated-electrode sensors are used to determine the conductivity of water. These sensors are small, easy to fabricate, and have simple principle and low cost compared to the toroidal sensors.

One of the objectives of this paper is to develop cost effective, accurate and sensitive hydrogel-based sensors, following the principle of the sensors developed in [39], [40] for online drinking water quality. The other objective is to fabricate a low cost flow and pressure reducing interface which can directly be connect to WDS. For this purpose, an interface was designed, simulated (Section 2), and fabricated and tested (Section 3). Then, the pH and conductivity sensors fabricated (see the details in Section 4) and tested in the static and dynamic conditions provided by the interface (see Section 5). Tests are also performed to determine the effect of the exposure time on the performance of each sensor. Finally the performance of the pH and conductivity sensors (in terms of precision, accuracy, and cost) is compared against the common commercial sensors (see Section 5).