Robust Chemiresistive Sensor for Continuous Monitoring of Free

4 days ago - Free chlorine is widely used in industry as a bleaching and oxidizing agent. Its concentration is tightly monitored to avoid environmenta...
0 downloads 10 Views 1MB Size
Subscriber access provided by READING UNIV

Article

Robust Chemiresistive Sensor for Continuous Monitoring of Free Chlorine using Graphene-Like Carbon Aditya Aryasomayajula, Caroline Wojnas, Ranjith Divigalpitiya, Ponnambalam Ravi Selvaganapathy, and Peter Kruse ACS Sens., Just Accepted Manuscript • DOI: 10.1021/acssensors.7b00884 • Publication Date (Web): 25 Jan 2018 Downloaded from http://pubs.acs.org on February 2, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Sensors is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 8 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

Robust Chemiresistive Sensor for Continuous Monitoring of Free Chlorine using Graphene-Like Carbon Aditya AryasomayajulaΨ, Caroline Wojnas#, Ranjith DivigalpitiyaΦ, Ponnambalam Ravi SelvaganapathyΨ,* and Peter Kruse#,* Ψ #

Department of Mechanical Engineering, McMaster University, Hamilton, Ontario, Canada

Department of Chemistry and Chemical Biology, McMaster University, Hamilton, Ontario, Canada

Φ

3M Canada Company, London, Ontario, Canada KEYWORDS: Graphene, free chlorine, redox sensor, chemiresistive sensor, water quality monitoring, PANI ABSTRACT: Free chlorine is widely used in industry as a bleaching and oxidizing agent. Its concentration is tightly monitored to avoid environmental contamination and deleterious human health effects. Here, we demonstrate a solid state chemiresistive sensor using graphene like carbon (GLC) to detect free chlorine in water. A 15-20 nm thick GLC layer on a PET substrate was modified with a redox-active aniline oligomer (phenyl-capped aniline tetramer, PCAT) to increase sensitivity, improve selectivity and impart fouling resistance. Both the bare GLC sensor and the PCAT-modified GLC sensor can detect free chlorine continuously and unlike previous chemiresistive sensors, do not require a reset. The PCAT-modified sensor showed a linear response with a slope of 13.89 (mg/L)-1 to free chlorine concentrations between 0.2 to 0.8 mg/L which is relevant for free chlorine monitoring for drinking water and waste water applications. The PCAT-modified GLC sensors were found to be selective and showed less than 0.5% change in current in response to species such as nitrates, phosphates and sulphates in water. They also were resistant to fouling from organic material and showed only a 2% loss in signal. Tap water samples from residential area were tested using this sensor which showed good agreement with standard colorimetric measurement methods. The GLC and PCAT-GLC sensors show high sensitivity and excellent selectivity to free chlorine and can be used for continuous automated monitoring of free chlorine.

Chlorine gas and sodium hypochlorite (bleach) are common sources of chlorine in water. When introduced in water, they form a pH-dependent equilibrium of hypochlorous acid (HOCl, dominant at intermediate pH) and hypochlorite (OCl-, dominant at high pH). Together, these two types of chlorine are referred to as free chlorine. Free chlorine is commonly used in many industries as both bleaching and oxidizing agent. The key to successful use of chlorine in industry is in maintaining the right concentration. A high free chlorine concentration results in unnecessary cost of a process, is harmful to the environment and also has implications for human health. On the other hand, low levels of chlorine will not properly disinfect the water. Free chlorine is critical to disinfecting water to prevent disease, but it also poses its own risk as a toxin to aquatic life and the environment.1 Chlorine is the most widely used disinfectant for municipal wastewater because of the low cost and maintenance. Other methods such as ozonation2 and ultraviolet3 disinfection require high equipment setup cost and skilled labor to operate. Amperometric analyzers4,5 and colorimetric tests4 are the two major types of methods used for free chlorine measurement. Colorimetric tests are inexpensive and simple to use. This method, however, requires reagents, manual operation and high maintenance. Amperometric analyzers do not require reagents but are expensive and easily fouled. They also require reference electrodes in order to operate, adding complexity and maintenance issues.

Solid state sensors are a good alternative for measuring free chlorine in water. They sense chemical changes in their environment by transducing them into changes in their electronic properties such as resistance. Carbon nanotubes (CNT)6 and graphene/graphite7-12 have been previously used as a sensing material in solid state sensors for detecting free chlorine. Although these sensors showed good sensitivity, they were either designed for single use or required a manual reset for continuous functioning of the sensor. Furthermore, the method of fabrication was not suitable for large scale manufacture of these sensors with reliability and repeatability. Here we report for the first time a solid state sensor with graphene-like carbon (GLC) as the sensing material functionalized with phenyl-capped aniline tetramer (PCAT) for continuous monitoring of free chlorine in water. This sensor distinguishes itself by not requiring a reset to measure free chlorine, making it more suitable for continuous monitoring systems. The GLC sensors show a linear response to free chlorine concentration in the 0.2 to 0.8 mg/L range. A simple, low-cost manufacturing process for the sensors is shown. The GLC sensors showed excellent response to free chlorine while having a minimal response to other interfering species in water. Tests on tap water samples showed good agreement with a standard colorimetric free chlorine test kit. This sensor is suited for continuous automated monitoring of free chlorine. Experimental Section

ACS Paragon Plus Environment

ACS Sensors 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Materials Sodium hypochlorite solution (10-15%, reagent grade, catalog # 425044) and humic acid (technical grade, catalog # 53680) were purchased from Sigma Aldrich Canada. Methanol (HPLC grade) and DI water (Millipore, resistivity 18.2 MΩ·cm) were used for all dilutions. Screen printing electrode paste (product# 112-48) and dielectric paste (product# 12418T) were purchased from Creative Materials Inc. USA. Kapton double-sided tape (catalog# PIT2SD) was purchased from Caplinq Inc. and polydimethylsiloxane (PDMS, Sylgard180 elastomer kit) was ordered from Dow Corning. A Cricut craft cutter (Provo Craft & Novelty Inc.) was used for cutting the double-sided adhesive tape and the mask for screen printing. Glass slides (VWR, catalog# 89085-399), PTFE tubing (item# RK-06417-31), silicon tubing (item# RK-9640013), 19G x 38 mm needles (Becton Dickinson), 10 ml and 20 ml syringes (Becton Dickinson) and the Hach free chlorine colorimetric kit (item# RK-99574-46) were ordered from Cole Palmer Canada. PCAT used in this work was synthesized as described elsewhere.13 Transfer tape and roller were purchased from Uline Canada. The GLC sheet was supplied by 3M Canada. GLC coating on polyester film is commercially available from 3M as a cover tape for use in semiconductor chip packaging.14 GLC is made of multi-layer graphene platelets on polyester without any binder, manufactured using a proprietary process. This static dissipative GLC layer is about 12 nm thick. GLC coating has a sheet resistance in the order of 100 kΩ/sq and has about 68% optical transmission at 550 nm. SEM and AFM analysis show that the GLC layer is composed of graphitic platelets in a sea of nano-crystalline graphite. X-ray diffraction data indicates that the graphitic platelets are oriented parallel to the substrate. These and more information on GLC are found elsewhere. 15,16,17 Sensor Fabrication Figure 1a shows the different components of the GLC sensor. The fabrication process of the GLC sensor is shown in Figure 1b. The GLC sheet was attached to one side of the double-sided tape (Figure 1b(i)) and pressed uniformly using the roller to remove air bubbles. It is then cut into rectangles (20.3 mm x 15.3 mm) using the craft cutter (Figure 1b(ii)). The rectangular cut GLC sheet is then carefully pasted on top of a glass slide such that it is approximately aligned in the center of the glass slide as shown in Figure 1b(iii). Next, transfer tape is cut according to the design (refer to supplementary information for drawing) to make the screen printing mask. A rectangular strip (2.5 mm x 75 mm) is cut from the transfer tape and stuck on top of the rectangular GLC sheet attached to the glass slide previously. This transfer tape strip protects the GLC sheet from the screen printing paste. Carbon based electrode paste is screen printed on top of the GLC sheet using the mask (Figure 1b(iv)). After applying the paste, the glass slide is heated at 65 °C for 20 minutes. After drying of the paste, the mask was peeled off carefully and a thick coat of dielectric paste is applied such that it covers the entire GLC sheet uniformly (Figure 1b(v)). The glass slide is then baked for 1 hour at 65 °C for complete drying. Once the dielectric paste is completely dry, the thin strip of transfer tape is carefully peeled off from the GLC sheet thereby exposing it. PDMS and curing agent are mixed in a 10:1 ratio and poured into the 3D-printed mold (refer to supplementary information for design) and cured for 2 hours at 85 °C in an

oven. Two pieces of silicone tubing (5 mm) are cut and inserted at the inlet and outlet of the 3d printed mold before the PDMS is poured. Once cured, the PDMS block is peeled off from the mold and attached to the GLC sheet using the wet bonding method18 (Figure 1b(vi)). The entire sensor is then cured for an additional 30 minutes to ensure good bonding between PDMS and the glass slide. PTFE tubing pieces are attached to the inlet and outlet ports for flow experiments. Figure 1c shows the photograph of a GLC sensor fabricated using this method. The PDMS block has a rectangular channel that allows for the reagents to flow over the GLC sheet to measure the change in resistance. The dielectric layer ensures that the electrodes are isolated from the reagents and the reagents are only in contact with the GLC sheet. This way the entire sensor response is assured to be a bulk response from the GLC sheet, without any contributions from the contacts.19 Five sensors were fabricated following the abovementioned procedure, with an average measured resistance of 45 ±10 kΩ. Three of them were functionalized using PCAT molecules by flowing a methanolic solution of PCAT over the GLC sheet for 10 minutes.7,8,9 The other two sensors were not functionalized.

Figure 1. GLC sensor. (a) Schematic of a GLC sensor showing the different components. (b) Fabrication process of a GLC sensor (c) Photograph of a GLC fabricated sensor.

Sensor Measurements The current resulting from the application of a constant 100 mV potential across the GLC film was measured with a Keithley 2636 source measuring unit (Tektronics Inc., USA). A syringe pump (NE-300U, Bio-Lynx Scientific equipment Inc.) was used to provide an analyte flow rate of 0.2 ml/min during the measurements. A 5 minute rest period was given at the start of all the experiments for the sensor to stabilize. Then methanol was flowed for 10 minutes to wet and clean the GLC surface. For sensors with PCAT functionalization, a methanolic solution of PCAT was flowed over the GLC sheet for 10 minutes followed by a 5 minute methanol wash before the DI water step. As the last step in preparing all sensors for measurement, DI water was flown to establish a baseline current. Free chlorine at different concentrations were then flown over the GLC sensor to measure the response by recording the current. Results and Discussion Free chlorine solutions at various concentrations (0.2, 0.4, 0.6 and 0.8 mg/L) were prepared by diluting hypochlorite solution in DI water. The concentrations for these solutions were confirmed using a colorimetric kit. The chosen concentrations mirror government regulations that limit the

ACS Paragon Plus Environment

Page 2 of 8

Page 3 of 8 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors free chlorine concentration in drinking water to a range between 0.2-2 mg/L.20 The experiment was started within one hour of solution preparation. Figure 2 shows the resulting current measured from a PCAT-GLC sensor exposed to different free chlorine concentrations in repeated cycles (low to high and high to low is defined as one cycle). The sensor responded to changes in free chlorine concentration from a low (0.2 mg/L) to high (0.8 mg/L) and high (0.8 mg/L) to low (0.2 mg/L) cycle which was repeated 4 times. To mimic realistic conditions during online measurements, the free chlorine concentrations were introduced in continuous flow directly following each other without reset or baseline calibration. Each measurement took close to 10 minutes to stabilize. The signal was considered stable once it varied by less than ±20 nA/s. To demonstrate stability, each free chlorine concentration was run for 30 minutes for the data shown in Figure 2, well past the minimum amount required for the sensor to stabilize. Additional data on several devices is included in the supporting information. Our experiments establish that the sensor can maintain its accuracy for several hours over many measurement cycles without requiring resets or baseline recalibration with DI water.

water reading and dividing this by the DI water reading. The DI water reading from each sensor was used as a reference to compare the sensor response due to variations in resistances between sensors during fabrication. Percent changes in current were first calculated for individual sensors for each concentration separately and then the combined average was used for plotting the calibration curves shown in Figure 3. Both GLC and PCAT-GLC sensors showed a linear response to free chlorine in the range of 0.2 to 0.8 mg/L. The slope for the PCAT-GLC sensor was higher (13.89 (mg/L)-1) than for the GLC sensor (9.71 (mg/L)-1). Using the calibration curve, we can calculate the free chlorine concentration of an unknown sample from the stabilized current value obtained for that sample.

Figure 3. Calibration curve for GLC and PCAT-GLC sensors. Both sensors showed linear response to free chlorine in the 0.2 to 0.8 mg/L range.

Figure 2. Measuring free chlorine using a PCAT-functionalized GLC sensor. Raw data showing the response of PCAT-GLC sensor for different free chlorine concentrations.

All the sensors (GLC with and without PCAT) were tested in the same way to measure their response to various free chlorine concentrations. Both type of sensors showed similar behavior for repeated cycles. In our previous work, we demonstrated free chlorine sensing using CNT6 and graphite pencil7,8 substrates. These sensors required a chemical or electrochemical reset after every measurement. GLC and PCAT-GLC sensors do not require a reset to operate, which is an important advantage over the other sensors and makes them suitable for continuous monitoring of free chlorine. Figure 3 shows the calibration curves obtained for both type of sensors, with error bars resulting from averaged measurements from 4 cycles for each of 3 sensors for GLC with PCAT and for each of 2 sensors for bare GLC. The error bars in the figure are very small due to the high level of reproducibility in the measurements. The stable current readings for each free chlorine concentration were used for calculating the percentage change in current (sensor response). The sensor response for a given analyte was calculated by subtracting its corresponding stable current reading from the baseline DI

The PCAT-GLC sensor showed a 30% change in current for 0.2 mg/L indicating that this sensor has the ability to detect much lower free chlorine concentrations than we were able to verify using the colorimetric kit. Sensor resolution can be calculated as 3 times the noise in the device. Noise in the GLC sensors is the variation measured for the same free chlorine concentration. From Figure 3, the sensor resolution was calculated as 0.084 mg/L. The sensor response for the PCATGLC sensor compared to GLC sensor is almost double. The increase in sensing response is due to the addition of PCAT molecules which may amplify the signal on the GLC sheet. The sensor responses for GLC and PCAT-GLC are an order of magnitude higher than previously reported PCAT-modified CNT and graphite pencil sensors.6,7 Not only are the GLC sensors more sensitive than the previously reported CNT network6 and pencil-drawn line7,8 sensors, they also show a different saturation behavior at higher free chlorine concentrations (Fig. 4). This can be understood if we consider that the GLC films are thinner and hence more two-dimensional. All the above mentioned sensors have in common that they constitute percolation networks of nanocarbon elements (nanotube bundles, graphite flakes, GLC fragments), a design feature that is responsible for their robust sensing response.21 Various forms of nanocarbon have been used in sensors22 because their facile doping by a variety of species.23,24 Of particular interest here is the ability of certain redox-active molecules such as polyaniline (PANI)25 or PCAT6,7,8 to modulate the degree of doping imparted on the

ACS Paragon Plus Environment

ACS Sensors 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

nanocarbon substrate based on variations of their oxidation state. PCAT has been demonstrated to adsorb onto graphitic substrates as a thin, uniform layer, and remain that way throughout changes of its redox state and acid/base state.8 It is not water soluble, but due to its strong non-covalent interactions with nanocarbons it will even stay in place over many cycles when rinsed with organic solvents in which it is soluble such as methanol.26 It mimics PANI in its electronic structure27 and has thus been demonstrated to be equally capable of redox-state dependent charge transfer.28

Figure 4. Saturation of sensor response around 1 mg/L free chlorine concentration for PCAT-GLC, and around 0.8 mg/L free chlorine for bare GLC.

A number of different mechanisms have been discussed for sensing responses of nanocarbon films, including electrostatic gating, changes in gate coupling, carrier mobility changes, and Schottky barrier effects at the interface between the nanocarbon film and the contact electrode.19,29 We can exclude interface effects in our sensors since only the central portion of the film is exposed to the analyte as a result of the microfluidic device design. In the case of percolation networks consisting of a mix of semiconducting and metallic carbon nanotubes, it has also been proposed that the modulation of Schottky barriers between different types of carbon nanotubes within the network may play a role in sensing,30 although that mechanism is unlikely to hold for GLC since the electronic properties of the network elements are not expected to be as distinct from each other as in the case of carbon nanotubes. Generally, electrostatic gating has been found to be the dominant cause for the sensing response of doping-based nanocarbon sensors.19,29 While the geometry of our sensors does not allow for the fabrication of a back-gate to probe the details of the sensor behavior, our observations are consistent with the electrostatic gating mechanism. The magnitude of the electrostatic gating effect is expected to drop off into the bulk of the nanocarbon film operating as a channel. Since the GLC films used in this work are only 15 to 20 nm thick, all free chlorine concentrations of above 0.8 mg/L are able to saturate the effect for the entire thickness of the film, while lower concentrations can be detected quantitatively. A thicker GLC film, or a less effective redoxswitchable dopant molecule layer would therefore be expected to lead to a sensing range beyond 0.8 mg/L. A thinner film may be suitable for further lowering the detection limit,

although that is not of relevance for the application presented here. A much thicker channel, such as a several hundred nanometer thick pencil-drawn film,8 will not be saturated in its response even at very high free chlorine concentrations above 100 mg/L, although the overall magnitude of the response will be lower due to the drop-off of the electrostatic gating effect beyond several tens of nanometers. Carbon nanotube bundles should in principle show a very high response, but are hampered by very slow percolation of the analyte into the pores of the nanocarbon film, resulting in a slow response and necessitating reset procedures to re-attain equilibrium at low free chlorine concentrations within a reasonable timeframe.6 The flat, 2D-like nature of the GLC film circumvents these challenges, resulting in a swift response without the need for a reset. There are several key differences between the GLC films and the previously studied nanotube6 or pencil-drawn film7,8 sensors which explain why GLC sensors no longer require a reset. In addition to the above-mentioned difference in morphology (GLC is thinner and smoother) of the film, the absence of clay or binders in the GLC film, and the change of substrate from glass to polyester also play a role. Clay and glass are silicate materials that can sustain charges and adsorbed ions at their surface, to the point where porous glass is used as an ion-sensitive membrane in pH electrodes. Clay in pencil lead further contains redox-active impurities such as iron.7,8 Many carbon nanotube products contain redox-active residual catalyst particles.23,24 The GLC films used here are free of binders and additives and reside on top of a non-ionic polymer surface.15 Thus they can directly respond to the redox-state of the PCAT molecules, without interference from trapped charges, ions or redox-active impurities. In order to test the effect of interfering species in water on the signal quality, sulfates (500 mg/L), phosphates (0.03 mg/L) and nitrates (45 mg/L) were tested on the GLC and PCAT-GLC sensors. The concentrations of these species were chosen as the regulation limits of these chemicals in the environment according to current drinking and surface water standards31,32,33 so that the maximal effect of these species on sensor performance can be determined. All the interfering species solutions were prepared in DI water and flowed over the sensors until the stabilization criterion (±20 nA/s) was achieved. This experiment was repeated for 3 times on both type of sensors. The percentage change in current was plotted for each species as shown in Figure 5.

ACS Paragon Plus Environment

Page 4 of 8

Page 5 of 8 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

Figure 5. Effect of interfering species on sensors. Both sensors showed very low increase in signal to interfering species. The PCAT-GLC sensor performed slightly better showing lower currents.

Both GLC and PCAT-GLC sensors showed excellent selectivity for free chlorine over interfering species. The average changes in current for PCAT-GLC and GLC sensors were found to be 0.1% and 0.14% for sulfates, for phosphates 0.24% and 0.31%, and for nitrates 0.08% and 0.3% respectively. The sensor response for interfering species is more than two orders of magnitude lower than the sensor response to free chlorine. While the PCAT-GLC sensors outperform the bare GLC sensors for selectivity in case of all three species tested, nitrates showed the most significant difference between the two types of sensors. The response of the GLC sensor to nitrates was 4 times higher than the PCATGLC sensor. This may be due to the stronger interaction of nitrate ions with graphene34 compared to sulfates or phosphates. Also, GLC sensors showed higher variations in response to interfering species compared to PCAT-GLC sensors. This experiment demonstrates the high selectivity of the PCAT-GLC sensors for free chlorine. One of the major challenges in water monitoring systems is sensor fouling due to natural organic matter which is inevitably present in both drinking and surface waters. Fouling causes deterioration of the sensor signal over time. We tested GLC and PCAT-GLC sensors for fouling resistance using humic acid (2 mg/L) as the model fouling agent. To test the fouling resistance of these sensors, first DI water was flowed for 10 minutes followed by 0.8 mg/L of free chlorine for 10 minutes. Then humic acid solution was flowed for 1 hour. Immediately after humic acid, DI water was flowed for 10 minutes followed by 0.8 mg/L free chlorine solution to quantify the loss in signal. Figure 6 shows the measurements from the sensors before and after humic acid exposure.

Figure 6. Fouling test on GLC and PCAT-GLC sensors. Humic acid was flown over the sensors. Signal values were measured before and after the humic acid flow for 0.8 mg/L of free chlorine solution. A 50 % and 2 % loss of signal was observed for GLC and PCAT-GLC sensors respectively. PCAT improves fouling resistance.

Sensor fouling data (Figure 6) clearly shows that the change in signal of the PCAT-GLC sensors before and after the humic acid exposure was not statistically significant while it was significantly reduced in the case of uncoated GLC sensors. The PCAT molecules non-covalently attach to the graphene surface and form a complete monolayer,8 therefore preventing fouling due to attachment of organic molecules such as humic acid from the aqueous analyte. To verify the performance of PCAT-GLC sensors using real world samples, tap water samples collected from a residential area in Hamilton, Ontario, Canada, were tested using the standard free chlorine colorimetric kit and the PCAT-GLC sensor. Tap water samples from four different sources in a residential unit (Sample 1: Kitchen; Sample 2: Garage; Sample 3: Bathroom; Sample 4: Garden) were collected from the same house. The sample vials were stored in a Styrofoam box filled with ice to slow the degradation during transport. All the samples were tested within 30 minutes of collection. Figure 7 shows the data tested using a standard colorimetric kit and the PCAT-GLC sensor. In addition, the tap water samples were spiked with 0.2 mg/L and 0.4 mg/L of free chlorine and tested using the colorimetric kit and the PCAT-GLC sensor.

ACS Paragon Plus Environment

ACS Sensors 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 7. Tap water sample test. Colorimetric free chlorine kit and PCAT-GLC sensors are compared for free chlorine detection in tap water samples. Tap water samples spiked with 0.2 mg/L and 0.4 mg/L are also tested using a colorimetric kit and PCATGLC sensors. Both the methods showed good data agreement.

The current values obtained from the PCAT-GLC sensor were converted to free chlorine concentration in mg/L using the calibration curve from Figure 3. The standard colorimetric kit showed 0.3 mg/L for all the tap water samples and was not sensitive enough to distinguish smaller variations. The PCATGLC sensor showed similar values as compared to the colorimetric kit but was sensitive enough to show small but consistent variations between tap water samples. This may be attributed to the internal software of the colorimetric sensor rounding off to the nearest 0.1 mg/L. A similar trend was observed for the spiked samples. The colorimetric kit recorded readings of 0.5 and 0.7 mg/L respectively the spiked samples whereas slight variations were observed for samples tested using PCAT-GLC sensor. For each individual set of measurements of a particular tap water sample (as collected, spiked with 0.2 mg/L, and spiked with 0.4 mg/L) the deviation of the PCAT-GLC sensor measurement from the colorimetric measurement was consistent, hinting at a possibly higher precision of the PCAT-GLC measurements. In contrast, the accuracy of the measurement is limited by the quality of the calibration curve, which was established using the colorimetric kit. For lack of a third, more accurate method to determine free chlorine concentrations, we cannot fully quantify the performance of our sensors. Nevertheless, the standard colorimetric kit and PCAT-GLC sensor data were in overall good agreement. Conclusion A solid state chemiresistive sensor using graphene like carbon is demonstrated for sensing free chlorine in water. A simple fabrication technique allows for the low cost of the sensor. The GLC sensors do not require a reset and are suitable for continuous monitoring of free chlorine. The GLC and PCAT-GLC sensors showed a linear response to free chlorine in the range of 0.2 to 0.8 mg/L. Both sensors showed good resistance to interfering species in water. The PCATGLC sensor showed excellent fouling resistance while the GLC sensor had a 50% loss of signal due to fouling. Tap water samples from a residential area were tested using a standard free chlorine colorimetric kit and PCAT-GLC sensors, with both methods showing good agreement. The GLC and PCATGLC sensors show high sensitivity with excellent selectivity to free chlorine in water and can be used for continuous monitoring of free chlorine in water.

ASSOCIATED CONTENT Supporting Information. The Supporting Information is available free of charge on the ACS Publications website. Drawing of screen printing mask for sensor fabrication, raw data for sensor measurements of several devices. (PDF)

AUTHOR INFORMATION Corresponding Authors * E-mail: [email protected] * E-mail: [email protected]

ORCID

Peter Kruse: 0000-0003-4051-4375 P. Ravi Selvaganapathy: 0000-0003-2041-7180 Notes The authors declare no competing financial interest.

ACKNOWLEDGMENT The authors would like to acknowledge Omar Sharif for the PCAT synthesis. The National Science and Engineering Research Council of Canada (NSERC) provided financial support through the Discovery Grant program. P. R. S. was also supported through the Ontario Research Fund, through a Discovery Accelerator Supplement award from NSERC, and through the Canada Research Chair program.

REFERENCES (1) Bull, R. J.; Birnbaum, L.; Cantor, K. P.; Rose, J. B.; Butterworth, B. E.; Pegram, R.; Tuomisto, J. Water chlorination: essential process or cancer hazard? Toxicol. Sci. 1995, 28, 155-166. (2) Langlais, B.; Reckhow, D. A.; Brink, D. R. Ozone in water treatment: application and engineering. CRC press: 1991. (3) Legrini, O.; Oliveros, E.; Braun, A. Photochemical processes for water treatment. Chem. rev. 1993, 93, 671-698. (4) Jensen, J. N.; Johnson, J. D., Specificity of the DPD and amperometric titration methods for free available chlorine: A review. J. Am. Water Works Assoc. 1989, 81, 59-64. (5) Pan, S.; Deen, M. J.; Ghosh, R. Low-cost graphite-based free chlorine sensor. Anal. Chem. 2015, 87, 10734-10737. (6) Hsu, L. H.; Hoque, E.; Kruse, P.; Selvaganapathy, P. R. A carbon nanotube based resettable sensor for measuring free chlorine in drinking water. Appl. Phys. Lett. 2015, 106, 063102. (7) Hoque, E.; Hsu, L. H.; Aryasomayajula, A.; Selvaganapathy, P. R.; Kruse, P. Pencil-Drawn Chemiresistive Sensor for Free Chlorine in Water. IEEE Sens. Lett. 2017, 1, 4500504. (8) Mohtasebi, A; Broomfield, A. D.; Chowdhury, T.; Selvaganapathy, P. R.; Kruse, P. Reagent-Free Quantification of Aqueous Free Chlorine via Electrical Readout of Colorimetrically Functionalized Pencil Lines. ACS Appl. Mater. Interfaces. 2017, 9, 20748–20761. (9) Hallaj, T.; Amjadi, M.; Manzoori, J. L.; Shokri, R. Chemiluminescence reaction of glucose-derived graphene quantum dots with hypochlorite, and its application to the determination of free chlorine. Microchim. Acta 2015, 182, 789-796. (10) Jamal, M.; Dutta, A. K.; Islam, J.; Shao, H.; Yousuf, M. A.; Razeeb, K. M. In Disposable Chlorine Sensor Based on Pencil Graphite Electrode, Meeting Abstracts, The Electrochemical Society: 2017; pp. 1947-1947. (11) Dong, Y.; Li, G.; Zhou, N.; Wang, R.; Chi, Y.; Chen, G. Graphene Quantum Dot as a Green and Facile Sensor for Free Chlorine in Drinking Water. Anal. Chem. 2012, 84, 8378-8382. (12) Kumar, D. R.; Kesavan, S.; Nguyen, T. T.; Hwang, J.; Lamiel, C.; Shim, J.-J. Polydopamine@ electrochemically reduced graphene oxide-modified electrode for electrochemical detection of freechlorine. Sens. Actuat. B: Chem. 2017, 240, 818-828. (13) Wang, W.; MacDiarmid, A. G. New Synthesis of Phenyl/Phenyl End-Capped Tetraaniline in The Leucoemeraldine and Emeraldine Oxidation States. Synth. Met. 2002, 129, 199–205. (14) 3M 2698 Cover Tape http://www.3m.com/3M/en_US/company-us/all-3m-products/~/3MConductive-Pressure-Sensitive-Cover-Tape2698?N=5002385+3293048514&rt=rud (15) Bauld, R.; Choi, D.-Y. W.; Bazylewski, P.; Divigalpitiya, R.; Fanchini, G. Thermo-Optical Characterization and Thermal Properties of Graphene-Polymer Composites: A Review. J. Mater. Chem. C 2018 DOI: 10.1039/c7tc01719g (16) Ahmed, M. S.; Ezugwu, S.; Divigalpitiya, R.; Fanchini, G. Relationship between electrical and thermal conductivity in graphenebased transparent and conducting thin films. Carbon 2013, 61, 595601.

ACS Paragon Plus Environment

Page 6 of 8

Page 7 of 8 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors (17) Ezugwu, S.; Ahmed, M. S.; Bauld, R.; Divigalpitiya, R.; Fanchini, G. Influence of the addition of graphene-like materials on the thermophysical properties of poly (3, 4-ethylenedioxythiophene): poly (styrenesulfonate) thin film nanocomposites. Thin Solid Films 2013, 534, 520-528. (18) Wu, H.; Huang, B.; Zare, R. N. Construction of microfluidic chips using polydimethylsiloxane for adhesive bonding. Lab Chip 2005, 5, 1393-1398. (19) Heller, I.; Janssens, A. M.; Männik, J.; Minot, E. D.; Lemay, S. G.; Dekker, C. Identifying the Mechanism of Biosensing with Carbon Nanotube Transistors. Nano Lett. 2008, 8, 591-595. (20) U.S. Environmental Protection Agency, “Disinfectant use in water treatment,” in EPA Guidance Manual Alternative Disinfectants and Oxidants, 1st Ed. Washington, DC, USA: U.S. Environ. Protection Agency, 1999, ch. II, pp. 1–54. (21) Snow, A. W.; Perkins, F. K.; Ancona, M. G.; Robinson, J. T.; Snow, E. S.; Foos, E. E. Disordered Nanomaterials for Chemielectric Vapour Sensing: A Review. IEEE Sens. J. 2015, 15, 1301-1320. (22) Llobet, E. Gas Sensors Using Carbon Nanomaterials: A Review. Sens. Actuat. B: Chem. 2013, 179, 32-45. (23) Moonoosawmy, K. R.; Kruse, P. To Dope or Not To Dope: The Effect of Sonicating Single-Wall Carbon Nanotubes in Common Laboratory Solvents on Their Electronic Structure. J. Am. Chem. Soc. 2008, 130, 13417-13424. (24) Moonoosawmy, K. R.; Kruse, P. Cause and Consequence of Carbon Nanotube Doping in Water and Aqueous Media. J. Am. Chem. Soc. 2010, 132, 1572-1577. (25) Klinke, C.; Chen, J.; Afzali, A.; Avouris, P. Charge Transfer Induced Polarity Switching in Carbon Nanotube Transistors. Nano Lett. 2005, 5, 555–558. (26) Hoque, E.; Chowdhury, T.; Kruse P. Chemical In Situ Modulation of Doping Interactions Between Oligoanilines and Nanocarbon Films. Surf. Sci. 2018 DOI: 10.1016/j.susc.2018.01.003

(27) Greiner, M. T.; Festin, M.; Kruse, P. Investigation of Corrosion-Inhibiting Aniline Oligomer Thin Films on Iron Using Photoelectron Spectroscopy. J. Phys. Chem. C 2008, 112, 1899119004. (28) Mohtasebi, A.; Chowdhury, T.; Hsu, L. H. H.; Biesinger, M. C.; Kruse, P. Interfacial Charge Transfer between Phenyl-Capped Aniline Tetramer Films and Iron Oxide Surfaces. J. Phys. Chem. C 2016, 120, 29248-29263. (29) Shirsat, M. D.; Sarkar, T.; Kakoullis Jr., J.; Myung, N. V.; Konnanath, B.; Spanias, A.; Mulchandani, A. PorphyrineFunctionalized Single-Walled Carbon Nanotube Chemiresistive Sensor Arrays for VOCs. J. Phys. Chem. C 2012, 116, 3845-3850. (30) Kim, K. K.; Kim, S. M.; Lee, Y. H. Chemically Conjugated Carbon Nanotubes and Graphene for Carrier Modulation. Acc. Chem. Res. 2016, 49, 390-399. (31) Health Canada, Guidelines for Canadian Drinking Water Quality - Summary Table. Water and Air Quality Bureau, Healthy Environments and Consumer Safety Branch, Health Canada, Ottawa, Ontario, 2017. (32) World Health Organization, Guidelines for Drinking-Water Quality, 3rd ed.; Geneva, 2008. (33) U.S. Environmental Protection Agency, “Ambient Water Quality Criteria Recommendations: Nutrient Criteria for Wetlands in Nutrient Ecoregion XIII" EPA document number EPA-822-B-00-023, 2000. (34) Tang, H.; Sundari, R.; Lintang, H. O.; Yuliati, L. In Detection of nitrite and nitrate ions in water by graphene oxide as a potential fluorescence sensor, IOP Conference Series: Materials Science and Engineering, IOP Publishing: 2016; p. 012027.

ACS Paragon Plus Environment

ACS Sensors 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 8

for TOC only

ACS Paragon Plus Environment

8