A low-cost automatic sensor for in-situ colorimetric detection of

Subscriber access provided by Kaohsiung Medical University

Article

A low-cost automatic sensor for in-situ colorimetric detection of phosphate and nitrite in agricultural water Beichen Lin, Jin Xu, Kunning Lin, Mingpo Li, and Miao Lu ACS Sens., Just Accepted Manuscript • DOI: 10.1021/acssensors.8b00781 • Publication Date (Web): 08 Nov 2018 Downloaded from http://pubs.acs.org on November 8, 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 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 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.

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 21 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

A low-cost automatic sensor for in-situ colorimetric detection of phosphate and nitrite in agricultural water Beichen Lina, Jin Xub, Kunning Linb, Mingpo Lia, Miao Lu a,* a

Pen-Tung Sah Institute of Micro-Nano Science & Technology, Xiamen University, Xiamen 361005, China

State Key Laboratory of Marine Environmental Science, College of the Environment and Ecology, Xiamen University, Xiamen 361005, China b

KEYWORDS: in-situ automatic sensing, water quality monitoring, agricultural sensor network, “Fish-Bite” reservoirs, phosphate and nitrite measurement

ABSTRACT: This study proposed a low-cost sensor for in-situ automatic monitoring of phosphate and nitrite in agricultural water environments, involving a series of “Fish-Bite” reservoirs, multiple reagent capsules, and a colorimetric sensor. The “Fish-Bite” reservoir is an alternative to the pumps, valves and filters that are widely used for water sample collection, and also offers a closed cell for chromogenic reactions afterwards. Up to two capsules can be embedded in each reservoir to support chromogenic reactions that use two different reagents in sequence. From the results of calibration tests in laboratory, the limit of detection (LOD) was found to be approximately 0.01 mg/L for both phosphate and nitrite, with linear range of 0.01 - 1.00 mg/L for phosphate and 0.01 - 0.20 mg/L for nitrite. Furthermore, an in-situ experiment was successfully carried out in an irrigation canal beside a farmland to demonstrate the practicability and robustness of the device. The averaged concentrations of phosphate and nitrite were 0.0113 mg/L and 0.0383 mg/L, respectively. The relative deviations were 20.2% and 11.7%, respectively, referred to results obtained by using the standard spectrophotometric methods. With the advantages of being robust, fast and low-cost, this in-situ device is promising for the formation of agricultural sensor networks.

The inorganic nutrients, including phosphate, nitrite, nitrate and ammonium, etc., are essential in agriculture for crop growth. 1,2 Moreover, the nutrients are also important environmental indicators. 3 Excessive nutrients leach from the agricultural soil not only reduce soil fertility and plant yield, but also potentially threaten the environment and human health. 4-8 For example, phosphate is important in biogeochemical processes and ecosystems. 9-12 It is a nutrient for crop growth in agriculture, while high concentration of phosphorus (over 0.2 mg/L) can lead to eutrophication and threaten water environment. 13, 14 Similarly, while nitrite plays an important role in global nitrogen cycle, extra amount of nitrite in water systems is also related to eutrophication. Even more, nitrite is toxic to most biota and carcinogenic to human beings. Therefore, it is necessary to monitor nutrient concentrations for both agricultural and environmental perspectives. Rather than those traditional lab-based standard methods to measure inorganic nutrients, 15-20 a growing interest in agricultural management is to build up widely distributed sensor networks. 8, 21 The sensor networks, which make the monitoring more rapid and

labor saving, contribute to better fertilizer efficiency, incremental crop yields, as well as to reduce excessive nutrients leaching into the natural water environments. The ideal devices used to build up such sensor networks should be autonomous, in-situ, robust and low-cost, also with integrated functions to detect various target nutrients. Although ultra-violet optical, 22 electrochemical 23 and fluorescent methods 24 are being developed with promising potential, the colorimetric methods are still most validated and favored for autonomous in-situ application. 25 Some insitu inorganic nutrient sensors based on colorimetric methods, with outstanding performance have been reported, 26-28 and there are also available commercial insitu sensors for choice. 29-31 However, while their performances (e.g. excellent limit of detection (LOD) and linear range) are overqualified in agricultural applications, their excessive price (usually above $20, 000) and relatively bulky size limit their applications in the agriculture. 25 To make the sensor more affordable and practical, some reported in-situ devices employ low-cost microfluidic components and detectors to reduce the cost and size. 32-35

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

Impressively, an in-situ total phosphorus analyzer with component cost less than $3,000 was developed in 2015. 36 However, they still require consumable filters, precise but costly pumps and valves, and delicate fluid channels that needed to be rinsed after every measurement. These devices often have problems of reagent storage, and inherent difficulties in completely cleaning the fluidic paths. Moreover, they often only aim at single or one type of target analytes (e.g. nitrite and nitrate, phosphate and total phosphorus). As a result of these, the devices are still not efficient and economical enough for the formation of agricultural sensor networks. In this study, we propose an automatic in-situ device for chromogenic detection of phosphate and nitrite. This pumpless and valveless device is composed of commercial servo motors, colorimetric sensor and low-cost ABS plastic materials, thus its cost is significantly reduced to approximately $200. The carefully designed “Fish-Bite” reservoir, which is the core component of the device, is employed along with a reservoir chamber. As an alternative to widely used pumps and valves, the “Fish-Bite” reservoir can automatically open then close underwater while travelling through the reservoir chamber, hence the water sample is collected in-situ. This robust sampling method avoids both clogging and bubble formation, which often happen in flow analytical systems. In addition, the use of disposable reservoirs eliminates the need for rinsing and the risk of cross contamination. Moreover, with two capsule locators in each reservoir, two different reagents can be applied, supporting more than one chromogenic reactions (e.g. Griess reaction for nitrite and phosphomolybdenum blue (PMB) reaction for phosphate determination). Also, an improved dynamic colorimetric sensing method is optimized for this device. An in-situ experiment was carried out in an irrigation canal beside a farmland. The device was successfully deployed to determine the concentrations of phosphate and nitrite in the canal water. The LOD was found to be approximately 0.01 mg/L for both phosphate and nitrite concentration, with linear range of 0.01 - 1.00 mg/L for phosphate and 0.01 - 0.20 mg/L for nitrite.

MATERIALS, DEVICE AND METHODS Design and preparation of the device. The materials and modules used to build the experimental device were: 3 mm and 5 mm black Acrylonitrile Butadiene Styrene (ABS) plate (Shenzhen Xinnuode Technology Co., Ltd., China), polydimethylsiloxane (PDMS) silicon rubber (SYLGARD 184, DOW CORNING, USA), a custom-made mini-sized colorimetric sensor module that integrated a colorimetric chip (TCS3200, TAOS, USA), two water-proof servo motors (SW-1210SG, SAVOX, China), a system circuit board (LYSTM32F103C8 V1.2, DOFLY, China) that contained a STM32F103x8 CPU (STMicroelectronics, USA), an IS32LT3175N/P constant-current LED driver (ISSI, USA), white light chip LED (ranged from 440 to 700 nm) and infrared chip LED (ranged from 750 to 910 nm) (Pak, China), a commercial Bluetooth wireless serial port

Page 2 of 21

module (HC-05, TELESKY, China), a 11.1 v 2200 mAh 3S LiPo battery pack (ACE, China), 1 mm quartz glass slides (HONGPU, China), various sizes of rubber bands and elastic rings (VYR, China), and nylon screws (BRT, China). With these materials and modules, the total cost of the device was approximately $200 (see Supporting Information for detailed price list and cost calculation). The device is shown in Figure 1a. The mechanical structures were made of machined ABS plastic plates and nylon screws. The waterproof servo motors and elastic rings were applied as underwater actuators and elastic components. The use of waterproof or stainless modules and materials guaranteed the underwater reliability and durability. It is worthy to mention that most of the structure had black lusterless surfaces, so that stray light was eliminated as much as possible, improving the detection accuracy. The device consisted of four main parts: an autoloader, a reservoir chamber, a capsule opener and a colorimetric detector, as shown in Figure 1b. The autoloader, included a reservoir pusher and a magazine, could continuously load the reservoirs into the reservoir chamber. The magazine was capable to hold up to 5 reservoirs. The servo-driven capsule opener was installed above the reservoir chamber. The colorimetric detector included two LEDs and a colorimetric sensor module. The white light and infrared chip LEDs were integrated into the capsule opener, driven by a constant-current LED driver. The colorimetric sensor module was mounted on the opposite side of the same axis facing to the LEDs. It is worthy to highlight our innovative “Fish-Bite” reservoir, as well as the ancillary reservoir chamber. The “Fish-Bite” reservoir was inspired by the swallowing action of fish underwater. A single reservoir consisted of three pieces: maxilla, tongue and chin, as shown in Figure 2a and Figure 2b. The tongue was the middle piece, supporting a volume of 2.4 mL. Both the maxilla and chin had cylindrical knobs on both sides and a transparent quartz glass optical window on the center for optical measurements. The maxilla was thicker than the chin, and had additional two capsule locators that used to hold capsule or plug. During each deployment period, the disposable reservoirs were not to be reused by the device. Before the underwater application, the three pieces of each reservoir were banded together with two elastic rings. Unless opened, the reservoir remained watertight due to the elasticity of the rings and surface tension of the water, hence the inner space of each reservoir would not be fouled over time until use. The operation procedures of the reservoir and relevant mechanisms are shown in Figure 2c. It is of note that all sampling and detection procedures took place underwater. The specific LED was turned on at the beginning, according to the target compounds (e.g. infrared LED for phosphate and white light LED for nitrite). When a reservoir was being loaded, the cylindrical knobs on both maxilla and chin slid into the slide rails on the walls of the reservoir chamber. The autoloader continued pushing the


Page 3 of 21 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 1. Experimental device, (a) experimental device; (b) schematic view. reservoir, then the maxilla and chin were opened by the slide rails while the tongue continued sliding forward without vertical motions. Meanwhile, the surrounding water instantly filled the reservoir without formation of any bubbles. After that, the reservoir closed itself, and was further pushed forward to the sensor position. The second half of the slide rails were stair shaped rather than sloped, to prevent the reservoir from directly sliding out. Afterwards, the reservoir, which was full of water sample, reached the sensor position. The servo motor drove the capsule pressing rods, pressed the capsules to release the chromogenic reagents. The reagents were then mixed and reacted with the water sample, and then the colorimetric detection was started. By repeating these procedures, replacing the previous reservoirs with new ones, continuous measurements were achieved. The detailed structure of the device and the dynamic illustration of the operating procedures were presented in the Supporting Information. The concept of reagent capsule was inherited from our previous studies. 37 It was proved that the use of reagent capsule was a promising, feasible and practical alternative to pipetting and injection. The capsules isolated the liquid reagents from the air and water, and were used for only one measurement. It could protect the reagent from degrading with time, or being contaminated or oxidized. In addition, the disposable capsules also acted as one-time use quantitative reagent injectors. This is advantageous as they are inexpensive and disposable, compared to costly and complex pumps, valves or other flow actuators that require post-sampling rinsing. The capsule shape and dimensions were optimized in this work. While the reservoir supported up to two reagent

capsules, a solid plug was introduced to block one of the capsule locators, for those reactions use only one reagent (e.g. Griess reaction for nitrite). The cylindrical reagent capsules were made of PDMS, with a volume of 0.24 mL and an outer dimensions of Φ10×5.5 mm. The fabrication processes of the capsules and solid plugs were presented in the Supporting Information. The chromogenic reagents were injected into finished capsules using syringes. A pinhole was on the bottom of each capsule. Due to the tiny size of the pinhole and good elasticity of the PDMS material, there was no reagent leakage hence the pinhole did not need to be sealed. When the capsule was being pressed, the reagent came out through the holes, then mixed and reacted with the water sample. The colorimetric sensor collected data from the R (red), G (green), B (blue) and C (clear) colorimetric channels. All the four colorimetric channels had different sensitivities to lights in various wavelengths. These channels were complementary to each other, therefore the colorimetric sensor overall covered a wide wavelength range of detection from 400 nm to 1050 nm. Furthermore, the LEDs with different light wavelengths were activated to cooperate with the sensor for determinations of different target compounds with different absorptions. In this study, the infrared LED and data from C channel were used for phosphate determination, while nitrite determination needed to use white light LED and data from G channel. With less modules and actuators applied, the power consumption of the device was remarkable. The theoretical standby power and measurement power were as low as 0.015 W and 0.265W, respectively (see Supporting


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

Information for consumptions).

detailed

calculations

of

power

Page 4 of 21

were specifically analyzed, because the C channel had the overall strongest sensitivity in this wavelength range. All the chemicals used solutions were purchased from

Figure 2. “Fish-Bite” reservoir, (a) explosive view; (b) reservoirs, capsule and plug; (c) operation procedures of the reservoir and relevant mechanisms. Calibration methods for phosphate. The popular and classic method for phosphate determination is the phosphomolybdenum blue (PMB) method, 38 where ammonium molybdate and potassium antimonyl tartrate react in acid medium with phosphate to form a phosphomolybdic acid, which is reduced by ascorbic acid to a blue-colored complex called molybdenum blue. With maximum absorption wavelength from 710 nm to 880 nm, the color is proportional to the phosphate concentration. It is worthy to point out that, in this experiment, the infrared LED was applied to adapt the absorption wavelength range of PMB. Also, the data from C channel

Sinopharm Chemical Reagent Co., China, and prepared with deionized (DI) water. The ammonium molybdate stock solution (14 g L-1), was prepared by dissolving 3.5 g (NH4)6Mo7O24·4H2O in 125 mL 5.52 mol L-1 H2SO4 (Guarantee grade) solution, then adding 0.075 g potassium antimony tartrate, and diluted to 1 L. The solution was kept in brown bottle. The ascorbic acid (AA) concentration was 100 g L-1 and prepared daily. 39 The molybdate and AA reagents were separately injected into different capsules. The reagents in capsules could adequately react with most water samples with phosphate concentration up to 1.00 mg/L collected in the reservoir


Page 5 of 21 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

(2.4 mL). The pH of the final solution was in the range of 0.5-1.0 with [H+]/[MoO2- 4] ratio about 70, as suggested by the references. 18, 19, 38, 40, 41 Then the capsules were installed into the capsule locators on the maxilla of each reservoir. A series of samples were prepared with concentrations of 0.01, 0.05, 0.10, 0.20, 0.30, 0.50, 0.70 and 1.00 mg/L, calculated as phosphorus, P, for making the calibration curve. In a single test, the experimental device was powered on, and then submerged into a dark sink filled with the sample. The LED was activated and kept on. Then a reservoir was loaded, travelled through the reservoir chamber, opened then closed to collect water sample, and finally reached the sensor position. Meanwhile, the capsule pressing rods pressed down onto the capsules to release the molybdate and AA reagent in sequence into the reservoir, and facilitated them to mix and react with the sample. Before the colorimetric sensor started to work, the molybdate and AA reagents were mixed well with the sample for 20 s. Afterwards, the data were collected every second. The total test time was Ttotal-phosphate = 200 s. Data of every second included the numbers of pulse n, received within a period of 50 ms, from the R, G, B and C channels. The data of received pulse number n were then normalized. For all the four channels, the curves of the normalized pulse number n were plotted against the test time, T. The measurement was repeated 10 times for each calibration sample. The design of reservoir magazine facilitated the repetition. Finally, a calibration curve for phosphate was established by analyzing these dynamic colorimetric data of various phosphate concentrations.

capsule of reagent (0.24 mL) was also adequate to react with the water sample in a single reservoir (2.4 mL). The pH of the final solution was in the range of 1.0-2.0, as suggested by the references. 18, 20, 43, 44 The nitrite concentrations in the calibration samples, calculated as nitrogen, N, were 0.01, 0.03, 0.05, 0.10 and 0.20 mg/L. The nitrite measurement was similar to the phosphate measurement except the nitrite chromogenic reaction involved only one reagent, and the other capsule locator was blocked using a solid plug. Also, as the Griess reaction was slower than the PMB reaction for phosphate, the set total test time was Ttotal-nitrite = 480 s. The experimental procedures, similar with described in previous section, were repeated 10 times for each water sample with different nitrite concentration. Finally, a calibration curve for nitrite was established. In-situ phosphate and nitrite measurements. To demonstrate the practicability and robustness of the device, an in-situ experiment was carried out in an irrigation canal beside a farmland (24°36'16.8"N 118°19'14.3"E) in Xiao Long village, Xiamen, China. The device in operation is shown in Figure 3. All of the electronics were packed in a waterproof box mounted above the foam buoy, and the main body of the structure was hung below the buoy (underwater). Data were collected in-situ underwater, sent to the microprocessor in the waterproof box and then wirelessly transferred to the laptop. Figure 3. Device in in-situ operation, (a) the canal and

Calibration methods for nitrite. The popular method for spectrophotometric determination of nitrite has been established based on the Griess reaction, 42 where nitrite reacts with an aromatic amine, sulfanilamide, to form a diazonium compound followed by coupling with N-1naphthylethylenediamine to form an azo dye. The maximum absorption wavelength of the purple red reaction product is 540 nm. In this study, the combination of LED and sensor channel for nitrite detection was different from what was used for phosphate detection. For nitrite detection, white light LED (440 - 700 nm) was applied. As previously mentioned, the G channel of the colorimetric sensor also had a peak sensitivity around 540 nm, which made it suitable for the application. Hence the data from the G channel were used for nitrite analysis. All the solutions were prepared with DI water. The chemical reagents were purchased from Sinopharm Chemical Reagent Co., China. To prepare a mixed reagent for nitrite color development, 2.0 g sulfanilamide (SAM) (Analytical Grade) and 0.2 g N-1-naphthylethylenediamine dihydrochloride (NED) (Analytical Grade) were dissolved in 400 mL of 0.6 mol/L HCl (Guarantee Grade) solution. 43, 44

Similar to that described in the previous section, the reagent was injected into the capsules for use, and a

experimental setup; (b) surface view; (c) underwater view. At the monitoring position, the water had been continuously analyzed 10 times for both phosphate and nitrite. The reservoirs that held capsules with different reagents for phosphate and nitrite were alternately loaded in the magazine for use. Then, the previously introduced methods were used of phosphate and nitrite, by alternatively turning on infrared and white light LEDs. As mentioned in the previous sections, each measurement took 200 s for phosphate and 480 s for nitrite, so that the total experimental period of the continuous measurements was approximately 2 h. At the end, with 20 reservoirs used,


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

10 sets of data were collected for both phosphate and nitrite measurements. As a reference, water samples were collected from the same position in the canal and analyzed in the laboratory using standard methods. 17 A commercial spectrometer (V1100D, MAPADA, Shanghai, China) and a cuvette with 5 cm path length were employed.

Page 6 of 21

the continuous data curve and coloration rate as a whole, rather than quantifying from single static data readings, hence the errors caused by the turbidity could also be partly eliminated. The presented natural exponential function was used to fit the data curves:

𝑦 = 𝑎𝑒 ―𝑏𝑥 (1)

RESULTS AND DISCUSSION Establishment of phosphate calibration curve. A total of eight graphs were plotted for water samples with different phosphate concentrations. Each graph contained 10 curves of normalized received pulse number from C channel, n, plotted against time, T. Pulse number, n, was normalized to unify the starting points of the curves and to improve the accuracy of quantification. The typical phosphate curves for normalized pulse number from the C channel, n, corresponded to each phosphate concentration, are shown in Figure 4. The pulse number n decreased as the time increased, due to the increased amount of reaction product, which absorbed more light and reduced the amount of pulses received by the sensor. The absolute values of the slopes were larger for higher phosphate concentration.

where the parameter b was named as the coloration rate index, to represent the coloration rate of the sample. By analyzing the coloration rate index b, the concentration of phosphate could be quantified before the end of the reaction, which made the detection faster. All curves from the C channel of different phosphate concentration were fitted using Eq. 1, so that the fluctuations in the data curves could be smoothed to avoid the uncertainties. Then, the coloration rate index b was derived from the formulas of the fitted plots, corresponding to every sample of different phosphate concentrations. The scatter diagram for b values with different initial phosphate concentration C0 are presented in Figure 5. As shown, there was a good linear relationship between phosphate concentration and the coloration rate index b, and the linear fitting equation was: bphosphate = 0.0022Cphosphate – 8.37565×10-5 (2)

Figure 4. Typical phosphate curves of normalized pulse number, n, against time, T, corresponding to different phosphate concentrations. It can be found in Figure 4 that the normalized pulse amount turned to be almost constant at the designed reaction time. Longer reaction time might lead to a more completed reaction to a small extent, on the other hand, it could also reduce the analytical speed. With the applied rapid dynamic colorimetric detection method, 37 the quantification could be realized even if the reaction was not fully completed. The quantification could be finished within 200 s with the method, while the complete PMB reaction took approximately 5 min. 17 Standard colorimetric methods used static intensity readings at the end of chromogenic reactions for target analyte quantification. However, in this filter-free sensor, the turbidity of the water sample introduced uncertainties into colorimetric data readings, so that the static detection was not applicable. The applied dynamic colorimetric detection method addressed this problem by considering

By applying this formula, the phosphate concentration of general water samples could be quantified within the linear range of 0.01 - 1.00 mg/L. Moreover, the LOD was found to be approximately 0.01 mg/L for phosphate concentration.

Figure 5. Calibration curve of phosphate concentrations Cphosphate against coloration rate index bphosphate. Establishment of nitrite calibration curve. The typical nitrite curves of normalized pulse number from the G channel, n, corresponding to each nitrite concentration, are shown in Figure 6. Similarly, the figure shows a monotone decrease in pulse number n as the test time T


Page 7 of 21 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

increased. The normalized pulse amount also turned to be almost constant at the end of the designed reaction time. Furthermore, there is an obvious gradient in relationships between curves at different nitrite concentrations. Similar to that described in the previous section, all curves from the C channel of different nitrite concentration were fitted using Eq. 1. After the coloration rate index b was derived for each data set, the calibration curve of Griess reaction was plotted (Figure 7). The linearity between nitrite concentration and coloration rate index b was remarkable, and gave a linear fitting equation: bnitrite = 0.0024Cnitrite – 1.95244×10-5 (3)

The nitrite concentration of general water samples could be quantified by using Eq. 3, within the linear range of 0.01 - 0.20 mg/L. The linearity in this curve was even better than that for phosphate curve (Figure 6), and the error bars were also significantly shrunk. This was because two reagents were used in the phosphate measurement which introduced more randomness and uncertainties into the mixing and reaction process.

The LOD was approximately 0.01 mg/L for nitrite concentration as seen, which was significantly lower than the threshold nitrite concentration limit for drinking water (maximum 3 mg/L recommended by WHO 6 and 1 mg/L as per the USA and Chinese standards 7, 45). Moreover, the device can also be used to evaluate the classes of ground water, i.e. 0.01 mg/L for Class II and 0.02 mg/L for Class III, as per the Chinese quality standard for ground water. 46 In-situ measurement results. The experiments were carried out as previously described. A total of 20 reservoirs were used, and 10 sets of data were collected from the insitu measurements for both phosphate and nitrite. These data were analyzed as described in previous sections. Curves of normalized pulse number, n, against time, T, were plotted for both phosphate and nitrite detections. Typical phosphate and nitrite curves are shown in Figure 8, with the data from C channel for phosphate and G channel for nitrite.

Figure 6. Typical nitrite curves of normalized pulse number, n, against time, T, corresponding to different nitrite concentrations. For the nitrite measurement, the dispersion of b and linear fitting goodness were overwhelmingly superior to our previous work 37 due to optimizations of the materials, reservoir shapes and software programs. Similar to the previous calibration experiments carried out in the laboratory, the downward trends of the curves were due to the coloration process during measurements. However, the in-situ curves for both phosphate and nitrite tended to be more fluctuant than those obtained in laboratory environment, which might due to the Figure 8. Typical curves of normalized pulse number against time from in-situ detections on (a) phosphate; (b) nitrite. uncertainties caused by water flow, turbidity and environmental light changes in the canal water. The water flow caused mild vibrations on the buoy of the device,

ACS Paragon Figure 7. Calibration curve of nitrite concentrations Cnitrite Plus Environment against coloration rate index bnitrite.

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

while the turbidity and environmental light changes slightly interfered the light path between the LED, reservoir and the colorimetric sensor. Moreover, the phosphate curve was even less smooth because the PMB reaction involved two reagents. The phosphate and nitrite curves in Figure 8 were fitted with Eq. 2, and their coloration rate indexes were derived. A total of 10 b values were derived from the experimental data for both phosphate and nitrite measurements. These coloration rate index b were then averaged, and the averaged values were substituted into linear fitting equations (Eq. 3 and Eq. 4). The results of substitutions were 0.0113 mg/L for phosphate and 0.0383 mg/L for nitrite. The water sample collected from the same position in the canal was also analyzed using the standard spectrophotometric methods. The results were 0.0094 mg/L for phosphate and 0.0343 mg/L for nitrite. Data quality control was processed in the laboratory, and the relative standard deviations (RSDs) were lower than 3% (n=5) for both phosphate and nitrite analyses. By taking these results as references, the relative deviations of the device were calculated as 20.2% and 11.7% for phosphate and nitrite measurement, respectively. The concentration variation in running water of the irrigation canal might resulted in a large detection variation for 10 different samples, which was the main reason that the data deviated from the result of one sample obtained using a standard method in the laboratory. Nevertheless, these in-situ experiments were carried out in the winter, which were farm slack seasons. Therefore, the nutrient concentrations in the water canal and farm field were much lower than those in spring and summer. The low concentration of target substances resulted in relatively large variations. Fortunately, the deviation issue could be compensated by increasing either sampling frequency or amount of sensors in the sensor network, 25 and the relative variations in measurements would be smaller in aquatic environments with higher phosphate and nitrite concentrations. It was noticeable that the reaction time depends on solution temperature. 47 Both of the in-situ measurements and in-lab calibrations were carried out under ambient conditions in January, and the indoor and outdoor temperature differences were insignificant because no heaters were applied in the laboratory located in southern China. In these in-situ measurements, there were also potential interferences of silicates, chromophoric dissolved organic matter (CDOM) and turbidity, etc. Silicate might interfere the determination of phosphate, since it could also react with the same colorimetric reagents for phosphate to form similar blue complex. To minimize the interference, reaction pH in the range of 0.5-1.0 and [H+]/[MoO2- 4] ratio about 70 was adopted as previously mentioned in the section of calibration methods for phosphate. Under these conditions, in the reservoirs the reaction kinetics was much more favored to phosphate than to silicate. 41, 48 The interference from turbidity should slightly affect the

Page 8 of 21

detection based on spectrometric techniques. Fortunately, the turbidity of the studied water body was less than 5 NTU, with which the interference was insignificant. CDOM absorbance is mostly within the range of shorter wavelengths (shorter than 450 nm), and in the longer wavelength the absorbance is usually below the detection limit of traditional spectrophotometers. 49, 50 The wavelengths adopted for determination of phosphate and nitrite were 710-880 nm and 540 nm, both in the range of longer wavelengths. Therefore, the influence of CDOM on the determination result of phosphate and nitrite could be ignored. Even more, the influence of CDOM and turbidity could be further eliminated to a certain extend using the developed quantification method. There were no re-calibration or internal calibration during the in-situ experiments. As previously described, the applied dynamic colorimetric detection method for quantification was different from traditional ones. This quantification method overall considered the continuous data curve and coloration rate, rather than only a single static data reading. Therefore, interferences such as signal drifting did not much affect the quantification method, and the in-situ calibration was not necessary. The concept of this prototype device was to avoid using the filters in order to simply the device, as well as to increase the robustness and extend the deployment time. Even in the worst case, the non-filter device would not easily get blocked or damaged by suspending particles in water, and was capable to keep recording the data. As summarized in Table S1 (see Supporting Information), while other reported in-situ devices for inorganic nutrient monitoring had their own remarkable merits in performances, all of them employed specialized reagent protection containers, consumable filters, and complex pumps and valves, which limited them to further reduce their price and cost of use. Compared with them, our sensor device was simpler, more economical and robust. The employments of rustless materials (e.g. ABS plastic structures and rubber bands as elastic components) made the device durable in underwater environments. The application of disposable “Fish-Bite” reservoirs and reagent capsules avoided the use of filters, pumps and valves, as well as protected the reagents from being contaminated, oxidized or degraded in a simpler method. The reservoirs stayed sealed until loaded and used, so that no particulates could get inside and foul the reservoirs before measurements. Moreover, there were no concerns of blockages and post-sampling rinsing of the flow tube or the formation of bubbles during sample collection, while these problems often happened to those flow systems with traditional components. The tolerances of each component were carefully controlled. Every gaps between the moving components were neither too loose nor too tight. Hence the device had smooth movements while the mechanisms were also not easily get stuck by the particulates. All these features were designed under the consideration of long-term deployments and robustness of the device. The applied dynamic sensing method was also more rapid than the standard methods because the


Page 9 of 21 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

coloration rate index b could be derived before the end of reaction. With such features, performance and cost of approximately $200, this device was ideal and promising for monitoring and rapid water quality screening in relatively harsh environments such as agricultural water. The scope of potential target analytes of the device could also be expanded with further improvements. Based on the basic concepts of the device, the autoloader, reservoirs and reagent capsules could be further optimized in shape, dimensions and materials, giving it a smaller size, larger magazine capacity and better reliability. By swapping light sources, detectors and reagents, the device could be expanded to measure even more kinds of target analytes.

ASSOCIATED CONTENT Supporting Information Supporting Information Available: The following files are available free of charge. The file “Supporting_Information_Rev.docx” includes: Detailed mechanical structure of the device; fabrication processes of the capsules and plugs; list of reported in-situ sensor devices or techniques for inorganic nutrients monitoring; detailed price list and cost calculation, and detailed calculations of power consumptions.

AUTHOR INFORMATION Corresponding Author

CONCLUSION AND PERSPECTIVE In this study, we designed a low-cost and compact device for in-situ phosphate and nitrite monitoring in agricultural water environment. “Fish-Bite” reservoirs, multiple reagent capsules and a colorimetric sensor were employed in this device. The LOD was found to be approximately 0.01 mg/L for both phosphate and nitrite concentration, with linear range of 0.01 - 1.00 mg/L for phosphate and 0.01 0.20 mg/L for nitrite.

*E-mail: [email protected].

ACKNOWLEDGMENT This work was supported by NSFC under the project No. 21127001 and Ministry of Science and Technology of People’s Republic of China under the project No. 2011YQ03012407.

The design of disposable “Fish-Bite” reservoirs eliminated a series of issues such as post rinsing, cross contamination, clogging and bubble formation. Furthermore, each of the reservoir has two capsule locators. Therefore, the device can carry out chromogenic reactions that use multiple reagents. With a series of reservoirs loaded in the magazine, multiple water samples can be alternately tested in a sequence for different substances, ultimately achieving in-situ automatic monitoring or screening. In addition, the capsule can property protect the reagents from being degraded, oxidized or contaminated. Overall, the device was cost effective due to the employments of commercial servo motors, sensor and low-cost ABS plastic materials. The unique dynamic colorimetric detection method is advanced in detection speed compared to the traditional manual methods, because it quantifies before the ends of the reactions. This feature made the method especially suitable for rapid screening. The effectiveness and robustness of the method and the device was demonstrated using a successful in-situ experiment carried out in an irrigation canal. The averaged in-situ measurement results of phosphate and nitrite concentrations were 0.0113 mg/L and 0.0383 mg/L respectively, with corresponding relative deviations of 20.2% and 11.7%. This performance proved that the device is a practical tool to monitoring or rapid screening for relatively harsh water systems such as agricultural water. The application range of this device is also promisingly expandable. With swaps of light sources and sensors, and minor further improvements on the reservoirs, autoloader and reagent capsules, the device can be capable of detecting more target substances for longer time period in aqueous environments in agricultural regions.


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

REFERENCES (1) Colla, G.; Kim, H. J.; Kyriacou, M. C.; Rouphael, Y. Nitrate in fruits and vegetables. Sci. Hortic. (Amsterdam, Neth.) 2018, 237, 221-238, DOI: 10.1016/j.scienta.2018.04.016. (2) Rodriguez, H.; Fraga, R. Phosphate solubilizing bacteria and their role in plant growth promotion. Biotechnol. Adv. 1999, 17 (45), 319-339, DOI: 10.1016/S0734-9750(99)00014-2. (3) Gray, S.; Hanrahan, G.; McKelvie, I.; Tappin, A.; Tse, F.; Worsfold, P. Flow analysis techniques for spatial and temporal measurement of nutrients in aquatic systems. Environ. Chem. 2006, 3 (1), 3-18, DOI: 10.1071/EN05059. (4) Carpenter, S. R.; Caraco, N. F.; Correll, D. L.; Howarth, R. W.; Sharpley, A. N.; Smith, V. H. Nonpoint pollution of surface waters with phosphorus and nitrogen. Ecol. Appl. 1998, 8 (3), 559568. (5) Cameron, K. C.; Di, H. J.; Moir, J. L. Nitrogen losses from the soil/plant system: a review. Ann. Appl. Biol. 2013, 162 (2), 145-173, DOI: 10.1111/aab.12014. (6) World Health Organization (WHO). Guidelines for drinking-water quality: recommendations, World Health Organization: 2004; Vol. 1. (7) US EPA, Drinking Water Contaminants - Standards and Regulations. Available at: https://www.epa.gov/dwstandardsregulations (accessed Aug 1, 2018). (8) Zia, H.; Harris, N. R.; Merrett, G. V.; Rivers, M.; Coles, N. The impact of agricultural activities on water quality: A case for collaborative catchment-scale management using integrated wireless sensor networks. Comput. Electron. Agr. 2013, 96, 126-138, DOI: 10.1016/j.compag.2013.05.001. (9) Karl, D. M. Aquatic ecology - Phosphorus, the staff of life. Nature 2000, 406 (6791), 31-33, DOI: 10.1038/35017683. (10) Cadee, G. C.; Hegeman, J. Phytoplankton in the Marsdiep at the end of the 20th century; 30 years monitoring biomass, primary production, and Phaeocystis blooms. J. Sea. Res. 2002, 48 (2), 97-110, DOI: Pii S1385-1101(02)00161-2 (11) Thingstad, T. F.; Skjoldal, E. F.; Bohne, R. A. Phosphorus Cycling and Algal-Bacterial Competition in Sandsfjord, Western Norway. Mar. Ecol. Prog. Ser. 1993, 99 (3), 239-259, DOI: 10.3354/Meps099239. (12) Karl, D. M.; Bjorkman, K. M. Phosphorus cycle in seawater: Dissolved and particulate pool inventories and selected phosphorus fluxes. Methods Microbiol. 2001, 30, 239-270, DOI: 10.1016/S0580-9517(01)30048-X. (13) Smith, V. H. Eutrophication of freshwater and coastal marine ecosystems - A global problem. Environ. Sci. Pollut. R. 2003, 10 (2), 126-139, DOI: 10.1065/espr2002.12.142. (14) Bennett, E. M.; Carpenter, S. R.; Caraco, N. F. Human impact on erodable phosphorus and eutrophication: A global perspective. BioScience 2001, 51 (3), 227-234, DOI: 10.1641/00063568(2001)051. (15) O’Dell, J. Method 353.2, Revision 2.0: Determination of nitrate-nitrite nitrogen by automated colorimetry. US Environmental Protection Agency, Cincinnati, Ohio 1993. (16) O’Dell, J. Method 365.1: Determination of phosphorus by semi-automated colorimetry. Rev. 2.0. Methods for determination of inorganic substances in environmental samples. US Environmental Protection Agency, Cincinnati, Ohio 1993. (17) Strickland, J. D.; Parsons, T. R. A practical handbook of seawater analysis. Fisheries Research Board of Canada, Ottawa 1972. (18) Ministry of Environmental Protection of China, The specification for marine monitoring - Part 4: Seawater analysis, GB 17378.4 -2007. 2007.

Page 10 of 21

(19) Ministry of Environmental Protection of China, Water quality-Determination of total phosphorus-Ammonium molybdate spectrophotometric method, GB 11893-89. 1989. (20) Ministry of Environmental Protection of China, Water quality-Determination of nitrogen (nitrite)-Spectrophotometric method, GB 7493-87. 1987. (21) Capella, J. V.; Bonastre, A.; Ors, R.; Peris, M. In line river monitoring of nitrate concentration by means of a Wireless Sensor Network with energy harvesting. Sens. Actuators, B 2013, 177, 419-427, DOI: 10.1016/j.snb.2012.11.034. (22) Gentle, B. S.; Ellis, P. S.; Grace, M. R.; McKelvie, I. D. Flow analysis methods for the direct ultra-violet spectrophotometric measurement of nitrate and total nitrogen in freshwaters. Anal. Chim. Acta. 2011, 704 (1-2), 116-122, DOI: 10.1016/j.aca.2011.07.048. (23) Gartia, M. R.; Braunschweig, B.; Chang, T. W.; Moinzadeh, P.; Minsker, B. S.; Agha, G.; Wieckowski, A.; Keefer, L. L.; Liu, G. L. The microelectronic wireless nitrate sensor network for environmental water monitoring. J. Environ. Monit. 2012, 14 (12), 3068-3075, DOI: 10.1039/c2em30380a. (24) Chatterjee, A.; Khandare, D. G.; Saini, P.; Chattopadhyay, A.; Majik, M. S.; Banerjee, M. Amine functionalized tetraphenylethylene: a novel aggregation-induced emission based fluorescent chemodosimeter for nitrite and nitrate ions. RSC Adv. 2015, 5 (40), 31479-31484, DOI: 10.1039/c4ra14765k. (25) Duffy, G.; Regan, F. Recent developments in sensing methods for eutrophying nutrients with a focus on automation for environmental applications. Analyst 2017, 142 (23), 4355-4372, DOI: 10.1039/c7an00840f. (26) Steimle, E. T.; Kaltenbacher, E. A.; Byrne, R. H. In situ nitrite measurements using a compact spectrophotometric analysis system. Mar. Chem. 2002, 77 (4), 255-262, DOI: Pii S03044203(02)00003-8 (27) Gentle, B. S.; Ellis, P. S.; Faber, P. A.; Grace, M. R.; McKelvie, I. D. A compact portable flow analysis system for the rapid determination of total phosphorus in estuarine and marine waters. Anal. Chim. Acta. 2010, 674 (2), 117-122, DOI: 10.1016/j.aca.2010.06.030. (28) Li, C.; Gui, J. C.; Sun, L. Y.; Cao, W. X.; Sun, Z. H.; Yang, Y. Z. In-Situ Instrument for rapid detection of nitrates and nitrites in Water. Spectrosc. Spect. Anal. 2013, 33 (6), 1714-1718, DOI: 10.3964/j.issn.1000-0593(2013)06-1714-05. (29) Sea-Bird Scientific. HydroCycle-PO4 Phosphate Sensor data sheet. Available at: https://www.seabird.com/hydrocyclepo-with-sdi-12/product?id=54627921731 (accessed Aug 1, 2018) (30) Systea. WIZ probe (Water in-situ analyzer) data sheet. Available at: http://www.systea.it/index.php?option=com_k2&view=item&id= 223:systea-spa-wiz-probe&Itemid=175&lang=en (accessed Aug 1, 2018). (31) EnviroTech Instruments. EcoLAB 2 Multi-Channel Analyzer System data sheet. Available at: http://www.labtech.com.mx/files/ecolab_2.pdf (accessed Aug 1, 2018) (32) Clinton-Bailey, G. S.; Grand, M. M.; Beaton, A. D.; Nightingale, A. M.; Owsianka, D. R.; Slavikt, G. J.; Connelly, D. P.; Cardwell, C. L.; Mowlem, M. C. A Lab-on-Chip analyzer for in situ measurement of soluble reactive phosphate: improved phosphate blue assay and application to fluvial monitoring. Environ. Sci. Technol. 2017, 51 (17), 9989-9995, DOI: 10.1021/acs.est7b01581. (33) Beaton, A. D.; Sieben, V. J.; Floquet, C. F. A.; Waugh, E. M.; Bey, S. A. K.; Ogilvie, I. R. G.; Mowlem, M. C.; Morgan, H. An automated microfluidic colourimetric sensor applied in situ to determine nitrite concentration. Sens. Actuators, B 2011, 156 (2), 1009-1014, DOI: 10.1016/j.snb.2011.02.042. (34) Beaton, A. D.; Cardwell, C. L.; Thomas, R. S.; Sieben, V. J.; Legiret, F. E.; Waugh, E. M.; Statham, P. J.; Mowlem, M. C.; Morgan, H. Lab-on-chip measurement of nitrate and nitrite for in


Page 11 of 21 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

situ analysis of natural waters. Environ. Sci. Technol. 2012, 46 (17), 9548-9556, DOI: 10.1021/es300419u. (35) Wang, H. L.; Zhang, S. W.; Guo, C. L.; Chu, D. Z. Development of an in-situ analyzer for nutrient monitoring in seawater. Aer. Adv. Eng. Res. 2017, 129, 550-553. (36) Taylor, E.; Bonner, J.; Nelson, R.; Fuller, C.; Kirkey, W.; Cappelli, S. Development of an in-situ total phosphorus analyzer, OCEANS’15 MTS/IEEE Washington, October 2015. pp. 1–4. DOI: 10.23919/OCEANS.2015.7404384 (37) Lin, B. C.; Xu, Z. Q.; Wang, J. Q.; Lu, M. A low-cost water quality monitoring prototype device with embedded chromogenic reagent capsules and dynamic colorimetric detection. Sens. Actuators, B 2017, 252, 24-29, DOI: 10.1016/j.snb.2017.05.056. (38) Nagul, E. A.; McKelvie, I. D.; Worsfold, P.; Kolev, S. D. The molybdenum blue reaction for the determination of orthophosphate revisited: Opening the black box. Anal. Chim. Acta. 2015, 890, 60-82, DOI: 10.1016/j.aca.2015.07.030. (39) Lin, K. N.; Ma, J.; Yuan, D. X.; Feng, S. C.; Su, H. T.; Huang, Y. M.; Shangguan, Q. P. Sequential determination of multinutrient elements in natural water samples with a reverse flow injection system. Talanta 2017, 167, 166-171, DOI: 10.1016/j.talanta.2017.01.063. (40) Drummond, L.; Maher, W. Determination of phosphorus in aqueous solution via formation of the phosphoantimonylmolybdenum blue complex. Re-examination of optimum conditions for the analysis of phosphate. Anal. Chim. Acta. 1995, 302 (1), 69-74, DOI: 10.1016/0003-2670(94)00429-P. (41) Zhang, J. Z.; Fischer, C. H.; Ortner, P. B. Optimization of performance and minimization of silicate interference in continuous flow phosphate analysis. Talanta 1999, 49 (2), 293304, DOI: 10.1016/S0039-9140(98)00377-4. (42) Griess, P. Bemerkungen zu der Abhandlung der HH. Weselsky und Benedikt „Ueber einige Azoverbindungen”.

Berichte der deutschen chemischen Gesellschaft 1879, 12 (1), 426428. (43) Feng, S. C.; Zhang, M.; Huang, Y. M.; Yuan, D. X.; Zhu, Y. Simultaneous determination of nanomolar nitrite and nitrate in seawater using reverse flow injection analysis coupled with a long path length liquid waveguide capillary cell. Talanta 2013, 117, 456462, DOI: 10.1016/j.talanta.2013.09.042. (44) Lin, K. N.; Ma, J.; Pai, S. C.; Huang, Y. M.; Feng, S. C.; Yuan, D. X. Determination of nitrite, phosphate, and silicate by valveless continuous analysis with a bubble-free flow cell and spectrophotometric detection. Anal. Lett. 2017, 50 (3), 510-529, DOI: 10.1080/00032719.2016.1182543. (45) Ministry of Environmental Protection of China, Standards for drinking water quality, GB 5749-2006. 2006. (46) Ministry of Environmental Protection of China, Quality standard for ground water, GB/T 14848-93. 1993. (47) Sjosten, A.; Blomqvist, S. Influence of phosphate concentration and reaction temperature when using the molybdenum blue method for determination of phosphate in water. Water Res. 1997, 31 (7), 1818-1823, DOI: 10.1016/S00431354(96)00367-3. (48) F. Koroleff, Determination of phosphorus. In Methods of seawater analysis; K. Grasshoff, M. Ehrhardt, K. Kremling, Eds.; Verlag Chemie, Weinheim, 1983, pp. 125–139. (49) DalL'Olmo, G.; Brewin, R. J. W.; Nencioli, F.; Organelli, E.; Lefering, I.; McKee, D.; Rottgers, R.; Mitchell, C.; Boss, E.; Bricaud, A.; Tilstone, G. Determination of the absorption coefficient of chromophoric dissolved organic matter from underway spectrophotometry. Opt. Express 2017, 25 (24), A1079-A1095, DOI: 10.1364/Oe.25.0a1079. (50) Ruiz, M. G.; Lutz, V.; Frouin, R. Spectral absorption by marine chromophoric dissolved organic matter: Laboratory determination and piecewise regression modeling. Mar. Chem. 2017, 194, 10-21, DOI: 10.1016/j.marchem.2017.03.012.


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 12 of 21

For TOC only

12 ACS Paragon Plus Environment

Page 13 of 21 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

399x249mm (300 x 300 DPI)


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

299x223mm (300 x 300 DPI)


Page 14 of 21

Page 15 of 21 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 2. “Fish-Bite” reservoir, (a) explosive view; (b) reservoirs, capsule and plug; (c) operation procedures of the reservoir and relevant mechanisms. 299x366mm (300 x 300 DPI)


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

299x225mm (300 x 300 DPI)

299x225mm (300 x 300 DPI)


Page 16 of 21

Page 17 of 21 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 4. Typical phosphate curves of normalized pulse number, n, against time, T, corresponding to different phosphate concentrations. 45x27mm (300 x 300 DPI)


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 5. Calibration curve of phosphate concentrations Cphosphate against coloration rate index bphosphate. 53x38mm (300 x 300 DPI)


Page 18 of 21

Page 19 of 21 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 6. Typical nitrite curves of normalized pulse number, n, against time, T, corresponding to different phosphate concentrations. 45x27mm (300 x 300 DPI)


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. Calibration curve of nitrite concentrations Cnitrite against coloration rate index bnitrite. 53x38mm (300 x 300 DPI)


Page 20 of 21

Page 21 of 21 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

530x717mm (96 x 96 DPI)


A low-cost automatic sensor for in-situ colorimetric detection of

Recommend Documents