Low-Cost, Robust, and Field Portable Smartphone ... - ACS Publications

Dec 1, 2016 - For development of the proposed sensor, we utilize the ambient light sensor (ALS) of the smartphone as light intensity detector and its ...
0 downloads 9 Views 2MB Size
Subscriber access provided by United Arab Emirates University | Libraries Deanship

Article

Low-cost, robust and field portable smartphone platform photometric sensor for fluoride level detection in drinking water Iftak Hussain, Kamal Uddin Ahamad, and Pabitra Nath Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.6b03424 • Publication Date (Web): 01 Dec 2016 Downloaded from http://pubs.acs.org on December 2, 2016

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.

Analytical Chemistry 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 12

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

Analytical Chemistry

Low-cost, robust and field portable smartphone platform photometric sensor for fluoride level detection in drinking water Iftak Hussain†, Kamal Uddin Ahamad‡ and Pabitra Nath†,* †

Applied Photonics and Nano-photonics laboratory, Department of Physics, Tezpur University, Assam, 784028, India



Department of Civil Engineering, Tezpur University, Tezpur, Assam 784028, India

ABSTRACT: Groundwater is the major source of drinking water for people living in rural areas of India. Pollutant like fluoride in ground water may present in much higher concentration than the permissible limit. Fluoride does not give any visible coloration to water and hence no effort is made to remove or reduce the concentration of this chemical present in drinking water. This may lead to serious health hazard for those people taking ground water as their primary source of drinking water. Sophisticated laboratory grade tools such as Ion selective electrodes (ISE) and portable spectrophotometers are commercially available for in-field detection of fluoride level in drinking water. However, such tools are generally expensive and require expertise to handle. In this paper, we demonstrate the working of a low cost, robust and field portable smartphone platform fluoride sensor that can detect and analyze fluoride level concentration in drinking water. For development of the proposed sensor, we utilize the ambient light sensor (ALS) of the smartphone as light intensity detector and its LED flash light as an optical source. An android application ‘FSense’ has been developed which can detect and analyze the fluoride level concentration in water samples. The custom developed application can be used for sharing of in-field sensing data from any remote location to the central water quality monitoring station. We envision that the proposed sensing technique could be useful for initiating fluoride removal program undertaken by governmental and nongovernmental organization here in India.

Presence of fluoride in drinking water within the controlled level is important for human health since it can prevent dental carries1. However, with an elevated concentration of fluoride in water may pose serious health hazard such as decaying of tooth enamel which subsequently can cause dental fluorosis2. Excessive intake of fluoride in drinking water could lead to skeletal abnormalities or damage3. Due to these health issues, World Health Organization (WHO) establishes a limit of 1 mg/L of fluoride in drinking water4. According to Bureau of Indian Standards the permissible value of fluoride in drinking water is 1.0 mg/L and in the absence of any alternate sources the maximum permissible limit may be as high as 1.5 mg/L5. It is estimated that 25 nations all over the world is affected by fluorosis6. Out of twenty-nine states, twenty states in India are found to be affected by endemic of fluorosis. Approximately 90 million inhabitants including 6 million children are in high risk or being affected due to elevated fluoride level concentration in drinking water7. In many parts of North-Eastern region of India and Assam in particular, fluoride level concentration up to 6.88 mg/L has been found in ground water8. Common people taking ground water as their primary source of drinking water in these regions are in a high risk

of serious health hazards. Due to non-availability of proper laboratory facility water samples from these regions are generally sent to the central water quality monitoring laboratory and the procedure of estimating fluoride concentration in water is an in-efficient and time consuming process. Several works related to the design and development of low cost fluoride monitoring systems in drinking water quality have been reported and demonstrated in the recent past9-11. These analytical methods requires visual inspection of the fluoride concentration, which may depend on the observers' perception thus, measured data may vary from observer to observer. Very recently, smartphone based fluoride concentration sensor has been reported12. This method is based on the RGB color model of the smartphone camera. According to Cisco Visual Networking Index (VNI), till 2015, 563 million smartphone connections were added globally13. This has been possible due to tremendous improvement both in hardware and software configuration of the smartphone has occurred in last few years. Due to the incorporation of high end sensors such as rare camera, ALS, user friendly graphical interface and real time data sharing facilities, researchers across the globe have

1

ACS Paragon Plus Environment

Analytical Chemistry

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

been actively engaged for last few years to convert the smartphone into a low cost and user friendly sensing tool for different physical, chemical and biological sensing applications14-23. The rare camera of the phone has been explicitly used for developing such diagnostic tools24-31. Most branded smartphones contain the Avago APDS9930 or ams AG (TAOS) TMD 2771 integrated ambient light and proximity sensor chip32, 33. Due to its large spectral response range 350 nm to 1000 nm and having high dynamic range from 0 –20,000 lux with a resolution of 0.01 lux, it is possible to convert the embedded ALS of the smartphone into a sensitive detector for various sensing investigations. Suitable android platform applications can be developed to access the ALS data in lux units without any further processing. Very recently, we have demonstrated the use of this specific detector for sensing of different parameters34-36. Herein, we have utilized the ALS of the smartphone and the LED flash to develop a low cost, easy to use and field portable fluoride sensor. The detection scheme is based on the standard SPADNS colorimetric method for fluoride level detection in which fluoride reacts with the zirconium dye and forms a colorless complex anion (ZrF62‑) and the dye37. The fluoride concentration proportionately forms complex anion in the solution. As the fluoride concentration increases it tends to bleach the dye to make it progressively lighter in colour. A specific wavelength from the LED flash of the smartphone is allowed to interact with the reagent treated water sample and the modulated transmitted optical signal from the sample is detected by the ALS of the smartphone. An android application is developed to convert the ALS response into a readable data. Initially, a characteristic calibration curve of the sensor is obtained by recording the sensor response for standard sample solutions and upon getting the calibration curve, fluoride level concentration of an unknown water sample can be estimated. Simple and easy to use user friendly interface has been developed for android smartphone so that the designed sensor can be handled by an inexperienced person. Further, by utilizing the existing mobile network facility, in-field sensing data can be shared with the laboratory located anywhere in the world. Thus, with the designed sensing scheme real time detection of fluoride level concentration in water samples can be possible. This eventually would be easier for governmental and non-governmental organizations to initiate fluoride removal program for a specific region.

Page 2 of 12

purification. The de-ionized (DI) water used for preparing of the sample solutions was obtained from a Milli-Q water purifying system. Preparation of calibrating solutions. The reagents have been prepared following the standard procedure38. 958 mg of SPADNS is dissolved in DI water and diluted to 500 ml to prepare SPADNS solution. Zirconyl acid reagent is prepared by dissolving 133 mg of Zirconyl chloride octahydrate in 25 ml of DI water. 350 ml of Hydrochloric acid is added to the solution and the final volume is diluted to 500 ml with DI water. Equal volume of SPADNS solution and Zirconyl acid has been mixed to produce SPADNS-Zirconyl acid complex reagent. Standard sodium fluoride solution has been prepared by dissolving 110.5 mg of Sodium fluoride in DI water and final volume is diluted to 500 ml with DI water. The standard sample solutions with known fluoride concentration in the range 0-3.0 mg/L have been prepared by diluting a specific volume of the standard Sodium Fluoride solutions with 50 ml of DI water. EXPERIMENTAL SECTION Spectrophotometric analysis of the test solution. Prior to start of the experimental investigation, spectrophotometric analysis has been performed to find the peak absorption wavelength condition for zirconium- SPADNS dye solution that contains water samples of different fluoride concentrations. It has been observed that for a given sample solution the peak absorption condition is seen in the wavelength range 560-580nm. Figure 1 shows the characteristics absorption spectrum for four different sample solutions

MATERIALS AND METHODS Reagents. SPADNS, sodium 2-(parasulfophenylazo)-1,8dihydroxy-3,6-naphthalene disulfonate and Sodium Fluoride were procured from SRL, India. Zirconyl chloride octahydrate were procured from Sigma-Aldrich. Hydrochloric acid has been procured from Rankem chemicals, India. All chemicals used in the present work are of analytical grade and were used as received without further

Figure 1. Absorption spectrum of zirconium-SPADNS dye mixed with water sample containing fluoride ion at different concentrations and inset shows the photo image of the corresponding samples.

2

ACS Paragon Plus Environment

Page 3 of 12

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

Analytical Chemistry

Figure 2. (a) Schematic of the smartphone based fluoride sensor; (b) Photograph of the designed sensor and (c) A screenshot image of the developed ‘FSense’ application for the present sensor.

From this preliminary sensing response curve, it can be seen that with the increase in fluoride concentration in water there has been a gradual decrement in the absorbance values of the transmitted light signal. This variation in modulated light signal can be easily detected by the ALS of the smartphone. Principle of smartphone based fluoride sensor. The sensing principle is based on Beer-Lambert law of absorption39,40. When a collimated optical beam propagates through a medium, the intensity of the beam gets attenuated due to absorption by the medium and the magnitude of absorption depends on wavelength of the incident light, path length and the concentration of the medium.

The intensity of the transmitted light IT can be expressed as

     (1) where,  is the intensity of the incident light propagating through the absorbing medium of path length , c is the concentration of the solution and  is the molar absorption co-efficient which is a function of wavelength of the incident light. Design of the sensing system. Figure 2(a) shows the schematic diagram of the designed smartphone fluoride sensor. In the present work, Sony Xperia E3 smartphone

3

ACS Paragon Plus Environment

Analytical Chemistry

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

has been used. The specification of the ALS for this smartphone can be found elsewhere34-36. Light signal from the LED flash is coupled to a multimode optical fiber of dimension 980/1000 μm and NA = 0.37 (Product id. 02534, Edmund optics) through a Plano-convex lens, (Lens 1) (7 mm diameter, focal length 11 mm, Edmund Optics 32-404). A narrow-band optical filter (Product id. 7-0302, Spectrogon) with peak transmission wavelength of 568 nm and full width at half-maximum (FWHM) 8 nm light signal from the optical source to the sample. The output light signal from the optical fiber is allowed to transmit through a pair of collimating (Lens 2) and focusing lens (Lens 3) arrangement and finally is coupled to the ALS of the smartphone. A quartz cuvette (Product id. 21Q-5, 5 mm path length, Spectrosil® Quartz) containing the water sample is placed in the optical path between the collimating and the focusing lens of the sensing set-up. All optical components used in the present work have been integrated in a custom development plastic holder and is attached to the smartphone. The LED flash of the smartphone used in the present work has very high intensity. It provides a luminance of 164 lux at a distance of 0.5 m and 700 mA forward current41. For the designed optical set-up, with a distance of 60 mm between the ALS and the output end of the optical fiber, the ALS measures the intensity of the collimated light beam as high as 1150 lux when the sample holder is filled with DI water. A high intensity light signal received by the ALS may lead to selfheating of the detector and this may subsequently cause unwanted fluctuations in the sensor reading. To mitigate this issue, a 3 mm thick nylon sheet has been used as a diffuser and is placed between the focusing lens and the ALS of the smartphone. It has been observed that with the incorporation of the diffuser, the intensity of the signal has decreased by a factor of 0.25. With these low intensity value the ALS gives desirable stable reading and low measurement error. An android application ‘FSense’ has been developed to detect the intensity reading of the ALS while measuring the fluoride level concentrations of water samples. By using the same application data analysis, storing and data transfer can be performed by the smartphone itself. Figure 2(b) shows the photo image of the designed sensor. The dimension of the plastic cradle is measured to be 75 mm length, 30 mm breadth and 20 mm in width and its weight is approximately 180gm without the smartphone. Development of ‘FSense’ in android platform OS. As already stated for simultaneous detection and analysis of fluoride level content in water samples, an android platform application ‘FSense’ has been developed and is installed in the smartphone. This application has been developed using MIT app inventor 2- a cloud based platform where one can design and develop applications according to its requirement42. Also, we have used the light sensor extension provided by the Pura Vida apps43 for detection of modulated light signal intensity received by the ALS of

Page 4 of 12

the smartphone. This application is compatible with any smartphones having android OS version greater than 2.3 (Gingerbread). Since the resolution of the ALS for the considered smartphone is 0.01 lux, we have programmatically limited the ALS response up to second decimal point. The workflow of the applications is shown in figure 3. Figure 3 (a) shows the screenshot image of display panel of the phone that contains the ‘FSense’ application. Upon clicking the ‘Fsense’ icon the user run the application and a new window pops-up which is shown in figure 3 (b). This window provides five optional buttons: ‘Instructions’, ‘Determine Fluoride concentration’, ‘Calibration’, ‘Tag location in Map’ and ‘Save information and share’. On clicking the instructions button, a new window will pop-up which contains stepwise instructions to operate the designed sensor. This has been shown in figure 3 (c). In the present investigation, we calibrate the designed sensor with known fluoride content samples in the range 0 mg/L- 2 mg/L. In the home screen of ‘FSense’ application when user clicks ‘Determine Fluoride concentration’ button, a new tab will pop-up and when the user inserts the fluoride content sample in the optical path of the setup, the smartphone measures the corresponding modulated light intensity in lux units. By using the precalibration data the sensing app estimates the fluoride level concentration of an unknown sample when the user clicks the ‘Find Fluoride concentration’ button in the same tab. This has been shown in figure 3 (d). Also, threshold level of fluoride concentration in water samples has been symbolically indicated in the same window of the application. We have set the permissible limit to be 1 mg/L as suggested by World Health Organization (WHO)4. The ALS output may fluctuate due to the variation of surrounding environment that may subsequently result measurement error of the designed sensor. To mitigate this affect the device should be re-calibrated for onsite sensing investigation. This can be done by clicking ‘Calibration’ button in the same application. Upon clicking this button the application provides options shown in figure 3 (e) to the user and if the user clicks ‘Yes’ it will pop-up the calibration tab where one can perform the onsite re-calibration of the designed sensor. Figure 3 (f) and (g) shows the normalized sensor response values for the standard sample solutions and the corresponding characteristics curve of the sensor respectively. If the user choose ‘No’ button in the Recalibration tab then the sensor considers the previously developed calibration curve and the home screen will pop-up. After determining the fluoride level concentration for a specific region, the user can mark the place in Google map by clicking ‘Tag location in Map’ which pop-up the Google map application shown in figure 3 (h). In the home screen when the user clicks the ‘Save information and share’ button it pop-ups another tab where one can write comments and information about the sensing data which will be stored as a text file in the phone. These have been shown in figure 3 (i) and (j). Figure 3 (j) shows the saved file as ‘Report.txt’

ACS Paragon Plus Environment

Page 5 of 12

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

Analytical Chemistry

Figure 3. Workflow of the custom developed android applications; figures (a)-(d) show the different steps of the detection process for pre-calibrated sensor, figures (e)-(g) show the optional calibration processes that may be required for in-field sensing investigations and figures (h)-(k) show the real-time data sharing and storing by the designed “FSense” application. The screenshot image in figure (h) is taken from Google map application, Google Inc., (used with permission) to indicate geographical location of the region where fluoride level detection in water has been performed.

ACS Paragon Plus Environment

Analytical Chemistry

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 6 of 12

in the device storage. Using the communication facility, in-field sensing data can be shared in a social network platform or by individual mailing. When the ‘Share’ button is clicked it will show all the communication applications present in the phone. For instance, we have shown the Gmail application shown in figure 3 (k) and the infield sensing data has been sent using this communication application. A video showing the operation of the designed sensor using the developed ‘FSense’ application has been provided in the supplementary information. Device calibration. The device has been calibrated with zirconium-SPADNS dye solution which contains water samples of known fluoride concentration. At first we have prepared solutions in the range 0 mg/L -3 mg/L with step incremental value of 0.2 mg/L. The sensor response is recorded to find out the linear range. Figure 4(a) shows the corresponding sensor response curve of the considered sample solutions. It can be seen that the sensor response is found to be linear when the fluoride level concentration in the sample varies in the range 0 mg/L to 2.6 mg/L and becomes non-linear beyond this level. Using the ‘FSense’ application a calibration curve is initially obtained and is stored in the phone database for its subsequent applications for detection of fluoride concentrations in unknown water samples. Figure 4(b) shows the calibrated sensor response curve for standard solutions with the fluoride concentration variation from 0 mg/L to 2.0 mg/L. The error bars shown in this figure represents the standard deviations for five consecutive measurements of each sample recorded by the designed sensor. The value of correlation-coefficient r = 0.99242 signifies the linearity of the detection schemes in measuring fluoride level concentration in water. From the regression analysis, following equation has been modeled which can be used to estimate fluoride concentration of an unknown water sample:

Figure 4. Response characteristics of the designed smartphone based fluoride sensor. (a) Sensor response curve for the range 0 mg/L – 3 mg/L. (b) Calibration curve prepared for 0 mg/L- 2 mg/L.

Unknown fluoride concentration 

 !"#$%& '(')$(* + 0.75 (2) 0.13

This relation between the normalized sensor response versus fluoride concentration of the standard fluoride solution has been used within the ‘FSense’ application to determine the concentration of fluoride in unknown water samples. If the fluoride concentration is above the calibration range then the successive dilution with distilled water is required. RESULTS AND DISCUSSION Sensitivity, Resolution and Limit of detection of the designed sensor. Sensitivity is defined as the change in the sensor response for per unit change of input stimulus. Resolution of a sensor system indicates the smallest

magnitude of input stimulus required to produce an observable change in the sensor response44. The sensitivity and resolution of the designed sensor has been estimated by measuring sensor response for Zirconium-SPADNS reagent treated water samples solution that contain fluoride concentration of 0 mg/L, 0.02 mg/L, 0.04 mg/L, 0.06 mg/L, 0.08 mg/L and 0.10 mg/L. Figure 5 shows the characteristic sensor response curve for the considered samples. The linear correlation coefficient is observed to be r = 0.99449. The error bars in this figure indicate the standard deviation for ten consecutive measurements of each sample solutions. Considering the linearity in sensor response the sensitivity can be determined using the following equation: Sensitivity (S) 

ACS Paragon Plus Environment

∆7

∆89

(3)

Page 7 of 12

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

Analytical Chemistry

Figure 5. Response characteristic of the designed sensor for samples with lower level of fluoride concentrations in the range 0 mg/L-0.1 mg/L.

Figure 6. Fluoride concentration measurement for field collected water samples. (a) Histogram representation of fluoride concentration estimation by a commercial spectrophotometer and by the designed sensor. (b) A screenshot image of one of the considered area (Hamren, Karbi-anglong) shown in Google maps application, Google Inc., (used with permission) to demonstrate the in-field data sharing ability of the designed sensor.

ACS Paragon Plus Environment

Analytical Chemistry

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

where, ∆S the change in sensor response and ∆CF is the change in concentration of the standard fluoride solutions. The sensitivity for the designed sensor is calculated to be 0.86 A.U./mgL-1.

Page 8 of 12

Sensor performance with different smartphones. All modern smartphones are in general equipped with ALS sensor, we have investigated the performance of the

The resolution of the designed sensing system depends on the resolution of the ALS which is 0.01 lux or 1.06 x 10-4 A.U. Resolution of the proposed sensor can be estimated as follows: 1 Resolution (R)  ; = x ?@A7 (4) < where, ?@A7 is the resolution of the ALS sensor and S is the sensitivity of the sensor. Using this equation the resolution of the sensor is estimated to be 1.23 x 10-4 mg/L. To estimate the Limit of Detection (LoD), we have considered the response characteristic of the designed sensor in the range 0 mg/L – 0.1 mg/L. We have used ICH guidelines45 to determine the LoD; according to which the LoD in an analytical measurement can be estimated as C D 

3σ (5)