Subscriber access provided by UNIV OF DURHAM
Article
Chromatographic separation and visual detection on wicking microfluidic devices: Quantitation of Cu2+ in surface-, ground-, and drinking water Gayan C. Bandara, Christopher Heist, and Vincent T Remcho Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.7b04087 • Publication Date (Web): 15 Jan 2018 Downloaded from http://pubs.acs.org on January 15, 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.
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 17 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
Chromatographic separation and visual detection on wicking microfluidic devices: Quantitation of Cu2+ in surface-, ground-, and drinking water Gayan C. Bandara, Christopher A. Heist, Vincent T. Remcho* Department of Chemistry, Oregon State University, Corvallis, OR 97331, USA Fax: 541-737-2062, Email:
[email protected] Abstract Copper is widely applied in industrial and technological applications, and is an essential micro nutrient for humans and animals. However, exposure to high environmental levels of copper, especially through drinking water, can lead to copper toxicity, resulting in severe acute and chronic health effects. Therefore, regular monitoring of aqueous copper ions has become necessary as recent anthropogenic activities have led to elevated environmental concentrations of copper. On-site monitoring processes require an inexpensive, simple, and portable analytical approach capable of generating reliable qualitative and quantitative data efficiently. Membrane-based lateral flow microfluidic devices are ideal candidates as they facilitate rapid, inexpensive, and portable measurements. Here we present a simple, chromatographic separation approach in combination with a visual detection method for Cu2+ quantitation, performed in a lateral flow microfluidic channel. This method appreciably minimizes interferences by incorporating a non-specific polymer inclusion membrane (PIM) based assay with a “dot-counting” approach to quantification. In this study, hydrophobic polycaprolactone (PCL)-filled glass microfiber (GMF) membranes were used as the base substrate onto which the PIM was evenly dispensed as an array of dots. The devices thus prepared were then selectively exposed to oxygen radicals through a mask to generate a hydrophilic surface path along which the sample was wicked. Using this approach, copper concentrations from 1 ppm to 20 ppm were quantified from 5µL samples using only visual observation of the assay device. Introduction Metal contaminants are widely considered to be among the most hazardous of environmental pollutants.1-3 Due to their valuable physical and chemical properties, these metals are utilized in a wide variety of modern applications, hence,
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
Page 2 of 17
the environmental risk is very high,4-6 as was seen recently in Portland, Oregon, where glass manufacturers were found to have released toxic metals into the environment in the course of producing raw materials for the decorative glass hobby/industry.7 Copper is one of the most broadly used of all metals, and is a mainstay in many industries and technologies, in addition to being one of the most widely used materials for domestic plumbing.8 This introduces a high risk of leaching of unacceptably high levels of copper into domestic water for human consumption, as recently evidenced in many public schools in Portland, Oregon.9 Exposure to low levels of copper is likely to be beneficial as it is an essential nutrient for proper health and development of humans and animals.10,11 However, exposure to higher environmental levels of copper, especially through drinking water, can lead to copper toxicity resulting in severe health effects such as jaundice, hemoglobinuria, kidney failure, liver damage, and potentially death.12,13 Therefore, to control copper in drinking water, the US Environmental Protection Agency (USEPA) has established a regulation stating that the dissolved copper concentration should not exceed 1.3 ppm in more than 10% of consumers’ water outlets.14 Current analytical techniques for copper analysis require sophisticated laboratory environments with trained personnel and benchtop instruments, such as atomic absorption spectroscopy (AAS), atomic emission spectroscopy (AES), and inductively coupled plasma optical emission spectroscopy (ICPOES).15-18 These techniques can require time consuming sample preparation and pre-treatment steps before analysis.15,16 Therefore, such techniques, while entirely suitable for onetime, highly precise and accurate analysis, are not practical for on-site rapid analyses or frequent field monitoring applications. Monitoring elevated levels of micronutrients/metals such as copper in drinking/environmental waters requires a new, inexpensive, simple and portable analytical technique that can generate reliable qualitative and quantitative measurements. Membrane-based wicking microfluidic devices with simple, stable assay chemistries are ideal as these devices facilitate rapid, inexpensive, and portable analysis.19,20 One of the major challenges in adapting a simple chemical assay to the microfluidic format is the elimination of interferences to enable element specific detection. Such processes usually require pre-treatment of the sample, highly specific assay chemistries, or specific detection mechanisms as well as instruments, all of which make the process more complicated, expensive and challenging to adapt to a portable microfluidic platform. Therefore, alternative methods are required to gain specificity in a microfluidic assay without sacrificing the quality of the measurement. Here we present an inexpensive and simple chromatographic separation approach on a lateral flow microfluidic channel, which radically minimizes interferences for detection of Cu2+ in water. The device incorporates a non-specific polymer inclusion membrane (PIM) based assay which, when coupled with a dot-counting quantification approach enables
2 ACS Paragon Plus Environment
Page 3 of 17 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
simple visual observation to generate quantitative information. Device fabrication is done using a hydrophobic polycaprolactone-filled glass microfiber (GMF) substrate on which the PIM is evenly dispensed as a dot array, followed by selective exposure to oxygen radicals (through a pattern mask) to generate a hydrophilic flow path that aligns with the linear paths of the dot array.21 The assay platform itself, prior to exposure to oxygen radicals, is hydrophobic; thus the finished device consists of a hydrophilic pattern within a hydrophobic field. Interaction of Cu2+ ions with the complexation ligand, 1-(2-pyridylazo)-2-naphthol (PAN) in the PIM assay results in the formation of a red colored complex. We have capitalized on the solution properties of the complex to chromatographically separate PAN-Cu2+ from rest of the PAN-Mn+ complexes, enabling facile, selective detection of Cu2+ on the lateral flow microfluidic platform without any interferences from other PAN-Mn+ species in solution. To our knowledge, this is the first report of a simple chromatographic separation approach on a lateral flow microfluidic channel that has been developed to overcome interferences in this way.
Materials and methods Materials. PCL was obtained from Perstorp (Perstorp, Warrington, UK). Whatman glass microfiber (GF/A) membranes (GE Healthcare Bio-Sciences, Pittsburgh, PA, USA) were used as the base substrate for all microfluidic devices. The different mask patterns were designed using SolidWorks (2013-2014 Education edition, Waltham, MA, USA) and cut on tape (i tape, Intertape Polymer Group, Marysville, MI, USA). Trimethylchlorosilane (TMCS) was purchased from Alfa Aesar (Ward Hill, MA, USA). Creek water was obtained from Oak creek (Corvallis, OR, USA) and spring water (Arrowhead 100% mountain spring water, Nestle water North America Inc., Tamford, CT, USA) was purchased from a local supermarket. Ore samples containing CuO were purchased from Thorn Smith laboratories (Beulah, MI, USA). All buffer solutions were prepared using ultrapure deionized water (Milli-Q Advantage A10, EMD Millipore, Billerica, MA, USA). Glacial acetic acid (Fisher chemicals, Fair Lawn, NJ, USA) was used in preparation of pH 4 and 4.5 buffers, 2-(N-morpholino) ethanesulfonic acid (MES) (Cal-biochem, West Chester, PA, USA) and pyridine (EMD Chemicals, Gibbstown, NJ, USA) were used in preparation of pH 6 buffer, Tris base (VWR international, West Chester, PA, USA) was used in preparation of pH 8 buffer and ethanolamine (Alfa Aesar, Ward Hill, MA, USA) for preparation of pH 10 buffer. The concentration of each buffer was 10 mM and the pH of the buffer solutions was adjusted using either 1M HCl (Macron Fine Chemicals, Center Valley, PA, USA) or 1M NaOH (Mallinckrodt, Paris, KY, USA). The following chemicals were also used in this study: 1-(2-pyridylazo)-2-naphthol (PAN) (Alfa Aesar, MA, USA ), N,N'-bis(3,5-di-tert-butylsalicylidene)-1,2-cyclohexanediaminomanganese(III) chloride (Jacobsen’s ligand) (Strem
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
Page 4 of 17
Chemicals, MA, USA), Bis(2-ethylhexyl) sebacate (TCI America, Portland, OR, USA) Analytical grade Toluene (Macron Fine Chemicals, Center Valley, PA, USA), technical grade Acetone, ACS grade CuCl2•2H2O, MnSO4•H2O, ZnCl2 (Mallinckrodt, Paris, KY, USA), Pb(NO3)2 (Sigma-Aldrich, St. Louis, MO, USA), Analytical grade NaCl, CoCl•6H2O, CdCl2 (Fluka Chemical Corpera-tion, Ronkonkoma, NY, USA), FeCl3 (98%, Nitric acid (EMD Chemicals, Gibbstown, NJ, USA), NiCl2 (99.99%), Formic acid (99%) (Alfa Aesar, Ward Hill, MA, USA), For-mic acid ammonium salt (Alfa Aesar, Ward Hill, MA, USA). Instrumentation. Laurell WS-400 (North Wales, PA, USA) spin coater was used in spin coating. The paper tape masks were designed using SolidWorks (2013-2014 Education edition, Waltham, MA, USA). VLS 3.50 laser cutter (Universal Laser Systems, Scottsdale, AZ, USA) was used as the cutting tool in mask preparation. The PIM was dispensed using a digital solution dispenser (HP D300 Digital Dispenser, USA). Evactron (Redwood City, CA, USA) decontaminator/RF plasma cleaner was used in oxygen radical exposure experiments. Agilent 240 AAS was used in Flame AAS experiments (Agilent Technologies, Santa Clara, CA, USA). PCL-filled GMF membranes. PCL solutions (w/v) were prepared by dissolving appropriate weights of PCL in appropriate volumes of toluene. Solutions were spin-coated at 2500 rpm for 30 seconds on GMF membrane followed by drying at 50 oC for 15 min. The initial weight percentage of PCL: GMF was approximately 50:50 under the above conditions. PIM assay. The PIM assay was prepared by dissolving 0.005g of PAN, 0.040 g of Jacobsen’s ligand, 0.100 g PCL, and 0.010 g Bis(2-ethylhexyl) sebacate in 10 mL of Toluene. The dispensed volume was determined by the required assay/experimental conditions. Surface modification of PCL-filled GMF. PCL-filed GMF was treated with 25% (v/v) TMCS and acetone at room temperature for 4 hrs. The treated membranes were soaked in 100% acetone to remove excess TMCS (1hr) followed by washing with excess acetone. The membranes were then dried at room temperature (>24hrs). Oxygen radical exposure. Oxygen radical exposure was conducted under conditions determined separately for each experiment. The pressure and forward RF power were maintained at constant values of 0.6 Torr and 13 W, respectively. Selective exposure to radicals – not to the plasma but only to radicals generated by the plasma – was accomplished by covering the area of the membrane intended to remain unexposed with a patterned mask, prepared as described above, and ensuring that the substrate was placed at a distance well below the plasma region.
4 ACS Paragon Plus Environment
Page 5 of 17 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
Unknown analysis. Solutions of ore samples were prepared by acid digestion (0.25-0.30 g with 30 ±1 mL of 3 M nitric acid) followed by diluting to 50.0 mL with DI water. 2.0 mL of this solution was then diluted to 250 mL to prepare the test solution for analysis by flame AAS. The test solution for microfluidic analysis was prepared by diluting 7.0 µL of the digested solution to 1.0 mL with 100 mM MES buffer (pH6.0).
Results and discussion Chemistry of the PIM. Conventional fabrication of PIMs focuses on element specific detection which often requires specific chemistries and conditions.22-24 Such requirements limit the applicability of PIMs in simple wicking microfluidic platforms. Hence, in the new approach described here, a colorimetric PIM assay was designed such that it included a simple and non-specific chemistry for complexation of metal ions. The active colorimetric complexation ligand in this case is PAN. The chemistry of PAN is well studied and understood: it forms discrete colored complexes with most of the common transition metal ions (PAN-Mn+). 23-25. These PAN-Mn+ complexes show different properties under different solution conditions, such as pH dependent complexation and variable solubility.25 According to the literature, Cu2+ reacts with PAN in 1:1 stoichiometry to form a red colored PAN-Cu2+ ligand complex.26,27. The organic PIM is hydrophobic, and therefore has limited interaction with aqueous solutions (figure S1 (a)). However, exposure to oxygen radicals alters the hydrophobic PIM surface chemistry, making it more hydrophilic, allowing it to interact readily with aqueous solutions as evidenced in figure S1 (b). As PAN is immobilized in the PIM, the Cu2+ ions in the aqueous sample must be transported into the PIM to form the PAN-Cu2+ complex. As shown in figure 1a, the Cu2+ (sample) ↔ Cu2+ (PIM) ↔ PAN-Cu2+ equilibrium should favor formation of the PAN-Cu2+ complex, thus producing a red color output. The time required to establish this equilibrium will determine the time of analysis. Schiff base ligands, such as Jacobsen’s ligand, are capable of forming coordinate complexes with metal ions through imine nitrogen and hydroxyl group coordination.28,29 Therefore, as evidenced in figure S2, the presence of excess Jacobsen’s ligand in the PIM indirectly promotes the above equilibrium shift towards the formation of PAN-Cu2+ by recruiting Cu2+ ions into the PIM. Achieving specificity in detection of Cu2+. Even though most of the PAN-Mn+ complexes are insoluble in water, the PAN-Cu2+ complex is soluble in water at neutral and acidic pH. Hence, PAN-Cu2+ can be extracted from the organic PIM region on a 2D wicking microfluidic channel fabricated on a PCL filled GMF membrane. Figure 1b shows the steps followed to fabricate a microfluidic device by selective exposure of the polymer-filled glass microfiber substrate to oxygen
5 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 17
radicals, using appropriate masks, leading to the production of a surface lateral flow microfluidic channel along which a linear array of PIM assay dots is distributed.21 This linear array arrangement allows for simple visual detection of the separated PAN-Cu2+ complex as it moves downstream with the sample flow. This results in a change in the coloration of the area between the brightly colored PIM dots, from white to pale red. A simple count of the number of red dots is then used to quantify the amount of Cu2+ in the system.
Figure 1. (a) Interactions of Cu2+ with the PIM and formation of PAN- Cu2+ complexes: The PIM has simple chemistry, PAN complexes with Cu2+ to form a red colored product. The Jacobsen’s ligand works as a transport enhancer for Cu2+ ions resulting increased transport of Cu2+ from the aqueous sample into the PIM. (b) Overview of the PIM assay microfluidic device fabrication process: (i) Pore space in the GMF substrate is filled with PCL by spin coating to obtain PCL-filled GMF (ii). (iii) PIM is dispensed as an array of dots on the PCL-filled GMF surface. (iv) Membrane is sandwiched between the top (facing towards the oxygen radical source) and bottom (facing away from the oxygen radical source) masks, and exposed to oxygen radicals (v). (vi) Tape masks are removed, revealing the hydrophilic surface pattern along the linear PIM array.
The specificity of this approach for Cu2+ detection was investigated by evaluating a variety of potential metal ion interferents in environmental/ drinking water. These experiments were conducted under various pH conditions to investigate the pH dependence of PAN-Mn+ solubility. As shown in figure 2, the assay itself is not specific for Cu2+, Co2+, Ni2+, Zn2+, Cd2+, or Mn2+, as these all formed colored complexes with PAN, most notably at alkaline pH (figure 2c and 2d). Interestingly, in the pH range of 3.0 – 7.0, only Cu2+ and Co2+ produced colored complexes. Under such conditions, the
6 ACS Paragon Plus Environment
Page 7 of 17 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
PAN-Cu2+ complex moved with the aqueous sample flow, and thus was separated from the PIM region. Therefore, despite the presence of other metal ions, the solubility of PAN-Cu2+ allowed for it to be differentially transported and therefore spatially separated, facilitating the detection of Cu2+ in the sample.
Figure 2. Assay results for a variety of common metal ions, illustrating solubility and pH dependence. A 5 µL aliquot of a 10 ppm Mn+ solution was applied to an individual assay channels to perform each assay. (a) pH 4 samples were applied. (b) pH 6 samples were applied. (c) pH 8 samples were applied. (d) pH 10 samples were applied. Migration of the red PAN-Cu2+ complex with the sample flow was observed under acidic pH conditions (a and b). The dimensions of a single PIM assay dot were 1.0 mm X 0.5 mm, and each dot was produced by dispensing two 25 nL aliquots of PIM solution using an HP D300 dispenser. The distance between dots, in which region color development was detected, was 1.0 mm. (* fabricated channels with no test solution, B: Buffer at respective pH).
It was observed that the efficiency of complexation and separation was improved by decreasing the pH of the sample from 6.0 to 4.0 (figure 2a and 2b). We postulate that the increased H+ concentration at the lower pH reduced the extent of interaction of Cu2+ with the negatively charged silanol groups on the GMF surface. Under such conditions, formation of the PAN- Cu2+ complex and its migration along the microfluidic path is more favorable. Minimization of undesirable secondary sorptive interactions is also desired. SEM images (figure 3a) and EDX spectroscopic surface
7 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 8 of 17
elemental composition analysis (Table S1) showed that the PCL coating on the GMF membrane covers a large proportion of the glass fiber surface (figure 3a ii) yet there is a considerable amount of glass surface (populated with silanol groups) still exposed. Therefore, an additional silylation step was added during membrane preparation for subsequent experiments. In order to further suppress secondary sorptive interactions, the GMF membrane medium was treated with a common silylating agent trimethyl chlorosilane (TMCS). TMCS reacted with exposed silanol groups on the GMF membrane to generate a more hydrophobic surface as shown in scheme 1 below:30
Scheme 1. Reaction of TMCS with silicate substrate.
This TMCS modification simplified the assay fabrication process and helped to define the boundaries of the hydrophilic channels. As shown in figure 3b, on the TMCS modified membranes, enhanced migration efficiency of the PAN-Cu2+ complex was observed at neutral to acidic pH as compared to unmodified membranes (figure 3b i). The unmodified membrane required a more acidic sample pH and excess solution volume to generate measurable color migration. The TMCS surface modified membranes delivered optimal assay performance.
8 ACS Paragon Plus Environment
Page 9 of 17 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. (a) SEM images of (i) raw GMF and (ii) TMCS treated PCL-filled GMF. Raw GMF consists of a network of borosilicate fibers31. (b) Assay results for Cu2+ on non-silylated (left) and silylated (right) media. TMCS surface modification masked the silanol groups and enhanced the efficiency of PAN- Cu2+ migration with the aqueous sample.
The concentration of Cu2+ in the sample was proportional to the distance travelled on the lateral flow channel32, 33
; the Cu2+ was transported with the surface flow until it was fully depleted from the sample by complexation with PAN.
Therefore, higher concentrations of Cu2+ traveled farther into the linear array of PIM dots, thereby increasing the number of red zones in direct proportion to the Cu2+ con-centration. As shown in figure 4a, the calibration curve yielded a linear relationship over the concentration range of 1.0-20.0 ppm with a sample volume of only 5µL. This device was fabricated with a dot resolution of 6.5 dots (1.0 mm X 0.50 mm) per 1.0 cm of channel length. At that dot resolution, the signal detection limit was 1.5 ppm (n=12)34 yet the visual limit of detection35 was 1 ppm. This device produced a 1.1 ± 0.8 ppm (n= 12) reading with 5 µL of 1.0 ppm (15.7 µM) of Cu2+ standard solution (in DI water). The highest detectable concentration for the device as constructed here was limited to 20 ppm as the capacity of the PIM is exceeded beyond that concentration. In contrast, non-TMCS modified devices produced a similar calibration curve with less precision due to the uneven flow characteristics of these devices (figure S3). This dynamic range is within the maximum allowable
9 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 10 of 17
concentration of Cu2+ in drinking water as stated by the US EPA. Therefore, this approach is highly suitable for detection of Cu2+ in drinking water as well as environmental water samples.
Figure 4. Device-resident calibration curves for the PIM/ dot assay for Cu2+. (a) A photograph of the actual developed assay with a concentration gradient of Cu2+. (b) Calibration data on a silylated device with an expanded view of the assay (in background).
Oxygen radical exposure effects on the chemistry of the PIM. As the fabrication process is conducted after the assay is dispensed and involves exposure of the assay reagents to oxygen radicals, there is a risk of altering the chemistry of the PIM. Such an effect could adversely impact quantitative outcomes. Therefore, ATR-FTIR spectra were collected and studied in an effort to identify any significant chemistry change in the PIM assay. Interestingly, nearly identical spectra were observed from 0 sec (unexposed) to 5 min (heavily exposed) finished assay devices (figure 5). This suggests that exposure of the PIM to oxygen radicals did not measurably alter the chemistry of the bulk PIM at the 12 sec. exposure time used for device fabrication. The Si-O stretching signal (e) was observed in the finger print region of every spectrum due to the overlap of background signal produced by the GMF substrate.
10 ACS Paragon Plus Environment
Page 11 of 17 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. ATR-FTIR spectra of PCL-filled GMF membranes at different oxygen radical exposure time intervals. (a) Asymmetric CH2 stretching, (b) Symmetric CH2 stretching, (c) C=O stretching, (d) C-O and C-C stretching, (e) Si-O stretching.36
Extending detectable concentration range, improving detection limits, accuracy and precision. As this simple method is based on counting dots, the accuracy of the test is dependent on the number of dots that represent a unit concentration. The accuracy and precision of the analysis can further be improved by using devices with high-resolution arrays of dots as shown in figure S4. In this improved scenario, each PIM dot is smaller, resulting in a lower amount of reagent per dot. This method allows for an increased number of dots to change color per unit concentration of analyte thereby increasing the accuracy of the measurement and affording the opportunity to enhance precision by including additional parallel channels. By using a high precision deposition tool, it was possible to demonstrate enhanced assay dot resolution and improved assay accuracy. In addition, it was shown that variable channel width could be harnessed as a means by which to accommodate samples having higher analyte concentrations (figure 6). Visual recognition of the color is critical for the counting dots detection approach, especially at lower concentrations. Given that visual acuity varies from person to person, we collected human perception data for a device with a dot resolution of 6.5 dots per 1 cm as depicted in figure S5. Based on the results, we determined that 90% of the population surveyed identified the color correctly. Of that 90%, 31% was able to identify color at the 0.5 ppm Cu2+ level, and 59% at concentrations of 1 ppm and above (figure S5 b).
11 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 12 of 17
Figure 6. Devices with non-uniform channel widths are appropriate over broader concentration ranges. Channels 1 and 2 consist of a uniform channel width of 1 mm while channels 3 and 4 have a combination of 1 mm and 1.5 mm widths. Channels 1 and 2 saturate with 20 ppm Cu2+, but varying the width of channels 3 and 4 prevents it from saturating at 20 ppm. The available assay area (amount) is lower on a channel with a smaller width than for a channel with a larger width, as the assay area (dot size) is limited by the edges of the microfluidic channel.
Validation of microfluidic detection approach and unknown analysis. A validation study was conducted to compare the microfluidic approach to Cu2+ analysis presented here with a widely accepted benchmark, flame AAS. Spiked drinking water and creek water samples were used to validate the accuracy of quantitative detection of Cu2+ ions on the microfluidic platform. Drinking water and creek water samples were spiked with a 200 ppm Cu2+ standard solution to obtain final concentrations of 2 ppm, 5 ppm, 8 ppm, 10 ppm, and 12 ppm. A non-spiked sample was used as the 0 ppm concentration. We limited our range of comparative study to 0-12 ppm to ensure high data fidelity and to correlate with the linear range of our AAS (Figure S6 a), though the linear range of the microfluidic approach continues to 20 ppm. As the initial pH of the test solutions was basic (creek water: 8.26, drinking water: 8.02), the PAN-Cu2+ complex was not mobile with the sample fluid flow but it was easily separated with a buffer flow (pH 4.5) produced by adding a 5.0 µL aliquot of buffer after dispensing the sample. There was no need for any other sample preparation step for the microfluidic analysis. However, for the AAS, creek water samples were filtered and both creek water and drinking water samples were required to be acidified before the analysis to generate reproducible data.
12 ACS Paragon Plus Environment
Page 13 of 17 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
The Cu2+ concentration values were calculated for the test solution by using the linear regression equation of calibration plots of each method (figure S6 b and c). As shown in figure 7, the concentration values for each method were plotted against each other to compare the accuracy of the microfluidic platform. A linear relationship between the newly developed microfluidic method and the benchmark flame AAS method was observed for both drinking water samples and creek water samples with R2 of 0.99 as shown in figure 7a and figure 7b respectively. Overall, this data demonstrates that this microfluidic approach is capable of producing accurate quantitative data for an unknown test solution.
Figure 7. A direct comparison between the flame-AAS and the microfluidic devices was performed by plotting Cu2+ concentration values for each method against each other (a) for spiked spring water, (b) for spiked creek water. Error bars for the flame-AAS measurements indicate much greater precision for that approach, at the expense of complexity and cost of instrumentation.
The applicability of the developed fabrication technique towards analyzing real world Cu samples was investigated by testing ore samples containing CuO. Multiple samples (n=6) were tested with the approach developed here and analyzed with a standard flame AAS method. The results obtained from the two methods were compared to the manufacturer’s provided analysis information. Solid samples were digested with nitric acid to dissolve all the copper resulting a highly acidic solution (below pH 3.0). Therefore, the pH of the test solutions was adjusted to pH 6.0 prior to the analysis by the microfluidic approach. The calibration curve (figure S-5 (c)) was also constructed using standards
13 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 14 of 17
prepared at pH 6.0 in order to maintain the integrity of the experiment. The microfluidic devices used for this experiment followed the same fabrication approach and PIM array as those shown in figure 1b. The results are summarized in table 1. The microfluidic analysis results are comparable to the results obtained by flame AAS (at the 95% confidence level) as well as the manufacturer’s data proving the practical applicability of the microfluidic assay.
Table 1. Comparison of Cu weight % in solid samples determined by different analytical approaches
CONCLUSION We have presented a simple separative approach on a lateral flow microfluidic channel fabricated on a silylated PCL-filled GMF membrane to specifically detect Cu2+ in environmental and drinking waters using a non-specific PIMbased assay. A unique quantitative approach – visual observation and counting of colored assay zones – proved comparable in quantitative outcomes to a conventional (flame AAS) benchmark. This methodology is supported by a new approach to fabrication of wicking microfluidic devices, which allowed for the microfluidic channel to be fabricated after the assay had been dispensed. Therefore, it was possible to demonstrate control over the exposure area and thus regulate both the concentration range of applicability of the assay and the accuracy of the assay for samples containing trace concentrations of Cu2+. This approach enabled optimal device performance and broader acceptable tolerances for dispensing of the assay reagents. As this assay does not require any instrumentation or any expensive reagents, it is a highly useful and practical platform for frequent monitoring of Cu2+ in environmental/drinking waters.
ACKNOWLEDGEMENT The authors acknowledge Owen T. Shellhammer for his assistance with device fabrication and optimization experiments, Teresa Sawyer of the Oregon State University Electron Microscope Facility and Matthew H. Kremer for their assistance with SEM imaging and EDX experiments, Yuanyuan Wu for her assistance with ATR-FTIR experiments, Neal
14 ACS Paragon Plus Environment
Page 15 of 17 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
Sleszynski and Kristi Edwards for provision of solid Cu unknowns, the students in CH 324 (quantitative analysis, Fall 2016) and for providing CuO samples and flame AAS results, as well as Sumate Pengpumkiat and Kelly Ramzy for helpful discussions.
ASSOCIATED CONTENT Supporting Information Oxygen radical exposure fabrication technique enhances the interaction of hydrophobic assays with aqueous solutions; Effect of Jacobsen’s ligand as the transport enhancer for the PIM assay; Variation of elemental composition of GMF with PCL filling; Performance of non-TMCS modified PCL-filled GMF membrane with PIM counting dot assay; Higher accuracy and precision with increased resolution of assay dot dispensing; Human perception data; Calibration curves for method validation experiment
REFERENCES (1)
Järup, L. Br. Med. Bull. 2003, 68 (1), 167–182.
(2)
Khan, S.; Cao, Q.; Zheng, Y. M.; Huang, Y. Z.; Zhu, Y. G. Environ. Pollut. 2008, 152 (3), 686–692.
(3) Tchounwou, P. B.; Yedjou, C. G.; Patlolla, A. K.; Sutton, D. J. In Molecular, Clinical and Environmental Toxicology; Springer, 2012; pp 133–164. (4)
Watson, M. L. J. Biophys. Biochem. Cytol. 1958, 4 (6), 727–730.
(5)
Zinkle, S. J. Phys. Scr. 2015, 2016 (T167), 014004.
(6)
Hou, D.; He, J.; Lü, C.; Ren, L.; Fan, Q.; Wang, J.; Xie, Z. Ecotoxicol. Environ. Saf. 2013, 93, 135–144.
(7) Oregonian/OregonLive, F. Z. | T. High levels of selenium near Bullseye Glass prompt new restrictions http://www.oregonlive.com/environment/index.ssf/2016/09/high_levels_of_selenium_near_b.html (Accessed December 20, 2017). (8) The balance, Learn about the common Uses of copper, Bell, T., 2017. https://www.thebalance.com/copperapplications-2340111 (December 20, 2017) (9) KGW.com, Portland, http://www.kgw.com/news/health/high-lead-levels-found-at-all-11-pps-buildings-tested-sofar/268102018. (Accessed December 20, 2017) (10)
Koreti, S.; Prasad, N. Micronutr. CONTENT BREAST MILK 2014, No. 3248.
(11)
Maryon, E. B.; Molloy, S. A.; Kaplan, J. H. Am. J. Physiol.-Cell Physiol. 2013, 304 (8), C768–C779.
(12)
Brewer, G. J. J. Trace Elem. Med. Biol. 2012, 26 (2), 89–92.
(13)
Brewer, G. J. Exp. Biol. Med. 2007, 232 (2), 323–335.
15 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 16 of 17
(14)
US Environmental Protection Agency April 1996.
(15)
Daşbaşı, T.; Saçmacı, Ş.; Ülgen, A.; Kartal, Ş. Food Chem. 2015, 174, 591–596.
(16)
Zhao, L.; Zhong, S.; Fang, K.; Qian, Z.; Chen, J. J. Hazard. Mater. 2012, 239, 206–212.
(17) 70–75.
Shoaee, H.; Roshdi, M.; Khanlarzadeh, N.; Beiraghi, A. Spectrochim. Acta. A. Mol. Biomol. Spectrosc. 2012, 98,
(18)
Karimi, M.; Sereshti, H.; Khojeh, V.; Samadi, S. Int. J. Environ. Anal. Chem. 2013, 93 (4), 401–415.
(19)
Loh, L. J.; Bandara, G. C.; Weber, G. L.; Remcho, V. T. Analyst 2015, 140 (16), 5501–5507.
(20) 1132.
Pengpumkiat, S.; Wu, Y.; Boonloed, A.; Bandara, G. C.; Remcho, V. T. Analytical Methods, 2017, 9 (7), 1125-
(21)
Bandara, G. C.; Heist, C. A.; Remcho, V. T. Talanta 2018, 176, 589–594.
(22)
Annane, K.; Sahmoune, A.; Montels, P.; Tingry, S. Chem. Eng. Res. Des. 2015, 94, 605–610.
(23)
Nagul, E. A.; Fontàs, C.; McKelvie, I. D.; Cattrall, R. W.; Kolev, S. D. Anal. Chim. Acta 2013, 803, 82–90.
(24)
Jayawardane, B. M.; Cattrall, R. W.; Kolev, S. D.; others. Anal. Chim. Acta 2013, 803, 106–112.
(25)
Pease, B. F.; Williams, M. B. Anal. Chem. 1959, 31 (6), 1044–1047.
(26)
Jaggi, S.; Gupta, U. Maced. J. Chem. Chem. Eng. 2013, 32 (1), 57–67.
(27) Parpot, P.; Teixeira, C.; Almeida, A. M.; Ribeiro, C.; Neves, I. C.; Fonseca, A. M. Microporous Mesoporous Mater. 2009, 117 (1), 297–303. (28)
Cozzi, P. G. Chem. Soc. Rev. 2004, 33 (7), 410–421.
(29)
Kemper, S.; Hrobárik, P.; Kaupp, M.; Schlörer, N. E. J. Am. Chem. Soc. 2009, 131 (12), 4172–4173.
(30)
Anderson, R. C.; Muller, R. S.; Tobias, C. W. J. Electrochem. Soc. 1993, 140 (5), 1393–1396.
(31)
GE Healthcare Life Sciences. General Electric Company 3013.
(32)
Cate, D. M.; Noblitt, S. D.; Volckens, J.; Henry, C. S. Lab on a Chip 2015, 15 (13), 2808–2818.
(33)
Pratiwi, R; Nguyen, M. P.; Ibrahim, S.; Yoshioka, N.; Henry, C. S. Talanta 2017, 174, 493–499.
(34)
Lewis, G. G.; DiTucci, M. J.; Phillips, S. T. Angew. Chem. Int. Ed. 2012, 51 (51), 12707–12710.
(35)
Harris, D. C. Quantitative Chemical Analysis, 8th ed.; Freeman, New York.
(36) 387.
Elzein, T.; Nasser-Eddine, M.; Delaite, C.; Bistac, S.; Dumas, P. J. Colloid Interface Sci. 2004, 273 (2), 381–
16 ACS Paragon Plus Environment
Page 17 of 17 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
For TOC only
17 ACS Paragon Plus Environment