VIETNAM NATIONAL UNIVERSITY – HO CHI MINH CITY INTERNATIONAL UNIVERSITY DEVELOPMENT OF a VERSATILE AND COMPACT PAPER BASED MICROFLUIDIC BIOSENSOR FOR DETECTION OF COPPER IN FOOD PRODUCTS

Abstract In this contribution, a versatile, inexpensive and compact paper-based microfluidic biosensor was developed to detect heavy metals in food products.. Urease enzyme was physicall

VIETNAM NATIONAL UNIVERSITY – HO CHI MINH CITY

INTERNATIONAL UNIVERSITY

DEVELOPMENT OF A VERSATILE AND COMPACT PAPER-BASED MICROFLUIDIC BIOSENSOR FOR DETECTION OF COPPER

IN FOOD PRODUCTS

A thesis submitted to The School of Biotechnology, International University

in partial fulfillment of the requirements for the degree of

B.S in Biotechnology

Student’s name: Nguyễn Lan Thảo - ID:BTIU09046 Supervisor: Dr Nguyễn Thái Lộc

June 2013

Acknowledgement

First of all, it is of great importance to deliver my thankfulness to Dr Loc

T Nguyen, who dedicated his generous time and sound knowledge to my

research as a mentor Truthfully, it was my absolute honor to work with Dr Nguyen Thanks to his insightful instruction, valuable experiences and

motivating encouragement, I became brave and confident enough to take on plenty of challenges Had it not been for his constant supports, I would not have successfully accomplished this study

Secondly, my sincere gratitude is expressed to Dr Khoi T Nguyen for consulting me with suitable techniques and stimulating ideas to deal with

difficulties that arose during the study of unfamiliar subjects That was his

inspiration which kept me persistent until the end

Moreover, I owe my gratefulness to all the laboratory managers and staffs, particularly MSc Long H Nguyen, MSc Dao Q T Tran, MSc Lieu B T Truong and BSc Anh L Dang, who devoted themselves to ensure a well-

equipped and safe laboratory and did not even mind working overtime or at weekend

I also would like to take this opportunity to say special thanks to Ms Han

B Nguyen, Ms Yen K T Dang and Mr Trong Q Luu for standing by me

throughout hardships, as well as cheering me up in various situations Besides, the supports from Mr Thien H Nguyen are much appreciated His interesting lessons enabled me to perform sophisticated image analysis as well as to turn ideas into vivid figures, all of which contributed to the best presentation I have ever made

It would be incomplete not to thank you all – my colleagues in lab 101, including Ms Dung M T Nguyen, Mr Dang D Nguyen, Ms Hien T T Nguyen,

Ms Khanh K T Nguyen, Ms My H T Nguyen, Ms Oanh K T Nguyen, Ms Vien

L Ngo and Ms Tram N L Nguyen for supporting and sharing with me no matter

in good times or in bad that made every moment priceless Together they

created the most memorable semester

Last but not least, it is my family who I want to convey the deepest appreciation to Thank my mom, Mrs Lan T Ta and my dad, Mr Hy D Nguyen for giving me such a favorable environment that I could concentrate and do the best in the study Thank my brothers for their sweet treats and assistance in delivering necessary research equipment No matter what the situations are, my family is always the greatest back-up that I can count on

For this moment, my heart is filled with endless love and sympathy from many other people that I cannot list here due to limited scope To all of those, I would love to personally say big thank you and promise to continuously improve myself

DEVELOPMENT OF A VERSATILE AND COMPACT

PAPER-BASED MICROFLUIDIC BIOSENSOR FOR

DETECTION OF COPPER IN FOOD PRODUCTS

Thao L Nguyen a , Loc T Nguyen a,*

a: School of Biotechnology, International University – Vietnam National University - HCMC

*: Corresponding author’s email address: ntloc@hcmiu.edu.vn

Abstract

In this contribution, a versatile, inexpensive and compact paper-based microfluidic biosensor was developed to detect heavy metals in food products The underlying principle was based on sensing of ammonia (NH3) released during urease-catalyzed hydrolysis of urea At the presence of heavy metals, the amount of NH3

produced diminished due to inhibition of urease enzyme Therefore, the concentration of target heavy metals could be indirectly determined from NH3, qualitatively, semi-quantitatively or quantitatively Hydrophobic barriers of functional areas the sensor were fabricated by paraffin-dipping method Urease enzyme was physically immobilized onto the sensor and phenol red was used for qualitative and semi-quantitative detection of heavy metal via image analysis The sensor was also integrated into an electrochemical system using simple screen-printed electrodes In the current study, Cu++ was used as a model heavy metal for testing performance of the sensor Qualitative results showed that a strong contrast between safe and unsafe sample, which was critical for the practical applicability of the sensor Relationship between color intensity and Cu++

concentration was characterized by a R2 of 0.98 and the linear range covered Cu++

concentrations from 0.01 – 1 ppm Detection limit was estimated to be 0.018 ppm which was well below standard limit established by WHO and EC for Cu++ Quantitative tests were still at beginning stage and were worth further investigations The findings from this study demonstrated that the proposed paper-based biosensor could be a promising platform to develop low-cost test kits for detection of heavy metals in foods

Keywords: Paper-based biosensor, wax-dipping, screen-printing electrodes, SPEs,

colorimetric test, amperometric test

1 Introduction

Heavy metals are ubiquitous pollutants and their presence in the environment can be attributed to natural or human activities Many heavy metals are carcinogens and may be involved in several dangerous diseases (Hossain et al., 2011) In general, heavy metals are not easily degraded and tend to accumulate in soils and sediments (Dominguez-Renedo et al 2013) Major sources

of soil contamination with heavy metals are wastewater irrigation, solid waste disposal, sludge applications, vehicular exhaust and industrial activities (Khan et

al 2008) Plants grown on contaminated soils may build up excessive content of heavy metals and eventually have negative effects on food quality and safety Therefore, it is of critical importance to monitor the contamination of heavy metals

in food products Currently, analytical methods for heavy metals usually rely on inductively coupled plasma/atomic, emission spectrometry (ICP/AES), inductively coupled plasma, mass spectrometry (ICPMS), atomic absorption spectroscopy (AAS), or wet chemical methods such as titrimetry, gravimetry, colorimetric assays, etc (Hossain et al., 2011) Despite their high sensitivity, selectivity, reliability, and accuracy, these methods are time-consuming, require sophisticated instrumentation, skilled personnel and complicated sample pretreatment (Turdean, 2011) Inherent drawbacks of these assays restrict their use mainly in centralized laboratory Thus, inexpensive, easy-to-use and portable test kits which can screen contamination of heavy metals for a large number of samples are highly desirable Recently, the emergence of paper-based microfluidic devices has become important resources for low-cost diagnosis The underlying principle of these devices is to pattern hydrophilic hydrophobic micron-size capillary channels

on paper using various methods including wax-printing, ink-jet printing, flexography printing, screen printing, etc (Li et al 2012) Paper has advantages

of being inexpensive, lightweight, available everywhere and compatible with biological samples (Martinez et al 2010) Paper-based devices can be easily disposed after use, require very small volume of sample and reagents, capable of analyzing multianalytes at the same time and easily mass produced (Nie et al 2010) Several paper-based assays with diverse formats developed for health care, environmental monitoring, food quality control and forensic science (Li et al 2012)

Paper-based devices can be applied for either qualitative or quantitative analysis When coupled with appropriate detection methods, paper-based devices can produce quantitative results with reasonable accuracy Popular detection methods such as colorimetric, electrochemical, electrochemiluminescence or

chemiluminescence were extensively reviewed by Li et al (2012) Paper-based microfluidic devices can also serve as an excellent platform for biosensor development Biosensors with exceptional performance such as high specificity and sensitivity, rapid response, low cost, compact size and easy-to-use were considered as important means in clinical, food and environmental monitoring (Amine et al 2006) One of the most widely used technique in monitoring pollutants and toxic compounds is enzyme inhibition-based biosensing in which concentrations of target compounds can be determined from the extent to which the enzyme is inhibited, signifying by the product concentration Popular enzymes used as bio-recognition elements for detection of heavy metals are horseradish peroxidase, urease, glucose oxidase, alcohol oxidase, glycerol 3-phosphate oxidase, invertase and acetylcholinesterase (Amine et al, 2006)

For analysis of heavy metals, several paper-based test kits with significant contribution in term of sensing methods, sensitivity, selectivity were developed for mercury (Gu et al., 2011; Hossain et al 2011; Torabi et al 2011; Aragay et al 2012), cadimium (Abe et al 2011; Hossain et al 2011; Marzo et al 2013), copper (Fang et al 2010; Hossain et al 2011), iron (Apilux et al., 2010), lead (Mazumdar

et al., 2010; Hossain et al 2011), chromium (Hossain et al 2011; Liu et al., 2012) and gold (Apilux et al., 2010) Considering the end-use of paper-based devices as low-cost diagnostic kits in developing countries, they should be able to perform analyses at different levels of complexity In certain situation, an inexpensive qualitative test requiring no advanced analytical skills is sufficient to make sure that heavy metals in foods, agricultural products or water sources are under a safe limit If more accurate results are expected, the same device can be integrated into colorimetric, electrochemical or other detection systems to produce semi-quantitative or quantitative results

Some authors (Fang et al 2010; Abe et al 2011; Gu et al., 2011; Torabi

et al 2011; Aragay et al., 2012; Liu et al 2012; Marzo et al 2013) propose approaches using fluorescence or strip reader to obtain quantitative results besides qualitative test However, these methods still require large sample volume, bulky and sometime complicated fabrication process Recently, methods such as wax printing, wax dipping appeared as alternative approaches for producing inexpensive micron size devices The fundamental principle of creating

a microfluidic device was to pattern hydrophilic channels bounded by hydrophobic barriers One of the simplest method was to use a printer to deposit patterns of solid wax on the paper which was then heated to enable the wax to penetrate into the entire thickness of paper, thus generating complete hydrophobic barrier An

alternative approach was wax dipping which was first proposed by Songjaroen et

al (2011) which are capable of both qualitative tests by visually observing color change or quantitative analysis using digital camera In this method, a mould was used to produce the hydrophobic areas The whole assembly of the mould and paper was quickly dipped and withdrawn from the melting wax Wax deposited on the uncovered parts, resulting in desired hydrophobic areas Paraffin was used in this study for its easy availability and low prices while having similar characteristics

as wax The fact is that digital cameras and scanners are not as selective and sensitive as conventional analytical instrumentation, nevertheless, highly selective and sensitive detectors are still required for low analyte concentrations (Dungchai

et al 2009) Electrochemistry-based method is attractive detection soucheme for paper-based biosensors due to its compact size, low-cost, high sensitivity and selectivity Using this method, Apilux et al (2010) successfully detected gold in waste stream Nonetheless, few studies have yet to develop paper-based biosensors capable of multiple testing schemes for detection of heavy metals in foods and food products

In this contribution, our objectives were a) to design and fabricate a based biosensor using wax-printing technology; b) to apply the produced biosensors for qualitative and semi-quantitative detection of heavy metals and c)

paper-to integrate the biosensors with electrochemical analyzer

2.1 Chemicals and solution preparation

Urease (type III, EC 3.5.1.5, 33U/mg) from Canavalia ensiformis (Jack

bean) was purchased from Sigma Aldrich (USA) Carbon ink (C-200) and Ag/AgCl inks (AGCL-375) were obtained from Hudson (USA).Tris-HCl (min 99.0%) was from HiMedia Labs (Mumbai, India) Urea, phenol red, CuSO4, KCl, NaOH and HCl were of analytical grade and used as provided Filter papers (60x60cm) and white pellet paraffin wax were sourced from local chemical stores

Stock solutions of enzyme (1000U/mL), Cu++ (680 ppm) were prepared in Tris-HCl buffer solution (50mM, pH 7.0) The enzyme stock solution was made on

a weekly basis and kept in refrigerator after use Urea (0.1M) and KCl (0.1M) mixture was prepared in distilled water

2.2 Fabrication of paper-based microfluidic biosensor

The design and dimensions of the sensor used in this study was illustrated

in Figure 1A The sensor was developed to accommodate both colorimetric and electrochemical tests simultaneously The design was made using CorelDraw x3 (Corel Inc., Mountain View, USA) Initially, reference, working and counter electrodes were patterned on reaction zone (Figure 1A) using screen printing method Carbon ink (C-200) was used for working and counter electrodes whereas silver/silver chloride ink (AGCL-375) was used for reference electrode The printed electrodes sensor was dried at 65oC in oven for about 30 min The printing patterns

of the electrodes were tested for continuity by a multimeter

Figure 1 (A)Schematic design of the paper-based biosensor illustrating sensing zones, loading area and conductive pads to interface electrochemical system (B)Steel mould used to pattern hydrophilic-hydrophobic channels on the sensor

In the next stage, the microfluidic channels were created A mould (Figure 1B) was cut from a 0.3 mm steel plate using computer numerical control (CNC) machine by a local workshop The mould can be used many times without deformation and decrease in resolution of the hydrophilic channels About 200g

of paraffin was melted in a 500 ml beaker using hot plate (IKA RH Basic 2) Experiment were conducted at different dipping temperature (55-80oC) and time (1-5s) to determine the optimal conditions for patterning hydrophylic barrier Prior

to wax dipping, the filter paper containing printed electrodes were cut into rectangles (50 mm width x 80 mm length) And sensor was sandwiched between

the mould and a glass slide The mould was positioned so that electrochemical reaction zone was aligned with the printed electrodes A permanent magnet was used to hold the mould against the glass slide The whole assembly was dipped into the melted paraffin and quickly withdrawn in predetermined time After the paraffin was cooled to ambient temperature, the mould was removed and the sensor was visually examined for any defections The hydrophilic channels should remain clear and sharp The paraffin needed to penetrate evenly into the filter paper Those sensors which did not meet the requirements were discarded The process was described in Figure 2

Figure 2 Fabricating procedures of paper-based biosensor includes: (1) Screen-printing working electrode (WE) and counter electrode (CE) using carbon inks, and (2) reference electrode (RE) and conductive pads with Ag/AgCl inks (3-4) The sensor was wax-dipped to create hydrophobic/hydrophilic pattern (5) Tape was attached to the back side

2.3 Qualitative and semi-quantitative analysis

The fundamental principle of the qualitative and semi-quantitative analysis was based on hydrolysis reaction of urea ((NH2)2CO) (1)

3 2

Urease 2

tested with Cu ions The testing protocol (Figure 3) basically involved 4 main steps a) immobilizing enzyme on the reaction zones; b) loading samples on the sensor; c) incubation and d) taking the readings after adding urea to induce color change

Figure 3 Qualitative and semi-quantitative testing procedure (A) Firstly the mixture of enzyme and indicator was loaded into the sensing zone and (B) sample was introduced to the loading area After incubating for 10 min (C) urea was added and ammonia was produced from the hydrolyzation (D) color develop after 3 min (E) The result was scanned and analyzed

2.3.1 Enzyme immobilization

In this study, the urease enzyme was physically immobilized on the paper substrate Urease enzyme stock solution (200U/ml) was prepared in Tris buffer (50mM, pH 7.0) and mixed with phenol red solution (0.4% in 50% ethanol) at ratio 1:1 (V/V) The enzyme-indicator mixture (ED) was used within 3 hours after preparation and kept on ice Prior to immobilization of enzyme on the sensor, one side of the sensor was covered with Scoth tape Then 1.5 µL of the enzyme-indicator mixture was placed on each reaction zones and allowed to air-dried for about 2 min The free side of reaction zones were now covered with another piece

of Scoth tape (Figure 4A) to control the flow of reagents eventually Periodically, enzyme immobilization was verified by adding 25 µL of CuSO4 solution at 0 and 5

mM concentrations to the loading zones of two different sensors The inhibitory reaction was incubated for about 10 min After that, 25 µL of 0.1M urea was loaded Since it will take times for color to develop, the reading time was standardized as 3 minutes afterward The enzyme immobilization was considered

successful if red color was observed for negative sample and no color change was noticed for positive sample

Figure 4 The attachment of scotch tape to the front and back sides of sensors

to prevent leakage and enhance uniformity of reagent distribution Tape covers the sensing zones on (A) front side and (B) back side

2.3.2 Qualitative analysis

The qualitative analysis was based on visual matching color of an unknown sample to a color scale of which variation of color intensity indicated concentration ranges of heavy metals present Experiments were conducted to record colorimetric response of the sensor to different concentrations of Cu++ (0.001-1ppm) Using an office scanner (Canon Co Ltd, Vietnam), the obtained results can served as the input to construct the color chart for qualitative analysis

2.3.3 Semi-quantitative analysis

Development of standard curve

Serial dilutions of CuSO4 (0.001-1ppm) solution were prepared from stock and used to develop standard curve for semi-quantitative analysis Samples (25µL) at different concentrations were loaded on the sensors and incubated for about 10 min Then, 25 µL of urea (0.1 M) was added and after 3 min, the results were recorded using the office scanner of which resolution was set at 600 dpi (Figure 3B-E)

The analysis of color developed on the loading zones was conducted using Adobe Photoshop CS2 (Adobe Inc., San Jose, USA) (Figure 3E) The rectangular

selection was fixed to 1000 pixel as to cover reacting regions As Histogram can automatically filters out the desire single channel color, it was used to measure the pink color directly from the RGB pictures Giving that the Green color was the major contributor to violet color, the Green channel value was measured

The standard curve was constructed by plotting the color data against concentrations of heavy metals used All data points presented were the average

of 9 replications The detection limits were calculated based on 3 times standard deviations of the negative control

Validation

To validate the results, an independent set of samples with known concentrations were prepared and tested using the developed sensor Colorimetric signals were obtained as described in the previous section and the concentration

of heavy metals were estimated from the calibration curves The recoveries percentage was calculated as below (Eq.2):

100%

) M , Cu of tration log(Concen

Calculated

) M , Cu of tration log(Concen

Actual

= (%)

of the electrodes on the sensor and the screw was maintained by a paper clip

Figure 5 Experimental set-up for electrochemical measurement

2.4.2 Urease-based amperometric response of the sensor to different concentrations of heavy metals

After the sensor was placed in the holder, a fixed voltage (-1.5 V) was applied and resulting current was measured for about 40 seconds The current intensity was dependent on the amount of NH3 produced in the reaction zone during urea hydrolysis Addition of heavy metals resulted in impaired catalytic ability of urease, hence indirectly affecting the amount of NH3 formed Therefore, amperometric response of the sensor could reveal the amount of heavy metals present in the samples Percentage of current reduction was used to estimate the contamination of heavy metals and was given by:

(%) reduction Current

% 100 I

With I1, I2 were the measured electrochemical currents before and after

inhibition with Cu++, in µA

To construct calibration curve, serial dilutions of Cu++ (0,0.001, 0.01 and 0.1ppm) were used Correlation equation describing relationship between current reduction and concentrations of Cu++ was obtained from regression analysis

2.5 Data analysis

Each data point represents the average result of six replications Using descriptive analysis, the results were showed by mean values plus standard deviation (SD) Next, the correlation between color intensity and Cu++

concentration was evaluated and the goodness of fit was considered using coefficient of determination (R2) For comparison purpose, the variation between measurements was evaluated using ANOVA and the Student t-test

3 Result and discussion

3.1 Fabrication of the paper-based microfluidic sensor

Effects of dipping temperature and time

Patterns of paraffin on paper at different dipping temperatures were presented in Figure 6

Figure 6 Effect of different dipping temperatures on the resolution of hydrophobic barrier with (a-f) corresponding to 55-80oC with 5oC increment

The consistency and resolution of the channels were directly influenced by melting temperature and dipping time As evidenced from the photos, the optimal temperature ranges were from 70-75oC Under these conditions, hydrophobic-hydrophilic boundaries were well-defined and the width of the channels was most consistent Lower temperatures led to the excessive build-up of paraffin on the paper surface When temperature of paraffin was greater than 75oC, melting paraffin tended to diffuse under the mold and, consequently, damage the channels, reaction zones and loading area Quick observations during the fabrication process revealed that dipping time should be limited within 1 second Longer exposure time would result in accumulation of paraffin on the surface, hence affecting the final finish of the sensor In a different study (Songjaroen et

al 2011), white beeswax pellets required significantly higher temperature of

125oC to achieve the optimal patterns However, dipping time was also recommended to be no longer than 1 s to prevent the excessive spreading of wax into the paper In this study, paraffin was selected to create hydrophobic barriers

of the microfluidic channels The cost of this material was lower than wax and it could produce a smooth surface after setting However, the barriers made by paraffin were not as stable as expected As illustrated from Figure 7, after a few times of exposure to testing reagents, paraffin boundaries weakened and diffusion

of solutions through the hydrophobic barriers was noticed

Figure 7 Leakage of reagent due to weakened of hydrophobic barrier after testing Comparison of (a) before and (b) after used sensor

Further studies should investigate on the effects of potential factors such

as melting point, hydrophobic-hydrophilic balance of dipping agent etc on the quality of microfluidic channels The information obtained would be valuable in selecting appropriate materials for patterning the paper-based sensor

Effects of fluid flow and evaporation

Performance of the paper-based microfluidic sensors was affected by the flow and evaporation of the fluids Due to capillary effects, fluids from loading area diffused through channels and entered the reaction zones It was found that, moving fluid leached and transported the immobilized enzyme and indicators to the borders of reaction zones As a result, the variation of color intensity was less visible for different samples (Figure 8) and this adversely affected the detection of color change by naked eyes or by optical devices such as cameras or scanners Several studies (Martinez et al 2007; 2008; Abe et al 2008; Bruzewicz et al 2008; Carrihlho et al 2009; Fenton et al 2009; Li et al 2012) experienced the same uneven distribution of color on paper-based microfluidic devices Fluid entering reaction zone from a channel spreads radially and carry coloring agents and lower near the entry port Visualization of fluid flow clearly confirmed the patterns of color distribution (Kauffman et al 2010) Abe et al (2008) suggested various solutions including using different geometrical arrangement and increasing printing cycles of sensing inks There was some improvement with pH and protein but not glucose sensing In this study, several geometrical arrangements, including 3D patterns, were investigated and none yielded satisfactory results The effects of fluid flow on uniformity of color development worsened when the sensing scheme required multiple loadings of samples and reagents In the current research, samples were loaded first and after incubation time, urea solution was again added to loading area to impart final color change The flows of sample and urea solution severely washed off the indicator and it was hardly differentiate color intensities of different samples