ARTICLE pubs.acs.org/ac
Electrogenerated Chemiluminescence Detection in Paper-Based Microfluidic Sensors Jacqui L. Delaney,† Conor F. Hogan,*,† Junfei Tian,‡ and Wei Shen‡ † ‡
Department of Chemistry, La Trobe University, Victoria 3086, Australia Australian Pulp and Paper Institute, Department of Chemical Engineering, Monash University, Clayton, Victoria 3800, Australia ABSTRACT: This paper describes the first approach at combining paper microfluidics with electrochemiluminescent (ECL) detection. Inkjet printing is used to produce paper microfluidic substrates which are combined with screen-printed electrodes (SPEs) to create simple, cheap, disposable sensors which can be read without a traditional photodetector. The sensing mechanism is based on the orange luminescence due to the ECL reaction of tris(2,20 -bipyridyl)ruthenium(II) (Ru(bpy)32þ) with certain analytes. Using a conventional photodetector, 2-(dibutylamino)ethanol (DBAE) and nicotinamide adenine dinucleotide (NADH) could be detected to levels of 0.9 μM and 72 μM, respectively. Significantly, a mobile camera phone can also be used to detect the luminescence from the sensors. By analyzing the red pixel intensity in digital images of the ECL emission, a calibration curve was constructed demonstrating that DBAE could be detected to levels of 250 μM using the phone.
n important emerging area in the field of chemical sensors is the development of simple, inexpensive sensors for medical diagnostics1-7 or other sensing applications.1-6,8,9 Such devices, if they are simple, affordable, rapid, and robust, can provide a means of bringing otherwise unattainable health care and other benefits to developing countries and remote communities. In order to be affordable, the sensors must be mass-producible from cheap, readily available starting materials and should preferably be readable without the aid of a dedicated scientific instrument. Whitesides et al.3 first described the fabrication of paper-based sensors in which microfluidic channels were defined by hydrophobic patterns produced using photolithography.1-4,8 These microfluidic paper-based analytical devices (or μ-PADs) have the significant advantage of not requiring any external means of fluid transport, as it occurs via capillary action. Also, they require only small sample volumes, the paper may filter or otherwise separate the sample, they are easy to store and transport, and they are readily disposed of safely via incineration. Importantly, these sensors can be read colorimetrically using a device which has become almost ubiquitous, even in the developing world: a mobile phone with a built-in camera. Cell-phone ownership is predicted to shortly reach 5 billion, with most growth occurring in the developing world.10 This has important implications for telemedicine, for example the data from such a sensor can be readily transmitted to a clinician anywhere in the world for interpretation. Other less expensive means of fabrication of paper microfluidics have since been explored. Printing is among the cheapest and most easily implemented means of mass production available, and inkjet printing7,11 of wax and other hydrophobic materials12 has recently been shown to be very effective in the production of paper-based sensors at minute cost.
A
r 2011 American Chemical Society
Apart from colorimetric based sensing,1-4,7,12,13 electrochemical detection has also recently been used for microfluidic paper-based sensors. Such devices have been referred to as μ-PEDs. Dungchai et al. demonstrated electrochemical detection of glucose, lactate, and uric acid8 and later Au(III)14 using a device in which the electrodes had been screen-printed onto the paper fluidic substrate. Nie et al.6 on the other hand used a screenprinted electrochemical sensor in face-to-face contact with a separate paper fluidic element to detect heavy metal ions as well as glucose. Carvalhal et al.15 used gold electrodes in contact with a paper fluidic channel capable of separating uric and ascorbic acid prior to their detection. Electrochemistry is particularly well suited for this type of technology because electrochemical detection is already simple and miniaturizable as demonstrated by the glucose sensing systems currently on the market. The technology for the high volume production of inexpensive screen-printed electrodes (SPEs) is also very well advanced, and many different designs are available commercially. Moreover, electrochemistry in paper appears to work rather well; the fibrous matrix does not impede diffusion to any significant degree. In fact, affixing a thin slab of liquid to the electrode surface provides some immunity from vibration and other convective interferences.6 Electrochemiluminescence (ECL), where a chemiluminescence reaction is initiated and controlled by the application of an electrochemical potential, has grown significantly in importance in analytical chemistry in recent years.16-20 ECL is attractive as a method of detection because it can combine the advantages of luminescence and electrochemical techniques as well as providing Received: September 8, 2010 Accepted: December 2, 2010 Published: January 19, 2011 1300
dx.doi.org/10.1021/ac102392t | Anal. Chem. 2011, 83, 1300–1306
Analytical Chemistry
ARTICLE
Figure 1. Fabrication and operation of a paper-based microfluidic ECL sensor. The paper microfluidics are produced in bulk using a conventional inkjet printer (a). The individual paper fluidic elements are cut to size and the hydrophilic portion filled with a 10 mM Ru(bpy)32þ solution before drying (b). The paper substrate is then aligned and fixed onto the face of the SPE by laminating with transparent plastic (c). A drop of sample is introduced through a small aperture in the plastic at the base of the channel, and when the detection zone is fully wetted, the sensor is placed close to the lens of the camera phone, a potential of 1.25 V is applied, and the resulting emission is captured and analyzed (d).
added selectivity. For example, the timing and spatial location of the luminescent reaction can be tightly controlled,16,17,20 and superior sensitivity may be achieved due to the low background.16,17,20-23 Tris(2,20 -bipyridyl)ruthenium(II), (Ru(bpy)32þ), is perhaps the best known ECL reagent, and the detection of many (so-called coreactant) species has been demonstrated by virtue of their ability to generate light on reaction with the oxidized form of the complex. NADH (nicotinamide adenine dinucleotide) is an example of an ECL coreactant which can be detected21-26 by ECL. NADH is an important biological compound and is found in over 250 biological pathways24 including the breakdown of ethanol to acetaldehyde.21-23,25,26,28 Here we present the first account of ECL-based sensing using paper-based microfluidic sensors. Using inkjet-printed paper fluidic substrates and screen-printed electrodes, we have constructed very low cost, disposable ECL sensors which may be read with a conventional photodetector or a mobile camera phone. Importantly, because ECL is performed in the dark, unlike the case with colorimetric detection it is independent of ambient light. As far as we know, this is the first instance of a mobile camera phone being used as a luminescence detector and certainly the first use of such a device to detect ECL emission.
’ EXPERIMENTAL SECTION Chemicals and Equipment. All solutions were made up in pH 7.5 0.1 M phosphate buffer unless otherwise stated. pH measurements were made using a MEP Instruments, Metrohm 827 pH Lab pH meter and a MEP Instruments Metrohm 6.0228.010 pH electrode. Tris(2,20 -bipyridyl)ruthenium(II) chloride hexahydrate (Ru(bpy)32þ) (99% grade), 2-(dibutylamino)ethanol (99%) (DBAE), and β-nicotinamide adenine dinucleotide (94% grade) in its reduced form (NADH) were obtained from Sigma-Aldrich. NADH solutions were made up fresh daily. Cyclic voltammetry was performed using a CH instruments (Bee Cave, TX) potentiostat (660B or 620C). Low-cost, screen-printed electrodes (SPEs) were manufactured by Zensor R&D (Taiwan). These consisted of a 3 mm diameter working electrode and an arc-shaped auxiliary electrode (both made of graphitic carbon powder) and a Ag/AgCl pellet reference electrode all on a 50 13 mm plastic substrate. ECL experiments were performed using an Eco Chemie, μ-Autolab type II potentiostat and an Electron Tubes Ltd. (model 98285B) photomultiplier tube in a custom-built
light-tight Faraday cage with a high voltage power supply biased at 500 V. The PMT signal was amplified using a custom-made amplifier and read via the auxiliary channel of the potentiostat. Preparation of Paper Microfluidics. The template for printing the paper fluidic pattern was prepared using Adobe illustrator. Figure 1 shows the shape of the pattern used. The channel was 7 mm long and 1.5 mm wide, while the detection zone was 8 mm in diameter. Alkenyl ketene dimer (Precis 900, Hercules Australia Pty Ltd.) was used as the cellulose hydrophobization agent. Alkenyl ketene dimer (liquid AKD) has two hydrocarbon chains of C16-C20 and a CdC double bond in each of its two hydrocarbon chains. Alkenyl ketene dimers are low-cost commodity materials used as sizing agents in the papermaking industry. When used for printing patterned paper fluidics on filter paper, the cost of liquid AKD for each paper fluidic device is around 0.000002 euro.29 Analytical grade n-heptane (Sigma-Aldrich) was used as the solvent to make solutions of the dimer. Whatman filter paper (No. 4) was used as the paper substrate. The electronically generated paper fluidic patterns were printed onto A4-size filter paper with an AKD-heptane solution (2% v/v) using a reconstructed commercial digital inkjet printer (Canon Pixma ip4500). Printing does not leave any visible mark on paper samples, which retained their original flexibility. The printed filter paper sheets were then heated in an oven at 100 °C for 8 min to cure alkenyl ketene dimer onto the cellulose fibers. After the curing process, the printed area sustains an apparent contact angle for water of 110°, whereas the unprinted channel is still wettable, allowing aqueous solution to wick. Fabrication of Paper Microfluidic Sensors. As illustrated in Figure 1, the printed paper microfluidics were cut to size and each loaded with 13 μL of 10 mM Ru(bpy)32þ solution. These were then left to air-dry. Once dry, each paper microfluidic substrate was laminated onto a Zensor screen-printed electrode (SPE) using a GBC Heatseal H212 office laminator and A5 GBC/Ibico document pouches (80 μm thickness). Many sensors could be laminated simultaneously in one pouch if desired. For sample introduction, a small incision is made in the lamination at the bottom of the channel and a drop is placed over it. The sample typically wicks to the detection zone in
