Optical sensing systems for microfluidic devices: A review

Author's personal copy a n a l y t i c a c h i m i c a a c t a 6 0 1 ( 2 0 0 7 ) 141–155

available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/aca

Review

Optical sensing systems for microfluidic devices: A review Bambang Kuswandi a , Nuriman a , Jurriaan Huskens b , Willem Verboom b,∗ a

Chemo & Biosensors Group, School of Pharmacy, University of Jember, Jember 68121, Indonesia Laboratory of Molecular Nanofabrication, MESA+ Research Institute for Nanotechnology, University of Twente, P.O. Box 217, 7500 AE Enschede, The Netherlands

b

a r t i c l e

i n f o

a b s t r a c t

Article history:

This review deals with the application of optical sensing systems for microfluidic devices.

Received 22 June 2007

In the “off-chip approach” macro-scale optical infrastructure is coupled, while the “on-chip

Received in revised form

approach” comprises the integration of micro-optical functions into microfluidic devices.

22 August 2007

The current progress of the use of both optical sensing approaches in microfluidic devices, as

Accepted 23 August 2007

well as its applications is described. In all cases, sensor size and shape profoundly affect the

Published on line 1 September 2007

detection limits, due to analyte transport limitation, not to signal transduction limitation. The micro- or nanoscale sensors are limited to picomolar-order detection for practical time

Keywords:

scales. The review concludes with an assessment of future directions of optical sensing

Microfluidic devices

systems for integrated microfluidic devices.

Lab-on-a-chip

© 2007 Elsevier B.V. All rights reserved.

Optical sensing Optical modes Absorbance Fluorescence Chemiluminescence

Contents 1. 2.

3.

4.

∗

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coupling of optical detection systems: off-chip approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. Absorbance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. Fluorescence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3. Chemiluminescence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Integration of optical functions: on-chip approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Absorbance with integrated detectors, waveguides, and light sources. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Fluorescence with integrated microlenses, waveguides, light sources, and detectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3. Chemiluminescence with integrated detectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4. Refractive index with integrated interferometers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusion and outlook. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Corresponding author. E-mail address: [email protected] (W. Verboom). 0003-2670/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.aca.2007.08.046

142 142 144 145 148 148 150 150 152 153 153 154 154

Author's personal copy 142

1.

a n a l y t i c a c h i m i c a a c t a 6 0 1 ( 2 0 0 7 ) 141–155

Introduction

Microfluidic devices are often described as miniature versions of their macro-scale counterparts. This analogy may be true for some aspects of microfluidic devices, since many phenomena do not simply scale linearly from large to small implementations, as e.g. increased surface area to volume ratio (s/v) and the omnipresence of laminar flow. Modern microfluidics [1] can be traced back to the development of a silicon chip based gas chromatograph at Stanford University [2] and the ink-jet printer at IBM [3] in the late 1970s. Though both these devices were quite remarkable, the concept of the integrated microfluidic device (which often fall under the broad categories of micrototal analysis systems or labs-ona-chip), as it is known today, was not developed until the early 1990s [4]. Since that time, the field has considerably grown and branched off into many different areas, of which a number of excellent specialized reviews and books have been published recently [5–15]. Originally, it was thought that the most significant benefit of these lab-on-a-chip devices would be the analytical improvements associated with the scaling down of the size [4]. Further development revealed other significant advantages including: minimized consumption of reagents, increased automation, and reduced manufacturing costs [16]. The latter of these has been perhaps the most important advancement as the field drifts from the relatively complex silicon and glassbased micromachining, originally developed in the electronics industry, to much simpler techniques and other materials [17–19]. As these manufacturing technologies are further and further advanced (both in terms of the potential complexity of an integrated device and the ease with which a simple prototype can be made) in parallel with analytical needs, the development of future integrated devices will inevitably be less expensive and faster than ever before. In the sensing system, the detection issues will arise when sensing systems are miniaturized. The reduced analysis volumes mean a reduction in detection volumes, decreasing the number of analytes available for detection and making it more difficult to detect them. Thus, the two main factors that affect the choice of the detection method for microfluidic devices are sensitivity and scalability to smaller dimensions. Electrochemical detection (e.g. conductivity, potential) does not fulfill all these conditions, where sensitive portable systems are required. Hence, there is a developing interest to couple and integrate optical components into microfluidic devices. The optical components used in these detectors are mainly lightemitting diodes (LEDs) or laser diodes as light sources, optical fibers, gradient refractive index lenses, and diffractive elements. These are assembled into compact detectors to develop a portable instrumentation based on microfluidic devices. A typical schematic representation of a complete microfluidic device with optical detector is given in Fig. 1. In principle, a wide variety of detection options in macro-scale optical infrastructure can be coupled to microfluidic devices [20,21]. Most reports rely on laser-induced fluorescence (LIF) [22]. Mass spectrometry (MS) also receives much attention to meet e.g. the requirements of biochemical analysis [23]. However, both LIF and MS need sophisticated and expensive instrumenta-

Fig. 1 – Typical schematic representation of a portable microfluidic device with optical detector. From: foodmicro.foodsci.cornell.edu/fmlab/sensors.html.

tion. Commercially available MS systems are not inherently portable and are more costly and less sensitive than LIF. Herein, we review optical sensing systems for microfluidic devices from the beginning of the year 2000 to present (for general microfluidic devices or historical details, readers are referred to the comprehensive set of reviews written by Manz and co-workers [24–27]. The main part deals with optical sensing systems that can be coupled to microfluidic devices. In the other part, we will focus on the integration of micro-optical components into microfluidic devices. Furthermore, the advantages and disadvantages of both approaches (off-chip and on-chip) are discussed.

2. Coupling of optical detection systems: off-chip approach Macro-scale optical detection, especially spectrometric detection is common because there is a wide range of applications, i.e. absorbance, fluorescence and chemiluminescence. In order to couple these macro-scale detection into micron-sized detection areas, the use of pinholes at focus points along the optical path or optical fiber is commonly called as “offchip approach”. The advantages of this approach are very low levels of background signal that can be combined with very sensitive photon detection techniques, such as photomultiplier tubes (PMTs) and charge-coupled devices (CCD) which in turn, result in very low detection limits. However, the reduction in path length within the device could also decrease the sensitivity of the method, in particular for absorbance measurements. Currently, the development of a wide range of intense light-emitting diodes and photodiodes that can be well coupled directly to microfluidic devices to provide a miniaturized technique of on-chip detection can easily be realized. The existing devices in the “off-chip approach” are generally well developed either as homemade detection system or commercial instrument system such as Shimadzu MCE-2010, Hitachi SV1100, and Agilent Bioanalyzer 21000. The analytical

Author's personal copy

Table 1 – Analytical performance of microfluidic systems using the “off-chip” approach Optical detection

Coupling technique Optical fibers, CCD detector set-up Optical fibers, microlens, slits, PMT detector set-up Optical fibers, a ball lens, CCD detector set-up Optical fibers, CCD detector set-up Optical fibers, PMT detector set-up Liquid-core waveguide (LCW) Microscope-based PMT detector set-up Microscope-based PMT detector set-up Optical path length, CCD detector set-up Scanner detector set-up Microscope-based PMT detector set-up Confocal microscope-channel-image set-up Confocal microscope- channel image set-up Optical fibers, pinhole filters, PMT detector set-up Optical channel, CCD camera set-up Spherical lens, filter, path length, PMT detector set-up

LIF LIF LIF LIF LIF LIF Fluorescence (LED-IF) Fluorescence Chemiluminescence Bioluminescence Chemiluminescence Chemiluminescence Chemiluminescence

Epifluorescence microscope-based PMT detector set-up Epifluorescence microscope-based PMT detector set-up Microscope-based PMT detector set-up Epifluorescence microscope-based CCD detector set-up Epifluorescence microscope-based CCD detector set-up Confocal microscope, a photodetector set-up Optical fiber, lens, PMT detector set-up LED fluorimetry with LCW Luminol reaction systems, optical fiber, PMT detector set-up Firefly luciferin–luciferase system, optical fiber, PMT detector set-up Luminol reaction system, optical fiber, PMT detector set-up Luminol reaction system, optical fiber, PMT detector set-up Luminal reaction system, PMT detector set-up

Chemiluminescence Electrochemiluminescence

Peroxyoxalate reaction system, PMT detector set-up ITO electrode, PMT detector set-up

Analyte Thiourea Fluorescein Ca(II) in urine Ca(II) in urine Ammonia Fe(II) Glutathione Riboflavin Flavin DNA Protein Pb(II) pH FITC–epinephrine FITC–dopamine FITC–amino acids Fluorescent dyes, fluorescamine-labelled protein Amino acids in green tea Fluorescein derivatives FITC–amino acid, FITC Protein Fluorescent dyes Fluorescent dyes Fluorescein, rhodamine FITC–arginine Cr(III), Co(II), Cu(II) ATP Catechol, dopamine Dansyl amino acids Cytochrome c, myoglobin, horseradish peroxidase Antioxidant Lincomycin, proline in urine

LOD

Ref.

167 ␮M S/N = 3 0.085 mM 2.68 × 10−5 M 2 ppm 1 ␮M – – ∼mM – 12 ␮g mL−1 – 2–11 1 × 10−7 M – 10−11 M, nM

[28] [29] [30] [31] [32] [33] [36] [37] [38] [39] [40] [41] [42] [43,44] [45,46] [47,48]

– – ∼nM, 7 × 10−10 M 10 ␮M – – – 10 nM ∼␮M ∼␮M ∼␮M ∼␮M 1.2 × 10−7 M, 1.6 × 10−7 M, 1.0 × 10−10 M – 9 ␮M

[54] [55,56] [57] [58] [59] [60] [61,62] [63] [64] [65] [66] [67] [68]

a n a l y t i c a c h i m i c a a c t a 6 0 1 ( 2 0 0 7 ) 141–155

Absorbance Absorbance Absorbance Reflectance Absorbance Absorbance Fluorescence (LIF) LIF LIF Fluorescence LIF Fluorescence Fluorescence LIF LIF LIF

[69] [70,71]

143

Author's personal copy 144

a n a l y t i c a c h i m i c a a c t a 6 0 1 ( 2 0 0 7 ) 141–155

performances of different microfluidic systems involving the “off-chip” approach are given Table 1.

2.1.

Absorbance

UV/vis absorbance detection is the most widely used detection method in common macro-structure sensing systems. However, despite its very wide range of possible applications, there are only a few examples of absorbance-based detection systems for microfluidic devices. The small dimensions of a microchip channel pose a severe problem for a sensitive and reliable absorbance measurement. Incorporating optical fibers into the microchip is a simple approach minimizing the number of required optical components. A simple fiber optics-based UV absorption system was used to monitor the separation of peptides [28]. The separation channel of the microchip was partly filled with a sol–gel-immobilized stationary phase (C4 -modified silica, 5 ␮m particles). UV absorption detection was carried out at the end of the channel without any stationary phase present. The chip was positioned between the ends of two optical fibers facing one another. The top fiber was connected to a deuterium–tungsten light source; the bottom optical fiber collected the transmitted light and guided it into a CCD array detector. The authors claim that the detection limit achieved for thiourea (167 ␮M) could further be improved by noise reduction and by using a second detection channel for reference monitoring. A more complex three-layer chip design comprises two integrated optical fibers, a microlens and a pair of slits for extended optical path length absorbance detection [29]. The slit channels were filled with black ink to absorb any scattered light and to ensure that only the transmitted light is collected by the detection fiber. As light from an optical fiber is highly divergent, both excitation and detection fibers usually need to get very close to yield sufficient irradiance. To overcome this problem, a cylindrical microlens in the PDMS material at the end of the excitation fiber has been created. In this way, the divergence was reduced so much that both fibers could still be operated at a distance of over 500 ␮m. This fact was utilized

by extending the optical absorbance path length. A separation channel (50 ␮m width) has been designed, yielding a Z-shaped detection cell with an optical path length of 500 ␮m, increasing the sensitivity by almost a factor of 10. A detection limit for fluorescein with a signal to noise ratio (S/N) of 3 has been achieved. Another example involves a simple optical fiber and a ball lens for coupling the signal from a microchip for the determination of calcium ions in urine [30]. A reflective mode was employed to increase the sensitivity of the assay (Fig. 2). The calcium ions interact with arsenazo III to form a complex with an adsorption maximum at 668 nm; the detection limit was 0.085 mM. The same approach was applied using arsenazo III immobilized on the surface of polymer beads [31]. The beads were positioned at the detection point of the microfluidic device with a fiber optic assembly for reflectance measurements. The sensor could be regenerated by rinsing with HCl solution. This device has a better detection limit (2.68 × 10−5 M) for calcium ions. A microfluidic device for the analysis of ammonia based on the reaction with indophenol using visible absorbance measurements has been described [32]. The device (with a size of 5 cm × 5 cm and a thickness of 1 mm) was coupled in-plane (x-axis) with an optical source and detector for ease and precise optical alignment for optical fiber guide in microchannels. Reaction rates for the reaction of ammonia with indophenol were remarkably increased using a thermoelectric heater (2 cm × 2 cm). For 2 ppm ammonia (at 308 K), the total reaction time was in the range of 100–150 s. A simple, robust, and automated microfluidic chip-based flow injection analysis (FIA) system with gravity-driven flows and liquid-core waveguide (LCW) spectrometric detection has been developed [33]. Sample loading and injection were performed by linearly moving the array of vials filled alternately with 50 ␮L samples and carrier, allowing the probe inlet to enter the solutions in the vials through the slots sequentially and the sample and carrier solutions to be introduced into the chip driven by gravity. The performance of the system was demonstrated using the complexation of o-phenanthroline with Fe(II). A 20-mm-long Teflon AF 2400 capillary (50 ␮m i.d., 375 ␮m o.d.) was connected

Fig. 2 – Design for absorbance detection incorporating optical fibers. Reprinted with permission from [31].

Author's personal copy a n a l y t i c a c h i m i c a a c t a 6 0 1 ( 2 0 0 7 ) 141–155

145

to the chip to function as a LCW detection with a cell volume of 40 nL and an effective path length of 1.7 cm. Linear absorbance response was obtained in the range of 1.0–100 ␮M Fe(II), with a good reproducibility of 0.6% R.S.D. (n = 18). The sensitivity of the system was much better than that of conventional FIA systems, which typically consume 10,000 times more sample. The highest sampling throughput of 1000 h−1 was obtained by using injection times of 0.08 and 3.4 s for sample and carrier solutions, respectively, with a sample consumption of only 0.6 nL for each cycle.

2.2.

Fluorescence

Fluorescence detection is still the most widely used optical method for microsensing systems, due to its superior selectivity and sensitivity [24,34]. Although a variety of excitation sources are available, laser-induced fluorescence is most easily adapted to the dimensions of microchips. The coherence and low divergence of a laser beam makes it easy to focus on very small detection volumes and to obtain very high irradiation, resulting in one of the lowest detection limits of any detection system [25,35,36]. Lamp-based excitation systems represent a less expensive but more flexible alternative in terms of the choice of the wavelength. Microscope-based detector set-ups using xenon or mercury lamps are applied to a variety of different analytical problems with impressive results. Furthermore, recent advances in laser technology have produced stable light sources that cover a rapidly increasing range of wavelengths from the ultraviolet to the infrared region of the electromagnetic spectrum [36]. They are relatively inexpensive, and can easily be focused onto micron-sized detection areas (Fig. 3). Laser-induced fluorescence has been used for the detection of glutathione (GSH) and the average separation efficiency was about a factor of 5 higher than that obtained with GSH standards using macro-scale device (pinched injection) [36]. LIF with photomultiplier tube detection has been used for the subsecond separation of three flavin metabolites [37]. Due to the native fluorescence of riboflavin, fluorescent labelling was not needed, and flavin mononucleotide and flavin adenine dinucleotide could be detected down to mid-nanomolar concentrations. Qin et al. [38] separated flavin metabolites and developed a very interesting information-rich detector set-up. They used a pulsed nitrogen laser pumping different dye solutions to obtain a tunable laser excitation. Fluorescence emission was guided to a spectrometer with an intensified CCD detector. This set-up potentially facilitates wavelength-resolved detection comparable to diode-array detection in UV/vis absorbance measurements. Preliminary results showed that under static conditions flavin metabolites can be detected in the lower micromolar range. The application of a radial capillary array electrophoresis microplate and scanner for high-throughput DNA analysis has been presented [39]. Detection was accomplished by a laser-excited rotary confocal scanner with four-color detection channels. Loading of 96 samples in parallel was achieved using a pressurized capillary array system. The practical utility and multicolor detection capability were demonstrated by analyzing 96 methylenetetrahydrofolate reductase (MTHFR) alleles in parallel using a non-covalent two-color staining method.

Fig. 3 – Common set-up used for the LIF detection system, including (a) laser; (b) excitation filter; (c) focusing lens; (d) microchip; (e) objective for collimated emission; (f) emission filter; and (g) PMT detector. With kind permission from Springer Science and Business Media [21].

A compact device combining UV filters, a 275 nm dichroic mirror and a UV-transparent microscope objective has been built [40]. Fiber-coupled to a 266 nm frequency-quadrupled Nd:YAG laser, this cube was mounted onto the objective holder of a commercially available fluorescence microscope. With this device, it was possible to detect the separation of proteins with native fluorescence in fused-silica microchips down to a concentration of 12.5 ␮g mL−1 (0.9, 0.5, and 0.5 ␮M for lysozyme, trypsinogen, and chymotrypsinogen). Reinhoudt and co-workers [41] used confocal microscopy to image a channel that has been functionalized with the TAMRA fluorophore for Pb2+ detection. Using confocal microscopy, the acetonitrile-filled channel was first imaged in the absence of analyte. A 0.1 mM acetonitrile solution of the perchlorate salt of Pb2+ was then flowed through, resulting in a 50% increase in fluorescence intensity. Reproducible changes in the fluorescence intensities were then obtained from a subsequent addition of the Pb2+ solution (Fig. 4). Another application of this sensing system involves the sensing of pH [42]. The internal surface of the network has been coated with a monolayer of Rhodamine B dye as a fluorescent sensing molecule. Upon rinsing the microchannels with acidic or basic solutions, it was possible to switch between the fluorescent and nonfluorescent forms reversibly. To widen the scope of optical sensing in microchannels an oregon green dye derivative was immobilized, which functions as a sensing molecule for pH differences in aqueous solutions. The fluorescence intensity directly correlated to the pH of the solution in contact, indicating the possibility of using such a system as a pH sensor in microfluidic devices.

Author's personal copy 146

a n a l y t i c a c h i m i c a a c t a 6 0 1 ( 2 0 0 7 ) 141–155

Fig. 4 – 70 ␮m × 70 ␮m confocal microscopy images of the channel filled with acetonitrile and filled with a 10−4 M solution of Pb2+ in acetonitrile. The image reveals an increase of 50% in the fluorescence intensity: (a) in acetonitrile; (b) in a 10−4 M solution of Pb2+ in acetonitrile. Copyright (2004) American Chemical Society [41].

Incorporating optical fibers into the microchip is an approach used to simplify the detection system by minimizing the number of required optical components. Lin and co-workers [43,44] etched an additional channel into the microchip, into which they introduced an optical fiber coupled to a blue diode-pumped laser. This channel ended 190 ␮m away from the separation channel and enabled excitation without the need for any focusing optical components. Beneath the chip they attached a 400 ␮m pinhole, a holographic notch filter (476 nm), and an interference filter (535 nm). The emission light was then detected directly with a PMT from below the chip without any collimating optics (Fig. 5). Fluorescein isothiocyanate (FITC)labelled epinephrine and dopamine could be detected in a concentration range of 2 × 10−4 –1 × 10−7 M with a linear response. The same group [45] developed a microchip with four parallel separation channels, enabling simultaneous LIF detection of four different analyte solutions. Using a cylindrical lens to form a laser line, a 20 mW solid-state laser (473 nm) was spread along the chip in order to excite all the channels at the same time. To conserve the spatial resolution between the different detection points, a CCD camera was used as a detector. Without the need to reprocess recorded video sequences, they developed a software program that directly showed the appropriate electropherograms for each channel on-line. Four different FITC-labelled amino acids could be separated from their labelling reagent. The initial unsatisfactory reproducibility observed when comparing the four channels was improved through the addition of an internal standard to an average of 7.2% R.S.D. However, the use of a 12-bit CCD camera with a

Fig. 5 – Set-up for the microchip capillary electrophoresis LIF detection system [44].

grayscale range from 0 to 4096 made the linear detection range very narrow. The feasibility of this approach was demonstrated in the rapid screening of chiral selectors [46]. A hand-held CE–LIF device with very impressive abilities has been presented [47]. The device has a modular design, including a separation platform with a multichannel highvoltage power supply, a battery pack, and a control panel equipped with an LCD panel for direct evaluation. The separation platform itself incorporates a microchip (2 cm × 2 cm) with two mirrored separation channels, onto which two 392 nm laser diodes are focused. Fluorescence emission was collimated by a spherical lens and was detected by a PMT after passing a 460 nm long-pass filter. The detection of both channels with one single PMT was achieved by asynchronously pulsing the laser diodes at 10 Hz and using a 50% duty cycle. Software deconvolution separated the data stream into electropherograms corresponding to each channel. Since it was developed for the detection of biological warfare agents, the device was successfully applied to the separation and detection of fluorescamine-labelled ricin and staphylococcal enterotoxin samples. Further improvement in the design of the instrument for the applicability and stability of the system, involved the development of a 16channel high-voltage power supply, a new cartridge-based fluid delivery system, and a redesigned microchip layout (2 cm × 3 cm) [48]. Using this instrument, LIF detection allowed a mid-picomolar (10−11 M) sensitivity for fluorescent dyes and a low nanomolar sensitivity for fluorescamine-labelled proteins. Following LIF detection, lamp-based approaches form the second largest group of optical detection techniques used for microsensing systems. Several examples of lamp-based excitation using epifluorescence microscopes combined with PMT detection have been reported [49–52], including the detection of subsecond chiral separation [53] and the fast separation of amino acids in green tea [54]. Chen and coworkers [55] used this detection method to investigate the potential of flow-through-based microchip electrophoresis for quantitative analysis. Using field-amplified stacking injection a 160-fold increase in sensitivity has been observed [56] in the case of fluorescein derivatives. The highly sensitive instrument developed by Cheng and co-workers [57] achieved detection limits for FITC-labelled amino acids in the low nanomolar range, comparable to results obtained with laser-induced excitation. The application of a two-channel photon counter behind the preamplifier

Author's personal copy a n a l y t i c a c h i m i c a a c t a 6 0 1 ( 2 0 0 7 ) 141–155

of the PMT enabled the detection of very weak signals. The detection limit for FITC was 7 × 10−10 M. The dynamic range of the system, however, was very narrow (1.5 decades) due to saturation effects in the photon counter. Although PMTs are the most frequently used detectors in this field by far, they cannot be used for applications requiring spatial resolution, since they are single-point detectors. In contrast, the two-dimensional chips in chargecoupled devices are sensitive detectors when used for these quantitative imaging purposes. Sodium dodecyl sulfatepolyacrylamide gel electrophoresis (SDS-PAGE) has been used [58] in the isoelectric focusing of proteins in short microchannels. A fluorescent protein marker sample (molecular mass range 20,000–200,000) was separated in less than 30 s in a channel with a length shorter than 2 mm. Using an epifluorescence microscope with mercury lamp excitation, the separation was imaged by a CCD camera. Protein concentrations from 10 to 100 ␮mol L−1 were observed. A CCD camera to observe continuous separation in free-flow electrophoresis (FFE) [59] has been successfully applied to separate two fluorescent dyes in only 75 ms. Short residence times and small sample flows make the system applicable to the fast monitoring of chemical and biochemical production lines. An acoustooptic tunable filter was used to detect multiple fluorescent signals on a fluidic microdevice [60]. A confocal laser-induced fluorescence detection set-up was applied to excite fluorescent dyes in glass microchannels, presenting a streamlined and robust detection system consisting of a narrow-bandwidth AO filter and a single photodetector. The flexibility of the filter was demonstrated by alternating between wavelengths for precise microchannel alignment and sweeping through a range of wavelengths for preliminary spectral characterization of subnanoliter probe volumes of target analytes. In order to demonstrate the multicolor capability of the system, 19-wavelength detection was performed during the separation of a three-dye sample mixture. LEDs are the most effective light sources available and require only low-power driving currents. These advantages, combined with their very compact dimensions, mean that they are perfectly suitable for integration into microfluidic devices. An approach to LED-induced fluorescence detection with an integrated LED and optical fiber (Fig. 6) has been presented [61]. The distance between the diode and the detection area plays a major role in the divergence of light emitted from LEDs. By separating a LED from its epoxy lens, a LED with a flat surface was obtained, which was incorporated into the microchip during fabrication. Although no focusing optics was necessary due to the extreme proximity of the LED, the broad emission spectra of the LEDs used made the introduction of a thin excitation filter necessary. An optical fiber was molded into the PDMS chip at right angles to the microchannel with an optimum distance of 100 ␮m. The emission light was guided through the optical fiber through a band-pass emission filter and was detected by a PMT. Detection limits were significantly improved compared to previous results [62]. For fluorescein and rhodamine dyes, they are in the mid-nanomolar range, but a microscope-based LIF system gave a 20-fold improvement in sensitivity. This is explained by the fact that the entire laser light is easily focused onto the microchannel.

147

Fig. 6 – Procedure used to fabricate the PDMS microchannel chip incorporating the LED and optical fiber (1, buffer reservoir; 2, sample reservoir; 3, sample waste reservoir; 4, buffer waste reservoir). With kind permission from Springer Science and Business Media [61].

The signal-to-noise level of LED fluorimetry using a liquidcore waveguide-based microfluidic capillary electrophoresis system was significantly enhanced using a synchronized dual wavelength modulation approach [63]. A blue and a red LEDs were used as excitation and reference source, respectively. The reference source was used for backgroundnoise compensation in a microfluidic system. A Teflon AF-coated silica capillary served as both the separation channel and LCW for light transfer. The two LEDs were synchronously modulated at the same frequency, but with a 180◦ -phase shift, alternatingly driven by a same constant current source. The LCW transferred the fluorescence emission, as well as the excitation and reference lights that strayed through the optical system to a photomultiplier tube. To test the system, fluorescein isothiocyanate-labelled amino acids were separated and detected. A five-fold improvement in S/N ratio was achieved by dual wavelength modulation, compared with single wavelength modulation. A detec-

Author's personal copy 148

a n a l y t i c a c h i m i c a a c t a 6 0 1 ( 2 0 0 7 ) 141–155

tion limit (S/N = 3) of 10 nM FITC-labelled arginine was obtained.

2.3.

Chemiluminescence

Chemiluminescence (CL) as a detection method for microsensing systems has the advantage of high sensitivity, low detection limits, and simple instrumentation compared with other spectrophotometric techniques, due to the exclusion of an external light source [35]. However, the drawback of this detection technique is the limited number of chemiluminescence reagents. Furthermore, since the chemiluminescence reagent needs to be mixed with the separated analytes before detection, a more complex microchip layout is required. Liu et al. [64] developed several chip layouts and evaluated their performance with respect to different chemiluminescence model systems, including the metal ion-catalyzed luminol-peroxide reaction and the dansyl species-conjugated peroxalate–peroxide reaction. The separation of Cr(III), Co(II), and Cu(II) ions as well as the chiral recognition of dansyl–phenylalanine enantiomers could be accomplished within one minute with low micromolar and even submicromolar detection limits. A comparison between the different chip layouts showed that the pattern with the Y-shaped junction was preferred by the luminol-peroxide CL system, while a V-shaped junction yielded better results with the peroxalate–peroxide system. In the Y-shaped junction, the peroxalate–peroxide could not be effectively transported by electro-osmotic flow. The V-shaped design was later applied for the submicromolar detection of ATP and ATP-conjugated metabolites using a firefly luciferin–luciferase bioluminescence system [65]. While a constant supply of the chemiluminescence reagent is crucial for reproducible detection, it is most frequently delivered by a micropump rather than by electro-osmotic flow. Using such a system, the separation and detection of catechol and dopamine [66] as well as dansyl amino acids [67], gave detection limits in the low micromolar range. An extremely simple CL set-up has been developed by directly monitoring a microchip onto PMT [68] (Fig. 7) without using any optics or filters. Most of the back of the chip was made opaque through the application of black tape; only a rectangular window (2 mm × 3 mm) was uncovered, which served as the detection cell. Using isoelectric focusing, cytochrome c, myoglobin, and horseradish peroxidase could be separated in less than 10 min with detection limits of 1.2 × 10−7 , 1.6 × 10−7 and 1.0 × 10−10 M, respectively. The measuring of antioxidant capacity involves another example of a microfluidic system incorporating chemiluminescence detection [69]. The detection is based on a peroxyoxalate chemiluminescence (PO-CL) assay with 9,10bis-(phenylethynyl)anthracene as the fluorescent probe and hydrogen peroxide as the oxidant. Antioxidant plugs injected into the hydrogen peroxide stream result in inhibition of the CL emission which can be quantified and correlated with the antioxidant capacity. The PO-CL assay is performed in 800-␮m-wide and 800-␮m-deep microchannels on a PDMS microchip. Of the plant-food-based antioxidants tested, ␤carotene was found to be the most efficient hydrogen peroxide

Fig. 7 – Schematic representation of the instrumentation for a simple CL set-up. Reprinted with permission from [68]. Copyright Wiley-VCH Verlag GmbH & Co.

scavenger followed by ␣-tocopherol and quercetin. Although the method is inherently simple and rapid, excellent analytical performance was afforded in terms of sensitivity, dynamic range, and precision, with R.S.D. values typically below 1.5%. Electrochemiluminescence (ECL) detection has been applied for microchip separations [70]. In ECL, an electroactive compound, in this case tris(2,2 -bipyridyl)ruthenium(II) ([Ru(bpy)3 ]2+ ), is oxidized by applying a voltage to additional electrodes in the separation channel, and it subsequently reacts with the analytes with the emission of photons. Thin-film indium tin oxide (ITO) electrodes were added during microchip fabrication [68]. The transparency of the ITO material makes these electrodes superior to the more widely used platinum-based electrodes, because none of the photons generated are absorbed or deflected. The method was applied to the detection of proline and the determination of lincomycin in urine down to a concentration of 9 ␮M [71], while the additional evaluation of the current at the ITO electrodes enabled their simultaneous electrochemical detection due to the catalytic effects of the oxidized [Ru(bpy)3 ]3+ [72]. The system was simplified by immobilizing the ruthenium complex on the ITO electrodes [73]. This did not affect the detection limits, however, since no addition of the complex to the running buffer was required, the [Ru(bpy)3 ]2+ consumption was drastically reduced.

3. Integration of optical functions: on-chip approach The integration of optical components or functions in a microfluidic platform that should be able to perform all chemical functions and detection in a single device, requires increased integration of not only fluidic elements, but also electrical or other types of elements. This method can be classified as “the on-chip approach”. The microelectromechanical systems (MEMS) world has demonstrated the integration of mechanical and electrical functionalities into small structures for diverse applications. Micromachining technologies have

Author's personal copy

Table 2 – Analytical performance of microfluidic systems using the “on-chip” approach Optical detection

Fluorescence Fluorescence Fluorescence Chemiluminescence Chemiluminescence Chemiluminescence Chemiluminescence Refractive index (RI) Refractive index Refractive index Refractive index Refractive index

Integrated optical elements CMOS imager CMOS for photodetector, advanced read-out Waveguide Doped SiO2 waveguide Doped SiO2 waveguide Array of waveguide Liquid dye laser, waveguides, cuvettes, photodiodes Microlens Microlens Multiple-2D planar microlenses SU-8 Waveguide Silicon nitride waveguide, diffraction grating Long-pass filters with dye molecules Waveguide OLEDs OLEDs OLED Blue LED, CdS filter, silicone photodiode Light-emitting silicon avalance diode, planar waveguide, Si photo-detector Vertical cavity surface emitting laser, PIN photodetectors, filters Single photon avalance diode Silicone photodiode, filters Photodiode Photodiode, luminal reaction system Thin-film organic photodiode, luminol reaction system Thin-film organic photodiode, luminol reaction system Four-channel Young interferometer LCORRs Mach–Zehnder interferometer 2D photonic crystal lattice waveguide Photonic crystal multichannel waveguide

Analyte Bromophenol blue, orange G Uric acid in urine Crystal violet Caffeine, paracetamol, ketoprofen, ascorbic acid 24 ␮M Cy5 Alexa fluor 633 dye Xylenol orange Cy5 FITC–albumin Fluorescent dyes Oxygen Antibody–antigen-binding events – Cy3 fluorescent, Cy3 tagged streptavidine Rhodamin B Fluorescein carboxyflorescein Alexa 532 fluorescent dyes Fluorescein Goldnanoparticle-labelled streptavidin

LOD

Ref. [76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91] [92] [93] [94]

IR dye 800 phosphoramide

– 0.33 mg dL−1 0.54 ␮M – S/N = 9:1 1 ␮M (Ab), 10 nM (Fl) – 3.3 nM – – 0.025 mM – – – – 1 ␮M 3 ␮M 1.2 × 10−7 M 3.8 pM (continuous), 13 pM (stopped-flow) 250 nM

23-mer-Cy-5 labelled oligonucleotide Fluorescein DNA Hydrogen peroxide Hydrogen peroxide Hydrogen peroxide Antihuman serum albumine/human serum albumin – Immunoreaction Air, methanol, isopropanol Air, isopropanol, xylene

6 pM 17 nM 0.9 ng mL−1 5 ␮M ∼1 mM 10 ␮M – – – – –

[96] [97] [98] [99] [100] [101] [102] [103] [104] [105] [106]

[95]

a n a l y t i c a c h i m i c a a c t a 6 0 1 ( 2 0 0 7 ) 141–155

Absorbance Absorbance Absorbance Absorbance Absorbance (also fluorescence) Absorbance (Ab), fluorescence (Fl) Absorbance Fluorescence (LIF) Fluorescence Fluorescence Fluorescence Fluorescence Fluorescence Fluorescence Fluorescence Fluorescence Fluorescence Fluorescence Fluorescence

149

Author's personal copy 150

a n a l y t i c a c h i m i c a a c t a 6 0 1 ( 2 0 0 7 ) 141–155

traditionally been silicon-based, since it makes the combination of mechanical and electrical functions in single devices possible. Recent effort has in fact focused on the fabrication of mechanical devices with passive optical functions, such as movable mirror arrays, refractive microlenses, and optical filters [74]. The term micro-optical electromechanical systems (MOEMS) has been considered as a branch of MEMS. It drives the development of micro-optical technologies to replace bulky, large and expensive macroscopic optical systems. In general, silicon and other semiconductor materials also exhibit active optical properties, such as the ability to absorb and emit light. Direct electronic optical sensors like the silicon photodiode, operate by converting absorbed photons directly into electronic carriers which are ultimately detected. Alternatively, many semiconductors and combinations thereof form the basis of laser diodes and light-emitting diodes, a class of devices producing light in different wavelength regions from the blue to the near infrared [75]. From both the materials and technology standpoint, the integration of optical functions (passive or active) into a microchip is very promising. However, the existing devices in the “on-chip approach” are generally still in its infancy. The analytical performances of different microfluidic systems involving the “on-chip” approach are given in Table 2.

3.1. Absorbance with integrated detectors, waveguides, and light sources A commercially available complementary metal oxide semiconductor (CMOS) imager chip bonded together with a microfluidic channel network cast in PDMS has been tested for absorbance measurements of bromophenol blue and orange G [76]. Since the CMOS imager contained an interference filter, fluorescence measurements were also possible. However, the detection limits for the dyes were not specified. The CMOS standard process has also been applied for the fabrication of photodetectors and advanced readout electronics [77]. The latter one contained an 8-bit analog-to-digital converter in order to have a bitstream output that is proportional to the detected light intensity. The CMOS chip was used together with a polystyrene microfluidic device for absorbance measurements using an optical path length of 1 mm. The detection limit was 0.33 mg dL−1 for uric acid in a urine sample. In another approach, a 3D microfluidic device has been inserted in a 5 mm long Teflon tubing (n = 1.29) with an internal diameter of 300 ␮m for absorbance measurements [78]. The Teflon tubing contained the detection cell and consequently reduced the transmission loss and increased the optical path length, since the light was guided across the channel. The detection limit for the dye crystal violet was 0.54 ␮M. Peterson et al. [79] used nitrogen-doped SiO2 (superior UV transmission) for waveguide fabrication in the absorbance mode at 254 nm with a 750 ␮m long detection cell for on-chip electrophoretic separation of caffeine, paracetamol, ketoprofen, and ascorbic acid. A doped SiO2 waveguide has also been used, in which a 1:16 waveguide beam splitter was integrated for parallel probing of an array of 16 microchambers [80]. The chip was designed for simultaneous fluorescence and absorbance measurements. Preliminary fluorescence mea-

Fig. 8 – Photograph of a lab-on-chip device with integrated microfluidic dye laser, optical waveguides, microfluidic network, and photodiodes. Reproduced by permission of the Royal Society of Chemistry from [82].

surements of 24 ␮M Cy5 showed a signal-to-noise ratio of 9:1. Integrated optical waveguide technology has been used for the fabrication of a miniaturized optical detection system for sensing in microchannels [81]. The design allowed the implementation of both absorption and fluorescence measurement methods. An array of optical waveguides was fabricated using spin-on polymer technology on a silicon substrate and monolithically integrated with a microfluidic channel and Vgroove fiber alignment. Total system losses with a water-filled microchannel were experimentally determined to be 10.45 dB. Both the fluorescence and absorption detection capabilities of the fabricated device were demonstrated using Alexa Fluor 633 dye, with detection limits of 10 nM and 1 ␮M, respectively. Individual functional components were assembled to higher integrated devices [82]. Five different components (liquid dye laser, waveguides, fluidic channels, measurement cuvettes, and photodiodes) were monolithically integrated on one substrate (one layer of SU-8 polymer) as well as photodiodes were embedded in the silicon substrate (Fig. 8). The emitted light of the dye laser at 576 nm is directly coupled into five waveguides, that bring the light to five different locations along a fluidic channel for absorbance measurements. The transmitted portion of the light is collected at the other side of the cuvette, again by waveguides, and finally detected by the photodiodes. A test with pure water and pure ammonia buffer gave essentially the same signal readout of the photodiodes (so I/I0 ≈ 1). A xylenol orange dye concentration of 0.06 mM gave I/I0 = 0.7 and a concentration of 0.12 mM gave I/I0 = 0.3. Although the evaluation of the chip is not sufficient to determine the detection limit and the linear range, it proves the successful integration of the individual components.

3.2. Fluorescence with integrated microlenses, waveguides, light sources, and detectors The integration of microlenses and planar waveguides in microfluidic devices is useful for improvement of the detection in sensing systems. For instance, by using a planar waveguide the optical path length can be increased for absorbance measurements, or by focusing the light in the channel to increase

Author's personal copy a n a l y t i c a c h i m i c a a c t a 6 0 1 ( 2 0 0 7 ) 141–155

151

Fig. 9 – Cross-sectional view of (A) 3D view of the experimental set-up of for fluorescence measurements. (B) Close view on the portable devices with two different microchannels fabricated in PDMS. Reproduced by permission of the Royal Society of Chemistry from [84].

the excitation power for fluorescence measurements. Roulet et al. [83] fabricated microlenses directly into a glass chip for the collection of fluorescence light, by melting islands of photoresist into a hemispherical shape. In order to have a compact system, a pinhole, an interference filter, and a photodetector were placed in close proximity to the microlens. The detection limit for Cy5 excited with a 10 mW He–Ne laser was 3.3 nM. In another approach, a microfluidic device in PDMS contained an insertion channel to accommodate an optical fiber for fluorescence excitation (Fig. 9) [84]. The insertion channel was curved in the form of a lens at the interface towards the fluidic channel to focus the excitation light. The detection limit for fluorescence measurements of FITC-labelled albumin was improved three times. Multiple 2D planar microlenses have been used to focus the light from a LED into a microfluidic channel [85]. This design enabled a reduction in the spot size and a seven-fold increase of the fluorescence signal. Chang-Yen and Gale [86] fabricated a SU-8 waveguide for evanescent wave excitation of fluorophores for oxygen detection in microfluidic channels. A single mask step procedure was used for the integration of planar waveguide and fiber-coupling structures. The detection relies on oxygen quenching of fluorophores attached to the surface of the waveguide. A multilayer spin-coating process for interpolyelectrolyte formation was used for the immobilization of the dye to the waveguide. Oxygen was detected in the range of 0.025–0.70 mM. Monitoring rabbit immunoglobulin G (IgG) antibody– antigen-binding events has been performed using 150 nm thick silicon nitride waveguides for evanescent field-based fluorescence excitation (Fig. 10) [87]. The antibodies sample flow contained Cy5-labelled anti-rabbit IgG as the antigens. A diffraction grating fabricated in the waveguide layer was used both for in-coupling of the excitation light and for outcoupling of the fluorescence signal. The microfluidic channel network was used for continuous replenishment of the sample to the binding site, thereby increasing the mass transport to the surface. Vertical confinement of the sample stream to the region close to the waveguide surface ensured low reagent

consumption. High quality monolithically integrated optical long-pass filters are used in disposable diagnostic microchips [88]. The filters were prepared by incorporating dye molecules directly into the microfluidic chip, providing a fully integrated system that removes the usual need for discrete optical filters.

Fig. 10 – Schematic representation of the sensor and microfluidic set-up for evanescence measurements. (A) The 3D PDMS microchip is attached to a waveguide with an immobilized zone of antibodies. For detection, laser light is coupled into the waveguide through a corrugation grating. The evanescent part of the guided wave can be employed for direct sensing of binding events on the sensor surface. (B) Detailed cross-sectional view of the microchip/waveguide interface. Copyright (2002) American Chemical Society [87].

Author's personal copy 152

a n a l y t i c a c h i m i c a a c t a 6 0 1 ( 2 0 0 7 ) 141–155

Using 1 mm layers of PDMS doped with 1200 ␮g mL−1 Sudan II, resulted in less than 0.01% transmittance below 500 nm (OD 4), >80% above 570 nm, and negligible autofluorescence. The filters are robust in use, showing only slight degradation after extended illumination and negligible dye leaching after prolonged exposure to aqueous solutions. Soda lime glass plates have been applied as a material for monolithic microsystems, containing both microchannels and micro-optics [89]. The channel optical waveguides were prepared using an ion exchange technique, while the microfluidic device was fabricated using standard photolithography and wet etching methods. The waveguiding optics was used to excite the fluorescence in the microchannels. The application of the device was demonstrated for the separation and detection of a mixture of Cy3 fluorescent dye and Cy3 tagged streptavidine. Organic light-emitting diodes (OLEDs) have been successfully integrated as lightweight flat panel light source. The big advantage of OLEDs compared to LEDs is their flat film-like shape. This makes it easy to incorporate them into microfluidic devices and to bring them into close proximity of the channel. However, their fairly broad emission spectra require the additional application of excitation filters. OLEDs placed on the rear side of a glass substrate and a channel cast in PDMS on the front side [90] showed the emission peak centered at 520 nm and had a relatively wide peak of 70 nm. However, only visual inspection of the excitation of the fluorescence dye Rhodamin B was performed. Edel et al. [91] used a polyfluorene-based OLEDs fabricated on a separate glass substrate, which subsequently was placed in close proximity to a glass chip. The emission peak was centered at 488 nm and could be operated with only 3.7 V. The dyes fluorescein and carboxyfluorescein were separated and detected at detection limits as low as 1 ␮M. OLEDs with a 0.3 mmthick interference filter combined with a 400 ␮m pinhole to achieve effective excitation with the appropriate wavelength have also been used [92]. Detection of the emission light was accomplished by an optical fiber coupled to a PMT, giving a detection limit of 3 ␮M for the Alexa 532 fluorescent dye. However, the sensitivity using OLEDs is several orders of magnitude less than that of common LIF detection systems. The current drawbacks of OLEDs are due to low irradiance and light purity. Chediak et al. [93] used a blue LED, a CdS optical filter, and a silicone photodiode integrated in the same microchip. The optical chip was used for detection in a microfluidic device that was placed in close proximity in order to get a very integrated compact system. An advantage of this configuration is that the microfluidics part is disposable, while the optic part can be reused. Of 0.04 ␮m diameter carboxylatemodified microspheres, the fluorescence signal was measured with a lowest detectable concentration of 1.2 × 10−7 M. A light-emitting silicon avalance diode in an intriguing device that consists of a single-mode planar waveguide, a Si photodetector, and a microfluidic channel cast in PDMS has been presented [94]. The configuration of the waveguide was encompassed a 90◦ bend towards the microchip for coupling to the LED and the photodetector. The device was tested with the biotin–streptavidin-binding assay giving detection limits of 3.8 and 13 pM for the goldnanoparticle-labelled strepta-

vidin concentration for continuous and stopped–flow assays, respectively. Thrush et al. [95] used vertical cavity surface emitting lasers for near-infrared fluorescence detection of fluorophores spun onto a poly(methyl methacrylate) (PMMA) substrate. Vertical-cavity surface-emitting lasers for 773 nm excitation, PIN photodetectors, and optical emission filters were integrated on a GaAs substrate. These optoelectronic components were optically coupled to a glass microfluidic channel (100 ␮m width and 45 ␮m depth) using a discrete microlens to form a complete sensor; the detection limit was 250 nM for the IR dye 800 phosphoramidite. Using similar hybrid assembled devices (the bottom chip was dedicated to photodetection, while a separate microfluidic chip was placed on top) a single photon avalance diode has been fabricated in a glass CE device [96]. The diode was associated with a monolithically integrated active quenching circuit, which together with Peltier cooling of the diode chip to −15 ◦ C enables efficient counting of photons. The detection limit for a 23-mer Cy-5 labelled oligonucleotide solution was estimated to be 6 pM. Another design comprises amorphous silicon photodiodes with filters on top of a glass microchip using lowtemperature thin film deposition techniques [97]. This design allowed making an aperture in the middle of the diode, from where the excitation light was transmitted onto the microfluidic device, thereby achieving a 180◦ separation between the excitation and emission axis. The detection limits for fluorescein were 17 nM and 680 pM, for the hybrid-integrated system and a confocal microscopy set-up, respectively.

3.3.

Chemiluminescence with integrated detectors

Photodiodes have been fabricated in the same microchip as the bottom of the microfluidic channels via a monolithic approach [98]. In order to lower the detection limit, the photodetector sensitivity was improved by fabricating diodes with a shallow junction to reduce the recombination of photocarriers and thereby to increase the quantum efficiency. The applicability of the device was shown for PCR amplification and capillary electrophoresis separation of a 100 bp DNA ladder labelled with the YOYO-1 intercalating dye; the detection limit was 0.9 ng mL−1 . Using a similar approach, a photodiode fabricated in the bottom of the channels etched in a silicone microchip has been reported [99]. All fluidic and electrical connections were placed on the backside of the wafer to facilitate easy sealing of the channels by anodic bonding of a Pyrex wafer. Since the devices were designed for chemiluminescent determination of glucose, the channel network contained an enzyme reactor, a mixer, and a detection region. The emission light from the chemiluminescent reaction of hydrogen peroxide and luminol was determined on chip, with a detection limit for hydrogen peroxide of 5 ␮M. Thin-film organic photodiodes have been applied as integrated optical detectors for microscale chemiluminescence [100]. The copper phthalocyanine–fullerene (CuPc–C60 ) small molecule photodiodes have an external quantum efficiency of ∼30% at 600–700 nm, an active area of 2 mm × 8 mm and a total thickness of ∼2 mm. Simple detector fabrication, based on layer-by-layer vacuum deposition, allows facile integration with planar chip-based systems. To demonstrate the efficacy of the approach, CuPc–C60 photodiodes were used to monitor

Author's personal copy a n a l y t i c a c h i m i c a a c t a 6 0 1 ( 2 0 0 7 ) 141–155

a peroxyoxalate based chemiluminescence reaction within a microfluidic device. Optimal results were obtained for reagent flow rates of 25 ␮L min−1 , yielding a CL signal of 8.8 nA within 11 min. The reproducibility was excellent with typical relative standard deviations (R.S.D.s) below 1.5%. Preliminary quantification of hydrogen peroxide yielded a detection limit of ∼1 mM and linearity over at least three decades. In another example of a thin-film organic photodiode, the active layer comprised of a 1:1 blend by weight of the conjugated polymer poly(3-hexylthiophene) [P3HT] and [6,6]-phenyl-C61 -butyric acid-methyl ester, a soluble derivative of C60 [101]. The device has an active area of 1 mm × 1 mm, and a broadband response from 350 to 700 nm, with an external quantum efficiency of more than 50% between 450 and 550 nm. The photodiodes have a simple layered structure that permits facile integration with planar chip-based systems. Quantification of hydrogen peroxide indicated an excellent linearity and yielded a detection limit of 10 ␮M.

3.4.

Refractive index with integrated interferometers

An optical sensing hybrid system has been obtained by bonding a microfluidic system to an integrated optical (IO) four-channel Young interferometer (YI) chip [102]. The microfluidic system implemented into a glass plate consists of four microchannels with cross-sectional dimensions of 200 ␮m × 15 ␮m. The microfluidic system was structured in such a way that after bonding to the IO chip, each microchannel addresses one sensing window in the four-channel YI sensor. Experimental tests showed that the implementation of the microfluidics reduces the response time of the sensor from 100 s, as achieved with a bulky cuvette, to 4 s. The successful monitoring of the anti-human serum albumine/human serum albumine immunoreaction demonstrates the feasibility of this microfluidic sensing system for immunosensing applications. The microfluidic sensing system shows an average phase resolution of 7 × 10−5 × 2 for different pairs of channels, which at the given interaction length of 4 mm corresponds to a refractive index resolution of 6 × 10−8 , being equivalent to a protein mass coverage resolution of 20 fg mm−2 . A miniaturized and multiplexed, on-capillary, refractive index (RI) detector using liquid core optical ring resonators (LCORRs) for capillary electrophoresis devices has been developed by Zhu et al. [103]. The LCORR employs a glass capillary with a diameter of ∼100 ␮m and a wall thickness of a few micrometers. The circular cross-section of the capillary forms a ring resonator along which the light circulates in the form of the whispering gallery modes (WGMs). The WGM has an evanescent field extending into the capillary core and responds to the RI change due to the analyte conducted in the capillary, thus permitting a label-free measurement. This LCORR architecture achieves dual use of the capillary as a sensor head and a CE fluidic channel, allowing for integrated, multiplexed, and non-invasive on-capillary detection at any location along the capillary. The LCORR’s label-free sensing mechanism accurately deduced the analyte concentration in real time at a given point on the capillary. A sensitivity of 20 nm RIU−1 (refractive index units) was obtained, leading to an RI detection limit of 10−6 RIU.

153

A Mach–Zehnder interferometer (MZI) in the integrated optics allows a high-detection sensitivity using optical transduction techniques in microfluidic devices [104]. The evanescent wave of an optical waveguide interacts with an adjacent layer, which is the basis of the recognition of e.g. biomolecules. In this design, the MZI sensor exhibiting single-mode behavior was tested for the direct detection of immunoreactions. The total phase change was about 0.9 due to the specific binding of streptavidin with a concentration of 1 ␮g mL−1 on the biotinylated sensor surface. A 2D photonic crystal lattice, with holes arranged in a triangular pattern, has been used to improve the detection limit since the resonant peaks are narrower [105]. The sensor gave good responses for air, methanol, and isopropanol. This is due to a refractive index shift which results in different transmission spectra of these compounds. An alternative approach involves the use of a photonic crystal multichannel waveguide [106]. The shift in relative transmitted optical power is detected when one arm is exposed to a chemical solution. In this way, the optical response of air, isopropanol, and xylene has been measured.

4.

Conclusion and outlook

The field of microfluidic devices has very strong roots in many disciplines, and the phenomena that are encountered touch many aspects of physics, chemistry, and biology. The coupling and integration of an optical sensing system in microfluidic devices has successfully been applied in the past few years, as demonstrated in this review. This is due to two fundamental reasons: first, we have now entered the domain where the sample size is in the order of magnitude of the device features. Second, based on micro- or nanofabrication the optical functions can now be incorporated via a monolithic approach inside the chip. What about the future? A few phenomena and forces have not yet been mentioned. The integration of optical sensing with nanoscience, however, involves a fundamental challenge. As optical sensing systems scale to the micronanodomain, sensitivity and performance are compromised, because there are simply too few sensing active sites to produce a sensitive optical signal. This may be addressed by exploring strategies to increase the sensor response by replacing the linear, single photon response of the present sensors with extremely nonlinear optical responses. To achieve this objective, a multi-pronged approach that combines photochemistry, materials science, and photonics is required in which the light emitting molecular centers of conventional optical sensors are substituted with a variety of imaging technologies including optical wave guides and fibers. Finally, due to the availability of new nanotechnological tools over the last two decades, it has been possible to integrate micro-optical functions, such as lenses, waveguides, light sources, and detectors, in microfluidic platforms for the detection of analytes in the nanodomain owing to advances in optical spectroscopy and microscopy. The new area of integrated microfluidics applying these tools for micro/nanosensing has already borne.

Author's personal copy 154

a n a l y t i c a c h i m i c a a c t a 6 0 1 ( 2 0 0 7 ) 141–155

Acknowledgment The authors gratefully thank the KNAW (The Royal Netherlands Academy of Arts and Sciences) for supporting this work via the SPIN Mobility Program 2006–2007.

references

[1] P. Gravesen, J. Branebjerg, O.S. Jensen, J. Micromech. Microeng. 3 (1993) 168. [2] S.C. Terry, J.H. Jerman, J.B. Angell, IEEE Trans. Electron. Devices ED-26 (1979) 1880. [3] K.E. Petersen, IEEE Trans. Electron. Devices ED-26 (1979) 1918. [4] A. Manz, N. Graber, H.M. Widmer, Sens. Actuators B 1 (1990) 244. [5] P.S. Dittrich, A. Manz, Nat. Rev. Drug Discuss. 5 (2006) 210. [6] D. Mijatovic, J.C.T. Eijkel, A. van den Berg, Lab Chip 5 (2005) 492. [7] T.M. Squires, S.R. Quake, Rev. Mod. Phys. 77 (2005) 977. [8] H.A. Stone, A.D. Strook, A. Ajdari, Annu. Rev. Fluid Mech. 36 (2004) 381. [9] C.H. Ahn, J.W. Choi, G. Beaucage, J.H. Nevin, J.B. Lee, A. Puntambek, J.Y. Lee, Proc. IEEE 92 (2004) 154. [10] P.S. Dittrich, A. Manz, Anal. Bioanal. Chem. 382 (2005) 1771. [11] H. Andersson, A. van den Berg, Lab-on-Chips for Cellomics, Kluwer Academic Publishers, Dordrecht, 2004. [12] M. Kock, A. Evans, A. Brunnschweiler, Microfluidic Technology and Applications, Research Studies Press, Hertfordshire, UK, 2000. [13] M. Brivio, W. Verboom, D.N. Reinhoudt, Lab Chip 6 (2006) 329. [14] P. Watts, S.J. Haswell, Curr. Opin. Chem. Biol. 7 (2003) 380. [15] P. Watts, S.J. Haswell, Chem. Soc. Rev. 34 (2005) 235. [16] H. Becker, L.E. Locascio, Talanta 56 (2002) 267. [17] J.M.K. Ng, I. Gitlin, A.D. Stroock, G.M. Whitesides, Electrophoresis 23 (2002) 3461. [18] J.C. McDonald, D.C. Duffy, J.R. Anderson, D.T. Chiu, H. Wu, O.J.A. Schueller, G.M. Whitesides, Electrophoresis 21 (2000) 27. [19] F.-U. Gast, H. Fiehn, Lab Chip 3 (2003) 6N. [20] P.J. Viskari, J.P. Landers, Electrophoresis 27 (2006) 1797. ¨ [21] S. Gotz, U. Karst, Anal. Bioanal. Chem. 387 (2007) 183. [22] J. Wang, Talanta 56 (2002) 223. [23] M. Brivio, N.R. Tas, M.H. Goedbloed, H.J.G.E. Gardeniers, W. Verboom, A. van den Berg, D.N. Reinhoudt, Lab Chip 5 (2005) 378. [24] D.R. Reyes, D. Iossifidis, P.A. Auroux, A. Manz, Anal. Chem. 74 (2002) 2623. [25] P.A. Auroux, D. Iossifidis, D.R. Reyes, A. Manz, Anal. Chem. 74 (2002) 2637. [26] T. Vilkner, D. Janasek, A. Manz, Anal. Chem. 76 (2004) 3373. [27] P.S. Dittrich, K. Kaoru Tachikawa, A. Manz, Anal. Chem. 78 (2006) 3887. [28] R. Jinda, S.M. Cramer, J. Chromatogr. A 1044 (2004) 277. [29] K.W. Ro, K. Lim, B.C. Shim, J.H. Hahn, Anal. Chem. 77 (2005) 5160. [30] P. Caglar, S.A. Tuncel, N. Malcik, J.P. Landers, J.P. Ferrance, Anal. Bioanal. Chem. 386 (2006) 1303. [31] N. Malcik, J.P. Ferrance, J.P. Landers, P. Caglar, Sens. Actuators B 107 (2005) 24. [32] J.-S. Park, K.-B. Park, K.-S. Shin, H.-D. Park, M.-C. Kim, J.-R. Kim, S.-J. Park, Y.-H. Song, Sens. Actuators B 117 (2006) 516. [33] W.B. Du, Q. Fang, Q.H. He, Z.L. Fang, Anal. Chem. 77 (2005) 1330.

[34] K. Hata, Y. Kichise, T. Kaneta, T. Imasaka, Anal. Chem. 75 (2003) 1765. [35] S.-H. Lee, S.I. Cho, C.-S. Lee, B.-G. Kim, Y.-K. Kim, Sens. Actuators B 110 (2005) 164. [36] J. Gao, X.-F. Yin, Z.-L. Fang, Lab Chip 4 (2004) 47. [37] B.-F Liu, H. Hisamoto, S. Terabe, J. Chromatogr. A 1021 (2003) 201. [38] J. Qin, Y. Fung, D. Zhu, B.-C. Lin, J. Chromatogr. A 1027 (2004) 223. [39] Y. Shi, P.C. Simpson, J.R. Scherer, D. Wexler, C. Skibola, M.T. Smith, R.A. Mathies, Anal. Chem. 71 (1999) 5354. [40] P. Schulze, M. Ludwig, F. Kohler, D. Belder, Anal. Chem. 77 (2005) 1325. [41] L. Basabe-Desmonts, J. Beld, R.S. Zimmerman, J. Hernando, ´ N.F. van Hulst, A. van den Berg, D.N. P. Mela, M.F.G. Parajo, Reinhoudt, M. Crego-Calama, J. Am. Chem. Soc. 126 (2004) 7293. [42] P. Mela, S. Onclin, M.H. Goedbloed, S. Levi, M.F. ´ N.F. van Hulst, B.J. Ravoo, D.N. Reinhoudt, A. Garc´ıa-Parajo, van den Berg, Lab Chip 5 (2005) 163. [43] H.-F. Li, J.-M. Lin, R.–G. Su, K. Uchiyama, T. Hobo, Electrophoresis 25 (2004) 1907. [44] H.-F. Li, Z.-W. Cai, J.-M. Lin, Anal. Chim. Acta 565 (2006) 183. [45] Z. Shen, X. Liu, Z. Long, D. Liu, N. Ye, J. Qin, Z. Dai, B. Lin, Electrophoresis 27 (2006) 1084. [46] Y. Gao, Z. Shen, H. Wang, Z. Dai, B. Lin, Electrophoresis 26 (2005) 4774. [47] J. Fruetel, R.F. Renzi, V.A. VanderNoot, J. Stamps, B.A. Horn, J.A.A. West, S. Ferko, R. Crocker, C.G. Bailey, D. Arnold, B. Wiedenman, W.-Y. Choi, D. Yee, I. Shokair, E. Hasselbrink, P. Paul, D. Rakestraw, D. Padgen, Electrophoresis 26 (2005) 1144. [48] R.F. Renzi, J. Stamps, B.A. Horn, S. Ferko, V.A. VanderNoot, J.A.A. West, R. Crocker, B. Wiedenman, D. Yee, J.A. Fruetel, Anal. Chem. 77 (2005) 435. [49] Y. Mourzina, A. Steffen, D. Kalyagin, R. Carius, A. ¨ Offenhausser, Electrophoresis 26 (2005) 1849. [50] S.J.O. Varjo, M. Ludwig, D. Belder, M.-L. Riekkola, Electrophoresis 25 (2004) 1901. [51] D.A. Wicks, P.C.H. Li, Anal. Chim. Acta 507 (2004) 107. [52] D.E. Starkey, Y. Abdelaziez, C.H. Ahn, J. Tu, L. Anderson, K.R. Wehmeyer, N.J. Izzo, A.N. Carr, K.G. Peters, J.J. Bao, H.B. Halsall, W.R. Heineman, Anal. Biochem. 316 (2003) 181. [53] N. Piehl, M. Ludwig, D. Belder, Electrophoresis 25 (2004) 3848. [54] M. Kato, Y. Gyoten, K. Sakai-Kato, T. Toyo’oka, J. Chromatogr. A 1013 (2003) 183. [55] C.-C. Lin, C.-C. Chen, C.-E. Lin, S.-H. Chen, J. Chromatogr. A 1051 (2004) 69. [56] M. Gong, K.R. Wehmeyer, P.A. Limbach, F. Arias, W.R. Heineman, Anal. Chem. 78 (2006) 3730. [57] Q. Yan, R.-S. Chen, J.-K. Cheng, Anal. Chim. Acta 555 (2006) 246. [58] J. Han, A.K. Singh, J. Chromatogr. A 1049 (2004) 205. [59] C.-X. Zhang, A. Manz, Anal. Chem. 75 (2003) 5759. [60] J.M. Karlinsey, J.P. Landers, Anal. Chem. 78 (2006) 5590. [61] K. Miyaki, Y. Guo, T. Shimosaka, T. Nakagama, H. Nakajima, K. Uchiyama, Anal. Bioanal. Chem. 382 (2005) 810. [62] Y. Guo, K. Uchiyama, T. Nakagama, T. Shimosaka, T. Hobo, Electrophoresis 26 (2005) 1843. [63] T. Zhang, Q. Fang, S.-L. Wang, L.-F. Qin, P. Wang, Z.-Y. Wu, Z.-L. Fang, Talanta 15 (2005) 19. [64] B.-F. Liu, M. Ozaki, Y. Utsumi, T. Hattori, S. Terabe, Anal. Chem. 75 (2003) 36. [65] B.-F. Liu, M. Ozaki, H. Hisamoto, Q. Luo, Y. Utsumi, T. Hattori, S. Terabe, Anal. Chem. 77 (2005) 573. [66] R. Su, J.-M. Lin, F. Qu, Z. Chen, Y. Gao, M. Yamada, Anal. Chim. Acta 508 (2004) 11.

Author's personal copy a n a l y t i c a c h i m i c a a c t a 6 0 1 ( 2 0 0 7 ) 141–155

[67] R. Su, J.-M. Lin, K. Uchiyama, M. Yamada, Talanta 64 (2004) 1024. [68] X. Huang, J. Ren, Electrophoresis 26 (2005) 3595. [69] M. Amatatongchai, O. Hofmann, D. Nacapricha, O. Chailapakul, A.J. deMello, Anal. Bioanal. Chem. 387 (2007) 277. [70] H. Qiu, J. Yan, X. Sun, J. Liu, W. Cao, X. Yang, E. Wang, Anal. Chem. 75 (2003) 5435. [71] X. Zhao, T. You, H. Qiu, J. Yan, X. Yang, E. Wang, J. Chromatogr. B 810 (2004) 137. [72] H. Qiu, X.-B. Yin, J. Yan, X. Zhao, X. Yang, E. Wang, Electrophoresis 26 (2005) 687. [73] Y. Du, H. Wei, J. Kang, J. Yan, X.-B. Yin, X. Yang, E. Wang, Anal. Chem. 77 (2005) 7993. [74] G.T.A. Kovacs, Micromachined Transducers Sourcebook, McGraw-Hill, New York, 1998, Chapter 4, pp. 357–528. [75] J. Seo, L.P. Lee, in: Y. Baba, S. Shoji, A. van den Berg (Eds.), Micro Total Analysis Systems 2002: Proceedings of the ␮TAS 2002 Symposium, Nara, Japan, vol. 1, November 3–7, Kluwer Academic Publishers, Dordrecht, 2002, pp. 284–286. [76] M.L. Adam, M. Enzelberger, S. Quake, A. Scherer, Sens. Actuators A 104 (2003) 25. [77] G. Minas, J. Martins, J. Ribeiro, R. Wolfenbuttel, J. Correia, Sens. Actuators A 110 (2004) 33. [78] M.P. Duggan, T. McCreedy, J.W. Aylott, Analyst 128 (2003) 1336. [79] N.J. Petersen, K.B. Mogensen, J.P. Kutter, Electrophoresis 23 (2002) 3528. [80] J.M. Ruano, A. Glidle, A. Cleary, A. Walmsley, J.S. Aitchison, J.M. Cooper, Biosens. Bioelectron. 18 (2003) 175. [81] L. Malic, A.G. Kirk, Sens. Actuators A 135 (2007) 515. [82] S. Balslev, A.M. Jorgensen, B. Bilenberg, K.B. Mogensen, D. Snakenborg, O. Geschke, J.P. Kutter, A. Kristensen, Lab Chip 6 (2006) 213. [83] J.-C. Roulet, R. Volke, H.P. Herzig, E. Verpoorte, N.F. De Rooij, R. Dandliker, Anal. Chem. 74 (2002) 3400. [84] S. Camou, H. Fujita, T. Fujii, Lab Chip 3 (2003) 40. [85] J. Seo, L. Lee, Sens. Actuators B 99 (2004) 615. [86] D.A. Chang-Yen, B.K. Gale, Lab Chip 3 (2003) 297. [87] O. Hofmann, G. Voirin, P. Niedermann, A. Manz, Anal. Chem. 74 (2002) 5243.

View publication stats

155

[88] O. Hofmann, X. Wang, A. Cornwell, S. Beecher, A. Raja, D.D.C. Bradley, A.J. deMello, J.C. deMello, Lab Chip 6 (2006) 981. [89] R. Mazurczyk, J. Vieillard, A. Bouchard, B. Hannes, S. Krawcyk, Sens. Actuators B 118 (2006) 11. [90] S. Camou, M. Kitamura, J.-P. Gouy, H. Fujita, Y. Arakawa, T. Fujii, Proc. SPIE 4833 (2002) 1. [91] J.B. Edel, N.P. Beard, O. Hotmann, J. deMello, D.D.C. Bradley, A.J. deMello, Lab Chip 4 (2004) 136. [92] B. Yao, G. Luo, L. Wang, Y. Gao, G. Lei, K. Ren, L. Chen, Y. Wang, Y. Hu, Y. Qiu, Lab Chip 5 (2005) 1041. [93] J. Chediak, Z. Luo, J. Seo, N. Cheung, P.L. Lee, T.D. Sands, Sens. Actuators A 111 (2004) 1. [94] K. Misiakos, S.E. Kakabakos, P.S. Petrou, H.H. Ruf, Anal. Chem. 76 (2004) 1366. [95] E. Thrush, O. Levi, W. Ha, K. Wang, S. Smith, J. Harris, J. Chromatogr. A 1013 (2003) 103. [96] I. Rech, S. Cova, A. Restelli, M. Ghioni, M. Chiari, M. Cretich, Electrophoresis 27 (2006) 3797. [97] T. Kamei, B.M. Paegel, J.R. Scherer, A.M. Skelley, R.A. Street, R.A. Mathies, Anal. Chem. 75 (2003) 5300. [98] V. Namasivayam, R. Lin, B. Johnson, S. Brahmasandra, Z. Razzacki, D.T. Burke, M.A. Burn, J. Micromech. Miroeng. 14 (2004) 81. [99] A.M. Jorgensen, K.B. Mogensen, J.P. Kutter, O. Geschke, Sens. Actuators B 90 (2003) 15. [100] O. Hofmann, P. Miller, P. Sullivan, S.T. Jones, J.C. deMello, D.D.C. Bradley, A.J. deMello, Sens. Actuators B 106 (2005) 878. [101] X. Wang, O. Hofmann, A. Cornwell, R. Das, E.M. Barrett, A.J. deMello, J.C. deMello, D.D.C. Bradley, Lab Chip 7 (2007) 58. [102] A. Ymeti, J.S. Kanger, J. Greve, G.A.J. Besselink, P.V. Lambeck, R. Wijn, R.G. Heideman, Biosens. Bioelectron. 20 (2005) 1417. [103] H. Zhu, I.M. White, J.D. Suter, M. Zourob, X. Fan, Anal. Chem. 79 (2007) 930. [104] J. Hong, J.S. Choi, G. Han, J.K. Kang, C.-M. Kim, T.S. Kim, D.S. Yoon, Anal. Chim. Acta 573 (2006) 97. [105] M. Loncar, A. Scherer, Y. Qiu, Appl. Phys. Lett. 82 (2003) 4648. [106] J. Topol’ancik, P. Bhattacharya, J. Sabarinathan, P.-C. Yu, Appl. Phys. Lett. 82 (2003) 1143.