A microfluidic device using a green organic light emitting diode as an integrated excitation source

Abstract

A simply fabricated microfluidic device using a green organic light emitting diode (OLED) and thin film interference filter as integrated excitation source is presented and applied to fluorescence detection of proteins. A layer-by-layer compact system consisting of glass/PDMS microchip, pinhole, excitation filter and OLED is designed and equipped with a coaxial optical fiber and for fluorescence detection a 300 µm thick excitation filter is employed for eliminating nearly 80% of the unwanted light emitted by OLEDs which has overlaped with the fluorescence spectrum of the dyes. The distance between OLED illuminant and microchannels is limited to ∼1 mm for sensitive detection. The achieved fluorescence signal of 300 µM Rhodamine 6G is about 13 times as high as that without the excitation filter and 3.5 times the result of a perpendicular detection structure. This system has been used for fluorescence detection of Rhodamine 6G, Alexa 532 and BSA conjugates in 4% linear polyacrymide (LPA) buffer (in 1 × TBE, pH 8.3) and 1.4 fmol and 35 fmol mass detection limits at 0.7 nl injection volume for Alexa and Rhodamine dye have been obtained, respectively.

---

Introduction

Although more and more measurement schemes besides laser induced fluorescence (LIF),¹ including electrochemistry (EC),² chemilluminescence (CL)³/electrical chemilluminescence (ECL),⁴ mass spectrumetry (MS),⁵ and nuclear magnetic resonance (NMR),⁶ have been developed for microfluidic or “lab-on-a-chip” (LOC) systems in the past decade, the performance of laser induced fluorescence detectors is still enormously important especially in the area of life science research. As opposed to the original, large 488 nm argon ion lasers, small, low cost laser diodes are now commonly being used as the excitation source as first reported by Harrison and co-workers¹ in order to produce a more compact overall system.

Further simplification of the fluorescence detection could be obtained by employing a light emitting diode (LED) presumably at a reduced sensitivity. Webster et al.⁷ presented a monolithic device fabricated by 13 steps of lithography which had integrated photodiodes built on a silicon substrate and optical interference filter with a parylene based microchip. Whitesides and co-workers⁸ also reported an integrated fluorescence detection system consisting of a microavalanche photodiode (µAPD), a thin film of polymeric colored filters and a PDMS microchip. Blue LEDs were employed as external light sources in both systems. Recently LEDs with a very high output power and shorter wavelength have become commercially available. Since LED is not very expensive and can be driven at low power, we can even use it as a disposable light source.

Compared with inorganic LEDs, organic light emitting diodes (OLEDs) have a flat surface which makes it easy to integrate with microfluidic devices and flexible to fabricate into any size and shape by photolithography techniques. A recent review showed high interests in this field since OLEDs offer the potential of on-chip light source arrays with controlled spectral characteristics and in principle are cheap to integrate on a microchip.¹⁰ Kopelman et al. reported a fluorescent chemical sensor platform integrating an OLED device light-source with a fluorescent probe for an oxygen sensor.¹¹ Fujii and co-workers first presented an integrated PDMS microfluidic device with a 510 nm (peak wavelength) OLED and optical fibers.¹² In order to minimize the distance between OLED and microchannel they placed a channel cast in PDMS directly on the rear side of the glass substrate of the OLED. Unfortunately, no fluorescence signal of Rhodamine B was obtained in their system and the design could be further optimized. Kim et al. reported an advanced and compact microchip coupled with a green OLED of 530 nm (peak wavelength) and a PIN photodiode for fluorescence detection which achieving a detection limit of 0.01 µM Rhodamine 6G as reported.¹³ Edel and co-workers presented a polyfluorene-based thin-film polymer light emitting diode (pLED) which had a peak emission wavelength at 488 nm as an integrated excitation source for microfabricated capillary electrophoresis.¹⁴ For fluorescein and 5-carboxyfluorescein detection concentrations as low as 1 µM were achieved with lock-in-amplifier equipment. Later they employed an organic photodiode for detecting and monitoring a peroxyoxalate based chemiluminescence reaction.¹⁵

OLEDs as a promising light source for integrated microfluidic devices and fluorescence measurements are attracting more and more attention. However, only a few groups have so far entered this field probably because OLEDs are not commercially available up to now. There are still several critically unresolved problems including spectrum purity and intensity which also exist in LED detection systems and block the way for a wide application. Scientists have made enormous efforts to improve the sensitivity of LED and OLED systems such as employing a lock-in-amplifier,^7,14 a liquid core waveguide,¹⁶ an emission interference filter⁷ and high sensitive confocal structures.¹⁷ Compared with lasers, the output power of LEDs and OLEDs is fairly low and has a wider bandwidth of emission spectrum. So it is very important to eliminate as much as possible the part of the excitation light which overlaps with the emission spectrum of analytes in order to supress the background interference and at the same time confirm that the distance between light source and microchannels is minimized. In our previous reported work, an argon ion laser and red diode laser were both employed as light sources for fluorescence detection in microfluidics^18–20 with high sensitivity yet large size. Now we are presenting a glass/PDMS microchip device using a green OLED which has a peak wavelength at 520 nm as the excitation source. Between the light source and microchannel a 300 µm-thick TiO₂/SiO₂ interference filter is inserted to get rid of unwanted excitation light. A conventional photomultiplier tube and optical fiber are employed for fluorescence detection of Rhodamine 6G and Alexa 532 dye on the microchip. The influence of pinhole size, excitation light filtering on detection sensitivity and stability of OLED at different driving voltage has been studied and under optimized conditions the obtained S/N ratio of 50 µM Rhodamine 6G and 7 µM Alexa 532 is 16.9 and 10.2 respectively. Using this system Alexa 532 and its bovine serum albumin (BSA) conjugates have been separated and fluorescence detected in modified microchannels and 4% linear polyacrylamide (LPA) buffer.

Experimental

Reagents and protein derivation

AlexaFluor^®532 carboxylic acid, succinimidyl ester (532/554 nm) was purchased from Molecular Probes (Eugene, OR, USA). Tris(hydroxymethyl)aminomethane (Tris), bovine serum albumin (BSA, 66 200Da), EDTA and Rhodamine 6G (526/555 nm) were all obtained from Sigma-Aldrich (St. Louis, MO, USA). Acrylamide monomer, and N,N,N′,N′-tetramethylethylenediamine (TEMED) were both bought from Promega (Madison, WI, USA) while ammonium persulfate (APS) was from Amresco (Solon, Oh, USA) and [γ-(methacryloyloxy)propyl] trimethoxysilane (MAPS) was a product of Fluka (Buchs, Switzerland). All other chemicals were of analytical reagent grade, and Milli-Q water (18.2 MΩ, Millipore, MA, USA) was used throughout.

The running buffer of 4% (w/v) LPA in 1 × TBE was prepared by dissolving 1.6 g acrylamide and 0.076 g APS in 10 ml of water, slightly different from the reported work of Schmalzing et al.²¹ and Gomis et al.²² Then the solution was mixed with 20 ml of 2 × TBE (89 mM Tris, 89 mM boric acid and 2 mM EDTA, pH 8.3) and the volume was adjusted to 40 ml. Immediately 24 µl of TEMED was added and the solution was degassed with an ultrasonic bath and left overnight at room temperature for complete polymerization.

For protein labeling, the Alexa 532 dye was stored and handled as instructed on the web site of Molecular Probes.²³ In brief, the dye was dissolved in dimethylsulfoxide (DMSO) at 10 mg ml⁻¹ (13.8 mM) and stored at −20 °C. For derivation, 5 µl of dye was added into 45 µl of 20 mg ml⁻¹ BSA (in 0.1 M bicarbonate buffer, pH 8.3) and immediately stirred gently for an hour at room temperature. The conjugate solution was then stored at 4 °C protected from light without further purification.

Microchip fabrication and coating

The glass substrate with microchannels used in the following experiments was designed and home-made by standard photolithography and wet chemical etching techniques.^19,24 The cover plate was a piece of 100 µm thick PDMS replica from a flat glass wafer which was silanized in 3% (v/v) octadecyl trichlorosilane (Sigma, St.Louis, MO, USA) in dry toluene for 2 h beforehand.²⁵ 10 ∶ 1 of the silicone elastomer and curing agent (Sylgard 184, Dow Corning, Midland, MI, USA) were mixed and poured onto the wafer after stirring and degassing. The solution was baked in a vacuum oven at 65 °C for 4 h. Immediately PDMS was sealed to the glass substrate after peeling off the wafer and then the microchip was exposed to ultraviolet light (UV) which had a peak emission at 253.7 nm (X-30G, Spectroline, USA) and average intensity of 1.85 mW cm⁻² with the PDMS side face up for 9 h for further combination and oxidization of PDMS as described elsewhere.²⁶

The final chip had a cross-linked microchannel which was 70 µm deep and 100 µm wide (at half depth) with 1.5 cm of sample channel and 3 cm of separation channel and a distance between cross channel to detection point of 1.5 cm. 5 µl of MAPS was added to each reservoir of S, SW and B (Fig. 1a) and 0.1 atm of vacuum was applied to the BW reservoir and the microchannels were soon filled with MAPS. After that it was left at room temperature for reaction overnight. This was followed by rinsing with methanol followed by water for 2 min and 10 min respectively and dipping in freshly prepared reaction buffer (3% acrylamide, 0.6% ammonium persulfate and 0.2% TEMED in water) for 3 h as reported by Han et al.²⁷ Finally the substrate was rinsed with water for 10 min and dried with nitrogen for 10 min, ready for usage.

Fig. 1 Optical set-up of OLED induced fluorescence detection system: (a) detailed arrangement of each component; (b) photograph of the microfluidics and OLED system; (c) side view of the structure with a coaxial optical fiber; (d) side view of the structure with a perpendicular optical fiber.

OLED fabrication

The OLEDs used in the experiments consisting of a typical p–n diode bottom emitting structure of ITO/NPB/Alq3/Mg∶Ag/Ag were fabricated by organic molecular beam deposition on a lithographically patterned indium tin oxide (ITO) coated glass substrate as described previously.^28–30 The ITO substrate was routinely cleaned by ultra-sonication in acetone, ethanol, rinsed in de-ionized water and isopropyl alcohol, and finally irradiated in an oxygen plasma chamber. Then, the organic films, 40 nm of α-napthylphenylbiphenyl (NPB) and 60 nm of tris(8-hydroxyquinoline) aluminium (Alq₃) were deposited on the ITO substrate layer by layer in high vacuum as the hole injection layer and the electron transport layer respectively. After deposition of the organic layers, the top cathode was prepared by sequential deposition of 100 nm Mg∶Ag and 50 nm Ag layers without breaking the vacuum. The sandwich structure of OLEDs is shown in Fig. 2.

Fig. 2 Bottom emitting structure of a typical p–n OLED which consists of a layer of 40 nm NPB and 60 nm Alq₃ as the hole injection layer and electron transport layer respectively. The OLEDs were sealed with a cover plate as protection from the air and 4.5–12 V was applied to the ITO anode and metal cathode for luminescence.

When 4.5–12 V direct current was applied to the metal cathode and ITO anode the energy barriers between the highest occupied molecular orbital and lowest unoccupied molecular orbital levels were about 0.4 and 0.9 eV, respectively, which were high enough to localize the holes in the NPB layer and electrons in the Alq₃ layer. Recombination of these charges occurred across the barriers, with holes primarily moving into Alq₃. The green OLED had a 0.5 mm thick glass substrate and an array of 250 µm × 250 µm illuminants controlled by parallels of deposited electrodes which emitted an intensity of 20000 cd m⁻² and irradiance of 7.5 mW cm⁻² (at 12 V driving voltage) green fluorescence with a peak emission at 520 nm and ∼60 nm bandwidth (FWHM).

Detection system

A compact OLED induced fluorescence detection system was established for measurement of Rhodamine 6G, Alexa 532 and its BSA conjugates. The optical set up is shown in Fig. 1a and b. On the top of the OLED was a piece of 0.3 mm short-pass interference filter (550 nm) designed and fabricated by Optical Coating Center of the Film Machinery Research Institute (Beijing, China) which consisted of 30 alternating layers of SiO₂ and TiO₂ and was about 4.5 µm thick. In order to limit the dimension of the detection point three pieces of 12 µm thick silver foil with a 50, 100 and 200 µm pinhole were respectively inserted between the excitation filter and microchip. Above the microchip a 500 µm-core-diameter optical fiber (Daheng Optical, Beijing, China) was inserted into a ∼1 mm deep hole drilled and polished in the microchip substrate which had ∼0.7 mm distance to the separation channel coincided with the pinhole (see Fig. 1c). When 4.5–12 V DC was applied to the anode and cathode electrodes of OLED it was illumined and the green emission transited the interference filter and pinhole layers in turn by which the unwanted excitation light was removed, exciting the fluorescent dyes or protein derivations in the microchannels. A perpendicular detection structure (see Fig. 1d) was also employed here and a comparision of the results was made. The fluorescence signals were collected by the optical fiber, then passed through a long-pass emission filter (555 nm, kindly presented by Beijing Yingxian Instruments, China) to eliminate the exciter light and other interferences focused by a group of lens onto a confocal pinhole (400 µm id, Daheng, China) and was finally detected by the photomultiplier tube (PMT, CR131-01, Beijing Hamamatsu, China). Fluorescence signals were digitized using a 400 kHz sampling frequency A/D card (AC6111, W&W Lab, China) and a program written with VC++ 6.0 was used for data acquisition and control of the multi-terminal power source (Northeastern University, Shenyang, China). Fig. 1b is a photograph of the actual system, while c and d is its section view. The total distance between OLED illuminants and microchannel was ∼1 mm which helped to improve sensitivity of the OLED fluorescence detection system.

Electrophoresis conditions and operation

The microchip was rinsed with water before use and 4% LPA solution contained 1 × TBE (pH 8.3) was used as working buffer throughout. For microchip injections, the floating sample loading model was employed³¹ in which the sample was driven by electrophoresis as electroosmotic flow was minimized and ignorable. Injections and separations were performed with field strengths of 250 V cm⁻¹ and 470 V cm⁻¹ under reverse polarity for Alexa 532 and its protein conjugates while forward polarity for Rhodamine 6G. For Alexa 532, as an example, dye in DMSO was diluted with running buffer to different concentrations needed for the sample solution. Before operation, the channels were rinsed with water, then filled with buffer. During injection (see Fig. 1a), the sample migrated from S (grounded) to SW (500 V) with B and BW remaining afloat for 50 s. Then the power supply switched and the sample zone was separated during the migration from B (grounded) to BW (1400 V) with S and SW afloat. After separation the microchannels were immediately rinsed with water to prevent the microchannels from blocking.

Results and discussion

Filtration of the excitation OLED emission

Although the micro photodiode and deposited interference filter had been successfully integrated into the microfluidic optical detection systems, LEDs were still employed as a detached light source from the chip and detector.^7,8,32 Moreover a lock-in-amplifier always had to be used in order to achieve enough sensitivity mainly because LEDs have a wider spectrum emission (40–60 nm)⁹ than lasers (5–10 nm) and a much lower output power. Recently OLEDs have been regarded as a promising alterative^10,33 being cheap and easy to integrate with a microchip as a thin film and flat surface and a photolithography fabrication method similar to microfluidics. However, as OLEDs have a wider spectrum emission (85 nm, FWHM)¹⁴ it was imperative to eliminate the excitation light that overlapped with the emission spectrum in order to obtain a sensitive detection. The commercially available filters were too thick (4–6 mm) to work for OLED fluorescence systems because the viewing angle of OLEDs was extremely large (about 170°) thus resulting in a decrease of luminance density per unit area with an increase of distance from it. So the microchannel should be in principle close to the OLED light source as much as possible for the sake of high intensity of excitation. A thin film (300 µm) interference filter which blocked excitation light higher than ∼555 nm was designed and fabricated for this purpose. Fig. 3 shows its spectrum and also that of the emission filter and green OLED (with and without excitation filter). The green OLED excited from 500 nm to 560 nm (FWHM) and there was about one fourth of the excitation light which was able to pass through the emission filter (blocking up to 545 nm) and overlapped with the fluorescence signals. By inserting the excitation filter the emission spectrum of OLED was redefined before reaching the microchannel and the unwanted light was removed from the excitation light (nearly 80%).

Fig. 3 Optical characteristics of the filters and emission spectrum of OLED: excitation filter (dash dot dash); emission filter (dot); OLED emission without excitation filter (dash) and OLED emission with excitation filter (solid).

Fig. 4 shows the electropherograms of 300 µM Rhodamine 6G with three detection structures: perpendicular structure (bottom), coaxial structure with (middle) and without (top) the excitation filter. Since no excitation filter has been employed with previously reported work on OLEDs, optical fiber collected fluorescence signals perpendicular to the light source and microchannel (see Fig. 1d) was accepted in order to supress the background interferences by preventing the detector from facing directly towards the OLED source, although there was still a part of interferences that reached the detector. When an appropriate excitation filter was designed and used for eliminating interference excitation light a coaxial framework as shown in Fig. 1c remarkably improves its sensitivity by collecting more fluorescence signals and less interferences, only if the distance between light source and microchannel is limited as much as possible. In this experiment, the achieved sensitivity of Rhodamine 6G with the excitation filter (middle spectrum in Fig. 4) was about 13 times as high as that without the filter (top) where a piece of 300 µm thick glass slide was inserted instead for keeping the distance and 3.5 times as the perpendicular detection structure (bottom). For Rhodamine 6G, a forward voltage was applied for injection and separation and it needed a longer migration time (150 s) due to a weak EOF existing in the microchannels or a small plus charge in the buffer.

Fig. 4 Electropherograms of 300 µM Rhodamine 6G in 4% LPA buffer (1 × TBE, pH 8.3) with three detection structures: perpendicular structure (bottom), coaxial structure with (middle) and without (top) the excitation filter. Floating injection and separation performed under forward voltage with a field strength of 250 V cm⁻¹ and 470 V cm⁻¹ respectively.

However, a layer of directly deposited interference filter onto the OLED surface was not achieved because of the difficulties in fabrication. A thin film filter (300 µm thick) sacrificed sensitivity to some extend, whereas, it could be kept on being used even when the OLED did not work and was abandoned. Furthermore, optical characteristics of the filter were better than when deposited.⁷

Microchip fabrication and modification

A glass/PDMS microfluidic chip was employed in this research because a common glass substrate available in our lab was not less than 1.7 cm thick which resulted in a long distance between OLED and microchannels and was disadvantageous for high sensitivity detection. A piece of 100 µm thick PDMS was polymerized and immediately sealed with the glass substrate with microchannels after which the channels were modified by MAPS. However, no methanol should be added to the modification solution as done usually because it would be absorbed by the PDMS material and make it fragile. When LPA solution is polymerized in the microchip it would fill the porous surface of PDMS and form a layer of physical coating as well as the chemical effect of modification of glass channels. This kind of glass/PDMS chip proved to be strong under normal pressure as well as rinsing the microchannels and could be used for several weeks.

Stability of OLEDs at different voltages

70 µM of Alexa 532 dye was filled into the microchannels in order to study the influence of driving voltages of OLED on fluorescence signals and stability. When output of the power supply increased every 5 min from 4.5 V to 13.5 V, the fluorescence intensity skipped upwards accordingly (see Fig. 5). When the driving voltage was as high as 13.5 V, fluorescence signals sharply declined 80% in 5 min which is mainly because the working current between metal cathode and ITO anode has exceeded its rated upper limit and excessive Joule heat production leading to irreversible damage. Therefore, in the following experiments 12 V of driving voltage was used as the power supply of the OLED unless stated otherwise for the sake of high sensitivity and several days stably output of OLEDs could be obtained at a 12 V driving voltage.

Fig. 5 Stability of OLED emission at different driving voltages where 70 µM Alexa 532 dye in 4% LPA buffer (1 × TBE, pH8.3) was used as analyte and was filled into the microchannels for studying. The power supply output increased from 4.5 V to 13.5 V every 5 min at a step of 1.5 V.

Optimization of the pinhole

Three pieces of silver foil with about 50 µm, 100 µm and 200 µm pinholes were prepared and inserted between excitation filter and microchip for confining the detection spot (see Fig. 1a). Fig. 6 illustrates the detection of 140 µM Alexa 532 dye solution with 50 µm, 100 µm and 200 µm pinhole respectively (from bottom to top). Two components of this dye (hydrolytes or photolytes) were successfully separated and fluorescently detected in the microchip. Different signal to noise (S/N) ratios of 21.1, 132.3 and 272.5 were obtained respectively according to the variety of the pinhole size. A pinhole with larger diameter (300 µm) was also studied in our experiments, however, it only slightly raised the signals (less than 10%), because the maximum width of the channel was 200 µm and the luminescent unit size of the OLED was 250 µm × 250 µm. Therefore 200 µm diameter of pinhole was finally selected in this system.

Fig. 6 Fluorescence detection of 140 µM Alexa 532 dye in 4% LPA buffer (1 × TBE, pH 8.3) with 50 µm, 100 µm and 200 µm pinhole respectively (from bottom to top); different S/N ratios of 21.1, 132.3 and 272.5 were achieved according to the increase of pinhole diameter.

Performance of the OLED induced fluorescence detection system

Under the optimized condition above, different concentrations (from 7 µM to 700 µM) of Alexa 532 and Rhodamine 6G was detected using this system (see Fig. 7) because these two dyes had almost the same maximum wavelengths of absorption and emission spectrum, 532/554 nm and 526/555 nm respectively. Alexa 532 has recently been regarded as an ideal dye for use with 532 nm excitation sources and has more prominent fluorescent characteristics than Rhodamine 6G³⁴ which is proven by the results of Fig. 7. The concentration and mass detection limit obtained by this system for Alexa 532 was about 3 µM (S/N = 3) and 1.4 fmol respectively at 0.7 nl of sample injection volume. System sensitivity was considerably improved compared with the previously reported work, where about 18.5 of S/N ratio was achieved in the electrophoregram of 10 mM fluorescein at 0.1 nl injection volume with lock-in-amplifier equipment.

Fig. 7 Detection of Rhodamine 6G and Alexa 532 dye in 4% LPA buffer (1 × TBE, pH 8.3) at different concentrations from 7 µM to 700 µM (n = 3) at 0.7 nl of injection volume. For Rhodamine 6G injection and separation conditions were the same as in Fig. 4, while for Alexa 532 only the direction of voltage was changed.

However, the result was roughly six orders of magnitude poorer than good laser-induced fluorescence detection in capillary electrophoresis³⁵ because of its low irradiance and purity. Therefore, the sensitivity of OLED induced fluorescence detection systems needs further improvements by employing high performance OLED for future application.

Electrophoresis and fluorescence detection of BSA conjugates

Alexa dyes are structurally related fluorescent molecules that are named according to the wavelength (nm) of the nearest laser excitation. Alexa 532 ready reacts with non-protonated aliphatic amine groups including the amine terminus of proteins, producing stable carboxamide bonds.³⁶ This dye has begun to be widely used in the research of proteins, nucleic acids and cells for biological purposes. Fig. 8 is the electrophoresis and detection diagram of the protein derivations without further purification where the dye molecules migrated to the detector earlier than the BSA conjugates in a reverse electric field in 4% LPA buffer (1 × TBE, pH 8.3). The electropherogram of BSA conjugates in this research successfully correlated with that obtained by capillary electrophoresis in 1 × TBE (pH 8.3) buffer, while the Alexa 532 dye migrated to the opposite direction.

Fig. 8 Electropherogram of BSA conjugates labeled by Alexa 532 dye in 4% LPA buffer (1 × TBE, pH 8.3) under optimized conditions as in Fig. 7.

Conclusions

In this study, a novel microfluidic device using a green organic light emitting diode as excitation source was established and the sensitivity of fluorescence detection was improved by inserting a thin film of excitation filter which could remove the part of excitation light which overlapped with the emission spectrum of the dyes. Moreover the sensitivity could be further improved if the interference filter could be deposited directly onto the surface of OLED substrate. This OLED induced fluorescence detection microfluidic system was applied to electrophoresis and detection of BSA conjugates labeled with Alexa 532. The results proved that OLEDs are promising light sources for microfluidic fluorescence detection systems which have a small size and are easy to integrate. However, the intensity and stability of its fluorescence is expected to become more powerful in the near future with enormous efforts being directed in this area.

Compared with laser and LEDs, OLEDs are also advantageous for accurate fabrication into various size and shapes by photolithography techniques other than integration. Therefore, it is convenient to fabricate a two dimensional light source for multiple detection using a CCD camera. Moreover micro detectors can also be employed in OLED fluorescence detection systems and lead to further miniaturization even to palm or thumb size which would resolve the problem of a small microchip coupled with a bulky laser source for microfluidic fluorescence detections. We plan to explore these avenues further.

Acknowledgements

This research was supported by National Science Foundations of China (Grant No. 20299036 and 20475031) projects. The authors would like to thank Dr Deqiang Zhang of Beijing Visionox Technology Co., Ltd for OLED fabrication and graduate students Peng Wei and Shiliang Han of Prof. Yong Qiu’s group for helping to test optical characteristics of the OLEDs.

References

1. G. F. Jiang, S. Attiya, G. Ocvirk, W. E. Lee and D. J. Harrison, Biosens. Bioelectron., 2000, 14, 861.
2. A. T. Woolley, K. Q. Lao, A. N. Glazer and R. A. Mathies, Anal. Chem., 1998, 70, 684.
3. S. D. Mangru and D. J. Harrison, Electrophoresis, 1998, 19, 2301.
4. H. B. Qiu, J. L. Yan, X. H. Sun, J. F. Liu, W. D. Cao, X. R. Yang and E. K. Wang, Anal. Chem., 2003, 75, 5435.
5. J. J. Li, P. Thibault, N. H. Bings, C. D. Skinner, C. Wang, C. Colyer and J. Harrison, Anal. Chem., 1999, 71, 3036.
6. H. Wensink, F. Benito-Lopez, D. C. Hermes, W. Verboom, H. J. G. E. Gardeniers, D. N. Reinhoudt and A. van den Berg, Lab Chip, 2005, 5, 280.
7. J. R. Webster, M. A. Burns, D. T. Burke and C. H. Mastrangelo, Anal. Chem., 2001, 73, 1622.
8. M. L. Chabinyc, D. T. Chiu, J. C. McDonald, A. D. Stroock, J. F. Christian, A. M. Karger and G. M. Whitesides, Anal. Chem., 2001, 73, 4491.
9. K. Uchiyama, H. Nakajima and T. Hobo, Anal. Bioanal. Chem., 2004, 379, 375.
10. K. B. Mogensen, H. Klank and J. P. Kutter, Electrophoresis, 2004, 25, 3498.
11. V. Savvate'ev, Z. Chen-Esterlit, J. W. Aylott, B. Choudhury, C. H. Kim, L. Zou, J. H. Friedl, R. Shinar, J. Shinar and R. Kopelman, Appl. Phys. Lett., 2002, 81, 4652.
12. S. Camou, M. Kitamura, Y. Arakawa and T. Fujii, 7th International Conference on Miniaturized Chemical and Blochemlcal Analysts Systems, ed. M.A. Northrup, K.F. Jensen and D.J. Harrison, Transducers Research Foundation, Inc., California, USA, p. 383.
13. J. H. Kim, K. S. Shin, K. K. Paek, Y. H. Kim, Y. M. Kim, Y. K. Kim, T. S. Kim, J. Y. Kang, E. G. Yang, S. S. Kim and B. K. Ju, Proceedings of micro TAS 2004 (8th International Conference on Miniaturized System for Chemistry and Life Science), ed. T. Laurell, J. Nilsson, K. Jensen, D.J. Harrison and J.P. Kutter, The Royal Society of Chemistry, Cambridge, UK, p. 428.
14. J. B. Edel, N. P. Beard, O. Hofmann, J. C. deMello, D. D. C. Bradley and A. J. deMello, Lab Chip, 2004, 4, 136.
15. O. Hofmann, P. Miller, J. C. deMello, D. D. C. Bradley and A. J. deMello, Proceedings of micro TAS 2004 (8th International Conference on Miniaturized System for Chemistry and Life Science), ed. T. Laurell, J. Nilsson, K. Jensen, D.J. Harrison and J.P. Kutter, The Royal Society of Chemistry, Cambridge, UK, p. 506.
16. F. Q. Dang, L. Zhang, H. Hagiwara, Y. Mishina and Y. Baba, Electrophoresis, 2003, 24, 714.
17. S. L. Wang, X. J. Huang, Z. L. Fang and P. K. Dasgupta, Anal. Chem., 2001, 73, 4545.
18. Y. Jin and G. A. Luo, Electrophoresis, 2003, 24, 1242.
19. B. Yao, G. A. Luo, X. Feng, W. Wang, L. X. Chen and Y. M. Wang, Lab Chip, 2004, 4, 603.
20. B. Yao, X. Feng, G. A. Luo and Y. M. Wang, Chem. J. Chin. Univ., 2005, 25, 43.
21. D. Schmalzing, A. Adourian, L. Koutny, L. Ziaugra, P. Matsudaira and D. Ehrlich, Anal. Chem., 1998, 70, 2303.
22. D. B. Gomis, S. Junco, Y. Expósito and M. D. Gutiérrez, Electrophoresis, 2003, 24, 1391.
23. www.probes.com.
24. X. F. Yin, H. Shen and Z. L. Fang, Chin. J. Anal. Chem., 2003, 31, 116.
25. L. Ceriotti, N. F. de Rooij and E. Verpoorte, Anal. Chem., 2002, 74, 639.
26. D. Q. Xiao, T. Van Le and M. J. Wirth, Anal. Chem., 2004, 76, 2055.
27. F. T. Han, Y. Wang, C. E. Sims, M. Bachman, R. S. Chang, G. P. Li and N. L. Allbritton, Anal. Chem., 2003, 75, 3688.
28. Y. Qiu, Y. D. Gao, L. D. Wang, P. Wei, L. Duan, D. Q. Zhang and G. F. Dong, Appl. Phys. Lett., 2002, 81, 3540.
29. Y. Qiu, Y. D. Gao, P. Wei and L. D. Wang, Appl. Phys. Lett., 2002, 80, 2628.
30. G. T. Lei, L. D. Wang and Y. Qiu, Appl. Phys. Lett., 2004, 85(22), 5403.
31. S. C. Jacobson, R. Hergenröder, L. B. Koutny, R. J. Warmack and J. M. Ramsey, Anal. Chem., 1994, 66, 1107.
32. V. Namasivayam, R. S. Lin, B. Johnson, S. Brahmasandra, Z. Razzacki, D. T. Burke and M. A. Burns, J. Micromech. Microeng., 2004, 14, 81.
33. E. Verpoorte, Lab Chip, 2003, 3, 42N.
34. N. Panchuk-Voloshina, R. P. Haugland, J. Bishop-Stewart, M. K. Bhalgat, P. J. Millard, F. Mao, W. Y. Leung and R. P. Haugland, J. Histochem. Cytochem., 1999, 47, 1179.
35. Y. F. Gheng and N. J. Dovichi, Science, 1988, 242, 562.
36. M. Brinkley, Bioconjugate Chem., 1992, 3, 2.

---