Isabel Rodriguez1

Marie Lesaicherre2

Yan Tie3

Quanbo Zou4

Chen Yu4

Janak Singh4

Lim T. Meng3

Sridhar Uppili4

Sam F. Y. Li2

PonnampalamGopalakrishnakone5

Zachariah E. Selvanayagam5

1Institute of MaterialsResearch and Engineering,3 Research Link,Singapore

2Department of Chemistry3Department of BiologicalSciences, The NationalUniversity of Singapore,Singapore

4Institute of Microelectronics,Singapore

5Department of Anatomy,The National Universityof Singapore,Singapore

Practical integration of polymerase chain reactionamplification and electrophoretic analysis inmicrofluidic devices for genetic analysis

An integrated system of a silicon-based microfabricated polymerase chain reaction(�PCR) chamber and microfabricated electrophoretic glass chips have been devel-oped. The PCR chamber was made of silicon and had aluminum heaters and tempera-ture sensors integrated on the glass anodically bonded cover. Temperature uniformityin the reaction chamber was �0.3�C using an improved novel “joint-heating” scheme.Thermal cycling was digitally controlled with a temperature accuracy of �0.2�C. Smalloperating volumes together with high thermal conductivity of silicon made the devicewell suited to rapid cycling; 16 s/cycle were demonstrated. For analysis of the PCRproducts, the chamber output was transferred to the glass microchip by pressure.Analysis time of PCR amplified genomic DNA was obtained in the microchip in lessthan 180 s. The analysis procedure employed was reproducible, simple and practicalby using viscous sieving solutions of hydroxypropylmethylcellulose and dynamicallycoated microchip channels with poly(vinylpyrrolidone). DNA fragments that differ insize by 18 base pairs (bp) were resolved. Analysis of genomic male and female ampli-fied DNA by �PCR was achieved in microchip, and application of the integrated �PCR-�chip for the identification of bird sex was tested. Genomic DNA samples from severalbird species such as pigeon and chicken were analyzed. Hence, the system could beused as well to determine the sex of avian species.

Keywords: Bird sexing / Integrated micropolymerase chain reaction-capillary electrophoresis /Microcapillary electrophoresis / Microfabrication / Micropolymerase chain reaction EL 5199

1 Introduction

Recent advances in molecular and cell biology have beenproduced in great part as a result of the development ofrapid and efficient analytical techniques. Due to miniatur-ization and multiplexing, techniques like gene chip or bio-chip enable the characterization of complete genomes ina single experimental setup [1, 2]. Among the biotechni-ques, PCR has become the most important analyticalstep in life sciences laboratories for a large number ofmolecular and clinical diagnostics, Sanger sequencingreactions and preparative step in hybridization bioassayslike microarrays. Important developments made in PCRtechnology like real-time PCR [3, 4], have led to rapidreaction processes compared to conventional methods.During the past several years, microfabrication technol-

ogy has been expanded to the miniaturization of the reac-tion and analysis system such as PCR analysis with theintention of further reducing analysis time and consump-tion of reagents. Several research groups have beenworking on the ‘lab-on-a-chip’ devices and have led to anumber of advances in the fields of miniaturized separa-tion and reaction systems [5–18].

Important advances have been made at the componentlevel; microfabricated separation columns or microchipshave been employed for analyzing a wide variety of sam-ples predominantly for the analysis of biomolecules suchas DNA, RNA, proteins, and cells [5–7]. Microreactorshave mainly targeted DNA-based reactions such as thePCR [8–10]. The first PCR and microchip-integrated de-vice was developed by Woolley et al. in 1996 [11]. Sincethen, progress at the integration level has attracted agreat deal of attention and several integrated systems forPCR amplification and analysis have been constructed[12–18]. Developments in this field seek primarily thereduction of the amplification time in conjunction with thevolume of sample used. Towards this aim, different strate-gies have been adopted such as the reduction of thethermal mass by reducing the volume of the reaction mix-ture [12,13]. Kandurina et al. [14] developed a system that

Correspondence: Z. Emmanuel Selvanayagam, PhD, Depart-ment of Pediatrics, University of Medicine and Dentistry of NewJersey, 125 Paterson Street, Room No. 7050 CAB, New Bruns-wick, NJ 08903, USAE-mail: selvanze@umdnj.eduFax: 732-235-6325

Abbreviations: CHD, chromodomain-helicase-DNA; HPMC,hydroxypropylmethylcellulose; PDMS, poly(dimethylsiloxane);TBE, Tris-borate-EDTA

172 Electrophoresis 2003, 24, 172–178

2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 0173-0835/03/01-0201–172 $17.50�.50/0

Electrophoresis 2003, 24, 172–178 Integrated �PCR-CE 173

allows sample preconcentration by means of a porousmembrane reducing significantly the number of cyclesnecessary to obtain a detectable amount of product.Another approach has been the decrease of the tempera-ture ramping rates. The temperature transition periodscontribute in great part to the total cycling time and haveno benefit on the PCR reactions but very negative effectson the formation on nonspecific amplification products.Very significant strategies have been adopted to diminishthis problem. Landers and co-workers [10, 15] achievedvery short thermal cycles employing IR radiation as heat-ing source and realization a more effective cooling. Thegroup of Manz [9] developed a multiple heated zonescontinue flow reactor. Recently, a nanoliter size reactorwhere the PCR mixture is cycled across the heating zoneshas also been demonstrated where the amplification timeis greatly reduced [19].

In this paper, we report a practical integration of a micro-fabricated silicon PCR reactor with a standard glassmicrochip. Our approach was to reduce the temperatureramping rates by optimizing the design of the PCR reactorcomponents. We also aimed at maximizing heating effi-ciency and temperature uniformity to be able to performPCR amplifications of genomic DNA in actual test. APCR-based bird-sexing test was chosen to verify theefficiency, rapid cycling and analysis capabilities of thePCR microchip integrated device. The test was based onthe amplification of two chromodomain-helicase-DNA(CHD) genes, CHD-W and CHD-Z. The CHD-W is locatedon the W-chromosome, therefore it is unique to female;the CHD-Z gene is found on the Z-chromosome andtherefore occurs in both sexes (female, ZW; male, ZZ).The test employs PCR with a single set of primers, whichamplify both genes and incorporates introns whoselengths usually differ in most of the bird species. Theelectrophoretic analysis of the PCR product gives a singleCHD-Z band in male, but females have a second, distinc-tive CHD-W band. The two bands are quite similar in sizeand hence, this test makes a good model to verify therobustness of the system. The choice of materials to buildthe PCR chamber and electrophoretic microchips wasaccording to the required properties for the specific pro-cess. The material of choice to fabricate �PCR deviceswas silicon as it provides excellent thermal conductivityrequired to achieve rapid thermal cycles and good tem-perature homogeneity. But glass is preferred for electro-phoresis processes due to its optical and electrical prop-erties. As such, devices were fabricated using differentmaterials and were then assembled prior to use by usinga molded poly(dimethylsiloxane) (PDMS) gasket. Theassembly allows for the accommodation of a grand vari-ety of microchip designs, which makes this type of inte-gration very practical and versatile.

2 Materials and methods

2.1 Devices fabrication and operation

The design and fabrication of the �PCR chamber wasperformed at the Institute of Microelectronics (IME). Thechamber was fabricated in silicon using micromachiningtechniques and was enclosed by anodic-bonding with aPyrex glass cover with ultrasonically drilled inlet and out-let holes. Aluminum thin film resistors and temperaturesensors were patterned on the glass top of the reactionchamber. The �PCR devices were fabricated at waferlevel and subsequently diced into single devices beforethey were packaged onto a thermally conductive printedcircuit board (PCB). Details of the reactor design studies,microfabrication process and characterization have beendescribed previously [20]. The microfabricated PCRdevices had a reaction chamber volume of approximately3 or 10 �L and a total chip size of 7�20 mm. The �PCRelectronic control system consisted of a digital controlunit and data acquisition unit. The chambers were ther-mally cycled with a Lab VIEW program (National Instru-ments, Austin, TX, USA). The thermocouple input volt-ages were passed through a signal-conditioning unit(National Instruments) and into a 12 bit ADC on a perso-nal computer. Temperature control was accomplishedthrough a percentage/integrator/differentiator (PID) mod-ule with the Lab VIEW program. The ADC output wasused to control the heater passed through a currentsource circuit to supply the power necessary to drive theheaters. The devices were operated at 8 V and had aheating power consumption of 1–2 W. Microchip devicesused were off-the-self chips produced by Micralyne(Edmonton, Canada) and fabricated by conventionalphotolithographic methods. The separation channel was8.5 cm in length (0.5 cm from buffer reservoir to theinjection cross and 8 cm from the injection cross to thebuffer waste reservoir). The effective separation lengthwas 6 mm. The injection channel was 0.8 cm (0.4 cmfrom the sample reservoir to the injection cross and0.5 cm from the sample waste reservoir to the injectioncross). The separation and injection channel width was50 �m. All the channels were 20 �m deep. The chipsused had an injection cross offset of 100 �m. The samplereservoir of the microchips was drilled through for con-necting to the PCR outlet. The �PCR chamber and �chipwere then coupled through a PDMS gasket that had thesame thickness as the PCB and with a center hole ofthe same diameter as the microchip sample reservoir.An illustration of the integrated device is shown in Fig. 1.Due to the good adhesion properties of PDMS to glasssurfaces, a good seal was obtained and no leak wasobserved throughout the experiments. For the operationof the integrated system, �PCR-�chip devices were pre-

Min

iatu

riza

tio

n

174 I. Rodriguez et al. Electrophoresis 2003, 24, 172–178

Figure 1. Schematic illustration of the integrated �PCR-�chip device showing in (A) a top view and in (B) a crosssection.

filled in advance with the required reagents. A stopperwas used to close the �PCR outlet to prevent samplemixing. After the thermal amplification was completed,the PCR-amplified product was transferred to the micro-chip sample reservoir by applying a small pressure with asyringe connected by a rubber tube to the inlet hole of thePCR chamber. Once the PCR product was in the samplereservoir, the entire �PCR-�chip device was placed in thedetection system and electrophoresis was performed.

Separations were monitored using a laser-excited con-focal fluorescence detection system. A Leica episcopicfluorescence microscope (DAS Mikroskop DMLS; Leica,Heerbrugg, Switzerland) and its accessories were modi-fied for this purpose. The excitation radiation source was488 nm from a 6 mW argon-ion laser (Omnichrome,Chino, CA, USA) and was directed perpendicular to themicrochannels by a 510 nm dichroic beam splitter andfocused within the detection point of the channel usinga Leica PL FLUOTAR long working distance objective(magnification, 63; numerical aperture, 0.70). A barrierfilter (BA 515 nm) filtered the fluorescence emission col-lected perpendicular to the channels through the sameobjective. The detector, a Hamamatsu photomultiplier,was assembled and loaned from CE Resources Pte Ltd.(Singapore). The analogue signal was sampled at 10 Hz.A computer-controlled high voltage power supply cap-able of delivering four independent potentials was ob-tained from CE Resources Pte Ltd. Electrical contactfrom the power supply and the electrolyte solution wasperformed with platinum electrodes inserted in the reser-voirs. Solutions were filled or flushed by applying a

10–20 psi pressure from a compressed purified N2 cylin-der to the reservoirs filled with solutions. For the injectionof a sample plug, 800 V potential was applied betweensample reservoirs for 100 s. Subsequently, analysis wascarried out between separation reservoirs by applying2000 V. The distance from the injection point to thedetection window was 48 mm.

2.2 PCR amplification and electrophoresisseparation conditions

�PCR amplification reaction was performed with plasmidas well as genomic DNA. Both plasmid and genomic DNA(from blood samples) were extracted using Qiagen (Valen-cia, CA, USA) DNA isolation kit according the manu-facturer’s instructions. PCR conditions for both plasmidand genomic DNA were as follows (in final concentra-tions): 50 mM KCl, 10 mM Tris-HCl (pH 8.8 at 25�C), 0.1%Triton X-100, 3 mM MgCl2, 0.2 mM each of the primers,1 �M each dNTP, 0.025 U/�L of Taq DNA polymerase,and 20–30 ng/�L DNA template. The �PCR thermalcycling conditions for amplifying plasmid DNA: initialdenaturation at 96�C for 22 s, 30 cycles of 96�C for8 s, 55�C for 8 s, and a final extension at 72�C for 270 s.Thermal cycling profile for both male and female birdgenomic DNA: initial denaturation at 94�C for 120 s,36 cycles of 94�C for 35 s, 48�C for 50 s, 72�C for 50 sand a final extension at 72�C for 480 s. Positive controlswere run in a conventional Peltier thermal cycler (MJ Re-search, Watertown, MA, USA) in order to ensure directcomparability with results obtained from the �PCR de-vice. Since silicon and glass have inhibitory effects onthe PCR yield, a surface passivation procedure was per-formed. For this purpose a BSA solution (1.0 mg/mL) wasplaced in the reaction chamber for 2–3 min and thenremoved from the well. The electroosmotic flow in thechannels was suppressed by coating them dynamicallywith 106 molecular weight poly(vinylpyrrolidone) (PVP;Polysciences, Warrington, PA, USA) which was addedat a concentration of 0.5% to the buffer electrolyte. Theseparation medium used consisted of TBE buffer: 89 mM

tris(hydroxymethyl)aminomethane (Tris; Bio-Rad, Rich-mond, CA, USA), 89 mM boric acid, 2 mM disodiumEDTA, 0.5% w/v hydroxypropylmethylcellulose (HPMC;Aldrich, Milwaukee, WI, USA) and 6% mannitol (Fluka,Buchs, Switzerland). For visualization of DNA molecules,the intercalating dye, thiazole orange (TO) from Aldrichwas added to the run buffer at a concentration range of0.1–0.25 �g/mL. Microchannels were conditioned for useby first filling with water and subsequently flushing with0.1 M NaOH, water and sieving buffer. Regeneration of thechannels between runs was not necessary. Microchannelswere stored dried after being rinsed with Millipore water.

Electrophoresis 2003, 24, 172–178 Integrated �PCR-CE 175

3 Results and discussion

Silicon was chosen as a base material because it haswell-known characteristics and its application in micro-fabrication processes are also well established. It allowsfor precise design and micromachining of componentswith well-defined thermal conductance. A novel thermalisolation design was implemented by etching throughthermally conductive silicon membrane, to eliminatecross talk between the reaction and the substrate and toreduce parasitic heat capacitance. In addition, a novelheating scheme, named joint heating has been developedand tested. As it is shown in Fig. 1, heating structureswere only placed at the sides of the chamber close to thesmall joints made by the silicon beams. As predicted bythe theoretical studies performed for this type of symme-trical thermal arrangement, the temperature deviationacross the chamber was extremely reduced [20]. Thenovel joint-heating scheme together with the fast heattransfer rates of silicon provided very fast heating andcooling rates as well as temperature homogeneity valuesof � 0.3�C. We firstly investigated the thermal character-istics and rapid thermal cycling capabilities of the micro-fabricated PCR. For this purpose, a rapid PCR testwithout definite extension point was carried out using aplasmid CHD-Z gene. The conditions were optimized toobtain a measurable amount of product in the minimumreaction time. The integrated sensors provided the tem-perature distribution in the reaction chamber and showeda temperature deviation under 0.3�C around the reactionchamber. Figure 2 shows the temperature profile at theset point and the temperature detected by the integratedsensors as a function of the reaction time. The total anal-ysis time was 13 min. Cooling from the melting tempera-ture to the annealing temperature took 3.7 s correspond-ing to a cooling rate of 11�C/s. Ramping rates betweenthe set points were of the same order. The cycling tem-perature measured also had excellent agreement with

Figure 2. Temperature profile for amplification of plasmidDNA. (A) Initial denaturation and cycles 1–5. (B) Initialdenaturation: 96�C for 22 s; 30 cycles at: 96�C for 8 s,55�C for 8 s; final extension: 72�C for 270 s.

the set point temperature, with a �0.2�C fluctuation. Thetemperature accuracy between different amplificationcycles was �0.2�C.

After the thermal cycling, the PCR product was trans-ferred to the sample reservoir and subsequently electro-phoretic separation was performed. Figure 3 shows theelectrophoretic analysis of an amplified CHD-Z plasmidDNA in a 3 �L PCR chamber. The peak of the amplifiedproduct can be seen at 4 min (separation voltage in thiscase was 1 kV). Peaks due to nonspecific products werevery small and could hardly be observed in the electro-pherograms. This indicates that the thermal transitionsachieved with the designed �PCR were remarkably effi-cient and rapid and allowed for very specific reaction to

Figure 3. Electrophoretic ana-lysis of a rapid amplification ofthe plasmid CHD gene of amale chicken. Sample injection:elektrokinetic, reverse polarity250 V/cm across the injectionchannel; separation, 214 V/cmacross the injection channel;buffer, TBE buffer (89 mM Tris,89 mM boric acid, 2 mM diso-dium EDTA) modified with 0.5%w/v HPMC and 6% mannitol assieving medium; dynamic coat-ing method, 0.1% w/v PVP.

176 I. Rodriguez et al. Electrophoresis 2003, 24, 172–178

be realized. These excellent performance characteristicsof the �PCR chamber prompted us to amplify, genomicDNA using our device.

Genomic DNA was extracted from chicken (Gallus gallus)blood that was used to specifically amplify the CHD gene.Generally, since amplification of genomic DNA is not asefficient as plasmid DNA, a longer cycling time with anextra annealing step was incorporated in our design toobtain a measurable amount of PCR product. The totalamplification time was increased to 78 min and tempera-ture consistency from cycle to cycle was observed withaccuracy in the order of �0.2�C. After amplification in thePCR chamber, the PCR product was directly transferredto the electrophoresis channel. Figure 4 shows the elec-tropherogram corresponding to the CHD gene amplifiedfrom a female (Fig. 4A) and male chicken (Fig. 4B). Thegenomic DNA of other bird species was also tested withthe microsystem as bird-sexing test tool. Following theamplification of genomic DNA from a pigeon (Columbalivia) we found specific products in the electropherogramthat corresponded to the CHD-amplified gene of a femalepigeon (Fig. 4C).

We further evaluated the amplification efficiency of the�PCR by performing different amplification cycles of theCHD genes relative to various predetermined amounts oftemplate. Ideally, the amount of product amplified canbe predicted by the equation P = S02n, (P is the productnumber after amplification, S0 is the substrate or targetnumber before amplification and n is the number of

cycles of amplification), in a linear relationship betweenthe logarithm of P and the amplification cycles. As shownin Fig. 5, a linear plot was obtained, and these resultsdemonstrated that amplification with the microfabricatedPCR system was efficient.

In microchip separations, fast, efficient, and automatablecoating and filling procedures are necessary for routinelaboratory application. To avoid lengthy surface passiva-tion procedures, a dynamic coating procedure with PVPwas adopted. Application of PVP as separation matrix incapillary electrophoresis, has been successfully demon-strated in DNA electrophoresis separations [21, 22].PVP-coated silica wall reduces EOF values significantly.We assayed the use of PVP as sieving matrix at concen-trations up to 1% w/v. We found that high concentrationsof PVP produced high viscosity thus making it difficult tofill the channels. Measurement of the EOF indicated thatit had been greatly suppressed by the adsorption of PVP.Therefore, the resolution of PVP alone was not sufficient inthe microchannel for both fragments of female chickenCHD genes, which differ by 18 bp.

To further address this issue, we introduced in the separa-tion buffer additional polymers to increase the sievingstrength, including hydroxyethylcellulose (HEC) andHPMC. We found that using a buffer system composedof 89 mM TBE, 0.5% HPMC, 6% manitol and 0.1% PVP,the PCR products of the two CHD genes can be resolvedefficiently (Fig. 4). This buffer combination has adequateviscosity for easy flushing and filling of the microchannels,

Figure 4. Electropherogram of genomic CHD gene of different birds species. (A) Female chicken; (B) male chicken; (C)female pigeon. Conditions as in Fig. 3.

Electrophoresis 2003, 24, 172–178 Integrated �PCR-CE 177

Figure 5. Plot of the logrithm of the PCR product peakarea as function of the cycle number for a male chickengenomic CHD gene. Data was obtained from three elec-trograms performed on-chip. Conditions as in Fig. 3. Forthe amount of starting template copies and reagentsused, a plateau region is not revealed and the log ofpeak area is a linear function of the thermal cycles per-formed.

and when boric acid in TBE was replaced by acetic acidin the Tris-acetate-EDTA (TAE) buffer, the PCR productsfailed to separate with our system. This observation isconsistent with previous results by Han et al. [23] support-ing that the dynamic sieving network between HPMC andmanitol molecules may be mediated by the tetraboratebridges.

4 Concluding remarks

Microfabricated devices have been used for plasmidDNA amplification however, few studies have validatedthe application of these devices with genomic DNA. Weshowed here the application of a fully integrated micro-fabricated system for the amplification and analysis ofsex-specific genes from various avian species usinggenomic DNA. Reduction of thermal mass, that is, reac-tion volume in combination with an efficient heat transfer,makes rapid PCR thermocycling possible in our system.We have demonstrated 30 cycles of PCR amplification in13 min. The novel joint-heating scheme together with thefast heat transfer rates of silicon provided homogeneoustemperature in the PCR chamber as well as rapid thermaltransitions rates that allowed for specific reactions to berealized. Heating and cooling rates of 11�C/s were de-monstrated. Consequently, the system was capable ofamplifying genomic DNA with high specificity. Genomicamplification and analysis of different bird species wasachieved in our integrated microsystem and was suc-cessfully tested in bird sexing.

We demonstrated that resolution and analysis of PCR-amplified DNA can be obtained in the microchip in lessthan 180 s. We showed further that this integrated systemcan also resolve DNA fragments that differ in size by18 bp. Our procedure with the use of a dynamic coatingand replaceable sieving matrix were simple and practical,allowing for automation of sequential and reproducibleanalyses. The integration assembly is simple and ver-satile, enabling easy accommodation of various chip de-signs and materials, thereby further enhancing its appli-cation for a wide range of genetic analyses. The inte-grated system is detachable thus allowing for efficientcleaning and recycling of the system, further reducingthe cost of analysis. In conclusion, an effective inte-grated system for PCR amplification and electrophoreticanalysis in microchips has been built and its applicationfor sex determination of various avian species promisesa highly accurate, automatable and reusable microchipsystem for routine laboratory application for variousgenetic analyses.

The authors thank Dr. Bill Chen and Prof. C. C. Hang,National Science and Technology Board, Singapore, forstarting the Biosensor Focused Interest Group initiativeacross the different research institutes and The NationalUniversity of Singapore as well as for providing financialsupport. The authors also thank Dr. K. V. Chin, The CancerInstitute of New Jersey, New Brunswick, New Jersey forcritical observation on the manuscript and Mr. Chong SerChoong, IME, for assembly of the �PCR devices. We alsothank CE Resources Pte Ltd for the loan of the photo-multiplier and for technical assistance.

Received May 29, 2002

5 References

[1] Krishnan, M., Namasivayam, V., Lin, R. S., Pal, R., Burns,M. A., Curr. Opin. Biotechnol. 2001, 12, 92–98.

[2] Jain, K. K., Science 2001, 294, 621–625.

[3] Foy, C. A., Parkes, H. C., Clin. Chem. 2001, 47, 990–1000.

[4] Belgrader, P., Young, S., Yuan, B., Primeau, M., Christel,L. A., Pourahmadi, F., Northrup, A. M., Anal. Chem. 2001,73, 286–289.

[5] Sanders, G. H. W., Manz, A., Trends Anal. Chem. 2000, 19,364–378.

[6] Figeys, D., Pinto, D., Electrophoresis 2001, 22, 208–216.

[7] Khandurina, J., Guttman A., J. Chromatogr. 2002, 943, 159–183.

[8] Lee, T. M. H., Hsing, I. M., Lao, A. I. K., Carles, M. C., Anal.Chem. 2000, 72, 4242–4247.

[9] Kopp, M. U., De Mello, A. J., Manz, A., Science 1998, 280,1046–1048.

[10] Huhmer, A. F. R., Landers, J. P., Anal. Chem. 2000, 72, 5507–5512.

178 I. Rodriguez et al. Electrophoresis 2003, 24, 172–178

[11] Woolley, A. T., Hadley, D., Laudre, P., De Mello, A. J.,Mathies, R. A., Northrup, M. A., Anal. Chem. 1996, 68,4081–4086.

[12] Burns, M. A., Johnson, B. N., Brahmasandra, S. N., Handi-que, K., Webster, J. R., Krishnan, M., Sammarco, T. S., Man,P. M., Jones, D., Heldsinger, D., Mastrangelo, C. H., Burke,D. T., Science 1998, 282, 484–487.

[13] Waters, L. C., Jacobson, S. C., Kroutchinina, N., Khandur-ina, J., Foote, R. S., Ramsey, J. M., Anal. Chem. 1998, 70,158–162.

[14] Khandurina, J., McKnight, T. E., Jacobson, S. C., Waters,L. C., Foote, R. S., Ramsey, J. M., Anal. Chem. 2000, 72,2995–3000.

[15] Giordano, B. C., Ferrance, J., Swedberg, S., Huhmer, A. F. R.,Landers, J. P., Anal. Biochem. 2001, 291, 124–132.

[16] Lagally, E. T., Simpson, P. C., Mathies, R. A., Sens. ActuatorsB 2000, 63, 138–146.

[17] Lagally, E. T., Medintz, I., Mathies, R. A., Anal. Chem. 2001,73, 565–570.

[18] Lagally, E. T., Emrich, A. C., Mathies, R. A., Lab. Chip 2001,1, 102–107.

[19] Liu, J., Enzelberger, M., Quake, S., Electrophoresis 2002, 23,1531–1536.

[20] Zou, Q., Sridhar, U., Chen, Y., Singh, J., Selvanayagam, Z. E.,Lim, T. M., Yan, T., Rodriguez, I., Lesaicherre, L. M., IEEEInternational Electron Devices Meeting, Washington, DC2001, 371–374.

[21] Gao, Q. F., Yeung, E. S., Anal. Chem. 1999, 70, 1382–1388.

[22] Han, F. T., Lillard, S. J., Anal. Chem. 2000, 72, 4073–4079.

[23] Han, F. T., Xue, J., Lin, B. C., Talanta 1998, 46, 735–742.