Phm

SPISFJHAs

1

TpclatssspjhmdompebuaiErm

1

J. Micro/Nanolith. MEMS MOEMS 9�1�, 013050 �Jan–Mar 2010�

J

arameters affecting electrospray performance ofot-embossed open-channel polymericrofluidic chips

ana Malahatio G. Iovenitti

gor Sbarskiwinburne University of Technologyaculty of Engineering and Industrial Sciencesohn Streetawthorn, Victoria 3122ustraliamalahat@swin.edu.au

Abstract. We investigate the effect of chip tip-to-counter electrodespacing �gap�, channel-size aspect ratio, onset voltage on electrosprayionization �ESI� performance of embossed polymer microfluidic chips.Planar polymer substrates were hot embossed using an electroformedtool and a laser machined tool. The embossed microchips were testedfor successful ESI demonstration with minimum sample consumptionand high throughput. Factors influencing the stability of electrospraywere analyzed by a series of experiments. Theoretically and experimen-tally, it was observed that the distance of the microchannel tip to thecounter electrode directly relates to the onset voltage applied. Further-more, the overall pattern observed is the decrease in Taylor cone sizewith a decrease in channel dimensions. The total ion current was ana-lyzed as a function of time and onset voltage at various gaps. Five elec-trospray modes, namely, drop mode, pulse mode, Taylor-cone jet mode,multijet mode, and oscillating jet mode recorded at the open channel tipswere identified and the electrospray cycle investigated. These results arein good accordance with theory, and comparisons are drawn with thefindings of other researchers. © 2010 Society of Photo-Optical Instrumentation En-gineers. �DOI: 10.1117/1.3309710�

Subject terms: electrospray; hot embossing; microfluidics; Taylor cone formation;polymers; electroforming process; laser machining process.

Paper 09093PRR received Apr. 30, 2009; revised manuscript received Dec. 10,2009; accepted for publication Dec. 17, 2009; published online Mar. 9, 2010.

Introduction

he electrospray ionization mass spectrometry �ESI-MS�rocess involves the emission of a pressurized liquid from aapillary at the input of a mass spectrometer. At the capil-ary exit, the liquid is subjected to an electrical potential,nd the high electrical field generated induces charges onhe surface of the liquid in the area of the spray tip. Thepraying of the fluid substance in the vicinity or area of thepray tip generally occurs when the coulombic forces aretrong enough to overcome the surface tension forcesresent in the liquid. This spray occurs in the form of a thinet of liquid at the tip of the Taylor cone.1 Electrospray �ES�as advantages over other atomization methods due toonodispersity of droplets sizes, flexibility to generate

roplet of different sizes, compatibility, flexibility, and easef operation.2 Because of its unique features, there areany applications of ES, which include nanomaterial

reparation,3 thin-film deposition,4 MS �MS�,5 drug deliv-ry, and inkjet printing. The ES for analytical studies haseen investigated for several decades.6 In 1917, Zeleny7

sed ES in industry for fuel or paint spray. In 1968, Dole etl.8 applied ES techniques to separate the large moleculesnto many small molecules. Yamashita and Fenn9 integratedS with MS �ESI-MS� technology since the ES can sepa-

ate samples without any damage to the samples. Further-ore, it was also to integrate with any other tool used in

932-5150/2010/$25.00 © 2010 SPIE

. Micro/Nanolith. MEMS MOEMS 013050-

Downloaded from SPIE Digital Library on 14 Jun 2010 to 1

sample analysis. In 1994, Wilm and Mann10 set up a math-ematical model for capillary ES. In 1997, Ramsey andRamsey11 published a paper on microelectrospray chips.Apart from using a capillary tube as a nozzle, they usedphotolithography fabrication techniques to etch a micro-channel on a glass substrate and the exit of microchannel issurface modified to become hydrophobic. In the same year,Desai et al.12 used microelectromechanical systems�MEMS� technology to fabricate the microelectrospraynozzle. This nozzle was microfabricated on a silicon sub-strate. Wang et al.13 used Parylene-C polymer to fabricatethe nozzle in 1999. This nozzle was flexible and would notbe deformed due to internal residual stress. Despite manyyears of study, ES is not fully understood, mainly due to thecomplicated nature of the spraying process.14 Grace andMarijnissen15 noted that there is not yet a complete theorythat fully describes electrostatic spraying as a function offluid and operational variables. These authors concludedthat further work is necessary, including elucidation of theeffect of space charges, quantification of the electric field inthe vicinity of the nozzle during the dynamic spraying pro-cesses, and development of dimensionless parameters thatmay be used to simplify understanding. Hence, electrostaticspraying of conductive liquids into a nonconductive me-dium remains a rich field for continued research.14 Much ofthe work has been aimed at developing stable spraying sys-tems for particular applications16,17 and understandingqualitatively the many modes of ES.18 Work by Tsouriset al.19 examined the effect of several controlling variables,

Jan–Mar 2010/Vol. 9�1�1

36.186.72.210. Terms of Use: http://spiedl.org/terms

iitTiecraMurtmrsctUtvintaEfiomfidsubnbssBiw

sdtflfvftbctbtdrimg

Malahat, Iovenitti, and Sbarski: Parameters affecting electrospray performance…

J

ncluding the geometry of the nozzle from which the fluids sprayed, the flow rate of the fluid, and the conductivity ofhe continuous phase. In their experimental investigations,souris et al.19,20 found that the key to electrostatic spraying

s the electric stress on the fluid-fluid interface. A stronglectric stress can render the interface unstable and, thus,an result in the emission of fine droplets or bubbles. It waseported that the electric stress on the fluid interface undern electric field plays a major role in electrostatic spraying.any groups have fabricated ESI interfaces on microchips

sing glass,11 silicon,21 and plastic22 as the substrate mate-ials. Previous methods such as attachment of a nanosprayip to the microchannel exit, has yielded promising experi-

ental results; however, extensive work is required to fab-icate and manufacture the chip. This approach, althoughuccessful, is not compatible with mass production pro-esses. It may also be difficult to avoid a dead volume athe interface between the chip and the spray capillary.23

ndesired wetting at the contact between the capillary andhe microchannel resulted in the formation of a liquid deadolume.24 The issues with conventional ESI-MS systemsnclude the contamination and clogging of the microchan-el tip due to the electrochemical process at the emitterips. Moreover, the difference in material used for capillaryttachments and microchips can vary the repeatability ofS demonstration. Furthermore, bubbles were sometimes

ormed at a high rate at the junction between the feed cap-llary and microchips, severely disturbing the ES.23 Inpen-channel flow, there is less susceptibility to bubble for-ation for liquid flow in the microchannels and during thelling of the microwells.25 However, the trade-off for thisesign is the enhanced evaporative loss of the fluidamples, especially for liquids with high vapor pressurender ambient conditions. The evaporative losses in closed/onded channel systems is lower; however, the microchan-els and reservoirs have a greater propensity to formubbles during filling. Bubble formation in these bondedystems can impede the capillary flow and can lead to cata-trophic failure for fluid actuation in these devices.25 Also,anerjee et al.25 verified from the simulations that the fill-

ng action is faster in open-channel systems and consistentith the analytical results.Extensive literature review showed that the aims of re-

earchers are to obtain more reproducible mass-fabricatedevices, to minimize system complexity along with reduc-ion in manufacturing cost and time, and to eliminate deaduid volume related to the external connections. Hence, toabricate an ESI-MS chip with single material with no deadolume is significant to the ESI field.26 The focus is onabrication of high-performance ESI-MS devices manufac-ured using mass production techniques, such as hot em-ossing and injection molding. The polymer microfluidichips are also easy to align to the orifice of the mass spec-rometer because less care is needed to avoid collision andreaking.23 Different improvements regarding the quality ofhe emitter tips are proposed in the literature. Reproducibleimensions can be achieved using an enhanced fabricationoute based on laser micromachining.27 More durable coat-ngs have also been reported.28 Most of these improve-

ents, however, do not solve the basic problems of clog-ing and the lack of reproducibility that affect the analysis

. Micro/Nanolith. MEMS MOEMS 013050-

Downloaded from SPIE Digital Library on 14 Jun 2010 to 1

conditions.29 Using microtechnology techniques, reproduc-ible and ready-to-use high-quality ES tips can be massmanufactured.29

In this work, disposable low-cost polymer microfluidicopen-channel tips were successfully produced using thehot-embossing technique on polystyrene �PS� and polycar-bonate �PC� substrates. The embossed chips were success-fully tested for ES demonstration with minimum sampleconsumption and high throughput as reported in our previ-ous study.30 Furthermore, the open structure limits the com-mon problems of air bubbles and clogging.29 This paperaims to extend research in this field by investigating thefactors influencing the performance of ES polymer micro-chips in terms of repeatability, practicality, onset voltages,and functionality. The microchip used in the investigationwas designed with minimum complexity to simplify fabri-cation by having an open reservoir and channel. The designalso eliminated dead fluid volume normally due to externalconnections to the chip by using capillary forces for fluidflow along the microchannel.

ES is very sensitive at the vicinity of the spray tip.Hence, the control of the ES mechanism and reproducibil-ity of ES results with stable and reliable performance hasnot been achieved.31 The primary interest to the ES MSpractitioners is to control the spray process in order to gainhigh and stable ion signals.32 Interesting phenomena takeplace when high voltage is applied to a conductivecapillary/needle containing a polar fluid.33 As voltage isincreased the liquid air interface becomes polarized with anexcess ionic charge producing a Taylor cone,34 as shown inFig. 1, which shows a schematic diagram of the ESI pro-cess and the circuit. The inset shows the Taylor coneformed in this work, at the microchip tip, where s denotesthe size of the Taylor cone. Droplets are emitted from thecapillary tip with a frequency increasing with voltage. At acertain voltage �onset voltage�, which depends on the diam-eter of the capillary, the distance to the counter electrodeand the frequency of droplet emission increases exponen-tially, and an ES is generated.33 The charged droplets emit-ted at the tip of Taylor cone undergo coulombic fissionsaround Rayleigh limit.35

The Taylor cone is a very dynamic system that deforms,oscillates, and contains convection flows of charge andmass.36 It was shown recently by Marginean et al.37 thatTaylor cone deformation plays a central role in the mecha-nism of electrostatic spraying. Furthermore, the onset volt-age required to start jet production is proportional to thesquare root of the surface tension multiplied by the radiusof the capillary orifice. The higher the onset voltage is, thegreater are the chances of an electrical discharge.36 Theliquid under the stable ES operation is often observed toform a conic meniscus at the outlet of the capillary, throughthe apex of which a fine liquid ligament is ejected. Thisliquid ligament breaks into droplets further downstream toform a dispersed spray. If the liquid is conductive, then theconic meniscus at the outlet of the capillary will deforminto a cone under the action of electric force.2 Taylor38

proved that the semivertical angle of the cone was49.3 deg, theoretically and experimentally, and this cone isnow called the Taylor cone. Furthermore, Taylor showedthat the instability of an elongated drop would not occurunless a pressure difference existed, and it was found that a

Jan–Mar 2010/Vol. 9�1�2

36.186.72.210. Terms of Use: http://spiedl.org/terms

diPeac

ccaecothfifpAcgacipdpaoC�stslzfssctati

Fs

Malahat, Iovenitti, and Sbarski: Parameters affecting electrospray performance…

J

rop, elongated by an electric field, becomes unstable whents length is 1.9 times its equatorial diameter. Cloupeau andrunet-Foch39 studied the droplet size distribution and theffects of flow rate and applied voltage on the ES by usinggranulometer. Later, Fernandez40 studied the effects of the

harged droplets on the Taylor cone.In this study, ES is produced at the hot-embossed micro-

hannel exit tip without a cover. The hot-embossed micro-hannel is attached to a hot-embossed open reservoir. Inddition, our work contributes to research on ES param-ters influencing the initiation and formation of the Taylorone shape and size for open-channel systems. Formationf the microchannel is important for delivery of the fluid tohe tip where the ES forms. The corner of the channel mayave significant contribution to the electric double-layereld, subsequently to the fluid flow field for ESormation.41 The microchannel must end in a sharp tip torovide a sufficiently high electrical field for producing ES.ny geometrical deformation leads to a nonoptimal Taylor

one configuration during ES.42 Shinohara et al.43 investi-ated the influence of hot-embossed bonded channel tipngle on Taylor cone formation. They stated that the suc-ess rate of Taylor cone formation increased with decreas-ng the tip angle ���. Arscott et al.44 concluded that theerformance of nib sources was seen to be linked to the slotimensions; the smaller slot dimension is, the better theerformance of SU-8 micronibs. While a high voltage ispplied, a Taylor cone could be formed at the tip.13 The sizef the cone depends on the geometry and wetted area.45

hiou et al.45 investigated the effect of the outside diametero.d.� of the tip on the size of the Taylor cone. The relation-hip between the o.d. of the capillary �D0� and the height ofhe Taylor cone was plotted and indicates that the geometrichape of the Taylor cone is a strong function of the D0. Thearger the capillary is, the bigger the size of the Taylorones. Kebarle46 proposed that the required ES voltage is aunction of the radius of spraying capillary as well as theurface tension of the spraying liquid. Using a sequence ofnapshots, Wang et al.13 showed that the size of the Taylorone is governed by the tip orifice geometry and the forma-ion of the cone is governed by electrical potential. Wang etl.13 suggested that the cone size and shape is also relatedo the flow rate and the distance between the tip and MSnlet.

ig. 1 Schematic diagram of ESI process and the circuit; inset showdenotes the size of the Taylor cone.

. Micro/Nanolith. MEMS MOEMS 013050-

Downloaded from SPIE Digital Library on 14 Jun 2010 to 1

Jaworek et al.47 classified the electrospray modes intothe following two groups:

1. Dripping modes, in which only fragments of liquidare ejected from the capillary outlet by the deforma-tion and detaching of the liquid meniscus. These frag-ments can be formed as large regular drops �drippingmode�, fine droplets �microdripping mode�, a singleor multiple spindles �spindle and multispindlemodes�, or irregular fragments of liquids.

2. Jetting modes by which the meniscus elongates into along fine jet. The jet can be smooth and stable �cone-jet mode� or can move in any regular way: rotatearound the capillary axis �precession mode� or oscil-late in a plane �oscillating mode�. Sometimes a fewjets on the circumference of the capillary can be ob-served �multijet mode�. The case when the jetbranches are known as a ramified jet.

Researchers have reported spray modes where the liquidmeniscus repeatedly forms a transient jet at constant fre-quency that depends on voltage.37,48 The jet formation andcollapse occurs on time scales of typically �1 ms; thoughthe time period decreases as the applied field increases.49

Sequential phenomena that occur when ES are triggered bysudden voltage steps were also reported by Paine.49 More-over, a number of authors50,51 have applied voltage pulsesof short duration �1–10 ms� to an ES source in an attemptto make ES drop on demand. A final point is the existenceof fluctuations in the ion signal even under good conditions,whose origin must lie in the solution parameters governingthe spray process.32 A conductive liquid at the tip of a wet-ted needle or a filled capillary will, as the voltage is in-creased, deform into a cone. Once a threshold voltage�called the “onset voltage”� is reached, the electric stress atthe apex of the liquid surface overcomes surface tensionand a spray of particles is ejected toward a counterelectrode.52 Krpouna and Shea52 presented a method toquantitatively determine the onset voltage of a single orarray of ES taking into account the real emitter-extractorgeometry and the surface tension of the liquid. Ieta et al.33

investigated the emission frequency of capillary droplets ofaqueous ESs near the onset voltage and stated that duringES, the droplets are emitted with a frequency increasing

aylor cone formed in this work at the microchip tip, where dimension

s the T

Jan–Mar 2010/Vol. 9�1�3

36.186.72.210. Terms of Use: http://spiedl.org/terms

wcdieioe

2

2Phtprsmciec2tottffl

aaocemfcnrmre

2Ftm

Fv�e

Malahat, Iovenitti, and Sbarski: Parameters affecting electrospray performance…

J

ith voltage, but it may often be difficult to determine pre-isely what is the onset voltage. The following sectionsescribes the fabrication of polymer microfludic chips, test-ng of microchips for ESI using the microdevice developed,valuation of performances of PS/PC chips for ESI, andnvestigation of influence of ES parameters, such as gap,nset voltage, and current on ES emitted from hot-mbossed polymer microfluidic chips.

Experimental Procedure

.1 Polymer Microfluidic ESI Tip Preparationlanar microfluidic ES tips were fabricated using polymerot embossing with electroformed and laser-machined mas-er tools on PS and PC substrates. For the hot-embossingrocess, the tool was designed with a microchannel andeservoir feature attached. The master tool is comprised ofix radially arranged tools on a single surface to produceultiple chips. Accordingly, six different dimension micro-

hips can be obtained in a single embossing run as reportedn our previous study.30 Figure 2 illustrates the scanninglectron microscope �SEM� images of a polystyrene micro-hip embossed with the laser-machined steel tool. Figure�a� shows the top view of embossed reservoir feature at-ached to the microchannel. Figure 2�b� shows the top viewf embossed polymer microchip tip, and Fig. 2�c� showshe side view of the embossed open-channel PS microchipip. The images show good surface finish and uniformlyormed microchannel and reservoir on the polymer micro-uidic substrates.

The embossed PS/PC substrates show smooth surfacesnd were tested for hydrophilicity by taking the contactngle measurements. The average contact angle measuredn the embossed surfaces was �40 to 42 deg, which indi-ates that the surface is moderately hydrophilic, and thembossed chips were tested for capillary action usingethanol/water �70 /30� as the test solution. Fluid is trans-

erred from reservoir to microchannel exit tip as a result ofapillary forces and electroosmosis, thus eliminating theeed for external pressure devices such as the use of sy-inge pumps, which are generally used for filling theicrochannel/capillary with the fluid sample in the systems

eported in the literature.34,53 The next step was to test thembossed chips for ES.

.2 ESI Experimentsor ES chip design, the critical points are the application of

he high voltage, ease of operation, and dead volume mini-ization. To simplify the experiment, the ESI interface de-

ig. 2 SEM images of hot-embossed PS microfluidic chip: �a� topiew of embossed reservoir feature attached to the microchannel,b� top view of embossed polymer microchip tip, and �c� side view ofmbossed open-channel PS microchip tip.

. Micro/Nanolith. MEMS MOEMS 013050-

Downloaded from SPIE Digital Library on 14 Jun 2010 to 1

veloped was operated at atmospheric pressure and roomtemperature. The design of chip and the ES device devel-oped simplifies the experiments by the elimination of theuse of syringe pumps, which was used in systems reportedin the literature. The use of syringe pumps was to force thefluid into the microchannel. This can lead to air-bubble for-mation in the channel, thereby leading to catastrophic fail-ure of the system. The trade-off for the open-channel sys-tems is the evaporation of the fluid in the microchannel.54

However, 1 �L of fluid would last for 5 min, as reportedearlier,55 and this is sufficient for MS/MS analysis. Further-more, the open-channel design of the microchips eliminatesthe use of processes such as bonding and gluing and theirimpurities.

In the experiment, the chip is stationary and positionedin front of a movable polished aluminium plate that acts asa counter electrode in the ESI circuit. The plate positionrelative distance to the microchip tip was adjustable by us-ing a digital vernier and viewed through a microscope. Fig-ure 3 illustrates the experimental setup, ESI circuit, andapparatus used for ES experiments. The inset shows an em-bossed polymer microchip set for ES testing with a positiveelectrode �stainless steel wire� inserted into the reservoirand the microchip tip positioned a few millimeters in frontof the counter electrode �polished aluminum plate� con-nected in circuit.

Prior to ES testing, the hot-embossed chips were cleanedin an ultrasonic bath using ethanol for 10 min. A high-voltage power supply was used to provide up to 5000 V dcfor the experiments. The positive terminal of high voltagewas connected to a stainless steel electrode inserted into thereservoir, and the negative was connected to the multim-eter, which was connected in series to the counter electrode�aluminum plate�. In the ES experiments, the first step wasto set the gap between the channel exit and the counterelectrode. Second, a fluid droplet was added to the reservoirby a micropipette and the high voltage was switched on.The fluid solution used for all the ES experiments wasmethanol/water �70 /30�. The solution travels from the res-ervoir to the channel tip by capillary action, and the voltageis increased in increments until ES is initiated at the tip ofthe channel. The video camera on the microscope recordedthe formation of the Taylor cone and ES. The onset andduration of the ES was observed by the measurement ofcurrent using a multimeter and the video recording. Fluidaccumulated at the aluminum plate as a result of the ESphenomenon. Next, the distance between the channel exitand counter electrode was changed and the same procedurewas carried out. A stainless steel electrode wire was se-lected to reduce contamination and inserted into the reser-voir of the chip.

Salata36 stated that the progress in ES technology wouldhave been impossible without a continuous development ofES apparatus. An example of a state-of-the-art research ap-paratus has been reported.36 The combination of hot-embossing fabrication and an open channel provide an im-proved approach to reduce impurities and simplifymanufacture. Fluid flow in the channel is critical for thechip’s performance and capillary action. Several micro-chips were fabricated and tested, and the ES experimental

Jan–Mar 2010/Vol. 9�1�4

36.186.72.210. Terms of Use: http://spiedl.org/terms

raa

2Edfs�dacE

3Tstacstmvg

3Tautsaet4o

FEt

Malahat, Iovenitti, and Sbarski: Parameters affecting electrospray performance…

J

esults reported. Stable and fine Taylor cones were recordedt the channel exits of hot-embossed PS and PC substrates,s described in Section 3.

.3 Optical ESI CharacterizationS behavior and Taylor cone geometry were observed un-er the microscope, with the corresponding total ion currentrom the emitter exit simultaneously measured. The fluidample used for all the experiments was methanol/water70 /30�. The polymer microchips were set for testing at theesired spacing from the counter electrode. The fluid wasdded into the reservoir of the microchip by using a mi-ropipette, and the high voltage was switched on. StableSs were recorded by using a video and microscope.

Results and Discussionhe results of this study are broadly classified into threeections investigating the effects of chip-to-counter elec-rode spacing �gap�, microchannel dimensions in terms ofspect ratio �depth-to-width ratio�, and length of micro-hannel on ESI performance of the microchips. The resultshow that the observed ESI performance follow expectedheoretical trends. The influence of onset voltage on ES

odes and the voltage-current characterization is also in-estigated. Furthermore, the output is compared to and is inood agreement with the findings of other researchers.

.1 Influence of Gaphe chip tip–to–counter electrode spacing �gap� was set todesired value. A fluid sample was added to the reservoir

sing a micropipette. The fluid travels from the reservoir tohe tip by capillary action, and the high voltage waswitched on. Figure 4 shows microscope time sequence im-ges of the top view of ES at the tips demonstrating anlongating Taylor cone �Figs. 4�a�–4�d��, jet emission ellip-ical cone �Fig. 4�e��, and cone-jet mode electrospray �Figs.�f� and 4�g�� at an onset voltage of 2000 V for a durationf �3 min at a chip-to-counter electrode spacing of

ig. 3 Experimental setup, ESI circuit, and apparatus used for electS testing with a positive electrode �stainless steel wire� inserted int

he counter electrode �polished aluminum plate� connected in circui

. Micro/Nanolith. MEMS MOEMS 013050-

Downloaded from SPIE Digital Library on 14 Jun 2010 to 1

1.2 mm. When the ESI voltage was applied and slowlyincreased, the Taylor cone elongated and formed into anelliptically shaped cone. As the voltage was further in-creased to an onset voltage, the ellipsoid spontaneouslychanged to a pointed cone and the ES started, as evidencedby the measurement of a jump in ion current between theESI chip and polished ground plate. Onset voltages of 2000and 2500 V were measured for 1.2 and 1.4 mm microchip-to–counter electrode spacing, respectively. As the counterelectrode spacing increased, the onset voltage increasedsince the electrical field exhibits a logarithmic dependenceon chip tip–plate distance.56 It was also observed and re-corded that as the applied voltage increased and the Taylorcone size decreased, with an associated decrease in jet di-ameter and jet breakup length. As observed in Fig. 4, thecones are well defined with virtually no lateral dispersionof the fluid at the microchannel exit. Compared to litera-ture, Wang et al.56 reported threshold �onset� voltages of

y experiments; inset shows an embossed polymer microchip set forservoir and the microchip tip positioned a few millimeters in front of

Fig. 4 Microscope sequential images of top view of ES demonstrat-ing: �a�–�d� an elongating Taylor cone, �e� jet emission ellipticalcone, and �f, g� cone-jet mode electrospray at an onset voltage of2000 V for a duration of �3 min at a chip-to–counter electrodespacing of 1.2 mm.

rosprao the ret.

Jan–Mar 2010/Vol. 9�1�5

36.186.72.210. Terms of Use: http://spiedl.org/terms

3e1ec

w

V

waooctae

V

TstPmmffbtuctfa

3Atdss

FvlEP

Malahat, Iovenitti, and Sbarski: Parameters affecting electrospray performance…

J

200 and 3520 V for 1.5 and 2 mm microchip-to–counterlectrode spacing and stated that the threshold voltage at.5 mm is smaller than that at 2 mm due to the strongerlectric field, and as expected, the threshold voltage in-reases with higher gaps.

A relation giving the dependence of onset voltage Vonas given by Smith,57 which is as follows:

on = A��2Trc cos �0

�0� ln

4d

rc�1�

here T is the surface tension of the fluid, �0 is the half-ngle of the liquid cone at the tip of the capillary, rc is theuter radius of the capillary, �0 is the electric permittivityf vacuum, d is the distance between the chip tip andounter electrode, and A is a proportionality constant. As inhese set of experiments, d is the variable, and onset volt-ge varies linearly with the logarithm of chip-to-counterlectrode spacing �gap�. Therefore,

on�d� = a ln d + b �2�

he experimental data are in agreement with relation �2�, ashown in Fig. 5, which shows the measured ES as a func-ion of onset voltage and gap for various chip prototypes.S and PC substrates were embossed using a laser-achined steel tool and an electroformed nickel tool. Sixicrochips were obtained in a single embossing run; there-

ore, more than 24 microchips were tested and evaluatedor ES demonstration. It was observed that as the distanceetween the channel exit and counter electrode increases,he onset voltage also increases. Wang et al.13 describedsing a sequence of snapshots of how the size of the Taylorone is governed by the tip orifice geometry and the forma-ion of the cone is governed by the electrical potential. Inact, the cone size and shape are also related to flow ratend the distance between the tip and MS inlet.

.2 Effect of Aspect Ratio (AR)R is defined as the ratio of depth of the microchannel to

he width of microchannel. Table 1 lists the microchannelimensions, which were embossed using a laser-machinedteel tool and an electroformed nickel shim on PS/PC sub-trates. These dimensions are reported in our previous

ig. 5 Measured ES as a function of onset voltage and gap forarious chip prototypes. The legends LMPS and LMPC denote

aser-machined tool embossed PS and PC microchips, respectively.PPS and EPPC stands for electroformed tool embossed PS andC chips.

. Micro/Nanolith. MEMS MOEMS 013050-

Downloaded from SPIE Digital Library on 14 Jun 2010 to 1

study.30 Table 1 also records the ES parameters for the PSchip formed using the laser-machined steel tool at a gap of1.2 mm from the counter electrode, and the images of theTaylor cones with brief comments. Similar results werealso observed for various PS/PC substrates formed usingthe laser-machined tool and the electroformed nickel shim.

The hot-embossed open channel attached to a hot-embossed open reservoir on PS and PC substrates was ex-perimentally observed to be hydrophilic because the fluid isdelivered by electroosmosis and capillary forces, and themicrochips can demonstrate good ESI performance at rela-tively low voltages. It was experimentally observed that foran open channel of smaller widths and depths, the narrowerthe channel is, the better the ES performance at low volt-ages is. The smaller the microchannel dimensions are, thehigher the spray performance is. For a wide channel, thefluid tends to accumulate at the channel tip before it issprayed. Hence, a bigger Taylor cone formation occurs be-fore initiation of ES. From Table 1, the pattern observed isthe decrease in Taylor cone size with a decrease in channeldimensions.

In this work, ES is produced at the open-channel poly-mer microfluidic chip designed with a microchannel exitangle of 90 deg. Our work is similar to Shiea and Yuan,58

who produced electrospray from open-channel exits fabri-cated by different tip angles. They suggested that there wasno significant difference detected in the ion signal stabilityor the onset voltage of the ES generated from an openchannel fabricated with exits sharpened at different tipangles �30, 60, 90, and 120 deg�. Shinohara et al.43 inves-tigated the influence of a hot-embossed bonded channel tip

Table 1 Electrospray experimental parameters recorded for variousmicrochannel dimensions.

Jan–Mar 2010/Vol. 9�1�6

36.186.72.210. Terms of Use: http://spiedl.org/terms

acipdpaoCtowTcKtfcswt1l�oti

attomvacwidctt

r

Fa

Malahat, Iovenitti, and Sbarski: Parameters affecting electrospray performance…

J

ngle on Taylor cone formation. They stated that the suc-ess rate of Taylor cone formation increased with decreas-ng the tip angle ���. Arscott et al.44 concluded that theerformance of nib sources was seen to be linked to the slotimensions; the smaller slot dimension is, the better theerformance of SU-8 micronibs is. While a high voltage ispplied, a Taylor cone could be formed at the tip.13 The sizef the cone depends on the geometry and wetted area.45

hiou et al.45 investigated the effect of the o.d. of the tip onhe size of the Taylor cone. The relationship between the.d. of the capillary �D0� and the height of the Taylor coneas plotted and indicates that the geometric shape of theaylor cone is a strong function of the D0. The larger theapillary is, the bigger the size of the Taylor zones is.ebarle46 proposed that the required ES voltage is a func-

ion of the radius of spraying capillary as well as the sur-ace tension of the spraying liquid. Furthermore, the fabri-ation and evaluation of microchips with covered andealed channels but with a coverless ES electrospray tipas presented by Svedberg et al.23 They concluded that the

ip length influenced the test results and stated that with a0 mM solution and 1.5-mm-long tip, myoglobin crystal-ized at the end of the tip causing total failure after

15 min of ES. With a 0.3-mm-long tip, no failure hadccurred after 110 min �11 refills� of spraying. Long openips can result in evaporation of the sample solution, caus-ng the substance to dry for low flow rates.23

In this study, the microchannel length was kept constants 12.5 mm. Figure 6�a� shows the measured ES as a func-ion of aspect ratio and onset voltage for various chip pro-otypes, and Fig. 6�b� shows the measured ES as a functionf aspect ratio and size of the Taylor cone �measured inicrons�. The common trend observed experimentally for

arious replicated microchips, embossed using the laser-chined and electroformed tool, was the decrease in Taylorone size from a wide to a narrow channel of minimumidth and depth. It was found the size of the Taylor cone

ncreases as the aspect ratio decreases. The onset voltageirectly relates to the chip dimensions. The narrower thehannel is, the lower the onset voltage required is for ini-iation of ES with minimum sample consumption and highhroughput.

In this work, ES from open-channel microchip tips ofelatively smaller dimensions are demonstrated in compari-

ig. 6 �a� Measured ES as a function of aspect ratio and onset volnd size of Taylor cone.

. Micro/Nanolith. MEMS MOEMS 013050-

Downloaded from SPIE Digital Library on 14 Jun 2010 to 1

son to the work of Shiea and Yuanet al.58 The advantage ofthe microdevice developed in this study lies in its simplic-ity. The combination of hot embossing and open-channeldesign leads to reduced complexity and simplifies themanufacturing process. Because it is an open reservoir, noair bubbles are formed in the liquid circuit. Higher spraycurrent and lower jet diameter indicate that the device canperform equivalent to nanospray emitters while using mi-croscale dimensions. This allows higher sample throughputand eliminates potential clogging problems inherent innanocapillaries.

3.3 Influence of Onset Voltage on ES ModesFive ES modes were identified and recorded at relativelylow voltages at open-channel hot-embossed polymer micro-fluidic tips, as shown in Fig. 7. Figure 7 shows images oftop view of ES at the polymer microfluidic open-channeltips demonstrating: Drop mode 1800 V, �Fig. 7�a�� cuspcone/pulse mode 1900 V, �Fig. 7�b��. Taylor cone-jetmode 2000 V onset voltage �Fig. 7�c��, multijet mode

r various chip prototypes and �b� measured ES as a function of AR

Fig. 7 Microscope images of top view of ES at the polymer micro-fluidic open-channel tips demonstrating: �a� Drop mode 1800 V,�b� cusp cone/pulse mode 1900 V, �c� Taylor cone-jet mode2000 V, �d� multijet mode 3000 V, and �e, f� oscillating jet mode3500 V.

tage fo

Jan–Mar 2010/Vol. 9�1�7

36.186.72.210. Terms of Use: http://spiedl.org/terms

�cfitcttcvvpfdcdc

omcweflost

3Esstttiocw

Kw

Fmr

Malahat, Iovenitti, and Sbarski: Parameters affecting electrospray performance…

J

3000 V �Fig. 7�d��, and oscillating jet mode 3500 VFigs. 7�e� and 7�f�� at respective voltages. For open-hannel ESI systems, these results are highly original andrst of its kind. The results are in accordance with the

heory. Leu and Teng6 used a fused silica glass-bonded mi-rochip and reported the ES mode photographed at tip ofhe, nozzle, namely dripping, pulsating, cone jet, and mul-ijet mode at voltages ranging from 6.5 to 9 kV. In theirase, the onset voltage was 6.5 kV. They stated that as theoltage increased, the Taylor cone shrank and when theoltage exceeded 9.0 kV, multijet mode �14.5 kV� ap-eared. The point of interest here is that the surface energyor the liquid near the tip of an open channel should beifferent from the surface energy near the tip of a closedhannel. Therefore, the onset voltage for these two differentesigns �multijet mode: 3 kV for open channel, 14.5 kV forlosed channel� is different.

In this work, it was experimentally observed that, as thenset voltage increased, the ES entered the Taylor cone-jetode and at a high voltage, multijet with small Taylor

ones, and nanospray emitted from the tip of Taylor coneere recorded. With a further increase in voltage, the ES

ntered in oscillating mode and the cycle stopped as theow rate decreased. However, on addition of the fluid andn application of high voltage, the cycle restarted. The nexttep was to evaluate the relationship between current I andhe voltage potential.

.3.1 Relation between current and voltageS was observed by the jump in the ion current. The mea-ured ES ion current was plotted as function of time, ashown in Fig. 8�a�. The stable total ion current was moni-ored for �2 min using a multimeter attached in series tohe counter electrode. Figure 8�b� shows the relation be-ween the ion current and the applied voltage. The currentncreased as the voltage increased and stable cone-jet ES isbserved with nearly the same ion current. The chip tip-to–ounter electrode spacing was 1.2 mm, and the fluid usedas methanol/water �70 /30�.The results reported in Fig. 8 are in accordance with

ameoka.1 The total ion current measured and recordedas also evaluated as a function of onset voltages at vari-

ig. 8 �a� Measured electrospray total ion current as a functionultimeter attached in series to the counter electrode and �b� meas

eservoir.

. Micro/Nanolith. MEMS MOEMS 013050-

Downloaded from SPIE Digital Library on 14 Jun 2010 to 1

ous gaps, as shown in Fig. 9. From Fig. 9, it can be con-cluded that for open-channel systems, the higher the chip-to–counter electrode spacing is, the higher the sprayvoltage and current are. If the counter electrode is placedclose to the chip tip, the ES can be initiated at low voltages.However, the ES plume was observed to be stable onlywhen there was sufficient flow at the channel tip.

The measured ES was also evaluated as a function ofonset voltage and size of the Taylor cone as shown in Fig.10. Figure 10�a� inset shows an image of the Taylor coneformed at the front of the tip, and Fig. 10�b� the relation-ship between the onset voltages versus the height of theTaylor cone. As the voltage increased, the height of theTaylor cone increased. Similar trends were reported by Linet al.26 They developed a constant-speed-pulling method tofabricate the ESI tip by pulling mixed PMMA glue using a30-�m stainless steel wire through the preformed microflu-idic channel.

In this work, the ES phenomena was initiated at a volt-age of 2000 V at the tip of a channel with dimension of100 �m width and 100 �m depth, and a chip-to–counterelectrode spacing of 1.2 mm. Kameoka1 concluded that thethickness of the emitter film did not influence the sprayingperformance and stated that the ESI activity was only de-

. The stable total ion current was monitored for �2 min using alectrospray total ion current as a function of applied potential to the

Fig. 9 Measured electrospray total ion current as a function of onsetvoltage for different gaps.

of timeured e

Jan–Mar 2010/Vol. 9�1�8

36.186.72.210. Terms of Use: http://spiedl.org/terms

pamnacdwpvchceoitmcbdcaceflTEdtpctst

3Advcw

Ffo

Malahat, Iovenitti, and Sbarski: Parameters affecting electrospray performance…

J

endent on the top of the two-dimensional tip. A crucialspect for successful ESI interface is the formation andaintenance of a stable ES. ESI stability depends on a

umber of factors, including flow rate, and surface tension,nd it is well known that a narrow or drawn capillary tipan be used to demonstrate stable ES.13 Furthermore, therops formed at an electrically connected nozzle or orificeill acquire a certain amount of induced charge, which de-ends on the imposed electric field strength/appliedoltage.59 Studies have also shown that the motion ofharge carriers, such as ions, in a fluid can cause electro-ydrodynamic �EHD� flow or Coulomb-drivenonvection14 flow during ES formation. Information aboutlectric charge �including free charge and induced charge�n the electroformed droplets and the electric-field intensitynside the experimental cell are also very important to studyhe EHD effects on the mechanism of drop formation and

otion during ES.60 Tsouris et al.14 studied the effect ofapillary-tip configuration and voltage on pressure appliedy the fluid on the grounded disk electrode, at an electrodeistance of 3 cm and conical-tip capillary. The pressure in-reases as the distance between the electrodes is decreased,nd it is approximately doubled when the distance is de-reased by a factor of 2. A graph showing the effect oflectrode distance and voltage on pressure applied by theuid on the grounded disk electrode was also presented.14

hey stated that the experimental results show that, due toHD flow, the pressure applied on the electrically groundedisk is higher for the conical-tip electrode, the shorter dis-ance between the electrodes, and the negative-polarity ap-lied voltage. In this work, ES was produced at the pointedhip tip with a tip angle of 90 deg. The pointed out that ESIip thus provides the higher local electric field needed forpraying of higher surface-tension liquids without any elec-ric breakdown effects.

.3.2 Voltage-current characterization in ES moderscott and Troadec measured the ionization current pro-uced during ES of a given test liquid using the current-oltage �I-V� technique. In terms of ES from a tip having aapillary-tube–based topology, the electric field at the tipas evaluated using61

ig. 10 �a� Inset shows an experimental image of the Taylor coneormed at the front of the tip and �b� the relationship between thenset voltages versus the size s of the Taylor cone.

. Micro/Nanolith. MEMS MOEMS 013050-

Downloaded from SPIE Digital Library on 14 Jun 2010 to 1

Etip = 2Vapp/rtip ln�4d/rtip� , �3�

where Vapp is the applied voltage, rtip is the radius of the tip,and d is the distance from the tip to the ground plane. Inaddition, the onset electric field �Eonset� for ES can be esti-mated using the following relationship:

Eonset = ��2� cos �/�0rtip� , �4�

where � is the surface tension of the liquid, � the half-angleof the Taylor cone, and �0 is the permittivity of free space.Thus, by combining Eqs. �3� and �4�, Eq. �5� was obtained,which estimates the ES onset voltage Vonset at a given outertip radius rtip,

61

Vonset � ln�4d/rtip���rtip� cos �/2�0� . �5�

Using the above model, Arscott and Troadec61 predictedan onset voltage of 100 V for a device having a tip radiuson the order of 150 nm. In this work, the experimentalonset voltage recorded is 2000 V for a microchip having achannel width and depth of 100 �m. Furthermore, Arscottand Troadec61 stated that, first, this model predicts for de-vices having submicron tip radii to demonstrate very lowESI onset voltages �200 V�, especially if the tip radius is onthe scale of a few tens or hundreds of nanometers. Second,it was noted that at low values of tip radius �e.g., 50 nm�,the distance between the tip and the ground plane has littleeffect on the onset voltage as the electric field at the tip isgoverned by the tip radius. They concluded that the mea-sured ES onset voltages for such small tip dimensions donot agree with this model because the tip-to-ground planedistance d has a greater effect in determining the value ofVonset. In this study, it is concluded that our experimentaldata are in quantitative agreement with the above model.The results also indicate that the lower the chip-to–counterelectrode spacing requires a lower onset potential, as shownin Fig. 5. The total ion current measured and recorded wasalso evaluated as a function of onset voltages at variouschip tip-to–counter electrode spacing �gaps� as shown inFig. 9, from which it can be concluded that for open-channel systems, the higher the chip-to–counter electrodespacing is, the higher the spray voltage and current are.

4 ConclusionsThe aim of the research was to investigate the factors in-fluencing the performance of ES polymer microchips interms of repeatability, practicality, onset voltages, and func-tionality. Microchips used in the investigation had an openreservoir with an open microchannel attached and were fab-ricated by the hot-embossing technique to avoid compli-cated fabrication procedures. The simple design eliminateddead fluid volume normally due to external connections tothe chip by using capillary forces for fluid flow along themicrochannel.

The open-channel chip design combined with the hot-embossing process is different from other attempts reportedin the literature. The ES performance of the emitter isevaluated with respect to its geometry, operating condi-tions, and ESI parameters, such as onset voltage and cur-rent. The results indicate that AR of the microfluidic chipsis one of the most important design factors that affects the

Jan–Mar 2010/Vol. 9�1�9

36.186.72.210. Terms of Use: http://spiedl.org/terms

EcttcTc

ATVva�t

R

1

1

Malahat, Iovenitti, and Sbarski: Parameters affecting electrospray performance…

J

S performance. An open-channel polymer microfluidichip with higher AR will result in lower sample consump-ion, a higher ES current, and requires lower onset poten-ial. The results also indicate that the lower the chip-to–ounter electrode spacing requires a lower onset potential.he correlation results are compared to the literature. Theonclusions can be summarized as follows:

1. As the distance between the channel exit tip andcounter electrode increases, the onset voltage also in-creases for open-channel ESI systems.

2. For open channels of smaller widths and depths, thenarrower the channel is, the better the ES perfor-mance is at low voltages.

3. The narrower the channel is, the lower the onset volt-age is required for initiation of ES with minimumsample consumption and high throughput.

4. Five ES modes, namely, drop mode, pulse mode,Taylor-cone jet mode, multijet mode, and oscillatingjet mode, recorded at the open-channel tips wereidentified, and the ES cycle was investigated alongwith the evaluation of ES parameters.

5. The measured ES ion current was evaluated as afunction of time and onset voltage at various chip-to–counter electrode spacing �gap�.

6. For open channel systems, the higher the chip-to–counter electrode spacing is, the higher the sprayvoltage and current are. As the voltage increases, thesize of the Taylor cone increases.

cknowledgmentshe authors thank MiniFAB �Aust� Pty. Ltd., Scoresby,ictoria, Australia, for help in hot embossing and for pro-iding the resources during the ES experiments. Thanks arelso extended to the Center for New ManufacturingCNM�, Swinburne Tafe, Victoria, Australia, for fabricatinghe laser-machined tool.

eferences

1. J. Kameoka, “Polymeric microfluidic and electrospray devices,”Ph.D. dissertation, Faculty of Graduate School, Cornell University,Michigan �2002�.

2. X. Chen, J. Cheng, and X. Yin, “Measurements of drop velocity inelectrospray by PIV,” in Optical Technology and Image Processingfor Fluids and Solids Diagnostics, G. X. Shen, S. S. Cha, F.-P.Chiang, and C. R. Mercer, Eds., pp. 181–187, SPIE Press, Belling-ham, WA �2002�.

3. J. P. Borra, D. Camelot, K. L. Chou, and P. J. Kooyman, “Bipolarcoagulation for powder production micro-mixing inside droplets,” J.Aerosol Sci. 30�7�, 945–958 �1999�.

4. G. K. Soo, J. Y. Kim, and H. J. Kim, “Deposition of MgO thin filmsby modified electrostatic spray pyrolysis method,” Thin Solid Films376, 110–114 �2000�.

5. J. B. Fenn and M. Mann, “Electrospray ionization for mass spectrom-etry of large biomolecules,” Science 246, 64–71 �1989�.

6. T. S. Leu and C.-H. Teng, “Design, fabrication and study of micro-electrospray chips,” Mater. Sci. Forum 505–507, 1249–1254 �2006�.

7. J. Zeleny, “Industry application of electrospray technology,” J. Phys.Rev. 10, 1–6 �1917�.

8. M. Dole, L. L. Mack, and R. L. Hines, “Molecular beams of macro-ions,” J. Chem. Phys. 49, 2240–2249 �1968�.

9. M. Yamashita and J. B. Fenn, “Electrospray ion source: anothervariation on the free-jet theme,” J. Phys. Chem. 88, 4451–4459�1984�.

0. M. S. Wilm and M. Mann, “Electrospray and Taylor-cone theory,Dole’s beam of macromolecules at last?,” Int. J. Mass Spectrom. IonProcess. 136, 167–180 �1994�.

1. R. S. Ramsey and J. M. Ramsey, “Generating electrospray from mi-crochip devices using electroosmotic pumping,” Anal. Chem. 69�6�,

. Micro/Nanolith. MEMS MOEMS 013050-1

Downloaded from SPIE Digital Library on 14 Jun 2010 to 1

1174–1178 �1997�.12. A. Desai, Y.-C. Tai, M. T. Davis, and T. D. Lee, “A MEMS electro-

spray nozzle for mass spectroscopy,” in Proc. of Int. Conf. on SolidState Sensors and Actuators, Transducers ’97, Chicago, pp. 927–930�1997�.

13. X.-Q. Wang, A. Desai, T. Yu-Chong, L. Licklider, and T. D. Lee,“Polymer-based electrospray chips for mass spectrometry,” in Proc.of 12th IEEE Int. Conf. on Micro Electro Mechanical Systems, pp.523–528, Orlando �1999�.

14. C. Tsouris, W.-T. Shin, and S. Yiacoumi, “Pumping, spraying, andmixing of fluids by electric fields,” Can. J. Chem. Eng. 76�3�, 589–599 �1998�.

15. J. M. Grace and J. C. M. Marijnissen, “A review of liquid atomizationby electrical means,” J. Aerosol Sci. 25, 1005–1019 �1994�.

16. M. Cloupeau, “Recipes for use of EHD spraying in cone-jet modeand notes on corona discharge effects,” J. Aerosol Sci. 25, 1143–1157�1994�.

17. K. Tang and A. Gomez, “Generation of monodispersed water dropletsfrom electrosprays in a corona-assisted cone-jet mode,” J. ColloidInterface Sci. 175, 326–332 �1995�.

18. M. Cloupeau and B. Prunet-Foch, “Electrohydrodynamic sprayingfunctioning modes: a critical review,” J. Aerosol Sci. 25, 1021–1036�1994�.

19. C. Tsouris, D. W. DePaoli, J. Q. Feng, and T. C. Scott, “An experi-mental investigation of electrostatic spraying of nonconductive fluidsinto conductive fluids,” Ind. Eng. Chem. Res. 34, 1394–1403 �1995�.

20. C. Tsouris, D. W. DePaoli, J. Q. Feng, O. A. Basaran, and T. C. Scott,“Electrostatic spraying of nonconductive fluids into conductive flu-ids,” AIChE J. 40, 1920–1923 �1994�.

21. L. Licklider, X.-Q. Wang, A. Desai, Y.-C. Tai, and T. D. Lee, “Amicromachined chip-based electrospray source for mass spectrom-etry,” Anal. Chem. 72�2�, 367–375 �2000�.

22. J. Wen, Y. Lin, F. Xiang, D. W. Matson, H. R. Udseth, and R. D.Smith, “Microfabricated isoelectric focusing device for direct electro-spray ionization-mass spectrometry,” Electrophoresis 21, 191–197�2000�.

23. M. Svedberg, M. Veszelei, J. Axelsson, M. Vangbo, and F. Nikolajeff,“Poly�dimethylsiloxane� microchip: microchannel with integratedopen electrospray tip,” Lab Chip 4, 322–327 �2004�.

24. V. Parmar, T. Redeby, W. v. d. Wijngaart, Å. Emmer, and G. Stemme,“Electro-osmotic flow through closed-open-closed microchannels: anapproach to hyphenation of capillary electrophoresis and MALDI,” inProc. of IEEE MEMS 2006, Istanbul, pp. 406–409 �2006�.

25. D. Banerjee, N. A. Amro, and J. Fragala, “Optimizing microfluidicink delivery for dip pen nanolithography,” Proc. SPIE 5345, 230–237�2004�.

26. C.-H. Lin, J.-T. Shiea, and Y.-L. Lin, “On-chip fabrication of none-dead-volume microtips for ESI-MS applications,” Mater. Sci. Forum21–123, 611–614 �2007�.

27. G. A. Valaskovic and F. W. Mclafferty, “Long-lived metallized tipsfor nanoliter electrospray mass spectrometry,” J. Am. Soc. MassSpectrom. 7, 1270–1272 �1996�.

28. M. S. Kriger, K. D. Cook, and R. S. Ramsey, “Durable gold-coatedfused silica capillaries for use in electrospray mass spectrometry,”Anal. Chem. 67, 385–389 �1995�.

29. S. LeGac and C. Rolando, “An open design microfabricated nib-likenanoelectrospray emitter tip on a conducting silicon substrate for theapplication of the ionization voltage,” J. Am. Soc. Mass Spectrom. 17,75–80 �2006�.

30. S. Malahat, P. G. Iovenitti, and I. Sbarski, “Electrospray from hotembossed polymer microfluidic chips formed using laser machinedand electroformed tools,” Proc. SPIE 7269, 726906 �2008�.

31. Y. Lin, J. Wen, X. Fan, D. W. Matson, and R. D. Smith, “Lasermicromachined isoelectric focusing device on polymer substrate forelectrospray mass spectrometry,” Proc. SPIE 3877, pp. 28–35 �1999�.

32. T. Dulcks and R. Juraschek, “Electrospray as an ionisation methodfor mass spectrometry,” J. Aerosol Sci. 30�7�, 927–943 �1999�.

33. A. Ieta, H. Gerlach, T. E. Doyle, A. Pallone, and R. Amundson,“Analysis of droplet formation near onset voltage of aqueous electro-sprays,” in Proc. of Electrostatics 2008, ESA Annual Meeting, pp.1–7 �2008�.

34. J. Kameoka, R. Orth, B. Ilic, D. Czaplewski, T. Wachs, and H. G.Craighead, “An electrospray ionization source for integration withmicrofluidics,” J. Anal. Chem. USSR 74�22�, 5897–5901 �2002�.

35. A. Gomez and K. Q. Tang, “Charge and fission of droplets in elec-trostatic sprays,” Phys. Fluids 6�1�, 404–414 �1994�.

36. O. V. Salata, “Tools of nanotechnology: electrospray,” Curr. Nanosci.1�1�, 25–33 �2005�.

37. I. Marginean, L. Parvin, L. Heffernan, and A. Vertes, “Flexing theelectrified meniscus: the birth of a jet in electrosprays,” Anal. Chem.76, 4202 �2004�.

38. S. G. Taylor, “Disintegration of water drops in an electric field,”Proc. R. Soc. London, Ser. A 280, 383–397 �1964�.

39. M. Cloupeau and B. Prunet-Foch, “Electrostatic spraying of liquids incone-jet mode,” J. Electrost. 22, 135–159 �1989�.

Jan–Mar 2010/Vol. 9�1�0

36.186.72.210. Terms of Use: http://spiedl.org/terms

4

4

4

4

4

4

4

4

4

4

5

5

5

5

5

5

5

5

5

5

6

Malahat, Iovenitti, and Sbarski: Parameters affecting electrospray performance…

J

0. D. L. M. Fernandez, “The effect of charge emission from electrifiedliquid cones,” J. Fluid Mech. 243, 561–571 �1992�.

1. C. Yang and D. Li, “Analysis of electrokinetic effects on the liquidflow in rectangular microchannels,” Colloids Surf., A 143, 339–353�1998�.

2. H. Becker and L. E. Locascio, “Polymer microfluidic devices,” Tal-anta 56, 267–287 �2002�.

3. H. Shinohara, T. Suzuki, F. Kitagawa, J. Mizunoa, K. Otsuka, and S.Shoji, “Polymer microchip integrated with nano-electrospray tip forelectrophoresis-mass spectrometry,” Sens. Actuators B 132, 368–373�2008�.

4. S. Arscott, S. LeGac, C. Druon, P. Tabourier, and C. Rolando, “Aplanar on-chip micro-nib interface for NanoESI-MS microfluidic ap-plications,” J. Micromech. Microeng. 14�2�, 310–316 �2004�.

5. C.-H. Chiou, G.-B. Lee, H.-T. Hsu, P.-W. Chen, and P.-C. Liao, “Mi-cro devices integrated with microchannels and electrospray nozzlesusing PDMS casting techniques,” Sens. Actuators B 86, 280–286�2002�.

6. P. Kebarle, “A brief overview of the present status of the mechanismsinvolved in electrospray mass spectrometry,” J. Mass Spectrom.35�7�, 804–817 �2000�.

7. A. Jaworek, A. Krupa, A. T. Sobczyk, M. Lackowski, E. T. Czech1,S. Ramakrishna, S. Sundarrajan, and D. Pliszka, “Electrospray nano-coating of microfibres,” Mater. Sci. Forum 140, 127–132 �2008�.

8. M. Alexander, M. Paine, and J. Stark, “Pulsation modes and theeffect of applied voltage on current and flowrate in nanoelectro-spray,” J. Anal. Chem. USSR 78, 2658–2664 �2006�.

9. M. D. Paine, “Transient electrospray behaviour following high volt-age switching,” Microfluid. Nanofluid. 6�6�, 775–783 �2008�.

0. R. Moerman, J. Frank, C. Marijnissen, T. Schalkhammer, and V. D.GWK, “Miniaturized electrospraying as a technique for the produc-tion of microarrays of reproducible micrometer-sized protein spots,”Anal. Chem. 73, 2183–2189 �2001�.

1. O. Yogi, T. Kawakami, M. Yamauchi, J. Ye, and M. Ishikawa, “On-demand droplet spotter for preparing pico to femtoliter droplets onsurfaces,” Anal. Chem. 73, 1876–1902 �2001�.

2. R. Krpouna and H. R. Shea, “A method to determine the onset volt-age of single and arrays of electrospray emitters,” J. Appl. Phys. 104,064511–064518 �2008�.

3. S. Ssenyange, J. Taylor, D. J. Harrison, and M. T. McDermott, “Aglassy carbon microfluidic device for electrospray mass spectrom-etry,” Anal. Chem. 76�8�, 2393–2397 �2004�.

4. D. Banerjee, N. A. Amro, S. Disawal, and J. Fragala, “Optimizingmicrofluidic ink delivery for dip pen nanolithography,” J. Microli-thogr., Microfabr., Microsyst. 4�2�, 023014 �2005�.

5. S. Malahat, P. Iovenitti, and I. Sbarski, “Design and fabrication ofpolymer microfluidic chip for ESI-MS,” Proc. SPIE 6799, 679910�2007�.

6. Y.-X. Wang, J. W. Cooper, C. S. Lee, and D. L. DeVoe, “Efficientelectrospray ionization from polymer microchannels using integratedhydrophobic membranes,” Lab Chip 4, 363—367 �2004�.

7. D. P. H. Smith, “The electrohydrodynamic atomization of liquids,”IEEE Trans. Ind. Appl. IA-22�3�, 527–535 �1986�.

8. J. Shiea and C. H. Yuan, “Sequential electrospray analysis usingsharp-tip channels fabricated on a plastic chip,” Anal. Chem. 73�6�1080–1083 �2001�.

9. W. He, M. H. I. Baird, and J. S. Chang, “The effect of electric field onmass transfer from drops dispersed in a viscous liquid,” Can. J.Chem. Eng. 71, 366–376 �1993�.

0. W. He, M. H. I. Baird, and J. S. Chang, “The effect of electric field ondroplet formation and motion in a viscous liquid,” Can. J. Chem.Eng. 69, 1174–1183 �1991�.

. Micro/Nanolith. MEMS MOEMS 013050-1

Downloaded from SPIE Digital Library on 14 Jun 2010 to 1

61. S. Arscott and D. Troadec, “A nanofluidic emitter tip obtained byfocused ion beam nanofabrication,” Nanotechnology 16�10�, 2295–2302 �2005�.

Sana Malahat received her BS in mechani-cal �production� engineering from DeccanCollege of Engineering and Technology, Os-mania University, Hyderabad, India, withfirst class distinction. Before moving to Mel-bourne, Australia, she had worked as aCAD engineer and as a lecturer in produc-tion engineering in India. She is currentlystudying for her PhD at Swinburne Univer-sity of Technology, Melbourne, under thesupervision of A/Prof. Pio Iovenitti and Dr.

Igor Sbarski. Her active research interests are CAD/CAM/CAE, de-sign and fabrication of a polymer microfluidic chip for ESI, finiteelement modeling, and simulation of the hot-embossing process.She is currently employed as a sessional staff in the MechanicalEngineering Department, Swinburne Tafe, FEIS, Swinburne Univer-sity of Technology.

Pio G. Iovenitti graduated in mechanicalengineering from RMIT University and sub-sequently received a MEngSc from MonashUniversity. He was later awarded a PhDfrom Swinburne University of Technology.He has industry experience in the appliancemanufacturing sector where he was in-volved in design, project management, andR&D. He joined Swinburne University ofTechnology in 1987, and has taught under-graduate and postgraduate students CAD/

CAM/CAE and advanced technology, and has also researchedthose areas. He has been involved in research in MEMS and mi-crofluidics since 1999, and was project leader of the microfluidicsproject in the CRC for Microtechnology. Past related research in-cludes the investigation of microstructures for improved passivemixing in microfluidic devices and CAD/CAM software for excimerlaser micromachining microstructures on polymers.

Igor Sbarski has more than 30 years of re-search and industrial experience in thefields of development of novel plastics,composites, polymer processing, and recy-cling of plastics. He is currently a senior lec-turer at Swinburne University of Technology.He has established a laboratory for pro-cessing and testing plastics and compositesand is currently supervising a number ofPhD and Master’s research students. Dr.Sbarski authored over 60 journal and con-

ference papers in his area of expertise.

Jan–Mar 2010/Vol. 9�1�1

36.186.72.210. Terms of Use: http://spiedl.org/terms