Development of a new hybrid technique for rapid speciation analysis by directly interfacing a microfluidic chip-based capillary electrophoresis system to atomic fluorescence spectrometry

Feng Li1

Dong-Dong Wang1

Xiu-Ping Yan1

Jin-Ming Lin2

Rong-Guo Su2

1Research Center for AnalyticalSciences,College of Chemistry,Nankai University,Tianjin, China

2Research Center forEco-Environmental Sciences,Academic Sinica,Beijing, China

Development of a new hybrid technique for rapidspeciation analysis by directly interfacing amicrofluidic chip-based capillary electrophoresissystem to atomic fluorescence spectrometry

This paper represents the first study on direct interfacing of microfluidic chip-basedcapillary electrophoresis (chip-CE) to a sensitive and selective detector, atomic fluo-rescence spectrometry (AFS) for rapid speciation analysis. A volatile species genera-tion technique was employed to convert the analytes from the chip-CE effluent intotheir respective volatile species. To facilitate the chip-CE effluent delivery and to pro-vide the necessary medium for subsequent volatile species generation, diluted HClsolution was introduced on the chip as the makeup solution. The chip-CE-AFS inter-face was constructed on the basis of a concentric “tube-in-tube” design for intro-ducing a KBH4 solution around the chip effluent as sheath flow and reductant forvolatile species generation as well. The generated volatile species resulting from thereaction of the chip-CE effluent and the sheath flow were separated from the reactionmixture in a gas-liquid separator and swept into the AFS atomizer by an argon flow forAFS determination. Inorganic mercury (Hg(II)) and methylmercury (MeHg(I)) were cho-sen as the targets to demonstrate the performance of the present technique. Bothmercury species were separated as their cysteine complexes within 64 s. The precision(relative standard deviation, RSD, n = 5) of migration time, peak area, and peak heightfor 2 mg?L21 Hg(II) and 4 mg?L21 MeHg(I) (as Hg) ranged from 0.7 to 0.9%, 2.1 to 2.9%,and 1.5 to 1.8%, respectively. The detection limit was 53 and 161 mg?L21 (as Hg) forHg(II) and MeHg(I), respectively. The recoveries of the spikes of mercury species in fourlocally collected water samples ranged from 92 to 108%.

Keywords: Atomic fluorescence spectrometry / Hyphenated technique / Mercury speciation /Microchip capillary electrophoresis / Miniaturization DOI 10.1002/elps.200410382

1 Introduction

The rapid acceptance of microfluidic chip-based capillaryelectrophoresis (chip-CE) system in separation sciencehas been motivated by a number of fundamental perfor-mance gains [1–4]. These include improved efficiencywith respect to sample size, response time, analyticalperformance, integration, and throughput. Of these, per-haps one of the most compelling advantages of the chip-CE system when compared to conventional counterparts,

high-performance liquid chromatography (HPLC) andconventional capillary electrophoresis (CE), is the dra-matic reduction in analysis time [2, 5, 6]. With the reduc-tion of volume of the manipulated flow, the separation inchip-CE can be completed in a faster mode with highefficiency. Besides its great advantages, the reduceddimensions of the chip resulted in detection problems.The small volumes (typically , nL) encountered in mostchip-CE systems have led to the development of a diver-sity of detection methods that provide high sensitivity andlow detection limits [7–19]. Due to the short optical length,the detection methodology via absorbance commonlyused in HPLC and conventional CE was not suitable [7]. Inthe past decade, laser-induced fluorescence (LIF) hasdominated the detection for chip-based devices [5, 6, 8–10]. However, the LIF technique suffered from significantdrawbacks, which include relatively high instrumentalcosts and the fact that the majority of molecular speciesdo not fluoresce or are not easily converted to fluorescentspecies, which prohibited its universal use. Recently,electrospray ionization-mass spectrometry, which could

Correspondence: Dr. Xiu-Ping Yan, Research Center for Analyti-cal Sciences, College of Chemistry, Nankai University, Tianjin,ChinaE-mail: [email protected]: 186-22-23504605Dr. Jin-Ming Lin, Research Center for Eco-EnvironmentalSciences, Academic Sinica, Beijing, ChinaE-mail: [email protected]: 186-10-62841953

Abbreviations: AFS, atomic fluorescence spectrometry; GLS,gas liquid separator; ICP, inductively coupled plasma

Electrophoresis 2005, 26, 2261–2268 2261

2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Min

iatu

riza

tio

n

2262 F. Li et al. Electrophoresis 2005, 26, 2261–2268

provide structural information and low detection limits,has received considerable attention [11–18]. With thedevelopment of a nanoelectrospray ionization source(nano-ESI), the chip-based mass spectrometry techniquehas shown great potential in fields such as DNA analysis,drug discovery, pharmaceutical screening, and medicaldiagnostics [11, 12, 14–17], where high throughput andstructure elucidation is highly desirable or even neces-sary. Inductively coupled plasma-mass spectrometry(ICP-MS) has also been coupled to chip-CE and appliedto speciation [19]. Undoubtedly, MS as a powerful detec-tion methodology will become of growing importance inchip-based analysis systems.

Atomic fluorescence spectrometry (AFS) is a very sensi-tive and selective method for the determination of anumber of environmentally and biomedically importantelements, such as mercury, arsenic, and selenium [20,21]. While exhibiting similar sensitivity to ICP-MS, AFSalso presents the benefits of lower instrument and runningcosts, shorter warm-up time prior to analysis, easyhandling [21], and the possibility of portability. In principle,the AFS is an ideal detector for a chip-based analysissystem and the marriage of microchip to AFS would be apromising technique for speciation analysis. However, toour knowledge, no work dealing with the combination ofchip-CE with AFS has been reported so far.

In the present work, an attempt was made to interfacechip-CE to AFS. Two environmental and toxicologicallyimportant species, inorganic mercury (Hg(II)) and methyl-mercury (MeHg(I)), were chosen as targets to demon-strate the performance of the developed hybrid tech-nique. Both species were baseline-separated within 64 sand excellent repeatability was achieved. The chip layoutdesign, consideration of the interface, optimization ofchip-CE separation, and analytical figures of merit of thenew hybrid technique were described and discussed.

2 Materials and methods

2.1 Chemicals

All of the reagents employed were at least of analyticalgrade. Purified water (18.2 MO?cm21), obtained from aWaterPro water purification system (Labconco, KansasCity, MO, USA), was used throughout this work. Boricacid (Beijing Chemicals, Beijing, China) and methanolwere used to prepare the electrolyte buffer. The pH of thebuffer solution was adjusted with 0.5 mol?L21 NaOH. Thebuffer was filtered with 0.45 mm Supor filters (GelmanSciences, Ann Arbor, MI, USA) prior to use. The stocksolution of 1000 mg?L21 Hg(II) was prepared by dissolvingmercury chloride (Alfar Aesar) in purified water directly.

MeHg(I) stock solution (1000 mg?L21, as Hg) was pre-pared by dissolving methylmercury chloride (Alfar Aesar)in suitable amount of methanol along with a subsequentdilution with purified water. Aqueous L-cysteine hydro-chloride 0.05% m/v (Sigma, St. Louis, MO, USA) wasused as chelating agent. All the solutions were stored at47C and protected against light.

A 0.22% m/v KBH4 solution was prepared by dissolvingKBH4 in 0.1% m/v KOH solution as the reductant. 1.1%v/v HCl was used as both the carrier of the chip-CE efflu-ent and the medium for subsequent volatile species gen-eration. River water and lake water samples were locallycollected. Immediately after sampling, the samples werefiltered through 0.45 mm Supor filters (Gelman Sciences)and analyzed.

2.2 Instrumentation

A model XCDY high-voltage power supply (ShandongChemical Engineering Institute, Jinan, China) and a ModelXGY-1011A nondispersive atomic fluorescence spec-trometer (Institute of Geophysical and GeochemicalExploration, Langfang, China) were employed through-out. A high-intensity mercury hollow cathode lamp (Ning-qiang Light Sources, Hengshui, China) was used asradiation source. The original quartz tube atomizer (7 mmID, 14 cm long) was replaced by a laboratory-madequartz tube (4 mm ID, 14 cm long), into which the volatilespecies and the hydrogen evolved from the reactor wereswept by an argon flow. A laboratory-made gas-liquidseparator (GLS), as described previously [22], was usedto isolate the gas from liquid. The argon flow was con-trolled by a rotameter. A Chromatographic Workstation(Nanjing Qianpu Software, Nanjing, China) was used fordata acquisition and data treatment.

2.3 Chip design

The schematic layout of the microchip is given in Fig. 1.This borosilicate glass microchip was fabricated bystandard photolithography, wet chemical etching, andheat bonding technology [23]. Reservoirs 1, 2, and 3(3.0 mm diameter) were for sample, waste, and buffer,respectively. In order to increase the capacity of thereservoirs, pieces of glass duct were attached to reser-voirs 1, 2, and 3. Reservoir 4 was designed as the inlet forthe makeup solution (1.1% v/v HCl), which met the sam-ple from separation channel at junction 5 and carried thesample through the chip exit port 6 to the chip-CE-AFSinterface. Reservoir 4 was epoxy-sealed with a siliconepad, into which a Teflon tube and a platinum electrodewere inserted for the makeup solution and electrical con-


Electrophoresis 2005, 26, 2261–2268 New hybrid technique for rapid speciation analysis 2263

Figure 1. Layout of the CE chip. (1) sample reservoir; (2)sample waste reservoir; (3) buffer reservoir; (4) make-upsolution reservoir; (5) junction of separation channel andmake-up solution channel; (6) chip exit port.

tact, respectively. The dimensions of the separation andinjection channels were 39 mm deep and 118 mm wide.The makeup solution channel (from 4 to 5, Fig. 1) wasenlarged to 350 mm deep and 900 mm wide, allowing effi-cient delivery of the separation channel effluent withoutsignificant dilution of the sample. The channel from 5 to 6has the same dimension as that of the makeup solutionchannel. The double tee injection cross was 200 mm long.The channel lengths from reservoir 1, 2, and 3 to theinjection cross were 1.0, 1.0, and 1.0 cm, respectively.The separation channel from injection cross to junction 5was 9.0 cm long while junction 5 was 0.5 cm away fromthe chip exit port 6.

2.4 Chip-CE-AFS interface

The efficiency of the chip-CE separation should not beimpaired by the interface. The sample must be trans-ported to the detector as efficiently as possible in a formthat can be read by the AFS. In this work, volatile speciesgeneration technique was used to convert the analyte intoits volatile species for AFS detection. As can be seen inFig. 2, a concentric “tube-in-tube” design was employedto couple the chip-based separation system with the AFS.A 100 mL of Eppendorf pipette (intermediate tube) was cutinto an appropriate length (25 mm), into which a piece ofTeflon tubing (0.5 mm ID, inner tube) was inserted. Oneend of the Teflon tubing was inserted into the chip exitport 6, which was manually enlarged to accommodate theTeflon tubing, and epoxy-sealed on the side wall of thechip. The big end of the pipette was plugged with an ap-propriate silicone gasket, into which the other end of the

Figure 2. Schematic diagram of the chip-CE-AFS inter-face (not to scale).

Teflon tubing was inserted until the end of the Teflon tub-ing reached a position approximately 0.1–0.2 mm to thesmall end of the pipette. The small end of the pipette wastightly inserted into a thicker wall silicone tube (outertube). Two holes were opened at appropriate places onthe walls of the intermediate tube and the outer tube tointroduce the KBH4 solution and argon flow into the inter-face, respectively. The chip effluent (the mixture ofseparation channel effluent and the makeup solution HCl)merged with KBH4 solution at the exit of the inner tube.The reaction mixture was swept by an argon flow into theGLS where the volatile species were separated from themixture and detected by AFS. The makeup solution andthe reducing agent KBH4 were supplied separately by thetwo peristaltic pumps of a model FIA-3100 flow injectionanalyzer (Vital Instrumental, Beijing, China). Obviously,the KBH4 solution and argon flow forced the chip effluentinto the GLS, so the effect of the backpressure due tohydrodynamic effect on the separation was minimized.

2.5 Chip-CE operation

The electrophoresis buffer consisted of 30 mmol?L21

boric acid and 8% v/v methanol (pH 9.36). Electrophore-sis separation was carried out in uncoated channels thathad been flushed with 0.1 mol?L21 NaOH for 30 min, withpurified water for 10 min, and finally with the buffer for40 min. No reconditioning was necessary between runs.Sample introduction was achieved by electrokineticinjection. For this purpose, 1000 V and 700 V were appliedto the sample reservoir and the buffer reservoir, while thesample waste reservoir was grounded and the makeupsolution reservoir floated. For chip-CE separation, 3000 Vwas applied to the buffer reservoir whereas the makeupsolution reservoir was grounded. 2100 V (70% of theseparation voltage) was applied to both the sample andwaste reservoirs to avoid the leakage of analytes from thesample channel to the separation channel. The experi-mental conditions used for the volatile species genera-tion, AFS, and chip-CE are summarized in Table 1.



Table 1. Operating parameters of chip-CE and AFS and volatile species generation

Parameter Setting

Chip-based capillary electrophoresis

Separation length 90 mmElectrolyte buffer 30 mM boric acid 1 8% v/v methanol (pH 9.36)Injection voltage Reservoir 1, 2, 3, 4: 1000 V, 0 V, 700 V, floatedInjection time 50 sSeparation voltage Reservoir 1, 2, 3, 4: 2100 V, 2100 V, 3000 V, 0 V

Atomic fluorescence spectrometer

Mercury hollow cathode lamp 60 mA (primary current)Quartz furnace temperature 1007CQuartz furnace height 6 mmNegative high voltage of photomultiplier 2270 V

Volatile species generation

Flow rate of 1.1% v/v HCl 0.35 mL?min21

Flow rate of 0.22% m/v KBH4 0.48 mL?min21

Carrier gas (argon) flow rate 200 mL?min21

3 Results and discussion

3.1 Chip layout

The schematic diagram of the microfluidic chip is illustrat-ed in Fig. 1. The double tee injection mode was employedhere to enable larger sample volume introduction than thatof commonly used cross injection mode. This geometryallows for high-efficiency sample injection and geometricdefinition of sample plug size [24, 25]. To enhance theseparation length, turns were introduced into the separa-tion channel. The elongation of the band, which resultedfrom the “effect of turn”, was reduced by adopting com-plementary pairs of turns and by increasing the turn radiusof curvature (5.0 mm diameter) [26]. Due to the flowincompatibility between chip and AFS, the makeup solu-tion of HClwas used to facilitate the sample transportation.The makeup solution channel was one quarter of a circle(5.0 mm diameter), and the separation channel was tan-gent to the makeup solution channel, so that the flow of themakeup solution was in the same direction as that of theelectroosmotic flow (EOF) in the separation channel andthe hydrodynamic effect of the makeup solution flow on theEOF was decreased. The channel dimension betweenjunction 5 and 6 was designed to have significantly lowerresistance to flow than the separation channel and also toreduce the negative effect of makeup solution on EOF.

3.2 Consideration of the chip-CE-AFS interface

The schematic setup for the chip-CE-AFS interface isshown in Fig. 2. To minimize the diffusion and mixing ofthe volatile species in the GLS and the quartz tube ato-

mizer, the volume of the GLS and the atomizer aredesigned as small as possible to get sharper peaks andhigher efficiency. The potential negative effect on theseparation due to the backpressure produced by thegenerated hydrogen and hydrodynamic effect in the rela-tively small interface must be minimized. In the presentdesign, HCl and KBH4 solutions merged at the exit of theintermediate tube. The generated hydrogen releasedimmediately forward to the outer tube, which had a largerinner diameter, so that the negative effect of the back-pressure on the separation was minimized. The positionof the outlet of the inner tube was found to be critical forthe chip-CE-AFS interface. When the outlet of the innertube was placed beyond the intermediate tube, poorpeak shape and repeatability were observed. The opti-mum position of the outlet of the inner tube located just0.1–0.2 mm inside of the intermediate tube.

The hydrostatic pressure resulting from the elevation dif-ference between the liquid surface in the reservoirs andthe chip exit would also exert an effect on the separation.The liquid surface in the reservoirs was higher (,3 mm)than that in the chip exit, so that the sample had a poten-tial to flow to the chip exit even when there was no elec-trical field. This hydrostatic pressure would shorten theanalysis time, but might impair the separation efficiency.To offset the effect of the hydrostatic pressure, the eleva-tion of the GLS was adjusted to where the liquid surface inthe sidearm of the GLS was at the same level with that inthe reservoirs. Compared with other chip-based systems[19, 27], the reservoirs in this work were not made air-tightso that the liquid surface was easy to adjust, and this wasvital for good repeatability.



3.3 Factors affecting chip-CE separation

To demonstrate the performance of the present chip-CE-AFS hybrid system, two environmentally and tox-icologically important species, inorganic mercury (Hg(II))and methylmercury (MeHg(I)), were chosen as targets.The optimized parameters affecting the chip-CE separa-tion of the mercury species include the buffer pH, theconcentration of methanol and boric acid in the buffersolution, and the applied voltage.

The pH of the buffer solution is one of the key parameters.Both mercury species could be baseline-separated in thepH range of 9.20–9.50 of the buffer solution (30 mmol?L21

boric acid and 8% v/v methanol), but the best separationefficiency was achieved at a pH of 9.36.

The effect of the concentration of boric acid on theseparation was tested from 10 to 50 mmol?L21 at pH 9.36.It was found that both species were baseline-separatedover the studied concentration range. The migration timeof mercury species along with electric current incre-mented as boric concentration increased. However, therepeatability was somewhat poor at lower boric con-centrations (, 20 mmol?L21). Therefore, a concentrationof 30 mmol?L21 boric acid was used to ensure a reason-able electric current and good repeatability.

Methanol, used as organic modifier inconventional CE, canimprove the resolution by decreasing the EOF. A range of 0–15% v/v methanol was tested. The resolution wasimproved with the increase of methanol concentration,whereas high methanol concentrations (.8% v/v) gavepoor peak shapes. Accordingly, 8% v/v of methanol wasincluded in the buffer solution to improve the separation ofthe mercury species.

The influence of the applied voltage on the separation ofthe mercury species was investigated from 1000 to5000 V. As expected, by increasing the applied voltage,the migration time and half-peak width dramaticallydecreased. Both species were baseline-separated be-tween 2000 and 3500 V. To ensure a good resolution, anapplied voltage of 3000 V was chosen for the separation.A typical electropherogram under 3000 V is shown inFig. 3. The separation efficiency was somewhat lower dueto the inevitable diffusion and mixing of the generatedvolatile species in the GLS and the atomizer.

3.4 Concentration and flow rate of the make-upliquid

In the current chip-CE-AFS system, a make-up liquidshould be employed not only to complete the electro-phoresis circuit and to facilitate the transportation of the

Figure 3. Electropherogram of 4 mg?L21 MeHg(I) and2 mg?L21 Hg(II). All other conditions as in Table 1.

effluent from the separation channel, but also to provide afavorable medium for the ensuing volatile species gen-eration. To this end, diluted HCl solution was chosen asthe make-up liquid. The effect of HCl concentration on thesignals of the mercury species was investigated at a flowrate of 0.35 mL?min21. The signal intensities of the twomercury species increased to a maximum as HCl con-centration incremented from 0.4 to 1.1% v/v, butdecreased with further increasing HCl concentration from1.1 to 5% v/v. Therefore, a 1.1% v/v HCl solution waschosen as the make-up liquid.

Studies on the flow rate of the make-up solution showthat the signal intensities of both species increased gra-dually as the flow rate increased from 0.15 to0.35 mL?min21 and then decreased with further increaseof the flow rate from 0.35 to 0.50 mL?min21. In addition,no detrimental effect on the separation caused by themake-up solution flow was observed due to no variationin the migration time with the make-up solution flow inthe range of 0.15–0.50 mL?min21. The intensity variationwith the flow rate of the make-up solution could beunderstood in terms of volatile species generation effi-ciency and dilution of the chip-CE effluent. Below a flowrate of 0.35 mL?min21, volatile species generation effi-ciency might dominate the signal intensity, whereas overthe flow rate of 0.35 mL?min21, the dilution of the chipeffluent probably controlled the signal. Accordingly, aflow rate of 0.35 mL?min21 was selected for the make-upsolution.

3.5 KBH4 concentration and flow rate

The effect of KBH4 concentration on the signals of bothmercury species was examined at a flow rate of0.48 mL?min21. The results are shown in Fig. 4. As KBH4

concentration increased from 0.03 to 0.22% m/v, the sig-nal intensity of the Hg(II) remained almost unchanged,



Figure 4. Influence of KBH4 concentration on fluorescentintensities of 4 mg?L21 MeHg(I) and 2 mg?L21 Hg(II). Allother conditions as in Table 1.

whereas that of MeHg(I) increased significantly. Furtherincreasing KBH4 concentration from 0.22 to 0.35% m/vresulted in a slow decrement of the signal intensity ofHg(II), but no obvious change in that of MeHg(I). There-fore, a 0.22% m/v KBH4 was used as the reductant forvolatile species generation.

The influence of the flow rate of 0.22% m/v KBH4 wasexamined with a make-up solution of 1.1% v/v HCl at aflow rate of 0.35 mL?min21. The optimum flow rate ofKBH4 solution ranged from 0.40 to 0.55 mL?min21 forboth species. Consequently, a flow rate of 0.48 mL?min21

for 0.22% m/v of KBH4 solution was used for furtherwork.

3.6 Argon flow rate

Studies on the effect of argon flow rate on the signalintensities of both mercury species exhibited that theoptimum argon flow rate ranged from 180 to220 mL?min21. No suction effect on the separation dueto the argon flow was observed because the migrationtime did not vary as the argon flow rate increased from140 to 400 mL?min21. Below the flow rate of180 mL?min21, the release of the volatile species fromthe reaction mixture was incomplete, resulting in lowersignal intensity. At higher flow rates (. 220 mL?min21),the dilution of the evolved volatile species in the atomizerwould be dominant, leading to the decrease of the signalintensities. Hence, an argon flow rate of 200 mL?min21

was adopted to maintain the maximum signal with goodprecision.

3.7 Figures of merit

The analytical figures of merit of the present chip-CE-AFStechnique for the speciation of both mercury species aresummarized in Table 2. The detection limits (3s) for Hg(II)and MeHg(I) based on peak height measurement were 53and 161 mg?L21, respectively. The recoveries of the spikesof Hg(II) and MeHg(I) from the natural water samplesranged from 92 to 108%. To demonstrate the repeatabilityof the developed chip-CE-AFS system, the electro-pherograms of the two mercury species with five con-secutive injections were recorded. The liquid surface inthe reservoirs was adjusted about every 2 runs. Norecondition of the channels was needed between runs.Excellent repeatability for 2 mg?L21 Hg(II) and 4 mg?L21

MeHg(I) (as Hg) was obtained with the precisions (RSD,n = 5) of the migration time, peak area and peak height inthe range of 0.7–0.9, 2.1–2.9, 1.5–1.8%, respectively.

To demonstrate the performance of the developed chip-CE-AFS technique, a comparison of the detection limitsand separation time obtained by several CE-hyphenatedtechniques for mercury speciation is made in Table 3. Oneof the most significant advantages of the present chip-CE-AFS hybrid technique is the remarkable reduction inseparation time for Hg(II) and MeHg(I) in comparison withthe conventional CE based hyphenated techniques (64 svs. 360–1200 s) [28–31]. Additional advantages of the

Table 2. Characteristic performance data of the chip-CE-AFS for mercury speciation

Hg(II) MeHg(I)

Precisiona) (RSD, n = 5) (%)

Migration time 0.7 0.9Peak area 2.1 2.9Peak height 1.5 1.8

Detection limits(mg?L21)

53 161

Calibrationfunctionb)

A = 17.15 C1 0.09 A = 4.50 C1 0.20

Correlationcoefficient

0.9997 0.9994

Recoveryc) (%)

River water 1 92 101River water 2 94 105Lake water 1 97 108Lake water 2 93 104

a) For 2 mg?L21 Hg(II) and 4 mg?L21 MeHg(I) (as Hg)b) A, peak height (mV); C, concentration (mg?L21) (as Hg)c) Recovery for spiking with 1 mg?L21 and 2 mg?L21 (as

Hg) for Hg(II) and MeHg(I), respectively



Table 3. Comparison of the developed chip-CE-AFS technique with conventional CE-hyphenated techniques in terms ofseparation time and detection limit for Hg species

Technique Ref. Sampleinjected(nL)

Interface Separationtime (s)

Detection limit (mg?L21)

Hg(II) MeHg(I)

Chip-CE-AFS This work ,0.7 Volatile species generation 64 53 161CE-Q-ICP-MS [28] 350 Conventional concentric Meinhard

nebulizer,600 81 128

CE-DF-ICP-MS [28] 450 Conventional concentric Meinhardnebulizer

600–1200 25 54

CE-Q-ICP-MS [28] 230 Volatile species generation .600 1 30CE-AFS [29] 150 Volatile species generation ,900 6.8 16.5CE-Q-ICP-MS [30] – Microconcentric nebulizer MCN-100 ,1200 170 80CE-Q-ICP-MS [31] 170 Microconcentric nebulizer MCN-100 ,360 6.0 13.6CE-Q-ICP-MS [31] 170 Cross-flow nebulizer ,360 112 149

Q-ICP-MS, quadrapole-inductively coupled plasma-mass spectrometryDF-ICP-MS, double focusing-inductively coupled plasma-mass spectrometry

developed methodology for mercury speciation are itslow instrumental and running costs and easy operation,as compared with CE-ICP-MS.

The concentration detection limit of the Hg(II) obtained bythe present technique is lower than those obtained usingCE quadrapole (Q)-ICP-MS with conventional Meinhardnebulizer [28], microconcentric nebulizer [30], and cross-flow nebulizer [31], but higher than those obtained by CE-Q-ICP-MS [28], CE-AFS [29], and CE double focusing(DF)-ICP-MS with conventional Meinhard nebulizer [28].For MeHg(I), the detection limit is comparable to thoseobtained by CE-Q-ICP-MS with conventional nebulizer[28] and cross-flow nebulizer [31], but higher than thoseobtained by CE-Q-ICP-MS [28], CE-AFS [29], CE-Q-ICP-MS with microconcentric nebulizer [30], and CE-DF-ICP-MS with conventional Meinhard nebulizer [28]. The mainlimitation for the concentration detection limit of thedeveloped chip-CE-AFS technique is the limited samplevolume injected (,0.7 nL) as compared with conventionalCE-based hybrid techniques.

4 Concluding remarks

This work for the first time demonstrated the successfulinterfacing of a chip-based separation system to AFS.The flow incompatibility between the microscale chip andthe macroscale AFS was overcome by directly etching amake-up solution channel on the glass chip. A concentric“tube-in-tube” design of the interface was adopted toeliminate the backpressure. The chip-CE-AFS opens upthe possibility for rapid speciation of toxic elements in anintegrated mode. The developed technique can also beapplied to hydride-forming elements without need of big

modifications. Further effort should be made to employthe present chip-CE-AFS technique for binding studies oftoxic metal species with biomolecules.

This research was supported by the National Basic Re-search Program of China (No. 2003CB415001), NationalNatural Science Foundation of China (No. 20475028,50273046), the National Key Technologies R&D Program(No. 2002BA906A28-2), and Specialized Research Fundfor the Doctoral Program of Higher Education (No.20040055038).

Received December 8, 2004

5 References

[1] Jacobson, S. C., Hergenroder, R., Koutny, L. B., Warmack,R. J., Ramsey, J. M., Anal. Chem. 1994, 66, 1107–1113.

[2] Woolley, A. T., Mathies, R. A., Anal. Chem. 1995, 67, 3676–3680.

[3] Woolley, A. T., Hadley, D., Landre, P., deMello, A. J., Mathies,R. A., Northrup, M. A., Anal. Chem. 1996, 68, 4081–4086.

[4] Service, R. F., Science 1998, 282, 399–401.

[5] Jacobson, S. C., Hergenroder, R., Koutny, L. B., Ramsey, J.M., Anal. Chem. 1994, 66, 1114–1118.

[6] Effenhauser, C. S., Paulus, A., Manz, A., Widmer, H. M.,Anal. Chem. 1994, 66, 2949–2953.

[7] Schwarz, M. A., Hauser, P. C., Lab Chip 2001, 1, 1–6.

[8] Vo-Dinh, T., Alarie, J. P., Isola, N., Landis, D., Wintenberg, A.L., Ericson, M. N., Anal. Chem. 1999, 71, 358–363.

[9] Jacobson, S. C., Moore, A. W., Ramsey, J. M., Anal. Chem.1995, 67, 2059–2063.

[10] Manz, A., Verpoorte, E., Effenhauser, C. S., Burggraf, N.,Raymond, D. E., Harrison, D. J., Widmer, H. M., J. HighResolut. Chromatogr. 1993, 16, 433–436.



[11] Xue, Q.-F., Foret, F., Dunayevskiy, Y. M., Zavracky, P. M.,McGruer, N. E., Karger, B. L., Anal. Chem. 1997, 69, 426–430.

[12] Figeys, D., Gygi, S. P., McKinnon, G., Aebersold, R., Anal.Chem. 1998, 70, 3728–3734.

[13] Shaffer, S. A., Tolmachev, A., Prior, D. C., Anderson, G. A.,Udseth, H. R., Smith, R. D., Anal. Chem. 1999, 71, 2957–2964.

[14] Zhang, B., Liu, H., Karger, B. L., Foret, F., Anal. Chem. 1999,71, 3258–3264.

[15] Wachs, T., Henion, J., Anal. Chem. 2001, 73, 632–638.

[16] Zhang, S., Van Pelt, C. K., Henion, J. D., Electrophoresis2003, 24, 3620–3632.

[17] Lion, N., Gellon, J. O., Jensen, H., Girault, H. H., J. Chro-matogr. A 2003, 1003, 11–19.

[18] Rossier, J. S., Youhnovski, N., Lion, N., Damoc, E., Becker,S., Reymond, F., Girault, H. H., Przybylski, M., Angew.Chem. Int. Edit. 2003, 42, 53–58.

[19] Song, Q.-J., Greenway, G. M., McCreedy, T., J. Anal. At.Spectrom. 2004, 19, 883–887.

[20] Cai, Y., Trends Anal. Chem. 2000, 19, 62–66.

[21] Armstrong, H. E. L., Corns, W. T., Stockwell, P. B., O’Connor,G., Ebdon, L., Evans, E. H., Anal. Chim. Acta 1999, 390,245–253.

[22] Yin, X.-B., Yan, X.-P., Jiang, Y., He, X.-W., Anal. Chem. 2002,74, 3720–3725.

[23] Fan, Z.-H., Harrison, D. J., Anal. Chem. 1994, 66, 177–184.

[24] Fu, L.-M., Yang, R.-J., Lee, G.-B., Liu, H.-H., Anal. Chem.2002, 74, 5084–5091.

[25] S-Lockyear, L. L., Colyer, C. L., Fan, Z.-H., Roy, K. I., Harri-son, D. J., Electrophoresis 1999, 20, 529–538.

[26] Culbertson, C. T., Jacobson, S. C., Ramsey, J. M., Anal.Chem. 1998, 70, 3781–3789.

[27] Deng, Y.-Z., Henion, J., Li, J.-J., Thibault, P., Wang, C., Har-rison, D. J., Anal. Chem. 2001, 73, 639–646.

[28] da Rocha, M. S., Soldado, A. B., Blanco, E., Sanz-Medel, A.,J. Anal. At. Spectrom. 2001, 16, 951–956.

[29] Yan, X.-P., Yin, X.-B., Jiang, D.-Q., He, X.-W., Anal. Chem.2003, 75, 1726–1732.

[30] Lee, T. H., Jiang, S. J., Anal. Chim. Acta 2000, 413, 197–205.

[31] Tu, Q., Qvarnström, J., Frech, W. G., Analyst 2000, 125, 705–710.


Documents

Development of a new hybrid technique for rapid speciation analysis by directly interfacing a microfluidic chip-based capillary electrophoresis system to atomic fluorescence spectrometry