Development of a low-cost microfluidic capillary-electrophoresis system coupled with flow-injection and sequential-injection sample introduction (review)

Microfabrication techniques used for the production of MEMS (micro electro-mechanical systems) have been successfully used to produce highly efficient microfluidic capillary electrophoresis chip systems. A limitation of this approach are the difficulties associated with the creation of the micrometer-sized structures in glass or other substrates, which currently involve specialized and expensive lithographic and etching processes. A further limitation is that hitherto most microfluidic chips are not designed for continuous introduction of a series of different samples, which limits the overall throughput of such systems. This article reviews the development of a microfluidic system for rapid CE separations, produced at a low cost of less than a dollar each, using equipment and materials readily available in the ordinary laboratory. Applications of the system, after coupling to flow-injection and/or sequential-injection sample introduction, for the determination of FITC- labeled amino acids by laser-induced fluorescence, trace metals by chemiluminescence, carbohydrates by amperometry, and inorganic and organic anions by indirect UV absorbance are exemplified to illustrate the performance and versatility of the microfluidic system. Received: 30 November 2000 / Revised: 13 February 2001 / Accepted: 23 February 2001     

An automated electrokinetic continuous sample introduction system for microfluidic chip-based capillary electrophoresis

An automated and continuous sample introduction system for microfluidic chip-based capillary electrophoresis (CE) was developed in this work. An efficient world-to-chip interface for chip-based CE separation was produced by horizontally connecting a Z-shaped fused silica capillary sampling probe to the sample loading channel of a crossed-channel chip. The sample presentation system was composed of an array of bottom-slotted sample vials filled alternately with samples and working electrolyte, horizontally positioned on a programmable linearly moving platform. On moving the array from one vial to the next, and scanning the probe, which was fixed with a platinum electrode on its tip, through the slots of the vials, a series of samples, each followed by a flow of working electrolyte was continuously introduced electrokinetically from the off-chip vials into the sample loading channel of the chip. The performance of the system was demonstrated in the separation and determination of FITC-labeled arginine and phenylalanine with LIF detection, by continuously introducing a train of different samples. Employing 4.5 kV sampling voltage (1000 V cm(-1) field strength) for 30 s and 1.8 kV separation voltage (400 V cm(-1) field strength) for 70 s, throughputs of 36 h(-1) were achieved with <1.0% carryover and 4.6, 3.2 and 4.0% RSD for arginine, FITC and phenylalanine, respectively (n = 11). Net sample consumption was only 240 nL for each sample.     

Continuous on-line derivatization and determination of amino acids by a microfluidic capillary electrophoresis system with a continuous sample introduction interface

An automatic, rapid and continuous on-line derivatization system coupled to microfluidic capillary electrophoresis (CE) for the determination of amino acids using o-phthaldialdehyde/N-acetyl-l-cysteine (OPA/NAC) as the derivative agents has been developed. By on-line derivatization, amino acids were automatically and reproducibly converted to the UV-absorbing derivatives, which were separated by capillary zone electrophoresis (CZE). Optimization of derivatization and separation condition was carried out to achieve both good sensitivity and separation efficiency. The separation could be achieved within 4 min and sample throughput rate can reach up to 16 h⁻¹. The repeatability (defined as relative standard deviation, R.S.D.) was 2.56, 2.85, 3.24 and 3.60% with peak area evaluation and 2.93, 3.12, 4.20 and 4.91% with peak height evaluation for arginine (Arg), phenylalanine (Phe), serine (Ser) and glycine (Gly), respectively. The limits of detection (S/N=3) were 10.46, 13.14, 34.39 and 44.79 μmol/l for Arg, Phe, Ser and Gly, respectively. Major advantages of the proposed method include improved precision and efficient automation of the derivatization by the FI system and the enhanced sampling frequencies by the combined FI-CE system.     

Multiplexed electrothermal vaporization sample introduction system for inductively coupled plasma spectrometry

A multiplexed electrothermal vaporization (ETV) system for sample introduction into an inductively coupled plasma was designed in an effort to increase sample turn-around time. Tungsten filaments (300 W), originally designed for overhead projectors, were chosen for use as ETVs to avoid the high power requirements associated with other ETV devices, e.g. graphite furnaces (2–3 kW). In short, we have multiplexed the thermal stages have been multiplexed such that a vaporization event can take place every 20 s. This represents a significant increase in the throughput typically associated with ETV-ICPMS, which is normally approximately 20–30 samples/h. Evaluated with respect to common figure of merit criteria, the performance of the multiplexed ETV system is similar to that seen with conventional graphite furnace ETV systems. However, several mass spectral interferences can be introduced by the presence of W into the plasma, which hinder the analysis of certain metals (Hg, Mo, etc.). Thus, other low power vaporizers (e.g. Re, Ta) should be considered for use in future systems.     

On-line stripping voltammetry of trace metals at a flow-through bismuth-film electrode by means of a hybrid flow-injection/sequential-injection system

In this work, we describe an automated stripping analyzer operating on a hybrid flow-injection/sequential-injection (FIA/SIA) mode and utilizing a bismuth-film electrode (BiFE) as a flow-through sensor for on-line stripping voltammetry of trace metals. The instrument combines the advantages of FIA and SIA and is characterised by simplicity, low-cost, rapidity, versatility and low consumption of solutions. The proposed analytical flow methodology was applied to the determination of Cd(II) and Pb(II) by anodic stripping voltammetry (ASV) and of Ni(II) and Co(II) by adsorptive stripping voltammetry (AdSV). The steps of the rather complex experimental sequence (i.e. the bismuth-film formation, the analyte accumulation, the voltammetric stripping and the electrode cleaning/regeneration) were conducted on-line and the critical parameters related to the respective analytical procedures were investigated. In ASV, for a accumulation time of 180 s the limits of detection for Cd(II) and Pb(II) were 2 and 1 μg l⁻¹, respectively (S/N = 3) and the relative standard deviations were 5.3% and 4.7%, respectively (n = 8). In AdSV, for a total sample volume of 1000 μl, the limits of detection for Ni(II) and Co(II) were 1 μg l⁻¹ (S/N = 3) and the relative standard deviations were 5.5% and 6.2%, respectively (n = 8). The measurement frequency ranged between 15 and 20 stripping cycles h⁻¹. The results indicate that the BiFE is well suited as a flow-through detector for on-line stripping analysis and, by virtue of its low toxicity, can serve as a viable alternative to mercury-based flow-through electrodes.     

Simultaneous electrothermal vaporization and nebulizer sample introduction system for inductively coupled plasma mass spectrometry

The novel analytical application of the combination of an inline electrothermal vaporization (ETV) and nebulization source for inductively coupled plasma mass spectrometry (ICP-MS) has been studied. Wet plasma conditions are sustained during ETV introduction by 200 mL/min gas flow through the nebulizer, which is merged with the ETV transport line at the torch. The use of a wet plasma with ETV introduction avoided the need to change power settings and torch positions that normally accompany a change from wet to dry plasma operating conditions. This inline-ETV source is shown to have good detection limits for a variety of elements in both HNO₃ and HCl matrices. Using the inline-ETV source, improved limits of detection (LOD) were obtained for elements typically suppressed by polyatomic interferences using a nebulizer. Specifically, improved LODs for ⁵¹V and ⁵³Cr suffering from Cl interferences (⁵¹ClO⁺ and ⁵³ClO⁺ respectively) in a 1% HCl matrix were obtained using the inline-ETV source. LODs were improved by factors of 65 and 22 for ⁵¹V and ⁵³Cr, respectively, using the inline-ETV source compared to a conventional concentric glass nebulizer. For elements without polyatomic interferences, LODs from the inline-ETV were comparable to conventional dry plasma ETV-ICP time-of-flight mass spectrometry results. Lastly, the inline-ETV source offers a simple means of changing from nebulizer introduction to inline-ETV introduction without extinguishing the plasma. This permits, for example, the use of the time-resolved ETV-ICP-MS signals to distinguish between an analyte ion and polyatomic isobar.     

Rapid enzymatic chemiluminescent assay of glucose by means of a hybrid flow-injection/sequential-injection method

This work reports a hybrid flow-injection analysis (FIA)/sequential-injection analysis (SIA) method for the rapid enzymatic assay of glucose with soluble glucose oxidase (GOD). The method relies on the sequential injection of segments of the sample and of a solution of enzyme by means of a multi-port selection valve in a flowing water stream. As the two zones are swept downstream, they overlap and merge so that the glucose in the sample is enzymatically oxidised. The generated hydrogen peroxide is merged with an alkaline luminol solution and the chemiluminescence (CL) intensity is monitored and related to the glucose concentration in the sample. The linear range of the method for glucose determination is 0.01-1 mmol L⁻¹, the relative standard deviation is 3.9% at the 0.08 mmol L⁻¹ level (n = 8), the limit of detection at the 2σ level is 4 μmol L⁻¹ glucose and the injection rate is 80 h⁻¹. The method was applied to the analysis of energy drinks and honey with relative errors in glucose determination in the range 100 ± 4.3%. The advantages of the proposed method are the wide linear range, the simple instrumentation used, the low consumption of sample and reagents, the elimination of catalysts and immobilised enzymes and the high sample throughput.     

Simultaneous quantification of 17 trace elements in blood by dynamic reaction cell inductively coupled plasma mass spectrometry (DRC-ICP-MS) equipped with a high-efficiency sample introduction system

A quadrupole inductively coupled plasma mass spectrometer (Q-ICP-MS) equipped with a dynamic reaction cell (DRC) and coupled with a desolvating nebulization system (APEX-IR) was employed to determine 17 elements (Al, As, Ba, Cd, Co, Cr, Li, Mn, Mo, Ni, Pb, Sb, Se, Sn, Sr, V, and Zr) in blood samples. Ammonia (for Al, Cr, Mn, and V) and O₂ (for As and Se) were used as reacting gases. Selection of the best flow rate of the gases and optimization of the quadrupole dynamic bandpass tuning parameter (RPq) were carried out, using digested blood diluted 1 + 9 with deionized water and spiked with 1 μg L⁻¹ of Al, Cr, Mn, V and 5 μg L⁻¹ of As and Se. Detection limits were determined in digested blood using the 3σ criterion. The desolvating system allowed a sufficient sensitivity to be achieved to determine elements at levels of ng L⁻¹ without detriment of signal stability. The accuracy of the method was tested with the whole blood certified reference material (CRM), certified for Al, As, Cd, Co, Cr, Mn, Mo, Ni, Pb, Sb, Se, and V, and with indicative values for Ba, Li, Sn, Sr, and Zr. The addition calibration approach was chosen for analysis. In order to confirm the DRC data, samples were also analyzed by means of sector field inductively coupled plasma mass spectrometry (SF-ICP-MS), operating in medium (m/Δm = 4000) and high (m/Δm = 10,000) resolution mode and achieving a good agreement between the two techniques.     

Batch-injection stripping voltammetry (tube-less flow-injection analysis) of trace metals with on-line sample pretreatment

The most essential limitation of batch-injection analysis (BIA) methodology compared to other flow methods (CFA, FIA, SIA) is the lack of possibility of on-line sample processing in the measuring system. Some procedures of on-line sample pretreatment in BIA are possible by changing the plastic tip of the automatic micropipette used for sample injection into a flow-through reactor, e.g. by packing it with a bed of a solid sorbent. This concept is employed in the voltammetric stripping determinations of trace metals using a bed of commercial chelating resin Chelex-100. It was found that, besides the electrochemical preconcentration of analytes in the form of amalgams on the surface of mercury thin film electrodes, an approximately 10-fold additional preconcentration can be achieved on the packed sorbent bed by using different volumes of aspirated sample solution and eluent. This procedure allows also efficient elimination of some matrix effects.     

The spectrophotometric determination of iron(III) in a flow-injection system with a mixed solvent

A fast and sensitive flow-injection procedure is described for the determination of iron(III). The complexing agent is ammonium diisopropyldithiophosphate in a 1:1 (v/v) isopropanol/water carrier stream. The linear range is 0.05–15 mg 1^?1 iron(III) with a detection limit of 0.01 mg 1^?1 and the injection rate is about 400 h^?1.     

Spectrophotometric determination of ethanol in blood using a flow-injection system with an immobilized enzyme (alcohol dehydrogenase) reactor coupled to an on-line dialyser

The determination of ethanol in whole blood without pretreatment using spectrotometric detection with an immobilized enzyme reactor coupled to an on-line dialyser is described. Optimum working conditions are given. The effects of the injection volume and the flow-rate on the peak height using the dialyser were investigated. A good linear calibration graph over the ethanol concentration range 3–40 μg ml^?1 was observed; these concentrations correspond to 0.3?4.0 g of ethanol per 1000 g of whole blood prior to dilution. For comparison, whole blood samples were analysed by gas chromatography with direct injection. The effect of temperature on the peak height was also studied in a system without the dialyser.     

Chemical vapor generation for sample introduction into inductively coupled plasma atomic emission spectroscopy: vaporization of antimony(III) with bromide

A new method for antimony determination in soils is proposed. It is based on the chemical vapor generation of Sb(III) with bromide, after a reaction in sulfuric acid media and transport of the gaseous phase into an inductively coupled plasma for atomic emission spectrometry. The experimental variables influencing the method were delimited by experimental design and the most important were finally optimized by the modified Simplex method. In optimized conditions the method involves the reaction of 579 microl concentrated sulfuric acid with 120 microl 5% w/v KBr and 250 microl antimony solution. Measurement of antimony emission intensity at 217.581 nm provides a method with an absolute detection limit of 3.5 ng and a precision (RSD) of 5.8% for the injection of five replicates of 175 ng Sb(III) (250 microl of 0.7 microg ml(-1) solution). The interference of common anions and cations on the antimony signal was evaluated. A 21% Sb(III) volatilization efficiency was calculated from the mean of six experiments at optimum conditions. The accuracy of the methodology was checked by the analysis of one standard reference soil after acid decomposition heating in a microwave oven.     

Capillary electrophoresis: An integrated system with a unique split-flow sample introduction mechanism

The performance of an integrated capillary electrophoresis system with a novel split-flow sample injection mechanism and special high sensitivity UV absorbance detector is described. Sample introduction into the capillary is accomplished with a standard HPLC-type microliter syringe. The injected sample is divided proportionally between the separation capillary and an adjustable splitvent. The volume of sample introduced into the capillary can be manipulated by varying the length or the i.d. of the splitvent tubing; or the volume of sample injected. Data are presented showing reproducibility of retention time, peak height, and peak area; minimum detectability; and operation at short UV wavelengths.     

Enzymatic chemiluminescent assay of glucose by sequential-injection analysis with soluble enzyme and on-line sample dilution

This work reports a sequential-injection analysis (SIA) method for the enzymatic assay of glucose with soluble glucose oxidase (GOD) and on-line sample dilution with chemiluminescence (CL) detection. A zone of sample was aspirated in the holding coil of the SIA manifold and, if necessary, was diluted on-line by means of an auxiliary dilution conduit. Then, a zone of GOD was aspirated adjacent to the sample zone and a stopped-flow period was applied to allow the enzymatic reaction to proceed with production of hydrogen peroxide. Then, zones of a catalyst (Co(II) solution) and alkaline luminol were aspirated into the holding coil. Finally, the flow was reversed and the stacked zones were sent to a flow-cell located in front of a photomultiplier tube (PMT) that monitored the CL intensity. The linear dynamic range was 1 × 10⁻⁵-1 × 10⁻³ mol L⁻¹ glucose, the coefficient of variation at 8 × 10⁻⁵ mol L⁻¹ of glucose was s_r = 3.1% (n = 8), the limit of detection at the 3σ level was c_L = 1 × 10⁻⁶ mol L⁻¹ and the sampling frequency was 28 h⁻¹. With on-line dilution by a factor of 1/200, the linear range could be extended up to 0.2 mol L⁻¹ glucose. The advantages of the proposed method are the simple manifold and instrumentation used, the scope for automated on-line dilution, the low consumption of sample and reagents and the elimination of enzyme immobilisation procedures. The method was applied to the analysis of commercial drinks and honey with percent relative errors in glucose determination in the range 100 ± 6.1%.     

Development of flow-injection liposome immunoanalysis (FILIA) for imazethapyr

Imazethapyr is the herbicide developed for use in leguminous crops. In this study, flow-injection liposome immunoanalysis (FILIA) has been shown to be capable of measuring imazethapyr in a buffered solution with a detection limit of 0.1 ppb through the optimization process. Protein A coated glass beads covalently conjugated with antibody were contained in a glass column, and this column was used as an immunoreactor. Liposomes which encapsulated a fluorescent dye, sulforhodamine B (SRB) or carboxyfluorescein (CF), generated the analytical signal. By loading larger volumes of sample onto the column, it was shown that the detection limit could be lowered. Liposomes containing carboxyfluorescein gave more sensitive response and a lower detection limit than those with sulforhodamine B. Also, improved response was obtained by using a smaller flow cell in the fluorescence detector due to the reduced dilution effect.     

Capillary electrophoresis system with flow injection sample introduction and chemiluminescence detection on a chip platform

A capillary electrophoresis (CE) system with chemiluminescence (CL) detection was combined with flow injection (FI) sample introduction on a chip platform. A falling-drop interface was applied to perform FI split-flow sample introduction while achieving electrical isolation from the CE high voltage. A tubular reservoir at the capillary outlet served as both the CL reaction and detection cell for the luminol-peroxide-metallic ion chemiluminescent reaction, with the luminol included in the separation buffer and CL reagent H2O2 continuously introduced into the outlet reservoir. An optical fiber was positioned within the outlet reservoir directly opposite, and 300 microns away from, the capillary outlet for collecting and transferring the generated CL to the PMT. The peak height signals and the separation efficiency were almost independent of the reagent flow-rate, making the system a robust one. The performance of the system was illustrated by the separation of Co(II) and Cu(II), achieving baseline separation in 60 s. Detection limits (3 sigma) were 1.25 x 10(-8) and 2.3 x 10(-6) mol dm-3 for Co(II) and Cu(II), respectively. Peak height precision was 1.9% RSD (n = 9) at the 10(-7) mol dm-3 Co level.     

The influence of the sample introduction system on signals of different tin compounds in inductively coupled plasma-based techniques

The influence of the sample introduction system on the signals obtained with different tin compounds in inductively coupled plasma (ICP) based techniques, i.e., ICP atomic emission spectrometry (ICP–AES) and ICP mass spectrometry (ICP–MS) has been studied. Signals for test solutions prepared from four different tin compounds (i.e., tin tetrachloride, monobutyltin, dibutyltin and di-tert-butyltin) in different solvents (methanol 0.8% (w/w), i-propanol 0.8% (w/w) and various acid matrices) have been measured by ICP–AES and ICP–MS. The results demonstrate a noticeable influence of the volatility of the tin compounds on their signals measured with both techniques. Thus, in agreement with the compound volatility, the highest signals are obtained for tin tetrachloride followed by di-tert-butyltin/monobutyltin and dibutyltin.The sample introduction system exerts an important effect on the amount of solution loading the plasma and, hence, on the relative signals afforded by the tin compounds in ICP–based techniques. Thus, when working with a pneumatic concentric nebulizer, the use of spray chambers affording high solvent transport efficiency to the plasma (such as cyclonic and single pass) or high spray chamber temperatures is recommended to minimize the influence of the tin chemical compound. Nevertheless, even when using the conventional pneumatic nebulizer coupled to the best spray chamber design (i.e., a single pass spray chamber), signals obtained for di-tert-butyltin/monobutyltin and dibutyltin are still around 10% and 30% lower than the corresponding signal for tin tetrachloride, respectively. When operating with a pneumatic microconcentric nebulizer coupled to a 50 °C-thermostated cinnabar spray chamber, all studied organotin compounds provided similar emission signals although about 60% lower than those obtained for tin tetrachloride. The use of an ultrasonic nebulizer coupled to a desolvation device provides the largest differences in the emission signals, among all tested systems.     

Determination of mercury in saliva with a flow-injection system

A simple, fast and reliable method, with a low detection limit, has been developed for the determination of total mercury in saliva samples. The method uses a brominating reagent, followed by on-line addition of KMnO₄ at room temperature to convert organically bound mercury to inorganic mercury ions, and determines mercury by flow-injection cold-vapour atomic absorption spectrometry. Using the method described, complete recoveries of five mercury compounds from saliva were attained. Results obtained on real samples using the new method were comparable to that obtained using the established method with batch system. The detection limit of this method, based on three standard deviations of the blank, is 0.05 μg l⁻¹ Hg in a saliva sample of 500 μl. A sample throughput of 80 measurements per hour is possible with the method. The calibration curves are linear up to 20 μg l⁻¹ and the dynamic range extends to 40 μg l⁻¹ Hg. At a concentration of 1μg l⁻¹ mercury in saliva, the relative standard deviation is about 2% for 11 replicates; a total volume of 0.5 ml saliva is required for three replicates.     

Extraction-spectrophotometric determination of anionic surfactants with a flow-injection system

Anionic surfactants are preconcentrated from 5-ml samples by extraction as the ion-pair with ethyl violet into toluene. The absorbance of aliquots of the toluene phase is measured at 610 nm in a flow-injection system. A phase converter is located prior to the injection valve to convert a water stream to, pumped with ordinary pump tubing, to a toluene stream. The working range was 0.01–1.0 mg l^?1 and the reproductibility (r.s.d, n = 10) was 2% for 0.4 mg l^?1 sodium dodecyl sulphate. The non-aqueous flow-injection system serves to miniaturize the extraction from separatory funnel (200 ml) to test tube (10 ml) scale without loss of precision or validity.