WO1992018857A1

WO1992018857A1 - Method and apparatus for batch injection analysis and trace metal testing

Info

Publication number: WO1992018857A1
Application number: PCT/US1992/002919
Authority: WO
Inventors: Joseph Wang; Ziad H. Taha
Original assignee: New Mexico State University Technology Transfer Corporation
Priority date: 1991-04-09
Filing date: 1992-04-08
Publication date: 1992-10-29
Also published as: EP0602043A4; US5292423A; EP0602043A1; AU1690192A

Abstract

Batch injection analysis comprises apparatus and method for injecting and transporting analytes toward a detector immersed in a confined, inert electrolyte. Passage of the analyte over the detector surface provides measurement of sample concentrations. Detectors normally comprise selective electrodes (11, 15, 16), such as biologically or chemically modified surfaces (11'), ion-selective probes, optical or thermal devices, thus eliminating conduits, valves, and pumps. Also disclosed are a method and apparatus for trace metal testing using mercury-coated screen printed electrodes. Both voltametric and potentiometric stripping analysis are used. Sample solutions were tested employing both stirring and non-stirring, as well as aeration and deaeration procedures. Microliter samples suitable for slide mounting were also employed.

Description

METHOD AND APPARATUS FOR BATCH INJECTION ANALYSIS AND TRACE METAL TESTING

BACKGROUND OF THE INVENTION Field of the Invention (Technical Field) :

This invention relates to apparatuses for measurement of analytes injected into a static electrochemical cell and a method for its use, as well as testing apparatus for heavy metals, and a method of using such apparatus.

Background Art:

Several different techniques for measurement of analyte concentration have evolved in the prior art. Notable among these techniques is flow injection analysis (FIA). In this measurement process, a sample to be analyzed is injected into a laminarly flowing carrier stream of solvent and reagents. Reproducible sample volumes are injected, so the reaction need not proceed to the steady state. The reaction is developed only to the point which permits recordation as the sample passes an appropriate detector. The output transient signals thus produced reflect the concentration of the injected analyte. Flow injection analysis is described in articles entitled "Flow Injection Analysis: New Tool for Old Assays—New Approach to Analytical Measurements," by Kent K. Stewart (Analytical Chemistry, Vol. 55, No. 9, August 1983) and "Flow Injection Analysis: From Test Tube to Integrated Microconduits," (Analytical Chemistry, Vol. 55, No. 11, September 1983); and in the Lachat brochure entitled "The FIA Concept;" the Tecator brochure on the FIAstar Flow Injection Analysis Bibliography; and the Control Equipment Corporation brochures entitled "The Applications: What it Does."

Obvious disadvantages are present with flow injection analysis. Pumps, valves, and tubing are required. Further, chemical reaction is often required to convert the analyte to a detectable species.

Other quantitative analysis apparatuses are taught by the prior art. U.S. Patent No. 4,865,992, entitled System and Method for Quantitative Analysis of a Solution , to Hach, et al., teaches such apparatus comprising continuous addition of reagent to a beaker containing a chemical species to be measured until an endpoint is reached. This procedure requires a chemical reaction, large samples, and involves a slow measurement of reaction product.

U.S. Patent No. 4,695,555, entitled Liquid Chromatographic Detector and Method, to O'Keeffe, involves "spray electrification" wherein droplets of an analyte acquire electric charges dependent upon the concentration of solute carried by the droplet. Measurement of concentration of solute is enabled by measurement of the amount of deviation of the droplets from a neutral path. U.S. Patent No.

4,003,705, entitled Analysis Apparatus and Method of Measuring Rate of Change of Electrolyte pH, to Buzza, et al., teaches specific measurements of C0₂ and chloride in blood, based on reaction with an electrolyte and subsequent pH measurement.

International Application No. PCT/DK89/00070, entitled A Method of Effecting NIR-Analyses of Successive Material Samples, and a System for Carrying Out the Method, to Johnsen, discloses a near- infrared reflection spectroscopy apparatus wherein the effect of remnant deposits is avoided by an advancing film between test chamber and optical unit.

Thus, it is seen that the prior art lacks a fast, repetitive, highly reproducible, versatile, and reliable analytical measurement system devoid of conduits, valves, and pumps.

Anodic stripping voltammetry (ASV) and potentiometric stripping analysis (PSA) have long been used in trace metal testing, as discussed in "Anodic Stripping Voltammetry as an Analytical Tool" (by Wang Environ. Sci . Techno 1. , Vol 16, No. 2 (1982)) and "Mercury- Coated Carbon-Foam Composite Electrodes for Stripping Analysis of Trace Metals" (by Wang, et al. , Analytical Chemistry, Vol. 64, (1992) . Anodic stripping voltammetry generally involves the reduction or electrolytic deposition of metals onto an electrode, termed preconcentration, followed by anodically reoxidizing and stripping the metals, thereby producing a plot of current as a function of voltage increasing in amplitude (the measurement step) , as discussed in "Anodic Stripping Voltammetry" (by Wang Journal of Chemical Education , Vol. 60, P. 1074).

Normally, ASV and PSA require laboratory conditions for optimum results (see Wang, "Anodic Stripping Voltammetry"). Beakers, nitrogen bubbling equipment, and stirrers are usually required.

Electrodes for ASV comprise a working electrode, reference electrode (usually Ag/AgCl) , and an auxiliary electrode, usually platinum.

Prior art working electrodes for ASV and PSA, such as those in U.S. Patent No. 4,804,443, entitled Method and Apparatus for the Determination of Electrochemically Active Components in a Process Stream, to Newman, et al., comprised hanging mercury drop and mercury-coated glassy carbon electrodes. The hanging mercury drop electrode requires laboratory conditions to insure stability and drop size of the drop. As discussed in "Mercury-Coated Carbon-Foam Composite Electrodes for Stripping Analysis of Trace Metals," by Wang, et al. , (Analytical Chemistry, Vol 64 (1992)) glassy carbon substrate electrodes also give better results under laboratory conditions.

Two articles entitled "Disposable Single—Use Sensors" and "Disposable Electrochemical Biosensors" (by Monika J. Green and Paul I. Hilditch, MediSense Inc . , Units 3 & 4) discuss single-use disposable sensors, also well-known to the prior art. Biosensors, for example, glucose monitors, may comprise a PVC substrate with a working (carbon) and reference (Ag/AgCl) electrodes coated thereon, as well as the enzyme. Such enzyme-coated electrodes are also described in parent application Serial No. 07/682,907, incorporated herein by reference. Other prior art applications of screen-printed electrodes are electrochemical measurements of ascorbic acid or reduced glutathione.

As disclosed herein, working electrodes, particularly flat or planar carbon paste electrodes, can effectively be modified. Chemical and biological modification involving selective electrode coatings or membranes, are disclosed. Also disclosed is the use of optical or thermal devices as sensors. Also disclosed herein are voltammetric and potentiometric measurements and measurement devices.

A requirement for decentralized testing of trace metals has evolved. Field or on-site trace metal testing further suggests a need for disposable single-use electrodes. However, despite the ready availability, low cost, and general convenience of screen- printed carbon electrodes, formerly used primarily as biosensors, they are nowhere in the prior art suggested for use In trace metal detection apparatus.

SUMMARY OF THE INVENTION

(DISCLOSURE OF THE INVENTION)

The present invention comprises a method and apparatus for measuring analyte sample concentration. The apparatus of the invention comprises a vessel, electrolyte confined within the vessel, an entry for introducing analyte samples to be analyzed into the vessel, a detector in the vessel for sensing the analyte samples to be analyzed, and an analyzer for analyzing the analyte samples.

In the preferred embodiment of the invention, the vessel further comprises a stirrer, and preferably a magnetic stirrer. The electrolyte is preferably inert relative to the analyte sample, and preferably comprises a solution such as potassium chloride, a phosphate buffer and sodium hydroxide, potassium dihydrogen phosphate and sodium nitrate, and CH₃C00Na, CH₃C00H, NaCl, and 1,2 diaminocyclohexane-N.N.NJN) tetraacetic acid, such as a solution comprising CH₃C00Na, CH₃C00H, NaCl; 1,2 diaminocyclohexane-N,N,N;N; tetraacetic acid; and aluminum sulfate. Preferably analyte samples are introduced into the vessel by injection, such as using a pipette. The detector preferably comprises an electrode. The electrode may be spherical or planar, may comprise carbon paste (which may further comprise glucose oxidase and/or ruthenium dioxide) , may be a pH electrode, or an ion-selective electrode (such as a chloride or fluoride electrode) . The analyzer for analyzing the analyte samples preferably comprises amperometric or potentiometric measuring devices. In most applications, the samples to be analyzed are introduced and sensed within 2 to 10 mm from each other. However, for other applications the distance is preferably less than 2 mm or greater than 10 mm.

The method of the invention comprises the steps of: a) providing a vessel; b) confining an electrolyte within the vessel; c) introducing samples to be analyzed .into the vessel; d) sensing the samples to be analyzed; and e) analyzing the samples.

In accordance with the present invention, there is further provided a method of analyzing trace metals comprising the steps of providing a plurality of flat printed electrodes, coating at least one of the plurality of flat printed electrodes with mercury, and analyzing a sample for heavy metal content with the plurality of electrodes. The method of the invention further comprises the step of providing at least one flat screen-printed Ag/AgCl reference electrode, and at least one flat screen-printed carbon electrode.

The step of analyzing a sample for heavy metal content comprises the steps of either voltammetrically or potentiometrically analyzing the sample. The preferred method of the invention further comprises the step of preconcentrating the heavy metal upon an electrode.

The preferred apparatus of the invention further comprises means for providing a plurality of flat printed electrodes, means for coating at least one of said plurality of flat printed electrodes with mercury, and means for analyzing a sample for heavy metal content with said plurality of electrodes. The means for providing a plurality of flat printed electrodes further comprises means for providing at least one flat screen-printed Ag/AgCl reference electrode and at least one flat screen-printed carbon electrode.

The preferred means for analyzing a sample for heavy metal content further comprises means for voltammetrically or potentiometrically analyzing the sample. The preferred apparatus of the invention further comprises means for preconcentrating the heavy metal upon an electrode and means for deaerating and stirring the sample. The preferred apparatus of the invention further comprises means for analyzing a microliter sample solution, which may be urine or drinking water.

A primary object of the invention is the provision of a highly reproducible and repetitive analyte measurement apparatus and method using a confined, inert, large-volume electrolyte.

Yet another ob ect of the invention is the provision of an analyte measurement apparatus and method which relies on specific sensing surfaces.

Still another object of the invention is to provide a batch injection analysis apparatus and method with performance equivalent to flow injecting analysis.

Another object of the invention is the provision of disposable single-use electrodes for trace metal detection.

Yet another object of the invention is the provision of portable, decentralized trace metal detection apparatus.

A further object of the invention is the provision of trace metal testing apparatus effective with microliter samples.

Still another object of the invention is the provision of inexpensive reusable electrodes for trace metal testing.

An advantage of the invention is the provision of a reliable analyte concentration apparatus totally devoid of pumps, conduits, and valves.

Another advantage of the invention is the provision of rapid "wash out" and dispersal of samples. Still another advantage of the invention is the provision of analyte sample apparatus and method amenable to amperometric or potentiometric measurements.

An advantage of the present invention is the ease of fabrication and low cost of the electrodes employed.

Another advantage of the invention is the lack of need for stirring and deaeration of samples.

A further advantage of the invention is the highly stable response and low cost attribute of screen-printed carbon electrodes for centralized operations.

Other objects, advantages, and novel features, and further scope of applicability of the present invention will be set forth in part in the detailed description to follow, taken in conjunction with the accompanying drawings, and in part will become apparent to those skilled in the art upon examination of the following, or may be learned by practice of the invention. The objects and advantages of the invention may be realized and attained by means of the instrumentalities and combinations particularly pointed out in the appended claims.

BRIEF DESCRIPTION OF THE DRAWINGS

The accompanying drawings, which are incorporated into and form a part of the specification, illustrate several embodiments of the present invention and, together with the description, serve to explain the principles of the invention. The drawings are only for the purpose of illustrating a preferred embodiment of the invention and are not to be construed as limiting the invention.

Fig. 1 is a schematic diagram of a batch injection analysis (BIA) vessel with amperometric detection;

Figs. 2(a)-2(c) depict sample injection in BIA;

Figs. 3(a) and 3(b) illustrate a comparison between BIA and FIA; . Figs. 4(a) and 4(b) depict repetitive Injections of hydroquinone and ferrocyanide in solutions;

Fig. 5 depicts five injections of ferrocyanide, each injection followed by massive additions of ferrocyanide;

Fig. 6 depicts the relative inertness of the electrolyte in BIA;

Figs. 7(a) and 7(b) depict responses for glucose using biologically and chemically modified electrodes;

Fig. 8 depicts effects of sample volumes upon different electrode distances;

Fig. 9 also depicts effects of sample volumes upon different electrode distances;

Fig. 10 is a schematic diagram of a BIA vessel used with potentiometric detection;

Fig. 11 is a comparison of responses obtained with high and low pH injections;

Figs. 12(a) and 12(b) are comparisons of the response for different injection rates;

Figs. 13(a) and 13(b) are comparisons of responses of spherical and planar electrodes;

Fig. 14 depicts responses for sample injections followed by massive addition of sample solutions;

Fig. 15 shows responses for sample injections of varying pH;

Fig. 16 shows responses for sample injections of common liquids of varying pH; Figs. 17(a) and 17(b) depict comparison responses for low and high concentration of chloride and fluoride injections;

Figs. 18(a) and 18(b) depict comparison results for chloride and fluoride solutions with massive additions of such chloride and fluoride solutions;

Fig. 19 shows effect of sample volume and tip-electrode distance upon response; and

Figs. 20(a) and 20(b) depict responses for repetitive injections of fluoride and chloride solutions.

Figs. 21(a)-21(c) depict voltammograms using a mercury-coated carbon strip electrode, a glassy carbon electrode, and a hanging mercury drop electrode, respectively;

Figs. 22(a) and 22(b) illustrate bare and mercury coated screen printed carbon electrodes;

Figs. 23(a) and 23(b) show voltammograms and potentiograms obtained with incrementally increased lead concentration;

Figs. 24(a)-24(c) depict voltammetric, potentiometric, and time plots obtained after different preconcentration times;

Figs. 25(a) and 25(b) show square-wave stripping voltammogram with carbon strip and glassy carbon electrodes, respectively;

Figs. 26(a) and 26(b) depict microliter potentiometric analysis in the same sample drop and in different sample drops;

Figs. 27(a) and 27(b) show voltammograms and potentiograms for repetitive stripping measurements; and

Fig. 28(a) through Fig. 29(b) depict voltammograms after testing urine and potentiograms after testing drinking water, respectively.

e.._SSTITUTESHEET DESCRIPTION OF THE PREFERRED EMBODIMENTS (BEST MODES FOR CARRYING OUT THE INVENTION

Reference is now made to Fig. 1 which depicts the apparatus of the preferred embodiment of the Invention. The apparatus comprises a cell 10 preferably made of a non—reactive material, such as

PLEXIGLAS^®, LEXAN^®, LUCITE^®, or the like. Working electrode 11 is inserted from the bottom of cell 10 and retained in position by a glass seal, for example, a Wilson glass seal, or the like. Pipette or micropipette 12, such as an Eppendorf standard pipette, or the like, is likewise retained in position on cell cover 13 by an indexed clamp (not shown) so as to assure a reproducible position, or more importantly, to assure a reproducible working electrode 11-pipette 12 tip separation distance 14. Counter electrode 15, usually platinum or other conductive noble metal, and reference electrode 16 are also mounted on cell cover 13 and complete the measuring circuit.

Working electrode 11 and electrode tip 11' preferably comprise carbon paste (55% graphite powder, 45% mineral oil) or platinum disks. Other electrodes, including ion—selective electrodes, biological and chemical modified electrodes, as well as optical (fiber-optic) or thermal sensors, can be used.

Aperture 17, closed by stopper 18, in cover 13 provides access for introducing and replenishing the cell solution, a relatively "inert" electrolyte. ("Inert," as defined herein with respect to the electrolyte, means that although the electrolyte is ionically dissociative, the electrolyte is for all practical purposes chemically unreactive with the analyte or sample.) The cell is drained through drain 20. Magnetic bar 19 provides agitation when cell 10 is placed upon an energized magnetic stirrer (not shown) . Alternatively, an internal stirrer or agitator could be provided.

Electrodes 11, 15, and 16 are connected In circuit to a voltammeter, for example, an EG&G PAR model 364A, or the like. The output of the voltammeter is displayed upon a strip-chart recorder, for example, a HOUSTON OMNISCRIBE, or the like (not shown) .

In operation, batch injection analysis (BIA) resembles flow injection analysis (FIA) in that an Injected sample is transported in reproducibly consistent fashion toward a detector. In batch injection analysis, this is accomplished by placing the injector outlet (the tip of pipette 12) in close proximity to the sensor or detector surface 11' . The sample thereby literally floods the detector surface.

In addition to reproducible and consistent transport over the detector surface, batch injection analysis requires effective "wash¬ out" characteristics. This is accomplished in the preferred embodiment of the invention by the use of a large-volume stirred cell solution. Thus, small samples (20-50 μL) are rapidly dispersed and greatly diluted (20,000-50,000 fold) over the entire cell volume, permitting a great number of repetitive tests with no "build-up" of analyte.

In the preferred embodiment of the invention, injection action of the sample through the pipette 12 provides convective transport toward the detector surface 11' . Accordingly, batch in ection analysis resembles a stopped-flow operation with samples flowing toward the detector surface 11' , but no flow of the cell solution or electrolyte occurs.

Figs. 2(a), 2(b), and 2(c) illustrate in an enlarged view, in succession, injection and transport of a sample from the tip of pipette 12 toward the surface 11' of detector electrode 11, and subsequent dispersion of the sample. Flat-surface electrodes, especially those of planar-disk configuration, appear quite effective in radial spread assuring activity only at the surface of the electrodes, even at relatively large pipette tip-electrode separations, as well as effective washout.

The present invention also employs voltammetric and potentiometric testing for trace metals. Differential pulse stripping voltammograms were obtained with an EG&G PAR 264A voltammetric analyzer, a PAR 303A static mercury-drop electrode, and a PAR 0073 X-Y recorder.

A Tracelab potentiometric stripping unit (PSU 20, Radiometer), with SAM 20 sample station (Radiometer) and an IBM PS/2 55SX, were used to obtain potentiograms. Square—wave stripping voltammograms were obtained with a BAS 100A electrochemical analyzer. Most voltammetric and potentiometric stripping experiments were carried out in 10- and 20-mL cells (BAS and Radiometer) , with the electrodes joining through holes In the cover.

The screen-printed electrodes (ExacTech Blood Glucose Strips of Medisense Inc, for example) were purchased. These strips comprise working (carbon) and reference electrodes printed on a PVC substrate (with carbon contacts on the opposite side) . One printed carbon contact served as a substrate for the mercury film electrode (since the original working-electrode target area is covered with enzyme/mediator layers) . The printed electrode (Ag/AgCl) from another strip served as reference during the voltammetric experiments. Potentiometric stripping work employed the conventional Ag/AgCl electrode of the TraceLab unit. Most experiments employed a platinum wire auxiliary electrode. Some experiments involved a two- electrode system and 100 μL sample drops. For this purpose, the strip was cut in the center, to allow placing of the carbon contact in direct proximity to the printed reference electrode (on a microscope slide) .

All solutions were prepared with double-distilled water. The metal atomic absorption standard solutions (1000 mg/L) were purchased. The supporting electrolyte was an acetate buffer solution (0.02 M, pH 4.8). Drinking water samples were collected from laboratory spigots. The urine samples were obtained from a healthy volunteer. Fumed silica was also obtained.

Anodic stripping voltammetry (ASV) and potentiometric stripping analysis (PSA) were performed in the following manner. The mercury film was preplated from a non-deaerated, stirred, 80 mg/L mercury solution (In 0.02 M HCl), by holding the carbon strip electrode at the deposition potential (-1.15 V for ASV or -0.90 V for PSA) for fifteen minutes. The potential was then switched to -0.20 V (ASV) or -0.05 V (PSA) for a two minute "cleaning" period.

Subsequent ASV and PSA cycles Involved the common metal deposition and stripping steps. Experiments were performed with both stirred and unstirred solutions (during the deposition) , as well as in the presence and absence of dissolved oxygen. The stripping step was performed with a quiescent solution. In ASV, the potential was scanned (usually with a differential pulse waveform) and stopped at -0.20 V. This potential was maintained for sixty seconds before the next measurement was performed. Potentiometric stripping was carried out by applying a constant oxidation current of +1.0 _/.A; the electrode was conditioned for fifteen seconds at -0.05 V before the next deposition-stripping cycle. The mercury film was removed by holding it at +0.40 V (vs. the printed reference electrode) for five minutes.

Screen-printed carbon and Ag/AgCl electrodes of disposable glucose strips were employed, as they are readily available, at a very low cost in connection with the ExacTech blood glucose meter. Since the target working electrode area for glucose testing is covered with the enzyme/mediator layer, the carbon contact — on the opposite side of the strip — was successfully used as substrate for the mercury film.

EXAMPLES (INDUSTRIAL APPLICABILITY) The invention is further illustrated by the following non- limiting examples.

EXAMPLE 1 Figs. 3(a) and 3(b) depict a current vs. time comparison of batch injection analysis (BIA) and flow injection analysis (FIA), respectively. The analyte comprised 20 μL samples of 1 x 10^"* M ferrocyanide. Different stirring rates for BIA were used: a) 0 rpm; b) 250 rpm; and c) 500 rpm. The flow rates for FIA were: a) .2 mL/min. ; b) .5 mL/min. ; and c) 1 mL/min. In both instances, the cell electrolyte was .1 M KCl; the electrode-tip distance (in BIA) was 2 mm. Both BIA and FIA resulted in sharp peak readouts with good resolution. A notable distinction, however, is evident in the peak broadening which occurred with FIA due to dispersion in the flow channel. In contradistinction thereto, the peak sharpness of BIA resulted from rapid "wash-out" due to stirring. The peak widths (after dispersion) are 1.4, 1.2, and 1.0 seconds for 0, 250, and 500 stirring rpm, respectively. The latter stirring speed corresponds to a sample injection rate of 720 samples per hour. Further, the data of the example indicated that batch injection analysis compared favorably with flow injection analysis in terms of sensitivity and detection limits.

EXAMPLE 2 Figs. 4(a) and 4(b) illustrate repetitive sample testing using ferrocyanide (Fig. 4(a)) and hydroquinone (Fig. 4(b)) solutions. Sixty repetitive samples of 5 x 10^"* M ferrocyanide and hydroquinone were injected into a cell containing .1 M KCl as electrolyte for the ferrocyanide and .1 M phosphate buffer for hydroquinone. The graphs indicated no apparent change in peak currents, response times, or baseline. The standard deviations, 1.6% for ferrocyanide and 1.1% for hydroquinone, compared favorably with the flow injection analysis technique. It is believed this consistency is largely attributable to micropipette-based injections, albeit manual. Consistency may be improved using robotic injection.

Further, this example demonstrated that the "build-up" of sample solutions does not affect batch injection analysis. This fact is attributed to the great dilution factor as well as the fact that the cell solution or electrolyte is inert relative to the electrode.

EXAMPLE 3 Fig. 5 further illustrates the fact that sample build-up does not affect batch Injection analysis performance. In this example,

20 μL of 50 mM ferrocyanide solution samples were injected over a 20 minute period (peaks 1-5 represent such injections) The electrolyte again was .1 M KCl and stirring rate was 250 rpm. Additionally, after each injection, a 2 mL volume of the same solution was added (A—E) from hole 18 in cover 13. Although each such addition was equivalent to 100 injections, the batch injection analysis peak currents remained essentially the same, and no baseline drift was observed. Deterioration of the baseline was observed only after 400— 500 equivalent injections (near point E) . Since 500 repetitive injections are normally standard under flow injection analysis, this represents a highly consistent and satisfactory performance. Drainage and replacement of electrolyte is thus indicated after approximately 500 repetitive injections, or by baseline deterioration.

EXAMPLE 4 Fig. 6 depicts a comparison of two injected samples of 2.5 x 10^" M acetaminophen, one sample (a) containing .01 M KCl electrolyte, the other not. The cell solution was .01 M KCl and stirring rate was 250 rpm. As clearly indicated, a well defined and sensitive response was observed for sample (b) . This example indicated great potential exists for assays of resistive or nonaqueous samples.

Batch injection analysis relies upon highly specific sensors or reactive sensor surfaces. With reactive sensor surfaces, reaction occurs at or in close proximity to the sensing surface. The analyte is thereby converted to a detectable species.

EXAMPLE 5

Figs. 7(a) and 7(b) illustrate the use of an enzymatic electrode (Fig. 7(a)) comprising glucose oxidase (10%), and a non- enzymatic inorganic electrode (Fig. 7(b)) comprising 20% ruthenium dioxide. Glucose concentrations in Fig. 7(a) represent three 50 μL injections ((1), (2), (3)) of .5 mM, 1.0 mM, and 1.5 mM, respectively. The electrolyte was .1 M phosphate buffer and 1 M NaOH. Fig. 7(b) demonstrates that inorganic electrodes are also effective with carbohydrates. Obviously, other chemical and biological modifications of electrodes are possible. Other schemes include providing electrodes with selective coatings or membrane coverings, as well as optical or thermal devices, which may allow dilution, filtration, dialysis, and the like, on the detector surface.

EXAMPLE 6 Fig. 8 demonstrates variations of peak sensed current vs. injected volume, as well as current variation dependence upon electrode—pipette tip distance. The analyte was 5 x 10- M hydroquinone: curve A corresponds to an electrode-tip distance of 2 mm; curve B corresponds to an electrode-tip distance of 5 mm; while curve C depicted results with a 10 mm tip-electrode separation. The electrolyte was .1 M phosphate buffer (pH 7.4). Currents increased rapidly to a volume of approximately 50 μL, then flattened somewhat. The data gave slopes of .4 on a log—log scale.

Fig. 9 corroborates this data. Curve A, representing a 10 μL sample of 5 x 10^~* M hydroquinone, and curve B representing a 50 μL sample of 5 x 10^-4 M hydroquinone, both indicated decreasing responses with increased electrode—tip distances. Similarly, responses for .5 mM ferrocyanide (not shown) increased linearly with Increasing radius (1-3 mm) of the electrode surface; again, the electrolyte was .1 M phosphate buffer.

Other factors affecting batch injection analysis response have been tested. Ten successive injections of 50 μL acetaminophen, hydroquinone, and ascorbic acid in increasing concentration (from 25 to 250 μM) were administered using a phosphate buffer electrolyte. In all cases, batch injection analysis response increased linearly with Increasing concentration.

Amperometric peaks for injections of 20 μL solutions of 2.5 x 10^-6 M ascorbic acid and hydroquinone were used to estimate the detection limits, with an applied potential of +0.9 V. Signal—to- noise ratios of 50 and 30 were obtained at this trace level.

Extrapolating to a signal—to-noise ratio of 3, these data correspond to detection limits of 5 x 10^~8 M hydroquinone and .3 X 10^~8 M ascorbic acid: these are lower values than those obtained with analogous flow injection analysis measurements.

While amperometric measurements have generally been shown as providing excellent responses, other detection schemes are available. Potentiometric measurements using ion—selective electrodes, optical, or thermal sensors are particularly appealing. Such electrodes require little or no sample pretreatment. The high specificity of ion—selective electrodes renders such electrodes very attractive for batch injection analysis, where the lack of solution handling requires active or selective detectors.

Fig. 10 depicts an electrochemical cell 10' suitable for potentiometric measurements. Ion—selective electrode 11' is inserted through the bottom of cell 10' . An aperture 17' in cover 13' was used for introducing, for example, a standard Eppendorf micropipette 12' , the tip of which is positioned a known distance (usually 2 mm) from ion-selective electrode 11' . Reference electrode 16' (Ag/AgCl, for example, a model RE-1 from BAS, Inc.) is also mounted in cover 13' . Buffer solution and cell solution are added through aperture 17' . Magnetic stirring bar 19' is rotated by a magnetic stirrer (not shown) below cell 10' .

Potentiometric measurements were performed with an amplifier having a gain range of 2.5-1,000; the millivolt outputs were recorded in an OMNISCRIBE strip-chart recorder.

Both spherical (for example, a BECKMAN model 39831) and planar (for example, a MARKSON, model 989B) electrodes were used for pH measurements. Fluoride and chloride electrodes (for example, ORION models 940900 and 941700, respectively) were used for fluoride and chloride concentrations, respectively.

All pH measurements were conducted with a cell solution of .1 M phosphate buffer containing .25 M potassium chloride (pH 7.00). Chloride measurements were performed in a .05 M KH₂P0₄ solution containing .25 M NaN0₃. The pH of such solutions was adjusted to 6 using NaOH. Fluoride experiments were performed in a solution containing .2 M sodium acetate, .17 M acetic acid, .35 M sodium chloride, and 1 gram/liter 1,2 diaminocyclohexane-N,N,NJN) tetraacetic acid (DCTA) . Some fluoride measurements were performed in the presence of .5 mM aluminum sulfate. Juice samples (e .g. , CAMPBELLS^® Tomato Juice and SPICY HOT V8™ Juice) were filtered through a .45 μ filter before measurement. The coffee sample was prepared by dissolving .3g of MOUNTAIN BLEND™ coffee in 25 mL of distilled water. Tap water was obtained from laboratory spigots.

EXAMPLE 7

Fig. 11 shows the responses obtained with sequential triple injections of 20 μL solution of pH 4 (A) and pH 10 (B) samples. Despite the great difference in hydrogen ion concentration, batch injection analysis shows no observable carryover. The sharp peaks and clear baseline indicate effective transport to and removal from the detector surface. Readout is within a few seconds following injection, indicating the possibility of high sampling rates. This is borne out by Figs. 12(a) and 12(b), which Indicate responses for pH 10 samples at a rate of 720 injections/hour (A) and 360 in ections/hour, respectively.

EXAMPLE 8 Figs. 13(a) and 13(b) illustrate the criticality of flat ion- selective detection electrodes. Forty repetitive injections of 20 μL pH 10 solution using spherical (Fig. 13(a)) electrodes and flat (Fig. 13(b)) electrodes gave the Indicated responses. The flat electrodes yielded a relative standard deviation of 1.3%, while the spherical electrodes yielded a relative standard deviation of 6.8%. The superiority of flat electrodes insofar as reproducible response is thus demonstrated; it is believed that flat electrodes facilitate rapid "wash-out," or dispersion of sample. Again, automation of the injection process can further Improve response.

EXAMPLE 9 Fig. 14 shows the results of three injections of 20 μL of pH 10 solution. At points (1) and (2) , 6 mL of pH 2 solution was added through the port 17' in cover 13' (see Fig. 10). Despite the fact that each such addition was equivalent to 300 injections, peak potentials remained essentially the same and no baseline drift was observed. This lack of "memory" effect is believed due to the tremendous dilution capability of batch injection analysis.

EXAMPLE 10

Fig. 15 depicts the results of a series of different pH solutions injected in triplicate, alternating between high and low pH levels (pH 2—11). The batch injection analysis responses were quick and sharp. The plot of peak potential vs. pH was linear (slope of 65 mV/pH, with a correlation coefficient of .998) not shown.

Fig. 16 depicts practical application of the above. CAMPBELLS® Tomato Juice (a) , tap water (b) , SPICY HOT V8 Juice (c) , and coffee (d) , gave the responses indicated against buffer standards of pH 10 (e) and pH 4 (f) . Results were consistent with conventional pH measurement techniques. EXAMPLE 11 Figs. 17(a) and 17(b) depict "carryover" effects using chloride (Fig. 17(a)) and fluoride (Fig. 17(b)) ion-selective electrodes, respectively. Injection concentrations were in triplicate and alternatingly high (100 mM) and low (5 mM) . The stirring rate for Fig. 17(a) was 300 rpm; for Fig. 17(b), 500 rpm. Sample volumes in all cases were 50 μL injected at a rate of 180/hour. The electrolyte solution for Fig. 17(a) was .05 M potassium dihydrogen phosphate (KH₂P0 ) and NaN0₃; the Fig. 17(b) cell solution was .2 M CH₃COONa; 0.17 M CHgCOOH, .35 M NaCl, are 1 g/L DCTA (1,2 diaminocyclohexane-

N,N,N|N; tetraacetic acid). The chloride electrode (Fig. 17(a)) gave faster responses, while the fluoride electrode (Fig. 17(b)) was more sensitive. The rapid "wash-out" characteristics and fall off to the baseline are attributed to the planar electrode configuration coupled with the huge dilution factor. These data indicate that batch injection analysis can tolerate numerous injections with little or no memory effects.

This is further borne out by the data of Figs. 18(a) and 18(b). Using the same cell solutions of Fig. 17, Fig. 18(a) depicts 100 mM chloride injections, while Fig. 18(b) shows the results of .5 mM fluoride injections. The arrows indicate massive additions of 2 mL of 100 mM chloride solution (Fig. 18. (a)), while in Fig. 18(b) the arrows indicate additions of 6 mL of .5 mM fluoride solution. Sample volumes in Fig. 18(a) were 50 μL; in Fig. 18(b), 20 μL. Additionally, the cell solution in Fig. 18(b) also contained .5 mM aluminum sulfate. Despite the "ramp" effect in Fig. 18(a), there was no effect upon the potential peaks. Similarly, in Fig. 18(b), the aluminum ion prevented a buildup of the fluoride ion concentration.

EXAMPLE 12 Fig. 19 depicts the effects of sample volume (A) and micropipette tip-electrode distance (B) on the response. Similar results had been obtained in voltammetric measurement. Peak potential response increased rapidly to 50 μL, then flattened. Response also increased sharply with tip—electrode distance (to 2 mm, after which response dropped rapidly) (Fig. 19(b)). EXAMPLE 13 Figs. 20(a) and 20(b) illustrate the high reproducibllity of batch injection analysis using ion-selective electrodes. Two series of 50 and 20 repetitive injections of 25 mM chloride (Fig. 20(a)) and 1 mM fluoride (Fig. 20(b)), respectively, yielded responses with standard deviation of only 2.4% and 1.4%, respectively, even when injected at rates of 100 and 50 samples/hour, respectively. This again demonstrated that batch injection analysis can tolerate the presence of analyte in the cell even after numerous injections.

Although not illustrated, repetitive in ections of 50 μL chloride and fluoride solution of concentrations of .5-10 mM and 10- 5,000 μM ranges, respectively, were used to assess linearity and detection limits. Plots of peak potential vs. log-concentration were linear (slopes of 57.8 and 56.8 mV/decade, respectively). Detection limits of .1 mM (.18 μg) chloride and 2 μM (2 ng) fluoride ions were estimated based on a signal—to-noise ratio of 3.

EXAMPLE 14 Fig. 21 compares stripping voltammograms for a solution containing 25 μg/L cadmium, 40 μg/L lead, and 35 μg/L copper, obtained under Identical conditions at a mercury—coated screen- printed carbon electrode (Fig. 21(a)), a glassy carbon electrode (Fig. 21(b)), and a hanging mercury drop electrode (Fig. 21(c)). Further, (a) and (b) used screen-printed Ag/AgCl reference electrodes, while a conventional Ag/AgCl electrode was employed In (c). Preconcentration was for three minutes at —1.15 V with stirred (400 rpm) deaerated solutions. A differential pulse waveform of 10 mV/s scan rate and amplitude of 50 mV was applied. The mercury—coated screen-printed electrode of the Invention exhibits well-defined, sharp stripping peaks, good resolution between neighboring signals, low background current, and a wide potential window. A relatively short (three minutes) preconcentration time apparently allows convenient quantitation of μg/L (parts per billion) concentration. Comparison to the traditional hanging mercury drop or glassy carbon electrodes indicates that sensitivity and overall signal—to-background properties are not compromised by the use of the screen-printed carbon substrate electrode of the present invention. It is further noted that the use of a screen-printed reference Ag/AgCl electrode resulted in approximately a 200 mV negative shift in peak potentials.

Figs. 22 depict scanning electron microscopy micrographs of bare (Fig. 22(a)) and mercury-coated screen-printed carbon electrodes (Fig. 22(b)). The bare carbon strip presents some roughness and discontinuity; the mercury deposition (Fig. 22(b)) resulted in numerous spherical microdroplets of 1-2 μm diameter, covering approximately 20% of the area. Under the same plating conditions, different microdistributions of the droplets occur on the carbon strip and glassy carbon electrodes; the strip exhibits a more favorable "array-like" behavior.

The microdistribution of mercury droplets in Figs. 22 enhances deposition efficiency from quiescent solutions. As a result of the nonlinear diffusional flux to the Individual droplets, and the distribution of the droplets, high ratios of current peaks in quiescent (ip, q) and stirred solution (ip ,s) are obtained. For example, an ip , q/ip , s value of .25 was estimated from the voltammetric stripping response for 30 μg/L lead following three minutes deposition. Analogous measurements at a mercury—coated glassy carbon surface yielded a value of 0.10. Apparently fewer surface sites for mercury plating are available on the carbon strip resulting in enhanced microarray character.

The data of Figs. 21 was obtained using common stripping conditions (a deaerated solution stirred during the deposition step) , decentralized stripping applications will usually require elimination of a nitrogen purge and convection (such as created by stirring) transport.

EXAMPLE 15 Figs. 23 illustrate the voltammetric (Fig. 23(a)) and potentiometric (Fig. 23(b)) stripping responses for screen-printed electrodes for quiescent (non-stirred) , non-deaerated solution of increasing lead concentration, from 20-100 μg/L (shown as 1-5) . Preconcentration was for 120 seconds at -1.15 V for Fig. 23(a) and -0.90 V for Fig. 23(b) with a quiescent non-deaerated solution. Contact current potentiometric stripping was employed at +1.0 μA. Despite these conditions and a short (two minutes) deposition period, well-defined peaks were observed. The five peaks depicted represented part of a series of ten 10 μg/L concentration increments. The calibration plots were linear over the entire range, with slopes of 17 nA • L/μg (Fig. 23(a))), and 1.37 mm² • L/μg (Fig. 23(b)), and a correlation coefficient of 0.999.

The sharper peaks and lower background response of the potentiometric stripping analysis (Fig. 23(b)) make it more attractive under quiescent, non-deaerated conditions.

EXAMPLE 16 Figs. 24 illustrate voltammograms (Fig. 24(a)) and potentiograms (Fig. 24(b)) for 50 μg/L lead in the presence of dissolved oxygen in a stirred, non-deaerated solution. Preconcentration was varied (1-7) in 40 second steps from 0 to 240 seconds. The larger the preconcentration period, the larger the response. However, even short preconcentration periods (40-80 seconds) exhibited well—defined peaks.

Fig. 24(c) depicts a plot of responses vs. preconcentration time plots, the voltammetric responses exhibit linear dependency while the potentiometric responses showed nonlinear dependency.

Larger preconcentration times allow convenient quantitation of s b-μg/L trace metal concentrations. Detection limits of 30 and 50 ng/L (parts per trillion) lead and cadmium, respectively, were estimated from voltammetric stripping measurements of 1.0 and

0.5 μg/L stirred and deaerated solutions of these metals following ten minutes preconcentrations. Analogous potentiometric stripping measurements of a non-deaerated solution yielded detection limits of 0.3 and 0.4 μg/L lead and cadmium, respectively.

Also evaluated was the effect of deposition potential over a range from -0.60 V to -1.20 V with a two minute deposition time for a non-stirred, non-deaerated 50 μg/L lead solution. The potentiometric response increased gradually between -0.60 V and -1.1 V, then levelled off. EXAMPLE 17 Figs. 25 illustrate square—wave stripping voltammograms with a mercury-coated carbon strip (Fig. 25(a) and a glassy carbon electrode (Fig. 25(b) for aerated (solid line) and deaerated (dotted line) solutions containing 30 μg/L lead. Preconcentration was for three minutes at -1.15 V with a stirred solution. The square wave amplitude was 30 mV with steps of 4 mV at a frequency of 30 Hz. The electrolyte was a 0.02 M acetate buffer with pH of 4.8. With the screen-printed mercury-coated electrode (Fig. 25(a)), the square-wave responses for both aerated and deaerated solutions were similar. In contrast thereto, the response at the mercury-coated glassy carbon electrode (Fig. 25(b)) clearly indicates a significant oxygen contribution. Apparently, the microarray character of the mercury- coated carbon strip electrode facilitates depletion of oxygen from its surface.

EXAMPLE 18 In view of the proposed field and decentralized uses of the preferred embodiment of the invention, microliter vice 10 mL solution stripping analysis, is particularly appropriate. Accordingly, the screen-printed carbon and Ag/AgCl reference electrodes were placed in direct contact on a microscope slide. Repetitive potentiometric stripping measurements of the two-electrode systems, as depicted in Figs. 26, comprised testing 50 μg/L lead solution in the same 100 μL drop (Fig. 26(a)) and in different drops (Fig. 26(b)). Preconcentration was five minutes at —1.15 V, and the solutions were quiescent and non-deaerated. The current employed was a constant +1.0 μA.

Well-defined peaks were observed despite the non-deaerated, unstirred, low concentration samples. The relative standard deviations for these series were 5.2 (Fig. 26(a)) and 3.9

(Fig. 26(b)). Testing of a three-electrode (not shown) revealed no apparent differences in responses.

EXAMPLE 19 Screen-printed electrodes also hold great promise for re-usable applications. Figs. 27 depict voltammograms (Fig. 27(a)) and potentiograms (Fig. 27(b)) for twenty repetitive stripping measurements of 100 μg/L and 50 μg/L lead solutions, respectively. Preconcentration was 120 seconds with an unstirred, non-deaerated solution.

In both stripping schemes, the peaks remained the same. The i relative standard deviation for these series was 2.4% (Fig. 27(a)) and 3.2% (Fig. 27(b)). Again, the potentiometric analysis provided a more favorable response in the presence of oxygen. The stable responses, coupled with the low cost, make screen—printed electrodes an attractive alternative to prior art electrodes.

EXAMPLE 20

Figs. 28 and 29 Illustrate the applicability of screen-printed electrodes to the analysis of urine and drinking water samples.

Voltammograms (Fig. 28(a)) for the urine sample comprised five minutes preconcentration at -1.15 V and pulse amplitude of 25 mV. The solution was deaerated and stirred. Successive concentration increments of 10 μg/L were added. Fumed silica was also added to the sample to "collect" organic surfactants.

The potentiograms of Figs. 29(a) and 29(b) were the result of a potentiometric stripping analysis of drinking water. The solution was also stirred but non-deaerated. Successive concentration increments (2,3) of 5 μg/L were added.

In both tests, well-defined peaks resulted. Lead sample values of 10.2 μg/L (Fig. 28) and 4.7 μg/L (Fig. 29) were calculated.

In conclusion, the above results demonstrate for the first time that screen-print electrodes are suitable for stripping measurements of trace metals. These extremely low cost electrodes function in a manner comparable to traditional stripping electrodes, with no compromise in performance. Neither deoxygenation nor stirring is ^• required; the electrodes hold great potential for decentralized (clinical, environmental, or Industrial) testing. These applications will ultimately require the development of small inexpensive portable stripping analyzers. Single-use applications will require complete stripping of the mercury prior to disposal of electrodes. Certain applications, for example, decentralized testing for blood lead level, will require the adaptation of simple and rapid sample preparation, for example, acidification. Additional coverage of electrodes with other layers may also be required.

The preceding examples can be repeated with similar success by substituting the generically or specifically described reactants and/or operating conditions of this invention for those used in the preceding examples.

Claims

CLAIMSWhat is claimed is:

1. An apparatus for measuring analyte sample concentration comprising: vessel means; inert electrolyte means confined within said vessel means; means for introducing analyte samples to be analyzed into said vessel means; detection means in said vessel means for sensing said analyte samples to be analyzed; and means for analyzing said analyte samples; wherein the ratio of analyte sample volume to electrolyte volume is within the range of 1/50,000 to 1/20,000.

2. The invention of claim 1 wherein said electrolyte means comprises at least one solution selected from the group consisting of potassium chloride, a phosphate buffer and sodium hydroxide, potassium dihydrogen phosphate and sodium nitrate, and CH₃C00Na, CH₃C00H, NaCl, and 1,2 diaminocyclohexane-N,N,N;NJ tetraacetic acid.

3. The invention of claim 1 wherein said detector means comprises planar carbon paste electrode means.

4. The invention of claim 1 wherein said means for analyzing said analyte samples comprises amperometric means.

5. The invention of claim 1 wherein said means for analyzing said analyte samples comprises potentiometric means.

6. A method for measuring analyte sample concentration comprising the steps of: a) providing a vessel; b) confining an inert electrolyte within the vessel; c) introducing samples to be analyzed into the vessel; d) sensing the samples to be analyzed; and e) analyzing the samples; wherein the ratio of analyte sample volume to electrolyte volume is within the range of 1/50,000 to 1/20,000.

7. The method of claim 6 wherein the step of sensing the analyte samples to be analyzed comprises the step of detecting the samples with a planar carbon paste electrode.

8. The method of claim 7 wherein the step of detecting the samples with an electrode comprises providing an ion-selective electrode.

9. A method of analyzing trace metals comprising the steps of: a) providing a plurality of flat printed electrodes; b) coating at least one of the plurality of flat printed electrodes with mercury; and c) analyzing a sample for heavy metal content with the plurality of electrodes; wherein said step of providing a plurality of flat printed electrodes further comprises the step of providing at least one flat screen-printed Ag/AgCl reference electrode, and at least one flat screen-printed carbon electrode.

10. The method of claim 9 wherein the step of analyzing a sample for heavy metal content further comprises the step of voltammetrically analyzing the sample.

11. The method of claim 9 wherein the step of analyzing a sample for heavy metal content further comprises the step of potentiometrically analyzing the sample.

12. The method of claim 9 wherein the step of analyzing a sample for heavy metal content further comprises the step of analyzing microliter sample solutions of urine and drinking water.

13. Apparatus for heavy metal trace testing comprising: 5 means for providing a plurality of flat printed electrodes; means for coating at least one of said plurality of flat printed electrodes with mercury; and means for analyzing a sample for heavy metal content with 10 said plurality of electrodes; wherein said means for providing at least one of said plurality of flat printed electrodes comprises means for providing at least one flat screen—printed Ag/AgCl reference electrode, and at least one flat screen-printed carbon electrode.

15 14. The apparatus of claim 13 wherein said means for analyzing a sample for heavy metal content further comprises means for voltammetrically analyzing said sample.

15. The apparatus of claim 13 wherein said means for analyzing a sample for heavy metal content further comprises means

20 for potentiometrically analyzing said sample.

16. The apparatus of claim 13 wherein said means for analyzing a sample for heavy metal content further comprises means for analyzing microliter sample solutions of urine and drinking water.