Printer Friendly

Hand-held voltammetric analyser for real-time monitoring of heavy metals in situ.

Introduction

Historical urban and industrial activity such as steel making, coal gas manufacture, mining and shipbuilding, have left a legacy of land and water contamination from highly toxic heavy metals which seriously affect human health and the environment. Apart from the direct effects on biological life and the ecosystem in general, water or land contamination can cause significant economic and financial damage. The Confederation of British Industry has estimated that as much as 200,000 hectares of land is contaminated in the UK and remediation of these areas could cost up to [pounds sterling]20 billion. Therefore, the necessity of monitoring pollutant levels at various points in industrial and recycling processes, in natural water, and at agricultural, urban and industrial sites is highly important.

Recent research on the detection of heavy metals using electrochemical measurements has focused on computer-based systems. The advantages of these systems may be summarised thus:

* they have the ability to manage experiments involving sophisticated measurement techniques

* large amounts of data may be stored

* data may be manipulated in complex ways,

* digital filtering may be performed, and

* the results can be present in a convenient format.

Several general-purpose electroanalytical instruments based upon computers are available commercially.--These instruments, however, are expensive, complicated to operate and bulky, and so they are impractical to use for on-site monitoring. It follows that a fast, reliable, relatively inexpensive, portable (hand held) and independent instrument which is capable of direct monitoring of heavy metal contaminants in-situ, is very desirable.

This paper describes the development of such an electrochemical instrument which is capable of gathering real-time quantitative data on a range of heavy metal contaminants. The unit has been developed for use on sites of contaminated land or water. In addition, it is able to determine the oxidation state of a metal, which is a measure of the metal's toxicity.

Instrumentation

The schematic diagram for the electrochemical instrument as developed is shown in Fig.1. It consists of five main units: the waveform generator, the potentiostat, the cell (sensor), the data acquisition system (comprising the Programmable Gain Amplifier, ADC and data interfaces) and the microcontroller.

[FIGURE 1 OMITTED]

Among the different electrochemical techniques suitable for heavy metal detection, differential-pulse anodic stripping voltammetry (DPASV) was chosen in preference to the other techniques, because it is a precise analytical method and has been widely used for the trace determinations of several heavy metals[1,2] with excellent limits of detection.

The excitation signal for DPASV consists of two parts. The first part is a pre-concentration step of a fixed duration (60sec) during which the application of a negative potential (-1.4V) causes the deposition of the analyte species onto the working electrode surface. The second stage is a stripping stage, in which the metal is oxidised back into solution by means of a time-controlled excitation waveform where the applied potential is scanned from the negative starting voltage of -1.4 V towards a positive end voltage of +1.0 V. The excitation waveform consists of small pulses (120ms period and 50ms duration) of constant amplitude (25mV) superimposed upon a staircase waveform of 2mV step potential. This waveform is provided by a specially designed waveform generator. The signal generator function is provided by the 16bit microcontroller which synthesises the excitation signal from its digital data equivalent. This approach allows signal generation to be very flexible the signal generator becomes very flexible, as all of the parameters can be re-programmed and a complex waveform can be generated very easily. The signal in digital form is converted to analogue using a digital to analogue converter (DAC) with a 12-bit resolution. The excitation signal is applied to the sensor via a potentiostatic circuitry, which controls the voltage potential applied to the cell. It also acquires and amplifies the current flowing through the cell. The current is sampled twice in each pulse period (once before the potential step, and then again just at the end of the pulse). The difference between these two current samples is recorded. This process is repeated for all the pulses of the signal.

The sensor (cell) uses solid (screen-printed) working electrodes[3] (fig.2) rather than the more common hanging mercury drop electrodes (HMDE) which are used by most traditional laboratory instruments. Apart from the size and cost difference, an important advantage of solid electrodes is that they do not require the use of toxic mercury.

[FIGURE 2 OMITTED]

In the data acquisition process, the electrochemical response current is firstly converted to a voltage, then amplified before being converted, at fixed time intervals, to digital data using a 16-bit analogue to digital converter (ADC). The resultant 16-bit binary words which convey the current information are stored in the microcontroller memory. The full scale range (FSR) and resolution of the data acquisition unit is set by a programmable amplifier under the control of the microcontroller. For very small signal amplitudes, the system FSR is set to 400 [micro]A corresponding to a 6.1 nA resolution; for higher amplitudes the FSA is set to 30 mA giving a resolution of 458 nA.

An important characteristic of the system is its portability, and so it is designed for battery operation. DC to DC converters supply internal logic (5 V) and linear ([+ or -] 12V) supplies from a single battery. The battery capacity of 1.4Ah delivers approximately eight hours of continuous operation at a power consumption of approximately 2 W.

Fig.3 shows the prototype design for the instrument, fitted within a box measuring 200 x 100 x 60mm.

[FIGURE 3 OMITTED]

Detection and Identification of Heavy Metals

An identification technique based on the probability density functions (PDF) of oxidation potential measurements has been developed. PDF curves have been used in the development of decision algorithms for feature selection in various applications[4,5]. Data are first arranged into a numerical order from which various statistical features are obtained, such as minimum, maximum, mean, median, and standard deviation. These are then used as a basis for classification.

The probability (p) for a feature which assumes a value between [f.sub.1] and [f.sub.2] is given by[6]:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (1)

where, pf) is the probability density function off.

In the case of Gaussian distribution, the probability density function pf) is given by:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2)

where [mu] is the mean value and [sigma] is the standard deviation.

As the potential of the excitation signal approaches the oxidation potential of one of the metals dissolved onto the electrode surface, the ions of that metal pass into the solution from the electrode. The current increases rapidly and reaches a maximum value (peak current) when the applied potential approximates to the metal's oxidation potential ([E.sub.p]). When an actual test is carried out, the oxidation potential [E.sub.p] is assessed and examined against a PDF to determine the probability of membership of that analyte with all analytes stored in a database of PDF measurements. The analyte representing the highest probability of likeliness is thus identified. Analytes identified in this way are automatically given a probability of likeliness, indicating the prediction accuracy. Fig.4 shows a block diagram for the whole process of identification by this means.

[FIGURE 4 OMITTED]

Measurements and Results

Test solutions of lead, cadmium, zinc, and copper were prepared at different concentration levels in the range of 1 to 50ppm. The four six ions selected for examination were chosen due to their importance as environmental pollutants. The solutions were prepared using ionised water with a supporting electrolyte of 0.1M sodium chloride (NaCl) and placed in turn in a sample reservoir with a 40ml capacity. The pre-concentration (deposition) time was 60sec at -1400mV potential. The scanning voltage range used was from -1400 to +1000mV, with step potential of 2mV, pulse height of 25mV, with a scan rate of 120ms and pulse duration of 50ms.

Thirty two independent measurements of the electrical potential and peak amplitude of all four metals were recorded. The instrument was also connected to a personal computer, and the results were monitored for comparison with the values obtained from the liquid-crystal display.

It was found that the lower possible concentrations of metals that can be detected by the instrument are 1ppm for [Lead.sup.II], 1ppm for [Cadmium.sup.II], 3ppm for [Zinc.sup.II], and 1ppm for [Copper.sup.II]. Fig.5 shows the differential-current voltammogram for each of the above metals at their lower detectable concentration level.

[FIGURE 5 OMITTED]

The calibration graphs for all of the metals were approximately linear, as shown in Fig.6, and relatively reproducible. The coefficients of the first order curve-fitting equations (calibration curves) were stored in system memory. Using these equations, if the metal is identified, the concentration level of that metal can be obtained.

[FIGURE 6 OMITTED]

It was found that the electric potential ([E.sub.P]) for every metal fits a Gaussian distribution with mean value [[micro].sub.E] and standard deviation [[sigma].sub.E]. The probability density function [G.sub.E]([E.sub.P]) (as given in the eq. 3 below) was then evaluated and stored for each.

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (3)

The statistical parameters of the electrical potential measurements are summarised in Table 1.

The above statistical measurement data was then used to develop a PDF-based identification technique. In this method, when an actual test is carried out, the electric potential corresponding to peak current is assessed and examined against a PDF to determine the probability of membership of that species with all metals stored in the database. The metal representing the highest probability of likeliness is identified. Metals identified in this way are automatically given a probability of likeliness, indicating the prediction accuracy.

The detection and identification of metals in multi-metal mixtures was also investigated. For this experiment one solution containing four different metals (Lead, Cadmium, Zinc and Copper) was prepared where the concentration of all three metals was 5ppm. Fig.7 shows the differential pulse voltammogram taken by the instrument. The four peaks represented the oxidation potential of the metals can be clearly seen. Using the identification algorithm, the system reported the following results: [Cadmium.sup.II] at 4ppm concentration, [Lead.sup.II] at 7ppm concentration, [Copper.sup.II] at 6ppm concentration, and [Zinc.sup.II] at 8ppm concentration. The four metals were thus easily identified by the instrument and their concentration was obtained with maximum error of 3ppm.

[FIGURE 7 OMITTED]

The instrument was also used to examine possible contamination of field samples. Five samples (given by SureClean Company, Aberdeen) were examined for heavy metal contamination. The only preparation of the samples was the addition of sodium chloride (NaCl) supporting electrolyte at a concentration of 0.1M. The samples were tested using the prototype instrument which gave the following results (Table 2):

These results have been verified with a commercial available electrochemical instrument (EG&E 394 trace analyser) which gave the differential voltammograms shown in Fig. 8.

[FIGURE 8 OMITTED]

As can be seen from this figure, the voltammograms of the four different samples verifies the results obtained by the portable instrument (presence of Zinc) due to the current peak positions which are close to the oxidation potential of [Zinc.sup.II]. It also can be seen that sample one has the highest Zinc concentration compared to the other three samples, and sample four the lowest. This also verifies the results obtained by the portable instrument.

Conclusions

A novel hand-held electrochemical instrument for detecting, identifying and measuring concentrations of heavy metals has been developed. It provides those facilities found in a traditional laboratory based instrument, but in a portable design. In contrast to existing commercial systems, it can stand alone without the need of a computer. The sensitivity of the system has been assessed and was found that metals can be detected at low concentrations up to 1ppm [Pb.sup.II], 1ppm [Cd.sup.II], 3ppm [Zn.sup.II] and 1ppm [Cu.sup.II]. An identification algorithm based on statistical information of oxidation potentials has been developed. The instrument capability of detecting metals in a multi-element solution has also been examined demonstrating good results. The system was also applied in a field test, where it was able to assay the contamination of four samples in a fast, easy and accurate procedure. These results were verified by a commercially available laboratory instrument.

References

[1] Wang J., "Stripping Analysis: Principles, Instrumentation and Applications", VCH Publishers, Deerfield Beach, FA, 1985, 119.

[2] Wang J. and Bruntlett C., "Advanced electroanalytical techniques versus atomic-absorption spectrometry, inductively-coupled plasma-atomic emission-spectrometry and inductively-coupled plasma-mass spectrometry in environmental-analysis", Analyst, 1994, Vol. 119, 219-232.

[3] McStay D. et all, "A Multi-capability Sensor for Hydrocarbons, Synthetic-based Fluids and Heavy Metals: Applications for Environmental Monitoring During Removal of Drill Cutting Piles", Journal of the Society for Underwater Technology, 2002, Vol.25(2), 69-75.

[4] Rose J.L.: "A 23 Flaw Sorting Study in Ultrasonics and Pattern Recognition", Materials Evaluation, July 1977, 87-92.

[5] Rose J.L., Y.H. Jeong and M.J. Avioli: "Utility of a Probability-density-function Curve and F-maps in Composite-material Inspection", Experimental Mechanics, April 1982, 155-160.

[6] Stanley L.T.: "Practical Statistics for Petroleum Engineers", The Petroleum publishing company, USA, 1973, 15-20.

* (1) Dr. Konstantinos Christidis, (2) Prof. Peter Robertson and (3) Prof. Pat Pollard

(1) Hellenic Civil Aviation Authority, Electronics Kos Airport, Kos, 85302, Greece.

(2) Centre for Research in Energy and the Environment, School of Engineering, The Robert Gordon University, Aberdeen, AB101FR, UK

(3) Centre for Research in Energy and the Environment, School of Engineering, The Robert Gordon University, Aberdeen, AB101FR, UK

* Corresponding Author E-mail: christidis_k@yahoo.co.uk
Table 1: Statistical parameters of electrical potential
measurements

METAL ELECTRICAL POTENTIAL

 Mean [[mu].sub.E] [mV] Std [[sigma].sub.E] [mV]

Lead(ii) -568 26.31
Cadmium(ii) -821 67.99
Copper(ii) -183 64.82
Zinc(ii) -1002 25.42

Table 2: Results using the prototype instruments

Sample No Metal present Concentration (ppm)

1 [Zinc.sup.II] 137
2 [Zinc.sup.II] 117
3 [Zinc.sup.II] 98
4 [Zinc.sup.II] 83
COPYRIGHT 2012 Research India Publications
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2012 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Christidis, Konstantinos; Robertson, Peter; Pollard, Pat
Publication:International Journal of Applied Environmental Sciences
Date:Jul 1, 2012
Words:2351
Previous Article:GIS-based multicriteria decision analysis applied for environmental issues; the Greek experience.
Next Article:Global warming: a review of the year our worst fears are exceeded by reality.
Topics:

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters