Cathodic stripping voltammetric determination at a hanging mercury drop electrode of the environmental heavy metal precipitant trimercapto-s-triazine (TMT)

E L S E V I E R Talanta 44 (1997) 497-500

Talanta

Short communication

Cathodic stripping voltammetric determination at a hanging mercury drop electrode of the environmental heavy metal

precipitant trimercapto-s-triazine (TMT)

Arnold G. Fogg a,,, Razali Ismail a, A. Rahim H.M. Yusoff a, Rahmalan Ahmad b, Flo r in G. Ban ica ~"

" Chemistry Department, Loughborough University, Loughborough, Leicestershire LEII 3TU. UK b Chemistry Department, Technical Universit)' ~[" Malaysia, Johor Bahru, Malaysia

Received 24 May 1996; received in revised form 20 August 1996 accepted 27 August 1996

Abstract

Trimercapto-s-triazine (TMT) is available commercially for precipitating heavy metals in effluents prior to discharge and for recovering silver and copper. The TMT content of an effluent for discharge is normally monitored down to about 2 ppm by means of its UV absorption at 285 nm. Indirect cathodic-stripping voltammetric methods of determining TMT at sub-ppb levels in standard solutions are reported here. These methods might prove suitable for the determination of TMT in effluent at levels lower than is currently possible. TMT can be accumulated and determined indirectly at pH 9.0 as its mercury salt down to sub-ppb levels. Accumulation is made at 0 V and the mercury TMT reduction peak is at - 0 . 47 V. Alternatively, by adding nickel(II), TMT can be determined optimally at pH 7.8, using the catalytic nickel peak at - 0 . 73 V and accumulating between - 0 . 1 0 and - 0 . 6 0 V: at this pH the HgTMT peak at - 0.47 V is small. At slightly higher pH (pH 8.6) the nickel TMT complex can be accumulated directly at - 0 . 4 0 V, but at this pH, however, a slightly increased sensitivity can be achieved by accumulating TMT as its mercury salt, at - 0 .1 V in the presence of nickel(II), the nickel TMT complex being formed during the potential sweep on the release of the TMT when the mercury salt is reduced. Unlike many other thiols TMT is not accumulated as its copper(I) salt on addition of copper(II) to the solution. 6! 1997 Elsevier Science B.V.

Keywords': Cathodic stripping voltammetry; Trimercapto-s-triazine; TMT: Catalytic nickel peak

1. Introduction

* Corresponding author. t Present address: Norwegian University of Science and

Technology, Department of Chemistry, N-7055 Dragvoll, Trondheim. Norway.

Tr i sodium tr imercapto-s- t r iazine (TMT) is available commercial ly as a 15% solut ion (TMT15; Degussa, Hanau , Germany) for the pre-

0039-9140/97/$17.00 (© 1997 Elsevier Science B.V. All rights reserved. PII S0039-9 t 40(96)02073-5

498 A.G. Fogg et a l . / Talanta 44 (1997) 497-500

cipitation of divalent heavy metals and silver from aqueous solutions. It is used in the treatment of effluent before discharge and in the recovery of metals, e.g. copper and silver, from electroplating and photographic baths and from electrical com- ponents (see Degussa literature).

A few years ago we were pleased to provide initial information on the voltammetric determination of heavy metals in samples provided by chemists at Powergen who were assessing the suitability of the technique for analysis of effluent from the Flue Gas Desulphurisation plant being installed at Ratcliffe on Soar Power Station. This plant, now in operation, removes sulfur dioxide from the flue gases. This is done by means of a limestone slurry, which is converted first into calcium sulfite, and then, through the injection of air, to calcium sulfate (gypsum), which is sold for plasterboard production and other purposes. Small concentrations of heavy metals enter the system from the limestone or the coal, via the flue gases: these are removed from the effluent initially with lime and then with TMT15.

Trade literature from Degussa gives methods for the determination of TMT. At higher levels (e.g. in TMT15), TMT can be determined acidi- metrically by potentiometric titration with hy- drochloric acid. At lower levels, e.g. in treated effluent, it can be determined by means of its UV absorption at 285 nm. Typical dosage levels of TMT15 are reported to be 50-100 ml per cubic meter of effluent (ie. 7.5 15 ppm TMT). Calibra- tion graphs are constructed from 1 10 ppm TMT using standard TMT solutions prepared in water having a similar salt composition to the effluent as high salt levels are reported to cause a reduction in the absorbance at 285 nm.

General studies are being made in our laborato- ries of the indirect determination of thiols by cathodic stripping voltammetry [1-7]. In particu- lar the first use of nickel and cobalt catalytic reduction peaks in CSV have been reported [1,7]. Comparisons are being made of the use of accumulation and determination of the mercury, copper(I), nickel and cobalt salts of thiols. In this communication we report on the cathodic stripping voltammetry of TMT in standard solutions using these methods, and give some indication of

possible interference from heavy metals, including copper and nickel, in the determinations.

2. Experimental

Cathodic stripping voltammetry was carried out with a Metrohm 646/647 VA Processor, using a multi-mode electrode in the H M D E mode. The three-electrode system was completed by means of a glassy carbon auxiliary electrode and an Ag/ AgC1 (3M KC1) reference electrode. All potentials are quoted relative to this reference electrode. Differential-pulse voltammetry was carried out with a pulse amplitude of 50 mV, a scan rate of 10 mV s-1 and a pulse interval of 1 s.

Samples of TMT15 were kindly provided by Powergen. Stock solutions of TMT (1.2 × 10 2 M) were prepared fresh every week: more dilute solutions were prepared daily by diluting these solutions. Standard copper(II), nickel(II) and other heavy metal solutions were prepared daily by diluting standard atomic absorption solutions from various manufacturers. Britton Robinson buffer (BRB) was prepared by dissolving boric acid (2.47 g), orthophosphoric acid (2.7 ml) and glacial acetic acid (2.3 ml) in water and diluting to 11. Appropriate volumes of this solution were adjusted to the required pH with sodium hydrox- ide solution (3 M).

The general procedure used to obtain cathodic stripping voltammograms was as follows: a 20 ml aliquot of buffer was placed in the voltammetric cell and the solution was purged with nitrogen for 6 min with the stirrer on. After an initial blank run, the required volumes of TMT and metal solutions were added by means of a micropipette. After forming a new mercury drop, accumulation was effected for the required time at the predeter- mined accumulation potential whilst the solution was stirred. The small mercury drop size was used on the Metrohm 647 VA stand. At the end of the accumulation period the stirrer was switched off and after 10 s had elapsed to allow the solution to become quiescent a negative-going potential scan was initiated. When further volumes of TMT solution or reagents were added the solution was deoxygenated for a further 20 s before producing further voltammograms.

A.G. Fogg et al. Talanta 44 (1997) 479 500 499

3. Resul t s

Typical cathodic stripping vol tammograms of T M T at pH 2-9 .4 are shown in Fig. 1. At the lower pH values (pH < about 7.5) two peaks are observed, which may be attributed to the reduction of mercury(l l )-TMT. The smaller second peak is absent at higher pH values. There is a shift in the potential of the main peak from - 0. l to - 0 . 4 7 V over this pH range, and the peak height reaches a maximum value at about pH 9.0. This pH was used subsequently. Calibration graphs were very reproducible, but were rectilin- ear up to 7 x 10 ~ M when a disproportionate increase of peak current was obtained, followed by a levelling off of the current. At this time we have no explanation for the disproportionate increase in current, although it may be associated with the accumulation of different layers of HgTMT. The final levelling off of the current is normally owing to saturation of the surface of the drop. The limit of detection (three standard devia- tions) was calculated to be 5 x 10 9 M.

The addition of copper(lI) removed the mercury T M T peak: a 7-fold excess of copper(II) eliminated the peak completely. No peak indicat- ing the accumulation of copper(I) T M T was observed (see Fig. 2). Addition of cadmium also removed the mercury T M T peak, and small peaks

i 511

(;

II

/

, i

/ I

" I! i

,, i ' 4 ..... - - ._ " - - 2 t L g . . ~ :7 ' v _ _ - - ~

u ! i 1 t I J

o 0( , -12 0 - 0 6 -12 v ~s Ag a,g('l

Fig. 1. The effect of pH on cathodic stripping vol tammograms of T M T at a hanging mercury drop electrode. Accumulation potential. 0.0 V; accumulation time, 60 s: T M T concentration, 2 .4× 10 VM. p H - A , 2.15:B, 2.50;C, 3.59;D, 4.44:E, 8.15; F, 8,69: (}, 9.06: H. 9.41.

600 F

3 0 0

' , I

I [ . . . . . i i ___

- 0 8 l! 4 ) 8

Fig. 2. Interference of copper(II) in the determination of TMT. Accumulation potential, 0.0 V; accumulation time. 60 s: T M T concentration, 2.4 x 10 7 M. pH = 9.(I. Copper(ll} concentration: A, 0 (buffer only): B, 0 (buffer + T M T only}: C, 3.14: D, 6.28; E. 9.44: F, 12.6: G, 15.7 × 10 7 M.

became evident between - 0 . 6 - 0 . 8 V indicat- ing some small accumulation of a cadmium T M T species (not shown). Lead, cobalt and zinc did not interfere at this pH: there was no sign of a catalytic cobalt peak but the possibility of its appear- ing at lower pH values needs to be investigated.

On the addition of a high concentration (e.g. 5 x 10 4 M) ofnickel(II) the catalytic nickel peak of N iTMT appeared at - 0 . 7 3 V in addition to the H g T M T peak at - 0 . 4 7 V, when accumulation was effected at - 0 . 1 V at pH 8.6. This indicates that T M T is accumulated still as H g T M T at 0.1 V, but is converted to NiTMT when the T M T is released when the H g T M T is reduced at - 0 . 4 7 V during the potential scan. T M T can be accumulated directly at - (/.4 V, but with a slightly reduced sensitivity. The opt imum pH for the production of the catalytic nickel peak is slightly lower (pH 7.8) than for the production of the peak of the mercury salt: at this pH the mercury peak is very small, even when accumulation is effected at 0.20 V. Typical cathodic stripping vol tammograms obtained for producing a calibration graph for the determination of T M T as its nickel complex with an accumulation time of 60 s are shown in Fig. 3.

500 A.G. Fogg et a l . / Talanta 44 (1997) 497-500

H

300

! D

0 A

0 -05 -1.0 0 -05 - 1 0 V vs Ag/AgCI

T S0nA T 150ha H

O (b) i ~

i C /J :fA lh I I I I I I

0 -0.5 -I 0 0 -0.5 -10 V vs A~AgCI

Fig. 3. Cathodic stripping voltammograms obtained for producing a calibration graph for the determination of TMT as its nickel complex (a) at pH 8.6 and (b) at pH 7.8. Accumulation potential, (a) -0.1 V and (b) -0 .2 V; accumulation time, 60 s. Nickel(ll) concentration: 5.1 x 10 4 M. TMT concentration: A, 0 (buffer only); B, 0 (buffer + Ni only); C, 0.3; D, 0.6; E, 0.9; F, 1.2; G, 1.8; H, 2.4 x 10 7 M.

De te rmina t ions o f the trace meta ls remain ing af ter T M T t rea tment of effluent are made by wel l -es tabl ished s t r ipping vo l t ammet r i c me thods af ter U V i r rad ia t ion to des t roy T M T remain ing in the effluent (J.A. Lees, Pr ivate c ommun ic a t i on and [8]). The trace meta ls r emain ing in the effluent could be present as dissolved or col loidal T M T complexes. The CSV me thods descr ibed in this c o m m u n i c a t i o n will de te rmine free T M T and soluble T M T complexes that can be accumula ted as H g 3 T M T 2 (solubi l i ty p r o d u c t = 1.4 x 10 4 7 ) ,

but appa ren t ly no t col lo idal T M T complexes.

Acknowledgements

The au tho r s are grateful to D r Jill Lees o f Powergen for her interest in this project . R.I . and A . R . H . M . Y . thank the Technical Univers i ty o f M a l a y s i a ( U T M ) for f inancial suppor t and leave o f absence.

References

4. Discussion

The values o f the solubi l i ty p roduc t s for

M 3 T M T 2 are quo ted in the Degussa l i te ra ture as being 1.0 x 10 18 (nickel) and 1.3 x 10 -31 (copper(II)) . This explains why the nickel CSV m e t h o d

can be appl ied bu t not the copper ( I ) method . The

nickel sa l t /complex is sufficiently soluble for the

nickel m e t h o d to be appl ied , whereas add i t i on o f copper ( I I ) to a T M T solu t ion o f a b o u t 1 x 10 -7

M causes prec ip i ta t ion , or ini t ial ly a co l lo ida l

so lu t ion o f the copper ( I I ) salt, which does not

a l low accumula t ion o f the copper ( I ) salt at the e lec t rode surface.

[1] F.G. Banica, J.C. Moreira and A.G. Fogg, Analyst, 119 (1994) 309.

[2] F.G. Banica, A.G. Fogg and J.C. Moreira, Analyst, 119 (1994) 2343.

[3] F.G. Banica, A.G. Fogg and J.C. Moreira, Talanta, 42 (1995) 227.

[4] A. lon, F.G. Banica, A.G. Fogg and H. Kozlowski, Elec- troanalysis, 8 (1996) 40.

[5] F.G. Banica, A.G. Fogg, A. Ion and J.C. Moreira, Anal. Lett., 29 (1996) 1415.

[6] A.G. Fogg, R. Ismail, R. Ahmad and F.G. Banica, Ana- lyst, 121 (1996) 1877.

[7] A.G. Fogg, R. Ismail, R. Ahmad and F.G. Banica, Ta- lanta, 44 (1997).

[8] 'The Determination of Twelve Trace Metals in Marine and other Waters by Voltammetry or AAS', H.M. Stationery Office, London, 1988.