iodide reaction

Analytica Chimica Acta, 217 (1989) 327-334 Elsevier Science Publishers B.V., Amsterdam -

327 Printed in The Netherlanda

DETERMINATION OF PICOMOLAR CONCENTRATIONS OF BROMIDE WITH A PIEZOELECTRIC DETECTOR BY CATALYSIS OF THE PERMANGANATE/IODIDE REACTION

S.-Z. YAO*, L.-H. NIB and Z.-H. MO New Material Research Institute, Department of Chemical Engineering, Hunan University, Changshu (China) (Received 12th January 1987)

SUMMARY The piezoelectric quartz crystal has been utilized to detect iodine produced by the bromidecatalyzed oxidation of iodine to iodate by permanganate in acidic solution. After extraction of iodine into toluene, the resulting frequency change caused by iodine adsorption on the crystal electrode is proportional to bromide concentration over the range 0.5-5 X lo-” M. Only silver (I), mercury (II) and large concentrations of chloride interfere significantly. The crystal detector is also used to indicate the end-point of a chloride titration with silver.

Kinetic methods provide some of the most sensitive analytical procedures. The catalytic effect of bromide on the oxidation of iodine to iodate by permanganate in a sulphuric acid medium has been utilized by a number of investigators for the determination of traces of bromide [l-3]. The reaction rate was followed by an extraction-spectrophotometric determination of the change in the concentration of the unreacted iodine as a function of time. The detection limit was 1 ng ml-’ [ 21. Serious interferences were caused by silver (I) and mercury (II), and other anions such as sulphide, thiosulphate, thiocyanate, cyanide and chloride also interfered [ 2,3]. Therefore, it is of interest to find a more selective method based on the catalytic action of bromide. The piezoelectric quartz crystal has not yet been utilized for kinetic analysis, though it has proved to be a powerful tool for the determination of traces of a number of substances in the atmosphere with high sensitivities [ 41. In the present paper, the piezoelectric quartz crystal is used as a detector for iodine [ 5 ] to follow the catalytic action of bromide on the above-mentioned oxidation reaction. A rapid method for the determination of pmol 1-l bromide is proposed.

0003-2670/89/$03.50

0 1989 Elsevier Science Publishers B.V.

EXPERIMENTAL

Apparatus and reagents The pieioelectric quartz crystals used were 9 MHz AT-cut quartz crystals, 12.5mm diameter (Peking Electronic Factory No. 707), with silver-plated electrodes (6-mm diameter) on each size. The crystal holder was directly connected to an integrated circuit oscillator made in this laboratory [6] and supplied with 5V by a JWY-30B d.c. variable supply (Shijiazhong Electronic Factory No. 4). The detection cell with the piezoelectric crystal and the crystal holder was set in a thermostated water bath at 25 t 0.5 “C!. All measurements were carried out in a air-conditioned room thermostated at 25°C. The frequency change was monitored by a digital counter (SS-3320 model, Shijiazhong Electronic Factory No. 4). A magnetic stirrer was used during the adsorption process. All chemicals used were of analytical-reagent grade. Double distilled water was used for the reagent preparation and throughout the procedure. The iodide stock solution (1 mg ml-‘) was prepared by dissolving 1.31 g of potassium iodide and 2 g of sodium carbonate in water and diluting to 1 1. The working solutions were prepared by successive dilution. The 0.01 M bromide and chloride stock solutions were prepared by dissolving their chlorides in water. Potassium permanganate solution (1 x 10T3 M) was prepared and stored in a brown glass bottle, and was used for no more than 3 days. Procedures Transfer 20 ml of the sample solution, 0.10 ml of 0.09 M sulphuric acid and 30 ~1 of potassium iodide solution containing 90 ,ug ml-’ iodide to a 60-ml separatory funnel. Mix, and let stand for 5 min. Start the reaction by adding 30 ~1 of 1 x 10m3M potassium permanganate, swirl to mix thoroughly and let the separatory funnel stand for exactly 3 min after the addition of the permanganate. Add 10 ml of toluene and immediately shake the separatory funnel for 15 s. After phase separation, transfer the organic layer to a lo-ml beaker. Stir the solution at a constant rate (450 r.p.m. ). Immerse the piezoelectric crystal, whose frequency has been measured to be F1, in the solution for exactly 10 min. Remove the crystal from the solution, and measure the frequency after it becomes constant (F2). Calculate the frequency change AF= (F2 - F, ) averaged over 2-3 measurements, and for the standards, plot the change against the bromide concentration. Between each run, remove the iodine species (mainly silver iodide [ 51) adsorbed on the electrode of the crystal by immersion in 0.05 M ammonia in acetone with stirring, until the frequency is constant, Determination of bromide in natural water. First, titrimetric determination of chloride in the water sample is done with the piezoelectric crystal as indicator. Transfer 20 ml of the water sample to a beaker in the water bath thermostated at 25 oC, and add 5 ml of ethanol. Stir at a constant rate (450 r.p.m. ) ,

329

and immerse the piezoelectric crystal in the solution. Titrate with 0.5 mM silver nitrate, and record the frequency of the crystal during the titration process. Plot the frequency against the titrant volume and find the end-point from the titration curve. Calculate the concentration of chloride in the water sample (Cc,). Prepare a solution of potassium chloride of concentration Cc, and determine the frequency change (&‘c,) using the procedure described as above. Then measure the frequency changes for the water sample and for other 20-ml portions of the sample with 50 and 100 ~1 of standard potassium bromide solution (0.25 ng Br- ml-‘) added to it (AF,, AF,). The concentration of bromide in the sample water is calculated as C, = C, V, ( dFCl- AF, ) /V, (AF,- dF, ), where C, is the concentration of bromide in the standard solution whose volume V, is added to the unknown sample of volume V,. RESULTS AND DISCUSSION

The bromide-catalyzed oxidation of iodide by permanganate ied by a number of authors [l-3]. It proceeds in two steps: lOI- + 2MnO;

+ 16H +-51Z

has been stud-

+2Mn2++8H20

Br-

1,+2Mn04+4H

+-210;

+2Mn2++2H20

The second, catalyzed oxidation reaction of iodine by permanganate is the rate-determining step and is usually followed by the extraction-spectrophotometric determination of the change in the concentration of the unreacted iodine as a function of the reaction time, e.g., by using iron (III) and mercury (II) thiocyanate, with a detection limit of 5 ,ug ml-’ iodine [ 31. When the piezoelectric quartz crystal is used as a detector for iodine in the organic extract [ 51, this detection limit can be lowered to 13 ng ml-’ (0.05 PM) iodine in the organic extract, equivalent to 0.6 ng ml-l (5 nM) iodide in the original aqueous solution. Optimization of experimental conditions Sulphuric acid. The concentration of sulphuric acid in the reaction mixture is an important factor affecting the oxidation reaction. In order to investigate the acidity effect and to find the optimum concentration of sulphuric acid, various volumes of 0.09 M sulphuric acid were added to 20-ml portions of blank samples containing 30 ~1 of potassium iodide solution (0.09 mg iodide ml-‘) and 30 ,~l of 1 mM potassium permanganate, and the resulting frequency changes were measured. The results are shown in Fig. 1. It can be seen that the frequency change vs. acidity plot has three parts; in the first, where no more than 0.1ml of the acid solution was added, the frequency change increased abruptly with increasing acidity until a maximum frequency change was reached. In the second part, there is a smooth slight decrease in frequency up

330

0

*.

1.5

3.0

I 3.25

0.09M HzSOL (ml)

Fig. 1. Effect of sulphuric acid on the frequency change. Fig. 2. Influence of the concentration

of permanganate on the rate of iodine production.

to 1.0 ml of acid added; in the third, where more acid was added, the frequency change decreased rapidly, showing that the oxidation of iodine was more significant and it would be difficult to carry out the bromide-catalyzed reaction in solutions of such acidities. Hence, 0.10 ml of 0.09 M sulphuric acid was chosen as most satisfactory condition for the catalytic determination of bromide. Iodide concentration. Investigation of the adsorption of iodine species on the silver-plated electrode of the piezoelectric crystal from the toluene extract [ 51 showed that the calibration graph of frequency change against iodine concentration [ I21 is linear over the range from 0.05 to 5 ,uM and is described by the equation [ I21 = (dF/201) PM, where AF is measured in Hz. Because of the dissolution of silver from the electrode during the removal of the adsorbed iodine, a very high concentration of iodine in the organic extract will inevitably lead to significant loss of electrode material and, hence, a shorter life for the electrode. Results showed that it is best to use a 1 PM solution of iodine in toluene, which corresponds to 30 ~1 of a 0.09 mg ml-’ iodide solution added to 20 ml of the sample. Then the reaction proceeds satisfactorily, resulting in a frequency change of 200-300 Hz, with good reproducibility for bromide determination and a relatively long life for the electrode. Potussiumpermanganute. In order to investigate the effect of permanganate in the absence of bromide, 30 ,~l portions of potassium permanganate solutions of different concentrations were added to a blank containing 30 ~1 of potassium iodide solution (0.09 mg ml-’ iodide) and 0.1 ml of 0.09 M sulphuric acid, and the frequency changes were measured after a 1-min reaction. As can be seen in Fig. 2, as the concentration of permanganate increases up to 1.5 PM, the reaction rate increases rapidly. Above 1.5 PM (MnO, :I- mole ratio> 1.5), the reaction rate apparently decreases linearly with increasing permanganate con-

331

centration. Thus although the rate of the uncatalyzed reaction is greatest in 1.5 ,uM permanganate, it is used for bromide determination because greatest sensitivity to bromide is also achieved under these conditions. Catalytic effect of bromide and chloride The catalytic effects of bromide and chloride on the oxidation of iodine to iodate by permanganate in acidic medium were investigated by following with the piezoelectric crystal detector. The rate curves were obtained by using 3 x lo-” and 5 x lo-” M potassium bromide solutions and a 2 x lo-” M potassium chloride solution, and 30 ~1 of 1 X 10m3 M potassium permanganate solution added to the sample solutions. A blank curve was also obtained by using double-distilled water. As shown in Fig. 3, the differences in frequency changes for the blank, bromide and chloride solutions increase with reaction time. A longer reaction time and hence a greater difference in frequency change would lead to a higher sensitivity of determination. However, because of the inevitable variation in the surrounding conditions, a longer reaction time may also yield a poorer precision. As seen in Fig. 3, a reaction time of 3 min results in a frequency change of 60 Hz for 3 x lo-l2 M bromide and 95 Hz for 5 x lo-l2 M bromide. As the piezoelectric detector can measure exactly the frequency change to within 10 Hz, a reaction time of 3 min is considered adequate for bromide determination. As seen in Fig. 3, the plots of frequency change vs. reaction time are all linear, indicating that the reaction between iodine and permanganate follows first 300-

: -F ;qqe. A\:: x

s

\\

200 -

x\

v

‘\\

x

‘3

,\ lOO0

1

2 Reaction

3 time

(mid

4 4

C er

(PM)

Fig. 3. Effect of bromide and chloride on the reaction rate: (1) blank; (2) 2 x lo-” (3) 3X10-12Mbromide; (4) 5X10-‘2Mbromide.

M chloride;

Fig. 4. Effect of bromide concentration on the frequency change: (1) without chloride present; (2) with 5 X lo-” M chloride added.

332

order kinetics at the various concentrations of the catalyst as well as in the absence of added catalyst. The rate constant increases with the concentration of bromide. When a number of reaction mixtures containing different concentrations of. bromide (O-7.75 x lo- l2 M) were prepared and their frequency changes measured, the frequency change vs. bromide concentration relationship was linear up to ca. 5 x lo-l2 M (Fig. 4). At higher bromide concentrations, there was a departure from linearity. Thus, the dependence of the reaction rate on the catalyst concentration is first order at low bromide concentrations. Chloride also appears to have a catalytic effect on the reaction, but the frequency change was 300-400 times smaller than bromide at the same concentration. Calibration, reproducibility and interferences The calibration graph (Fig. 4) of frequency change against bromide concentration is described by the equation: [Br-] = [11.2-

@F/21.5)]

~1O-l~ M

where AF is measured in Hz. The standard deviation was 10 Hz (5.6%) for 5 determinations of 3.1 x lo-l2 bromide. TABLE 1 Effect of other ions Ion

Added as

Cont. (l~lo--~M)

None NOaS2s,o,2SCNCNCl-

NaNO, Na,S Na&Os KSCN KCN KC1

Fe3+ A13+ M8+ Pb*+ cu2+

NH4Fe(SW2 AIz(SW3 MgSO4 Pb(NW2 cuso4

Hg2+

HgCI2

Ag+

AgNO,

1 1 1 1 1 2x10-4 5x10-4 1 1 1 1 1 1 3x10-8 5x10-2 1x1o-2 0.1

cd*+

case,

Frequency change (Hz) 2x10-“M Reagent Brblank 240 240 240 235 230 235 225 200 240 240 235 235 245 235 225 170 230 110

215 215 215 210 210 210 220 170 215 215 210 210 220 210 200 150 205 100

333

The effects of other ions on the determination of bromide were investigated. No significant interferences were caused by most anions and cations such as nitrate, sulphide, thiosulphate, thiocyanate, cyanide, iron (III), aluminium, magnesium, lead, copper (II), and cadmium. Chloride, silver and mercury (II) did interfere. The latter two cations form complexes or precipitates with iodide, thus decreasing the iodide concentration, resulting in a lower frequency change. Nevertheless, mercury(I1) and silver do not interfere at concentrations < 3 x lo-l2 M (mercury) and 10 x lo-l2 M (silver). The results are summarized in Table 1. Determination of bromide in natural waters Natural waters usually contain chloride. As a great excess of chloride may cause interference, the influence of chloride on the bromide determination was investigated. As can be seen in Fig. 4, though the bromide calibration graph is slightly less sensitive when a great excess of chloride is present, the graph remains linear over the same concentration range, so the bromide can still be determined if the concentration of the concomitant chloride has been determined first, and calibration solutions prepared at such a chloride concentration. In order to determine the chloride concentration, a silver titrimetric method is proposed based on the novel use of the same piezoelectric quartz crystal which was used for the kinetic method of determining bromide to follow the titration procedure. Because the frequency change of the piezoelectric crystal does not vary significantly with the presence of various inorganic salts [ 61, the rate of change of frequency in the titration is very slow at first due to the progressive substitution of chloride ion by nitrate, but increases rapidly as the equivalence point is reached because of the silver and nitrate ions added in excess and the proportionality between the frequency change of the piezoelectric crystal and the total concentration of salts in the solution in which the crystal is immersed [ 61. In the case of natural waters where the chloride concentration is much greater than bromide, the argentimetric result can be at-

Fig. 5. Titration curve for chloride in a sample of natural water containing 20% ethanol with 0.5 mM silver nitrate solution, with the piezoelectric crystal ae detector.

334 TABLE 2 Determination

of bromide in water samples

Water sample”

Vol. 0.25 ng ml-’ bromide solution added (~1)

Frequency change (Hz)~

Bromide found (pg ml-‘)

Spring water from The Yuelo Mountain

0 50 100

247 218 195

0.41 0.46

Synthetic sample containing 0.4 pg ml-’ bromide and 15.2 ng ml-’ chloride

0 50 100

248 221 191

0.42 0.39

*Chloride content: 15.2 ng ml-‘. Frequency change for a 15.2 ng ml-’ chloride solution measured by the procedure described in the text was 266 Hz. bAverage of 3 measurements.

tributed to the chloride alone without significant error. A titration curve is shown in Fig. 5. A standard addition technique was used to determine the bromide content of the water sample. The results are listed in Table 2. The method was checked using a synthetic sample containing the same concentration of chloride (15.2 ng ml- ’ ) and approximately the same concentration of bromide (0.4 pg ml- ’ ) . The results shown in Table 2 demonstrate the reliability of the method. The method is satisfactory for the determination of bromide in natural water. The error is about 2 0.02 pg ml-l at a bromide concentration of 0.4 pg ml-‘. The authors record their thanks for financial assistance from the National Commission of the Natural Science Funds of the People’s Republic of China in support of this work.

REFERENCES 1 2 3 4 5 6

M.I. Fishman and M.W. Skougstad, Anal. Chem., 35 (1963) 146. M. Shiota, S. Utsumi and I. Iwasaki, Nippon Kagaku Zasshi, 80 (1959) 753. K. Takahashi, M. Yoshida, T. Oxawa and I. Iwasaki, Bull. Chem. Sot. Jpn., 43 (1970) 3159. J.F. Alder and J.J. McCaIium, Analyst, 108 (1983) 1291. S.-Z. Yao, Z.-H. MO and L.-H. Nie, Anal. Chim. Acta, 215 (1988) 79. S.-Z. Yao and Z.-H. MO, Anal. Chim. Acta, 193 (1987) 97.