Determination of total dissolved sulfide in the pH range 7.5 to 11.5 by

(14) Erbelding, W. F. Anal. Chem. 1975, 47, 1983. (15) Krejci, M.; Posplsilova, N. J. Chromatogr. 1972, 73, 105. (16) Vespalec, R. J. Chromatogr. 1975...
0 downloads 0 Views 440KB Size
Anal. Chem. 1983,55, 1731-1734 (3) Stankoviansky, S.;Cicrnanec, P.; Kaniansky, D. J. Chromatogr. 1975, 106, 131. (4) Jackson, A. J. Chem. Educ. 1965, 42, 447. (5) Svoboda, V.; Marsai, J. J . Chromatogr. 1978, 148, 111. (6) Kambara, T.; Tachikawa, T. J. Chromatogr. 1968, 3 2 , 728. (7) Duhne, C.; Sancher, 01.Anal. Chem. 1962, 3 4 , 1074. (8) Jupille, T.; Togami, D.; Burger, D. Abstract 242, Pittsburgh Conference, 1982. (9) Evans, B.; Stoiz, J. Abstract 247, Pittsburgh Conference, 1982. (10) Keller, J. M. Anal. Chem. 1981, 5 3 , 344. (11) Haderka, S.J . Chrom,stogr. 1971, 5 7 , 181. (12) Vespalec, R.; Hana, K. J. Chromatogr. 1972, 6 5 , 53. (13) Poppe, H.; Kuysten, J. J . Chromatogr. 1977, 132, 369. (14) Erbelding, W. F. Anal. Chem. 1975, 47, 1983. (15) Krejci, M.; Pospisilova, N. J . Chromatogr. 1972, 7 3 , 105. (16) Vespalec, R. J . Chronoatogr. 1975, 108, 243. (17) Krejci, M.; Vespalec, R.; Sirec, M. J. Chromatogr. 1972, 6 5 , 333.

1731

(18) Haderka, S.J. Chromatogr. 1974, 9 1 , 167. (19) Alder, J. F.; Thoer, A. J. Chromatogr. 1979, 178, 15. (20) Aider, J. F.; Drew, P. K. P.; Fieiden, P. R. J . Chromatogr. 1981, 212, 167. (21) Aider, J. F.; Drew, P. K. P.; Fieiden, P. R. J . Chromatogr. 1983, 5 5 , 256. (22) Pungor, E. "Osciilometry and Conductometry"; Pergamon Press: Oxford, London, Edinburgh, New York, Paris, 1965. (23) Cruse, K.; Huber, R. "Hochfrequenztitration"; Verlag Chemie: Weinheim, 1957. (24) Reilley, Ch. N. "High Frequency Methods In New Instrumental Methods in Electrochemistry"; Deiahey, P., Ed.; Interscience: New York, London, 1954.

RECEIVED for review April 19, 1982. Resubmitted February 8, 1983. Accepted May 10, 1983.

Determination of Total Dissolved Sulfide in the pH Range 7.5 to 11.5 by Ion Selective Electrodes Hugo Guterman andl Sam Ben-Yaakov*

Department of Electrical and Computer Engineering, Ben-Gurion University of the Negeu, Beer-Sheua, Israel Aharon Abeliovich

Laboratory for Environmental Applied Microbiology, The Jacob Blaustein Desert Research Institute, Ben-Gurion University, Sede Boquer Campus 84990, Israel

A total dlssolved sulfldie meter was designed and tested over the concentratlon range to lo-' mol/L. The Instrument applies a sulfide Ion aotivlty electrode and a pH glass electrode, whose potential6 are measured agalnst a double Junctlon reference electrode. The potentials are processed by an electronic analog clrcult to obtaln an output voltage that Is proportional to total dlssolved sulfide. The proposed Instrument-adjustment procedure ellmlnates mutual dependence of the adjusted1 controls. The total dissolved sulfide readings of the Instrument were found to be pH independent over the pH range pH 7.5 to pH 11.5.

The importance of sillfide compounds in biological processes has been widely demonstrated ( 1 , 2 )and it is well-known that the generation of sulfide is linked to a number of vital, chemical, physical, and biochemical processes (3). The classical analytical methods for sulfide determination like the Methylene Blue or the iodomotric methods (4) are rather cumbersome, requiring elaborate sample handling and preparation and lengtlny calibration. In situ measurement of sulfide activity has bleen made possiblle through the introduction of ion-selective membrane electrodes (4). However, since these electrodes are sensitive to sulfide activity, determination of total sulfide can be accomplished only after the samples are buffered tri high pH by a higlh ionic strength buffer such as SAOB I1 (5). Total sulfide can be calculated from a simultaneous measurement by sulfide and pH electrodes. A suitable microcomputer interface for implementing such an instrumentation system has been recently described by Ben-Yaakov et al. (6). Frevert and Galster (7) proposed an analog method for direct determination of total sulfide by using a sulfide ion selective electrode coupled to a pH glass electrode. This method is limited to solutions with pHs p H

pS2- for pH

< pK1’

> PKl’

(2)

(3)

> pK2’

(4)

where p H = -log aH+ Equation 3 which is valid for the range pH 7 to pH 13 can be implemented by ion selective electrodes in conjunction with an analog electronic circuit of the form given in Figure 1. Taking into account the liquid junction potentials of the reference electrode (Ej), the electrodes output voltages will be (8, 9) VpH = EOpH SpH-pH Ej (5)

+ + VpS2- = EopS2-+ SpS2-*pS2-+ E. 1

,

(6)

where EOpH and EopS2-are constants, and SpH and SpS2are the slopes of the glass and sulfide electrodes, respectively. For ideal electrodes the slope should follow the Nernst expression

s = RT -

nF where R, T , F , and n are as usual. The deviation of the practical ion selective electrode from ideal behavior can be expressed as

(7)

where qpH and qpS2- are constants whose numerical value is normally smaller than one. The output voltage of the proposed electronic circuit given in Figure 1 will thus be

m) - VpK,’

(9)

where z and m represent the slope and offset adjustments of the electronic processor (Figure 1). By combining eq 5-9 one obtains VPST = EopS2- sps2-*ps2- E.1 z ( E 0 p H S p H - p H + Ej + m) - VpK2’

+

+

+

+m ~ -(VpHz + m VpSTz = V P S ~ -

VPSTl = VpSi2- - (VpHI

p x = -log [XI

VPST = v p s 2 - - z(VpH

Since the expression of VpST (eq 12) is similar in form to eq 3, the output voltage of the proposed circuits will be linearly proportional to pST in the range of pK1’ < pH < pK,/. The proposed adjustment method for z applies two standard solutions having the same S T (whose value may be unknown) a t different pHs (pH1 and pH2). The output voltage of the instrument in the two standard solutions will be

.

M - - VpK,’ (10)

) -~VpK2’

(12)

) -~VpK2’

(13)

If m (“offsetadjustment” of Figure 1)is initially adjusted, so that m = -VpH,, i.e., V’pHl = 0 (Figure 1)the output voltages will be VPSTl = VPSl2- - VpK2’

(14)

V P S T ~= VPS2’- - (VpH2 - VpHJz - VpK2’ (15) Final adjustment is made in the standard pH2 (“slope adjustment” of Figure 1)so that the output voltage is identical with the one registered for the standard of PHI, Le., V ~ S T ~ = VPST1. The required z value will be z =

v p g - - vps12VPHz - VPHl

Note that the adjustment of z does not change VpH1 which was earlier made equal to zero. The adjustment procedure thus requires only two steps: adjustment of m to obtain V’pHl = 0 when the electrodes are inserted in the standard solution of pH1, and adjustment of z to obtain VPST~= VPSTl when the electrode assembly is inserted in the standard solution of pH2. Once these adjustments are made the output voltage VpST should be linear with pS2- even if the specific ion electrodes exhibit a nonideal slope. Following m and z adjustment the assembly can be calibrated by conventional techniques such as the standard addition method. EXPERIMENTAL SECTION Electrodes. The pH electrode was a combination double

junction type 9092 (Broadly James Inc.). The pS2- electrode was a sulfide ion activity electrode, type F1212S (Radiometer, Denmark). Reagents and Solutions. All reagents except for sodium sulfide (Na2S.9Hz0)were analytical grade. Initial adjustments were made with two NazS M) solutions prepared in a pH 7 buffer (NBS Standard Buffer) (10) which was then adjusted t o about pH 9 by NaOH. Electrode calibration was performed on pH 7 buffer solution to which concentrated (2 M) Na,S solution was added incrementally by buret and the pH was adjusted by NaOH (2 M) addition.

ANALYTICAL CHEMISTRY, VOL. 55, NO. 11, SEPTEMBER 1983

1733

r i 1, ~

1-1

1-1

#-I

DIGITAL

DPM ICL7106 E V / K i t

Figure 2. Electronic clrcult diagram of the Instrument used In the present study.

Apparatus. The basic configurations of Figure 1 needed to implement the relationslhip of eq 11 were tested by an analog instrument (Figure 2). ‘The electronic design applies two high input impedance, low input current operational amplifiers (RCA CA3140) and a general purpose operationdl amplifier (National LM324) to realize the required differential and summing operations (8, 9, 11). It should be noted that the sulfide and pH electrodes are connected in opposite polaritieri to the corresponding differential amplifiers to compensate for the opposite sign of the slopes (SpH and SpS2-). The functions of the trimmers and potentiometers (P1-P5)are as follows: P1and Pz are used for optimizing the common mode rejection ratio of the two differential amplifiers. P, is the offset adjustment (m)and P, is used for slope adjustment, i.e., the z adjustment discussed above. P4 is used to control the magnitude of the voltage which simulate pK4 (eq 11). This potentiometer in used in effect as a general ouput voltage offset to ensure compatilbility between signal voltage range and the dynamic range of the digital panel metler (DPM) used. The latter was a 31/2 digit printed circuit module (INTERSlL ICL706EV). Instrument Adjustments and Calibration. The initial adjustments and calibration consisted of the following steps: (1) A buffer solution with gH in the range pK< + 1 < pH < p&‘ - 2 and known NazS concentration was prepared (in this work pH 8 and 1 M, respectivelly). (2) By use of Pz,V’pH1was adjusted to zero. (3) VpSTl was registered. (4) The pH of the standard solution was increased biy about one pH unit using 2 M NaOH solution. (5) VPST~ was made equal to VPSTl by setting P,. (6) The output signal level was adjusted by P4to bring it to within the midscale of the DPM. ( 7 ) The calibration curve of pST = f (VpST)was determined by incremental addition of concentrated (2 M) NazS solulion.

RESULTS AND

DISCUSSION

The theoretically expected error of STdetermination by the proposed method, as a function of pH, was evaluated by comparing the S T values calculated by eq 1to those obtained from the approximation of eq 3. The simulation was carried out for the range of pH 3 to p H 13 assurning that the values of pK,’ and pK4 are 6 and 14, respectively (7, 12, 13). The theoretical error of this approximation (Figure 3) was calculated to be less than 5% in the p H range 7.5-12.5 and should be less than 1% in the range pH 9 and pH 11. The pH range for which the approximation of eq 3 holds depends on the numeriical values of pK,’ and pKi. The expected variation of pK,’ and p K i with ionic strength can be estimated by relating the apparent constants to the thermodynamic dissociation constants K,, K2through the application

0 3

4

5

6

7

8

9

IO

II

1213

PH

Flgure 3. Expected error of S, determinatlon by the proposed method as a function of pH, assuming pK,’ = 6 and pK2’ = 14.

of the extended Debye-Huckel equation (EDHE) (14). The apparent constants can be expressed in the form (15)

PK,’ = pK1 + log ( Y T H ~ S-) log (YTHS-) (16) P K ~ PK, + log (YTHS-) - log (yTS2-) (17) where YT is the total activity coefficient, i.e., the ratio between single ion activity and its total concentration (both free and complexed) (16). Neglecting ion pair formation (i.e., assuming 7~~ = ri) the last two terms in each of the above expressions can now be approximated by EDHE (14)

0.54

pK,’ = pKi - 1+&

(18)

Although the EDHE is valid only for low ionic strength (g < 0.1) the upper limits for apparent constants variations can be derived by assuming that p m. -+

ApR1’,,,

= (pK1’ - P K ~ ) E, -0.5 ~~

APKzlmax = (PK~’ - pK2)max Y -1

(20) (21)

1734

ANALYTICAL CHEMISTRY, VOL. 55, NO. 11, SEPTEMBER 1983

Table I. Results of an Evaluation Test of the Proposed Total Sulfide Metera ST = 6.66 X 10.' mol/L pH 3.88 4.83 5.55 6.14 8.00 9.49 10.53 10.80 10.94 11.12 11.30 11.39 11.52 a

ST =

error, %

pH

0.44 1.33 2.97 5.60 6.65 6.65 6.65 6.65 6.65 6.65 6.65 6.65 6.65

-93.29 -80.0 -55.29 -15.86 -0.034 -0.034 -0.034 -0.034 -0.034 -0.034 -0.034 -0.034 -0.034

4.06 5.69 6.19 7.09 8.58 10.39 10.89 10.94 11.12 11.39 11.43 11.52 STM =

10'4STM, pH

mol/L

mol/L re1 error, %

4.69 5.42 6.19 8.00 8.85 9.85 10.71 10.89 10.98 11.25 11.53

0.445 1.11 2.22 3.52 3.32 3.52 3.14 3.32 3.32 3.32 3.32

-86.63 -66.46 -33.14 5.88 -0.028 5.88 -5.61 -0.028 -0.028 -0.028 -0.028

1 0 ' 4 S T ~ , re1 mol/L error, %

mol/L

actual concentration.

ST = 3.33 x

mol/L

ST = 1.66 X

1 0 - 5 S T ~ , re1

0.11 0.79 1.11 1.98 1.57 1.57 1.48 1.57 1.67 1.67 1.67 1.67

-92.85 -52.34 -32.71 19.54 -5.0-5.0 -10.31 -5.0 0.610 0.610 0.610 0.610

ST = 1.113 x pH

mol/L

mol/L re1 error, %

5.46 5.82 6.05 6.41 6.82 7.81 9.67 10.39 10.71 10.98 11.16 11.43 11.57

0.395 0.527 0.702 0.884 1.113 1.113 1.113 1.113 1.113 1.113 1.113 1.05 1.05

-64.45 -52.62 -36.85 -20.52 0.017 0.017 0.017 0.017 0.017 0.017 0.017 -5.57 -5.57

10-3STM,

measured concentration.

r -

sensitive to pH within the expected range (Table I and Figure 4). Tests were not carried out in solutions with pH values higher than 11.5 to avoid marked variation of total ionic strength. The discrepancy between the experimental data points and the theoretical response (Figure 4) below p H 7 should be attributed to the uncertainty in the value of pK; used in the evaluation of the theoretical curve. Since the numerical values of apparent constants are a function of ionic strength and ionic composition (15) it is unlikely that the values which were originally cited for seawater (7,12,13) are directly applicable here. Fortunately, the proposed method does not require a knowledge of these constants as long as calibration and measurements are made in the "safe" pH range: pH 7.5 to pH 11.5.

' 7

PH

Flgure 4. Response of experimental instrument to pH variation for different total sulfide concentrations (ST): (circles) ST = mol/L; mol/L; (crosses) ST = 1.66 X mol/L; (triangles) ST = 2 X (squares) S T = 6 X mol/L; (solid lines) model calculations assuming pK,' = 6; pKp' = 14.

As already discussed, the usable pH range of the instrument for a 5% error limit is pK1'

+ 1.5 C pH C pK,'

- 1.5

(22) Hence, the safe range is from about p H 7.5 to about pH 12. At higher ionic strength the lower p H limit may reach the value of pH 7. Another possible error source is the variation of the liquid junction potential between the calibration and test solution. This error is lower than in the usual case, due to the partial cancellation of the liquid junction potential by the analog computation scheme. The effect of the liquid junction potential is reduced by the factor (1 - z ) (eq lo), which assumes the numerical value of 1/2, if the slope of the sulfide ion selective electrode is half the slope of the glass electrode. Both the p K i and Ejerrors could be circumvented by calibrating the instrument in the test solution using the standard additive method. Under constant composition and ionic strength, the instrument exhibits an excellent linearity as long as the pH is held between pH 8 and pH 10. It should be noted that sulfide ion activity at the lower end of this calibration curve is about M which is apparently within the "mud" level of the sulfide electrode used (12, 17). The proposed adjustment procedure of the instrument (see above) was found to be effective in balancing out slope nonideality of the electrodes. The adjusted instrument was in-

ACKNOWLEDGMENT We thank A. Yehieli for assistance in the construction of the hardward involved in the present work. R e g i s t r y No. Sulfide, 18496-25-8; water, 7732-18-5.

LITERATURE CITED Silverman, M. P. I n "The Encyclopedia of Geochemistry and Environmental Sciences"; Falrbridge, R. W., Ed.; Van Nostrand Reinhold: New York, 1972;pp 1132-1134. Kaplan, I.R. I n "The Encyclopedia of Geochemistry and Environmental Sclences"; Falrbrldge, R. W., Ed.; Van Nostrand Reinhold: New York, 1972;pp 1148-1is1. Sekerka, I.; Lechner, J. R. Anal. Chim. Acta 1977, 93, 139-144. "Standard Methods for the Examination of Water and Wastewater", 15th ed.; APHA-AWWA-WPCF, 1960: pp 442-450. "Determination of the Total Sulfide Content in Water"; Orion: Cambridge, MA, Applicatlon Note A12. Ben-Yaakov, Sam; Raviv, Roni; Guterman, Hugo: Dayan, Alfred; Lazar, Boaz Talan 1982, 29, 267-274. Frevert, V. T.; Galster, H. Scbweiz. 2.Hydro/. 1978, 4 0 / 1 , 199-205. Brand, M. J. D.; Rechnltz, G. A. Anal. Chem. 1970, 42, 616-622. Wilde, P.; Rodger, P. W. Rev. Scl. Insfrum. 1970, 41, 356-363. Bates, R. G. "Determination of pH"; Why: New York, 1965. Graeme, Y. G., Tobey, G. E., Huesman, L. P., Eds. "Operational Amplifier: Design and Applications"; McGraw-Hill, New York, 1971. Heu, R. G.: Rechnitz, G. A. Anal. Chem. 1968, 4 0 , 1054-1060. Goldhaber, M. B. Ph.D. Thesis, Unlversity of California, Los Angeles,

1974. Laitinen, H. A. "Chemical Analysls"; McGraw-Hill: New York, 1960. Ben-Yaakov, Sam; Golbhaber, M. B. Deep Sea Res. 1973, 2 0 ,

67-92. Garrels, Robert M. I n "Glass Electrodes for Hydrogen and Other Cations, Prlnclples and Practlce"; Einsenman, George, Ed.; Marcel Dekker: New York, 1967;Chapter 13. Radlometer "F1212S Sulphide Seiectrode", Application Note.

RECEIVED for review January 17,1983. Accepted June 2,1983. This work is part of an M.Sc. thesis submitted by H.G. to Ben-Gurion University of the Negev, Israel.