Anal. Chem. 1985, 57, 2567-2570 (2) Morse, J. W.; Berner, R. C. I n “Chemical Modeling in Aqueous Systems”; Jenne, E. A., Ed.; American Chemical Society: Washington, DC, 1979; ACS Symposium Series No 93, Chapter 24. (3) Stumm, W.; Morgan, J. J. “Aquatic Chemistry”, 2nd ed.; Wiley: New York, 1981. (4) Baes, C. F., Jr.; Mesmer, R. E. “The Hydrolysis of Cations”; Wiley: New York, 1976. (5) Morel, F. M. M.; Morel-Laurens, N. M. L. I n “Trace Metals in Sea Water”; Wong, C. S., Boyie, E., Bruiand, K. W., Burton, J. D., Goldberg, E. D., Eds.; Plenum Press: New York, 1983; pp 841-869. (6) Kester, D. R.; Byrne, R. H.; Liang, Y. I n “Marine Chemistry in the Coastal Environment”; Church, T. M., Ed.; American Chemlcal Soclety: Washington, DC, 1975; ACS Symposium Serles No. 18, pp 50-79. (7) Culberson, C. H. I n “Marine Electrochemistry”; Whitfield, M., Jagner, D., Eds.; Wiiey: New York, 1981; pp 187-261. (8) Bates, R. G. “Determination of pH Theory and Practice”, 2nd ed.; Wiley: New York, 1973. (9) Sverdrup, H. U.; Johnson, M. W.; Fleming, R. H. “The Oceans: Their Physics, Chemistry and General Biology”; Prentlce-Hall: Englewood Cliffs, NJ, 1942. (10) Miliero, F. J. I n “The Sea”; Goldberg, E. D., Ed.; Wiley: New York, 1974; Vol. 5, Chapter 1. (11) Montgomery, R. B. Deep-sea Res. 1958, 5 , 134-148. (12) Plckard, G. L.; Emery, W. J. “Descriptive Physical Oceanography”, 4th enlarged ed.; Pergamon Press: Oxford, 1982. (13) Brewer, P. G. I n “Chemical Oceanography”, 2nd ed.; Riley, J. P., Sklrrow, G., Eds.; Academic Press: London, 1975; Vol. 1, Chapter 7. (14) Bruland, K. W. I n “Chemical Oceanography”; Riley, J. P., Chester, R., Eds.; Academic Press: London, 1983; Voi. 8, Chapter 45.
2567
(15) Byrne, R. H.; Acker, J. G.; Betzer, P. R.; Feeiy, R. A,; Cates, M. H. Nature (London) 1984, 372, 321-326. (16) Betzer, P. R.; Byrne, R. H.; Acker, J. G.; Lewis, C. S.; Joiiey, R. R.; Feely, R. A. Science 1984, 226, 1074-1077. (17) Morrls, M. J.; Byrne, R. H. Trans. Am. Geophys. Union 1982, 63 (3), 101. (18) Byrne, R. H.; Kester, D. R., J . Solution Chem. 1978, 7 , 373-383. (19) ie Noble, W. J.; Schiott, R. Rev. Sci. Instrum. 1976, 4 7 , 770-771. (20) Byrne, R. H. Rev. Sci. Instrum. 1984, 55, 131-132. (21) le Noble, W. J.; Schiott, R. Rev. Sci. Instrum. 1984, 55, 132. (22) Ramette, R.; Culberson, C. H.; Bates, R. G. Anal. Chem. 1977, 4 9 , 867-870. (23) Marquardt, D. W. J . SOC. Ind. Appl. Math. 1963, 7 1 , 431-441. (24) SAS Instltute, Inc. “SAS Users Guide: Basics”, 1962 ed.; SAS Institute Inc.: Cary, NC, 1962. (25) Sendroy, J., Jr.; Rodkey, F. L. Clin. Chem. (Winston-Salem, N.C.) 1981, 7, 646-654. (26) Byrne R. H.; Young, R. W.; Miller, W. L. J . Solution Chem. 1981, 10, 243-251.
RECEIVED for review February 11,1985. Accepted June 17, 1985. This research was supported by a University of South Florida faculty research grant, by a grant (NA80RAD00020) from the National Oceanic and Atmospheric Administration Air Resources Laboratory, and by the Gulf Oceanographic Charitable Trust (G. Robert-Baldo Graduate Fellowship).
Performance Tests for the Measurement of pH with Glass Electrodes in Low Ionic Strength Solutions Including Natural Waters ,William Davison* and Colin Woof Freshwater Biological Association, The Ferry House, Ambleside, Cumbria, United Kingdom
The performance of different commerclal reference eiectrodes, used to measure the pH of low Ionic strength natural waters, has been tested agalnst a renewable free diffusion Junction. Comparatlve measurements on synthetlc solutions were partially successful In predlctlng performance. Errors determlned In dilute buffers, dilute acids, and distilled water were simllar to those observed in natural waters, but the response In NBS buffers falled to reveal any problems. Poor electrodes depressed the pH from Its true value, more so in stlrred rather than quiescent solutions. The shlR In pH upon stirrlng was largest for the worst electrodes In the most dilute solutions. Better electrodes were characterized by high flow rates of Internal fllling solution through the Junction. As no one test could guarantee electrode performance, the adoption of a standard reference procedure to test electrodes was recommended.
Use of normal electrodes and established procedures for the measurement Of the pH Of low ionic does not guarantee the accuracy (1-6). The prime source of error is associated with the liquid junction of the reference electrode. Great care has been taken in establishing well-defined pH scales. If measurements are made with cell 1
I
reference KC1 unknown ( X ) or electrode > 3 . 5 mol dmW3 standard (S)
H+ion responsive electrode
,1 ,1 1,
they obey the conventional definition of pH and so they are
assumed to be “correct” (7). The measurement is calibrated by substituting a standard buffer solution ( S )for the unknown solution (X), where the pH of the standard buffer is conventionally assigned. This paper is concerned with the reproducibility of the pH measurement according to this convention. Interpretation in terms of the concentration or activity of H+ is left to others (8). Problems which arise are largely associated with the irreproducibility of the liquid junction (represented by the double line in cell 1). This is partly because definitions of pH do not explicitly state the type of junction which should be used.for practical measurements. However, as long ago as 1930 Guggenheim (9) tested various junctions and concluded that only a free diffusion junction (FDJ) could be relied on to give a reproducible potential approaching the ideal value. Free diffusion junctions are in fact incorporated into the pH scales. The BS scale uses such a junction for determining the pH of different standard buffers. The NBS scale uses it to check that the residual liquid junction potential between different buffers is minimal.. Thus a F D j is an obvious choice as a standard junction for precise measurements and it has been recently used as such by Culberson (IO), Illingworth ( I I ) , Brezinski (121,and Covington et al, ( I 3 ) . But what of other junctions? Most commercial electrodes use restrained junctions such as a ceramic or fiber plug or a sleeved joint. Although these have been shown to sometimes give very inaccurate results (11-14), some examples apparently perform well (4). At present there is no simple test which can be applied as a quality control procedure to assess the accuracy of a measurement made witha restrained junction. In this work a wide range of performance characteristics of a variety
0003-2700/85/0357-2567$01.50/0 0 1985 American Chemical Society
2568
ANALYTICAL CHEMISTRY, VOL. 57, NO. 13, NOVEMBER 1985
Table I. Composition and pH of solutions at 20 "C solution
molality, mol kg-l
reagent
buffer 1 buffer 2
potassium hydrogen phthalate disodium hydrogen phosphate potassium dihydrogen phosphate buffer 3 disodium hydrogen phosphate potassium dihydrogen phosphate buffer D1 potassium hydrogen phthalate buffer D2 disodium hydrogen phosphate potassium dihydrogen phosphate dilute acid HZS04 Mosedale see text for composition
0.05
0.03043 0.008695 0.025 0.025 0.01 0.003043 0.0008695 0.0000925 I = 0.0004
pH 4.001" 7.429" 6.881" 4.112b 7.621* 4.O4Oc 5.491d
1
Mosedale
see text for composition
I = 0.0004 5.006d
see text for composition
I = 0.0004 5.974d
2
Wrynose
"Values assigned according to ref 7 . Values assigned according to ref 6. Calculated using Guntleberg approximation for activity coefficient. Mean of measured values usinn FDJ. of commercial electrodes are compared with their ability to accurately measure the pH of natural waters. EXPERIMENTAL S E C T I O N Equipment and Reagents. All potentiometric measurements were performed in a water-jacketed (20 0.1 C)Metrohm titration vessel of 30-mL capacity. A Radiometer PHM 64 research pH meter, used exclusively in millivolt mode, was connected to a chart recorder to provide a continuous record. Although the meter was only specified to *lo0 FV, it proved to be stable to A10 KV allowing chart recorder sensitivity to be set at 0.1 mV cm-'. pH was calculated to f0.001 and then rounded to k0.01 for clear presentation. Buffer solutions (Table I) were prepared from AnalaR reagents according to the procedures in Bates (8). Multiple checks against replicate preparations and commercial products verified an accuracy of better than hO.01. Dilute buffers were prepared gravimetrically from the standard solutions. The dilute acid was prepared by weight from a volumetric stock solution. Water was single distilled. Stream samples were collected in polythene containers and filtered (Whatman GF/F) to remove particulate components. The ionic balance of these waters is dominated by five ions with typical composition (15)as follows (pM): Na+, 150; CaZr, 25; Mg2+, 25; C1-, 150; SO-,: 50. All solutions were equilibrated in a water bath (20 0.1 "C) prior to measurement. A Corning 003 11 lOlJ triple purpose glass electrode, stored in pH 4 buffer, was used for all measurements. Table I1 gives details of the reference electrodes. They were filled with the solution, usually saturated KC1, recommended by the manufacturers. The most rapid electrode response in dilute solution is obtained after storage in the same medium, which reduces memory effects associated with storage in buffers or distilled water (16). The FDJ, which has been described previously (13),was formed
*
*
in a glass T. Reciprocating ganged syringes facilitated the simultaneous replenishment of KCl and withdrawal of microliter quantities of the measured solution. By this means a new junction was formed for each measurement without contaminating the solution with KC1. Procedure. Measurement procedures were designed to simulate the normal routine of a careful laboratory. The electrodes and cell were copiously rinsed with distilled water and prerinsed with the solution to be measured. Buffers 1, 3, 2, D1, and D2, dilute acid, and distilled water were measured in that order using each electrode pair. The millivolts reading was recorded after 4 min of stirring (250 rpm) with a magnetic stirrer and then after 4 min without stirring. In almost all cases the reading was steady after about 1min. Results of buffers 1and 2 were used to calculate the Nernstian slope and the pH of the other solutions. The pH of the natural waters was measured in a different experiment. Unpublished observations showed that the pH of these waters was not perfectly stable and could drift by as much as 0.05 pH units within a day. The FDJ was assumed to give the correct pH. Measurements were performed on buffers 1 and 3 and the two natural waters using the FDJ, allowing sufficient equilibration time to ensure that true equilibrium was established (50.1 mV min-l) on both stirred and quiescent solutions. Similar measurements were then made with an alternative reference electrode but this time allowing 4 min of equilibration for stirred and quiescent values. The procedure using the FDJ was then repeated. Measurements made using the alternative electrode were compared with the mean values obtained with the FDJ. The rate of flow of filling solution from the reference electrode was determined from conductivity measurements. A bridge circuit was used to measure the resistance of the junction. Frit areas were measured with a low powered microscope.
RESULTS A N D DISCUSSION As this work is concerned with assessing errors, Table I11 presents the difference between the measured pH and the correct value. Table I gives the accepted p H for the buffers and dilute acid. Assuming that the partial pressure of COz is 3.3 X atm (13,the p H of air-equilibrated distilled water at 20 "C can be calculated to be 5.64. The stirring shift, that is the difference between values obtained for stirred and quiescent conditions, is also given for each solution. For the natural waters the value obtained using the FDJ was assumed to be correct. Some general points are immediately apparent (Table 111). Measurements performed on quiescent solutions produced results closest to the correct answer. The observed errors were almost exclusively negative; that is, the true p H was higher than the measured value. Stirring usually depressed the p H even further. Good results obtained using the FDJ indicate that the glass electrode was working well. The Nernstian slopes, measured using the FDJ, were 99.3% (stirred solutions) and 99.6% (quiescent solutions) of the theoretical value of 58.17 mV.
Table 11. Details of the Reference Electrodes no.
manufacturer
1
own
2" 3
Beckman Radiometer Orion Radiometer
4" 5"
8"
Russe11 Corning Russe11
9
Radiometer
6
I
design double junction reference single junction ref single junction ref single junction ref combination single tap junction single junction ref single junction ref combination double junction combination single junction
type
junction area, mm2 length, mm
code
renewable
free
diffusion quartz fiber ceramic frit sleeve
0.03 0.76
ceramic
1.07
9
1.82 2.05 2.02
O.lb
frit ceramic frit ceramic frit
ceramic frit Ag/AgCl
ceramic frit
history
3
39416
6
K 401
new
new
90-01-00 GK2501C
new
CRR
new new new
6
00311 602H
O.lb
CWL/LCW GK2401C
4-month lab use
1 year gen lab
" Claimed by the manufacturer to be particularly suitable for low-ionic strength solutions. Estimate rather than measurement.
ANALYTICAL CHEMISTRY, VOL. 57, NO. 13, NOVEMBER 1985
2569
Table 111. Performance of the Electrodes Described in Table 11“ electrode 5
6
7
8
+0.04 +0.04 0.00
-0.08 -0.03 -0.05
-0.15‘ -0.07‘ -0.08b
-0.17 -0.03 -0.14
-1.99 -1.68 -0.31
-0.11 0.00 -0.11
-0.06 -0.04 -0.02
-0.12 -0.07 -0.05
-0.12 -0.06
-0.17 -0.09 -0.08
-2.01 -1.54 -0.47
57.81 57.78
57.67 57.67
57.70 57.70
57.35 57.46
57.32 57.55
54.31 55.42
57.61 57.46
49.88 48.57
0.00 0.00
+9.01 +0.01 0.00
+0.01 +0.01
+0.01
0.00
+0.05 +0.04 +0.01
+0.06
+0.01 0.00
+0.02 +0.02
0.00
+0.01 0.00 +0.01
+0.01
+0.06 +0.01
-0.01 -0.01 0.00
0.00 0.00 0.00
-0.03 -0.01 -0.02
-0.03 -0.03 0.00
-0.07 -0.03 -0.04
-0.04 -0.02 -0.02
-0.45 -0.38 -0.07
-0.09 -0.08 -0.01
-0.01 +0.01 -0.02
+0.01 +0.01 0.00
-0.02 -0.02 -0.01
-0.05 -0.02 -0.03
-0.05 -0.04 0.00
-0.13 -0.05 -0.09
-0.04 -0.04 0.00
-0.45 -0.43 -0.01
-0.18 -0.12 -0.06
-0.39 -0.35 -0.03
+0.04 +0.05 0.00
0.00
0.00 0.00
-0.05 +0.01 -0.06
-0.05 -0.05 0.00
-0.11 -0.02 -0.09
-0.10 -0.02
-0.05 -0.06
-0.08
-1.25 -1.14 -0.12
+0.01
0.00 +0.05 -0.05
-0.01 -0.02 0.00
-0.04 -0.04 0.00
-0.18 0.00 -0.18
-0.17 -0.16 0.00
-0.31 -0.12 -0.18
-0.34 -0.10 -0.24
-2.62 -2.31 -0.31
-0.34 -0.34 0.00
-0.23 +0.04 -0.27
1.7 1.4
11.1 0.2
2.6 0.6
0.11 6.1
2
3
0.00b -0.01 +0.01
-0.11 -0.01 -0.10
-0.05 -0.05 0.00 57.78 57.93
0.00 0.00 0.00
1 ApH Mosedale St ApH Mosedale Q
stirring shift ApH Wrynose St ApH Wrynose Q
stirring shift Nernstian slope St (mV) Nerstian slope Q (mV) ApH buffer 3 St ApH buffer 3 Q
stirring shift ApH buffer D1 St ApH buffer D1 Q
stirring shift ApH buffer D2 St ApH buffer D2 Q
stirring shift ApH dil acid St ApH dil acid Q
stirring shift ApH dist H20 St ApH dist H20 Q
stirring shift flow rate, p L h-’ junction resistance, k0
107 6.9
1.5 0.9
4
30.-300 0.7
-0.08
9
0.1pH) in quiescent solutions, are easily detected. Measurements of dilute buffers or acids will deviate by more than 0.05 pH from the theoretical value. Proving electrodes for precise work is more difficult. Although tests using dilute acids are more sensitive than those using dilute buffers, the latter are better buffered and so they are less prone to preparation or contamination errors. Measurements on distilled water may also help but no single solution can be recommended to provide a definitive check. If measurements are to be made quite quickly in fairly large volumes of solution, contamination by
KC1 will not be a problem. An electrode with a large flow rate and flux can then be selected. Reliance should not be placed on a single series of tests because electrodes are prone to change due to aging or contamination. One method of quality control (19) is to check the pH of the distilled water each time the electrodes are rinsed. If a stock of air-equilibrated distilled water is reserved for pH measurement, experience will show that within a single laboratory the pH stays within quite narrow (k0.05) limits. Although it would be prudent to check that the electrode slope is near theoretical (>98%) and that the stirring shift is minimal, neither measurement will ensure the correct answer. However, negligible stirring shift is advisable for other reasons. The glass electrode exerts a buffering effect on the solution in its immediate vicinity, and so it can contribute errors in poorly buffered quiescent media (20). Checking that the stirring error is insignificantly small, verifies the absence of errors from both glass and reference electrode. Moreover, it permits measurements to be made on stirred solutions, which generally equilibrate more quickly. Direct comparison with results obtained in natural water using a more ideal junction, such as the FDJ, should provide the best test for any electrode. Evidently there is a need to establish a design for a simple reliable FDJ which can be universally accepted and so serve as a standard reference procedure for the measurement of pH. Such a system would have to be comprehensively tested before it can be recommended for general use.
ACKNOWLEDGMENT We thank Ed Tipping, Nigel Hetherington, and Elisabeth M. Evans for their assistance. Registry No. H,O, 7732-18-5. LITERATURE CITED (1) Analytical Quality Control (Harmonlsed Monitoring) Committee Analyst (London) 1984, 109, 431. (2) Herczeg, A. L.; Hesslein, R. H. Geochim. Cosmochim. Acta 1984, 48, 837. (3) Galloway, J. N.; Cosby, B. J.; Likens, G. E. Limnol. Oceanogr. 1979, 24, 1161. (4) McQuaker, N. R.; Kluckner, P. D.; Sandberg, D. K. Environ. Sci. Techno/. 1983, 17, 431. (5) Tyree, S. Y. Atmos. Environ. 1981, 5 , 57-60. (6) Covington, A. K.; Whalley, P. D.; Davison, W. Analyst (London) 1983, 108, 1528. (7) Covington, A. K.; Bates, R. G.; Durst, R. A. Pure Appl. Chem. 1983, 55, 1467. (8) Bates, R. G. "Determination of pH"; Wiley: New York, 1964. (9) Guggenheim, E. A. J. A m . Chem. SOC. 1930, 5 2 , 1315. (IO) Culberson, C. I n "Marlne Electrochemistry": Whitfield, M., Jagner, D., Eds.; Wlley: New York, 1981. (11) Illingworth, J. A. Biochem. J. 1981, 195, 259. (12) Brezinski, D. P. Analyst(London) 1983, 108, 425. (13) Covington, A. K.; Whalley, P. D.; Davlson, W. Anal. Chim. Acta 1985 169, 221. (14) Koch, W. F.; Marinenko, G. ASTMSpec. Tech. Pub/. 1983, No. 823, 10-17. (15) Carrick, T. R.; Sutcllffe, D. W. Occas. Pub/. Freshwater B i d . Assoc. 1983, 21. (16) Metcalf, R. C. Analysf (London) 1984, 109, 1225. (17) Stumm, W.; Morgan, J. J. "Aquatic Chemistry", 2nd ed.; Wiley: New York. 1981. (18) Westcott, C. C. "pH Measurement"; Academic Press: New York, 1978. (19) Wells, D.; Harrlman, R., manuscript in preparation. (20) Midgley, D.; Torrance, K. Analyst (London) 1979, 104, 63.
RECEIVED for review March 18,1985. Accepted May 21,1985. This work, stimulated by the Royal Society, was partly funded by the Natural Environment Research Council.
