- Email: [email protected]
Isotachophoretic determination of chromium(VI) at low parts per billion concentrations
Journal of Chromatography,
390 (1987) ill-120 Elsevier Science Publishers B.V., Amsterdam -
Printed in The Netherlands
CHROM. 19 124
ISOTACHOPHORETIC PARTS PER BILLION
I. ZELENSKP
DETERMINATION CONCENTRATIONS
OF CHROMIUM(VI)
AT LOW
and V. ZELENSKA
Department of Analytical Chemistry, Komensk? (Czechoslovakia)
University, Mlynskri Do&a
CH-2, CS 842 15 Bratislava
and D. KANIANSKY* Institute of Chemistry, slovakia)
Komenskj
University, Mlynskci Dolina CH-2,
CS 842 IS Bratislava (Czecho-
SUMMARY
Factors important in the isotachophoretic determination of CrV’ at low ppb concentrations were studied. To increase the selectivity of the analysis, the use of photometric detection at a 405 wavelength was preferred. Losses of Cr” due to adsorption were found to be the main potential source of analytical errors at the concentration levels of interest. The presence of sulphate in the sample solution at ca. 10e4 M concentration eliminated the losses due to adsorption on the walls of the sample handling glassware. In analysis with a low pH of the leading electrolyte, adsorption of Cr” on the walls of the separation compartment also played a role; addition of naphthalene- 1,3,6-trisulphonate to the sample solution considerably reduced this disturbance. When the precautions concerning adsorption were taken, the detection limits for Cr”’ were in the range 45 ppb (depending on the pH of the leading electrolyte) for a 30-~1 sample volume. The calibration graphs were linear over the concentration range lo-‘-5 . lop6 M with good correlation coefficients. The reproducibilities of the determinations within this concentration range were 2-3% or better. The practical utility of capillary isotachophoresis in the trace determination of CrV’ in drinking water and wastewater samples seems promising.
INTRODUCTION
The widespread industrial use of chromium (e.g., in electroplating, manufacture of metallic alloys, corrosion inhibitors) provides several routes for entry of CrV’ into the environment. As it is very toxic, with proved carcinogenecity’, chromium is on the list of priority pollutants of the U.S. Environmental Protection Agency*. For obvious reasons, CrV’ is currently monitored in different samples of environmental importance3 and a variety of analytical methods have been developed for this purpose (for reviews, see, e.g., refs. 3-7). It has been found that separation 0021-9673/87/$03.50
0
1987 Elsevier Science Publishers B.V.
I. ZELENSKk,
112
V. ZELENSKA,
D. KANIANSKY
and concentration methods can be advantageously combined, mostly with spectral and electrochemical methods of quantitation when CrV’ needs to be determined at low ppb* concentrations or less. Separation and concentration effects are inherently combined in isotachophoresis (ITP) and the advantages of these features of ITP in the determination of CrV’ were clearly demonstrated by Taglia and Lederer*. To our knowledge, with the exception of model separations9, no attention has been paid to the use of capillary ITP (c-ITP) for the determination of CrV’, in spite of the fact that it is often present in anionic test mixtures used in this method (see, e.g., refs. 10 and 11). The aim of this work was to investigate the capabilities of c-ITP in the trace determination of Cr”, with emphasis on its determination in water. To achieve detection limits at the low ppb level with an adequate load capacity of the separation compartment and a high overall selectivity of the analysis, we employed a column coupling configuration of the separation unit l*, l 3 with photometric detection of the analyte. EXPERIMENTAL
Instrumentation A CS Isotachophoretic Analyser (VVZ PJT, SpiSsk6 Nova Ves, Czechoslovakia) was assembled in the column coupling configuration of the separation unit using modules provided by the manufacturer. The analyser was equipped with a UVD 01 photometric detector (VVZ PJT) in the analytical column. This fixed-wavelength detector can be used for detection at 254 and 405 nm and its detection cell is placed upstream of the conductivity sensor. Peak areas from the detector were evaluated with an HP 3390 A reporting integrator (Hewlett-Packard, Avondale, PA, U.S.A.). Chemicals Chemicals used for the preparation of the leading and terminating electrolytes were obtained from Lachema (Brno, Czechoslovakia), Reanal (Budapest, Hungary) and Sigma (St. Louis, MO, U.S.A.). Some of them were purified by conventional methods. Naphthalene-1,3,6-trisulphonic acid (NTS) was purchased from K & K Labs. (Plainview, N.Y., U.S.A.) and hydroxyethylcellulose 4000 (HEC) and methylhydroxyethylcellulose 30000 (MHEC) from Serva (Heidelberg, F.R.G.). Other chemicals were also obtained from the above suppliers. HEC, after purification on a mixed-bed ion exchanger, was used as an additive to the leading electrolyte at a concentration of 0.2% and MHEC served for coating of the walls of the separation compartment as described previously14. RESULTS AND DISCUSSION
Choice of working conditions As photometric detection can be employed for quantitation when the analytes are migrating in the interzonal boundary layers* 5’l 6, we preferred its use in this work. l
Throughout the article the American billion (log) is meant.
ITP OF CHROMIUM(V1)
AT LOW ppb CONCENTRATIONS
113
This approach to quantitation is very convenient when determinations at trace concentration levels are required. In our introductory experiments with practical water samples spiked with Cr”’ we found that c-ITP has potential for the trace determination of this constituent and both wavelengths of the detector (see Experimental) are suitable for this purpose. Although the absorptivity of Cr”’ at 405 nm is lower than that at 254 nm (see Fig. l), we preferred the former wavelength as in this way we could increase the selectivity of the analysis. The isotachopherograms shown in Fig. 2 clearly illustrate the advantages of this choice. To minimize the number of migrating constituents and thus increase the selectivity of the separation conditions, we preferred to work a’t a lower pH of the leading electrolyte while the ratio of the absorptivities at 405 nm (Fig. 1) favours the use of operational system No. 2 (Table I). The choice of suitable spacers for the determination of Cr”’ in the above mode was one of the tasks of this part of the work. As the choice was not straightforward (see below), possible alternatives are discussed and the final proposal is given in the section devoted to the analysis of practical samples. Disturbances in the ITP analysis of 0”’ and means for their elimination In the introductory experiments we observed losses of Cr”’ owing to its adsorption on the walls of the volumetric flasks made of Simax glass. Different substances were applied as additives to the stock solutions of chromate and to the solutions used for the measurement of the calibration graphs. Whereas some of them were effective for only a short time (e.g., citric acid and salts of phosphoric acid), sodium sulphate when added at a concentration of lop4 h4 prevented the adsorption for at least 48 h. With these precautions in the sample handling, the quantitation of Cr”’ was trouble-free using the operational system at pHL = 6.9 with the expected value of the detection limit (see below). As already stated, one of the aims of this work was to increase the selectivity of the ITP analysis by working at a low pH of the leading electrolyte and operational system No. 1 (Table I) was chosen for this purpose. Experimentss carried out with this system, however, indicated a high detection limit, the reproducibilities of the
0.6
Fig. 1. UV-visible absorption spectra of a 10m4M solution of K2Cr04 in the leading electrolyte solutions (see Table I) at (1) pH 3.5 and (2) pH 6.9. The spectra were measured relative to the electrolyte solutions without Cr” (HEC was not added to the solutions for the spectral measurement). The cell pathlength was 2 cm.
I. ZELENSK-?, V. ZELENSKA,
114 a
dr(Vf)
D. KANIANSKY
b
:f
Fig. 2. Isotachopherograms from the determination of Cr”’ with photometric detection at (a) 254 and (b) 405 nm. Operational system No. 1 (Table I) was used. The CrV’ concentration was IO” M and (1) sulphate, (2) NTS and (3) methanesulphonate were added to the sample at lw4 Mconcentrations; Picrate (4) present in the sample at cu. lD_’ Mconcentration served as a marker for the identification of the migration position of Cr”‘. The sample was injected by a valve (30 ~1) and the driving currents were 250 and 45 ,uA for the pre-separation and analytical columns, respectively. A = Increasing light absorption; f = increasing time.
determinations were lower than those with the system at pHL = 4.9 and CrV’ was detected in the analytical column even in experiments when none of the separands was allowed to enter this column after the preseparation stage (see Figs. 3 and 4). Obviously, the above precautions concerning the sample handling were followed in these experiments and good performance of the equipment was revealed by repeating some runs at pHL = 6.9. Isotachopherograms from the photometric detector for a series of runs at pHL = 3.5 (Fig. 3) illustrate this unexpected behaviour of CrV’. These findings could be explained by a pH-dependent adsorption of CrV’ on the walls of the separation compartment (with the exception of 2-3 mm long channel in the bifurcation block the sample in this compartment came into contact only with polytetrafluoroethylene and a copolymer of fluorinated ethylene and propylene) and, therefore, we carried out an experimental investigation along this line. Adsorption of the separands is known in zone electrophoresis on carriersl’ and in capillary zone electrophoresis l *#lg. Its disturbing role in c-ITP seems less obvious, as the time course of the field strength along the separation compartment TABLE I OPERATIONAL
SYSTEMS
Parameter
Sysrem No.
Solvent Leading anion Concentration (m&f) Counter ion Additive to the leading electrolyte pH of leading electrolyte Terminating anion l
1
2
Water Cl10 BALA* HEC (0.1%, w/v) 3.5 Acetate
Water Cl10 BTP* HEC (0.1%. w/v) 6.9 Citrate
BALA = B-Alanine; BTP = 1,3-bis[tris(hydroxymethyl)methylamino]propane.
ITP OF CHROMIUM(V1) a
‘;
I
c
b
1
d
~-j--L- -L1
1
A
r
115
AT LOW ppb CONCENTRATIONS
&
CrWl)
Fig, 3. Isotachopherograms from the photometric detector (405 nm) for the determination of Crv’ in operational system No. .l (Table I). A 30-,ul volume of a solution containing Cr”’ at 1OV M concentration was injected in all the runs. Ethylenediaminetetraacetate (EDTA), phosphate, citrate and succinate present in the sample solution at 10m4Mconcentrations served as spacers. Fen-EDTA complex (1) at a 2. 1OV h4 concentration served as an internal standard for the evaluation of Cr” present in the analytical column. (a) Complete train of the zones was introduced into the analytical column; (b) sulphate, Crv’ and half of the EDTA zone (the constituents migrated in this order) were led out the separation compartment after the pre-separation stage (no Crw should be present in the analytical column); (c) as in (b) and in addition also the remainder of the EDTA zone and half of the phosphate zone were led out of the separation compartment after the first stage; (d) all separands and part of the terminating zone were eliminated as in (b) and (c) (Fe”‘-EDTA was not present in the sample in this experiment). The amount of Crw found in run (d) was cu. 30% of the amount detexcted in (a). For the driving currents and the meaning of the other symbols, see Fig. 2.
and displacement effects of the less mobile separands act against it. This feature of ITP has found practical use in desorption isotachophoresiszO. Nevertheless, a strong memory effect of the walls of the separation compartment to additives acting against electroosmosis10J4,21 can be explained only in terms of their preferential adsorption, Adsorption must also be considered in the ITP separation of some proteinsz2. In our case anionic constituents which were expected to exhibit adsorption on the walls of the separation compartment were added to the samples containing constant concentrations of CrV’ (for details see the legend to Fig. 4) and ITP runs with such samples were evaluated by the photometric detector. When naphthalene-1,3,6trisulphonate (NTS) was present in the sample by far the highest recovery of Cr” was achieved. The isotachopherograms in Fig. 4 show the efficiency of this constituent migrating before Cr ” . It is also apparent that the amount of NTS injected was important. Cr ‘I lost in the separation compartment could be recovered when the required amount of NTS was injected with some time delay after the injection of CrV’. All these results provide strong evidence of reversible adsorption of Crv’ on the walls
i&
trcvo
&“I)
&VI,
Fig. 4. Role of naphthalene-1,3,6_trisulphonate in the determination of CrV’ at a low pH of the leading electrolyte. Operational system No. 1 (Table I) was used. The concentration of Cr”’ was 2 . 1OV M in all experiments. Sulphate (1) and methanesulphonate (2) were present in the sample solutions at 10e4 M concentrations; the concentrations of NTS were as follows: (a) 0; (b) 1V; (c) 2 . 1OV; (d) 5 10m5;(e) 1c4 A4. The peak areas for Cry’ in runs (d) and (e) agreed within 2%. The sample volume and the driving currents were as in Figs. 2 and 3. For the meaning of the other symbols, see Fig. 2.
I. ZELENSKY,
116
V. ZELENSKA,
D. KANIANSKY
of the separation compartment at a low pH of the leading electrolyte. It is also apparent that neither a high field strength in the terminating zone nor displacement effects of the less mobile separands were effective in this instance. Therefore, a detrimental effect of the adsorption was eliminated in subsequent experiments by the use of NTS. As this constituent did not exhibit a long-term memory effect, it was added to the samples at the required concentration (see Fig. 4 and below). Discrete spacers for the determination of CrV’ in operational system No. 1 were chosen according to their effective mobilities relative to the chromate zone among the constituents proposed for this system 23. For some spacers migrating behind the chromate zone, with effective mobilities close to that .of the analyte, we observed typical chromate tailinglo*’ l (see also Fig. 5b) also when NTS was used. The tailing was detected when the amount of Cr”’ injected was sufficient for the formation of its own zone. On the other hand, when the amount injected was such that CrV’ migrated in the interzonal layer its position in the train of separands indicated its lower effective mobility under otherwise identical conditions (see Fig. 5~). A possible explanation of this migration behaviour can be found in the chemical equilibria. At a pH of the leading electrolyte (pH 3.5) the following equilibria must be considered4: _ Cr,O$ - + Hz0 = 2 HCrOi
HCrO;
(1)
Y? H+ + CrOf-
(2) T
a
JQL
I
R i3 ;k&l
I
A
2 9
v
:
$--LL i3c (VI) 4
:
‘L
I--
/3 /2
1
/L
t Fig. 5. Migration behaviour of CrV’ in the operational system at pHL = 3.5. (a) Blank run with (1) sulphate, (2) NTS and (3) sulphamate injected at 1W M concentrations; (b) as in (a), but Cr”’ was present in the sample at 10e4 ii/l concentration [the sensitivity of the detector (405 nm) was five times lower in comparison with runs (a) and (c)l; (c) as (a), but the concentration of sulphamate (3) was doubled and Cr”’ was present in the sample at a 10m6M concentration. L and T = leading and terminating zones, respectively; R = increasing resistance in the conductivity detector. For the meaning of the other symbols and for further experimental parameters, see Fig. 2.
ITP OF CHROMIUM(VI)
AT LOW ppb CONCENTRATIONS
117
and three ionic forms of CrV1coexist under such conditions. At lower concentrations of CrV’ (at cu. 10e4 M or less), HCrO, is the main ionic form, whereas at higher concentrations dichromate will dominate. (ref. 4, p, 56). When we consider the ionic mobilities of the forms involved 24, the migration behaviour is understandable. When Cr” is injected in a smaller amount it is present at a lower concentration in the interzonal layer in the preferred HCrO; form (having lower ionic mobility), whereas when it is migrating in its own zone the form Cr 20 =- (having a higher ionic mobility) prevails at the ITP steady-state concentration. Hence the tailing of Cr” (its presence in the sulphamate zone in Fig. 5b) can be explained in the following way. When for any reason (diffusion, thermal convection, wall adsorption, etc.) CrzO$ - is entering the sulphamate zone, the equilibrium (1) prefers the HCrOi form. At the same time HCrO; species present in the chromate zone bleedz5 into the sulphamate zone. As both sulphamate and HCrOi behave as anionic constituent of strong electrolytes (for the pK values see, e.g., ref. 24) they must migrate in a normal modez6 under the separation conditions employed. Consequently, their migration order according to the ionic mobilities is decisive and HCrO; will accumulate behind the sulphamate zone (see Fig. 5~). The accumulation of HCrO; must lead to the formation of CrzO$- species which are forced to migrate into the chromate zone. Under the conditions of ITP steady-state an equilibrium in this CrV’ transport is established and the zone stacked between the migration positions of chromate will resemble a steady-state mixed zone 27,28. The trace of the photometric detector in Fig. 5b indicates labile kinetics of the equilibrium (l), otherwise a different photometric profile would be expected (see, e.g., ref. 29). For obvious reasons, discrete spacers behaving in this way were excluded in the quantitative work. In the analysis of practical samples, however, some constituents can act similarly and this possibility must be considered when a universal detector indicates the presence of constituent(s) with effective mobility(ies) intermediate between those of HCrO; and CrzO?-. Methanesulphonate was found to be a suitable spacer for the determination of Cr” as it migrates behind HCrOi with an effective mobility only slightly lower than that of sulphamate (see Figs. 3 and 4). Quantitative aspects of the determination of Cry’ and its applicability in the analysis of practical water samples
With the above findings in mind, we added to the samples in quantitative work at pHL = 3.5 the following constituents: (1) sulphate at 10m4Mconcentration (to prevent the adsorption of CrV’ on the walls of the volumetric flasks); (2) naphthalene-1,3,5-trisulphonate at 2 . 1OV M concentration (to eliminate the loss of Cr” in the separation compartment and to space chromate from the side of the leading zone); and (3) methanesulphonate at lop4 M concentration (a spacer from the side of the terminating zone). In the work at pHL = 6.9 the use of NTS was not necessary (see above) and sulphate served as a spacer from the side of the leading zone. Parameters of the regression lines describing the calibration graphs for the determination of CrV’ in both operational systems are given in Table II. The relative
I. ZELENSKY, V. ZELENSKA,
118
D. KANIANSKY
TABLE II REGRESSION EQUATIONS AND CORRELATION 1O-6 M CONCENTRATION RANGE Operational system No.
Regression equation*
1 1
yA = 70 308 + yu = 1.77 +
2 2
yA = 71575 y, = 2.00
633 240 17.88 + 1225 100 + 36.53
x x x x
COEFFICIENTS
FOR Crv’ IN THE IO-‘-5 .
Correlation coejicien t
Number of data points
0.9983 0.9982 0.9934 0.9993
14 14 15 15
l yA, yu = Peak area (counts of the integrator) and peak height (mm), respectively; x = concentration (pmole/l).
standard deviations were 3.4 and 2.6% for the determination of 10 and 250 ppb concentrations, respectively, in the operational system No. 1 (Table I) when the peak areas were measured. The peak-height measurements (using a ruler on the isotachopherogram) gave reprodueibilities of 0.52% for the concentration range studied a b
b
c
‘r
XVI)
1 A
,,Cr(VI) A
Fig. 6. Determination of Cr”’ in drinking water with photometric detection (405 nm). (a) Sample of tap water with (1) NTS and (2) methanesulphonate at low4 Mconcentrations [picrate (3) present at a 2 . 10m7 M concentration, served as a marker; see Fig. 21; (b) as (a), but the sample was spiked with Cr”’ at a 2 ’ lo-’ A4 concentration and the concentration of picrate was lo-’ M. The analysis was carried out in operational system No. 1 (Table I). For the driving currents and the meaning of the other symbols, see . Fig. 2. Fig. 7. Analysis of wastewater from a chromium electroplating process. The working conditions were identical with those in Fig. 6. (a) Wastewater sample diluted 1:10; (b) as (a), but with a dilution of 1:100; (c) analysis of sample (b) after addition of ascorbate. The concentration of picrate (3) was ca. lo-’ M.
ITP OF CHROMIUM(V1)
AT LOW ppb CONCENTRATIONS
119
(lo-250 ppb). The determinations using the operational system No. 2 were characterized by similar relative standard deviations. The detection limits (peaks three times higher than the short-term noise’l) were 5 and 4 ppb of CrV’ for operational systems Nos. 1 and 2, respectively. The injection volumes were 30 ~1 in both instances. When the detection limits provided by c-ITP are compared with those in the literature30 for the determination of Cr by other methods employed in the trace analysis of this element (5-7000 ppb), it is apparent that c-ITP is a suitable alternative at low ppb concentration levels. Ion chromatography (IC) combined with photometric detection at 365 nm provides a 10 ppb detection limit for CrV1 present in a 275 ,MIinjection volume3r. When the detection limits claimed for ITP and IC, the absorptivities at 365 and 405 nm (see Fig. 1) and the injection volumes are compared, the advantages of the concentrating power of ITP are apparent. In the analysis of practical water samples we found the operational system at pHL = 3.5 to be more suitable than the alternative system at pHL = 6.9 as it eliminates the migration of carbonate, which can be present in some water samples at high concentrations. The isotachopherograms in Fig. 6 were obtained in the determination of CrV’ in drinking water. The isotachopherogram in Fig. 6a is from the analysis of tap water and that in Fig. 6b was obtained for the same sample spiked at 10.4 ppb (2. lop7 M) with Cr “’ . An example of the use of c-ITP in the determination of CrV1 in waste water is shown in Fig. 7. In this instance CrV’ discharged into an industrial effluent was determined; a very dilute sample of the effluent could be reliably analysed (see the legend to Fig. 7). The identity of Cr” was confirmed by spiking the sample with this constituent. Disappearance of the CrV1 peak on the addition of ascorbate to the sample (Fig. 7c) also confirmed the identity of the analyte and at the same time showed the purity of the peak. CONCLUSIONS
This work shows the promising possibilities of c-ITP in the trace determination of Cr”‘. Photometric detection at 405 nm and the separation at low pH make the analysis sufhciently selective at low ppb concentrations. The limit of detection (4-5 ppb for 30-~1 samples volumes) depends on the pH of the leading electrolyte and can probably be decreased further by increasing the sample volume and/or by detection at a wavelength close to the absorption maximum (see Fig. 1). Such an improvement, however, must take into a consideration the limits of the load capacity32 and a decrease in the detection selectivity. Reliable analytical results were achieved when adsorption of the analyte on the walls of the sample-handling glassware and on the walls of the separation compartment (at pHL = 3.5) was prevented. The latter disturbance could be effectively eliminated by adding naphthalene-1,3,6-trisulphonate to the sample solution. The results of the analyses of practical water samples indicate the applicability of c-ITP to the determination of CrV’ in materials of environmental importance. REFERENCES 1 J. Marhold, PFehled PrCmyslod Toxikologie, Avicenum, Prague, 1980, p. 275. 2 H. E. Wise, Jr., and P. D. Fahrenthold, Environ. Sci. Technol., 15 (1981) 1292.
120
I1 12 13 14 15 16 17 18 19 20 21
I. ZELENSKY,
V. ZELENSKA,
D. KANIANSKY
A. K. Lavrukhina and L. V. Youkina, Analiticheskaya Kfiimya Khroma, Nauka, Moscow, 1979. W. J. Williams, Handbook of Anion Determination, Butterworths, London, 1979. J. Fries and H. Getrost, Organic lieagentsfor Trace Analysis, E. Merck, Darmstadt, 1977, p. 104. M. J. Fishman, D. E. Erdmann and J. R. Garbino, Anal. Chem., 55 (1983) 102R. J. R. Garbino, T. R. Steinhemer and H. E. Taylor, Anal. Chem., 57 (1985) 46R. V. Taglia and M. Lederer, J. Chromatogr., 77 (1973) 467. P. I. Bresler, I. A. Ivanova, 0. V. Oshurkova and G. A. Shtilerman, Zh. Anal. Khim., 36 (1981) 593. F. M. Everaerts, J. L. Beckers and Th. P. E. M. Verheggen, Isotachophoresis: Theory, Instrumentation and Applications, Elsevier, Amsterdam, Oxford, New York, 1976. P. HavaHi and D. Kaniansky, J. Chromatogr., 325 (1985) 137. F. M. Everaerts, Th. P. E. M. Verheggen and F. E. P. Mikkers, J. Chromatogr., 169 (1979) 21. D. Kaniansky, Thesis, Komensky University, Bratislava, 1983. D. Kaniansky, M. Koval and S. Stankoviansky, J. Chromatogr., 267 (1983) 67. L. Arlinger, J. Chromatogr., 91 (1974) 785. M. Svoboda and J. Vacik, J. Chromatogr., 119 (1976) 539. W. Ostrowski, in Z. Deyl (Editor), Electrophoresis: A Survey of Techniques and Apphcations, Part A, Elsevier, Amsterdam, Oxford, New York, 1979, p. 69. S. Hjerten, J. Chromatogr., 347 (1985) 191. H. H. Latter and D. McManigill, Trends Anal. Chem., 5 (1986) 11. V. Ka%ka and Z. Prusik, J. Chromatogr., 273 (1983) 117. J. C. Reijenga, G. V. Aben, Th. P. E. M. Verheggen and F. M. Everaerts, J. Chromatogr.. 260 (1983) 241.
22 Z. Prusik and V. KaSiEka, in P. Bo&k (Editor), Basic and Advanced Course of Isotachophoresis, ITP-84, September 24, 1984. Hradec Kralove, Czechoslovakia, VVZ PJT, SpiSskL Nova Ves, 1984, p. 49. 23 V. Madajova, D. Kaniansky and J. Marak, presented at the 5th International Symposium on Isotachophoresis, Maastricht, September 3-5, 1986. 24 T. Hirokawa, M. Nishino, N. Aoki, Y. Kiso, Y. Sawamoto, T. Yagi and J.-I. Akiyama, J. Chromatogr., 271 (1983) Dl. 25 P. Gebauer, P. BoEek, M. Deml and J. Janak, J. Chromatogr., 199 (1980) 81. 26 P. Gebauer and P. BoEek, J. Chromatogr., 267 (1983) 49. 27 J. P. M. Wielders, Thesis, University of Technology, Eindhoven, 1978. 28 J. P. M. Wielders and F. M. Everaerts, in B. J. Radola and D. Graesslin (Editors), Electrofocusing and Isotachophoresis, Walter de Gruyter, Berlin, New York, 1977, p. 527. 29 P. Gebauer and P. Bo&k, J. Chromatogr., 299 (1984) 321. 30 E. Wlnninen, in L. Niinistrii (Editor), EuroanaZysis IV, Reviews on Analytical Chemistry, Akademiai Kiadb, Budapest, 1982, p. 157. 31 Detection of Low ppb Levels of Chromiumj Vr) in Wastewater and Drinking Water, Application Note 51, Dionex, Sunnyvale, CA, 1985. 32 F. E. P. Mikkers, F. M. Everaerts and J. A. F. Peek, J. Chromatogr., 168 (1979) 293.
390 (1987) ill-120 Elsevier Science Publishers B.V., Amsterdam -
Printed in The Netherlands
CHROM. 19 124
ISOTACHOPHORETIC PARTS PER BILLION
I. ZELENSKP
DETERMINATION CONCENTRATIONS
OF CHROMIUM(VI)
AT LOW
and V. ZELENSKA
Department of Analytical Chemistry, Komensk? (Czechoslovakia)
University, Mlynskri Do&a
CH-2, CS 842 15 Bratislava
and D. KANIANSKY* Institute of Chemistry, slovakia)
Komenskj
University, Mlynskci Dolina CH-2,
CS 842 IS Bratislava (Czecho-
SUMMARY
Factors important in the isotachophoretic determination of CrV’ at low ppb concentrations were studied. To increase the selectivity of the analysis, the use of photometric detection at a 405 wavelength was preferred. Losses of Cr” due to adsorption were found to be the main potential source of analytical errors at the concentration levels of interest. The presence of sulphate in the sample solution at ca. 10e4 M concentration eliminated the losses due to adsorption on the walls of the sample handling glassware. In analysis with a low pH of the leading electrolyte, adsorption of Cr” on the walls of the separation compartment also played a role; addition of naphthalene- 1,3,6-trisulphonate to the sample solution considerably reduced this disturbance. When the precautions concerning adsorption were taken, the detection limits for Cr”’ were in the range 45 ppb (depending on the pH of the leading electrolyte) for a 30-~1 sample volume. The calibration graphs were linear over the concentration range lo-‘-5 . lop6 M with good correlation coefficients. The reproducibilities of the determinations within this concentration range were 2-3% or better. The practical utility of capillary isotachophoresis in the trace determination of CrV’ in drinking water and wastewater samples seems promising.
INTRODUCTION
The widespread industrial use of chromium (e.g., in electroplating, manufacture of metallic alloys, corrosion inhibitors) provides several routes for entry of CrV’ into the environment. As it is very toxic, with proved carcinogenecity’, chromium is on the list of priority pollutants of the U.S. Environmental Protection Agency*. For obvious reasons, CrV’ is currently monitored in different samples of environmental importance3 and a variety of analytical methods have been developed for this purpose (for reviews, see, e.g., refs. 3-7). It has been found that separation 0021-9673/87/$03.50
0
1987 Elsevier Science Publishers B.V.
I. ZELENSKk,
112
V. ZELENSKA,
D. KANIANSKY
and concentration methods can be advantageously combined, mostly with spectral and electrochemical methods of quantitation when CrV’ needs to be determined at low ppb* concentrations or less. Separation and concentration effects are inherently combined in isotachophoresis (ITP) and the advantages of these features of ITP in the determination of CrV’ were clearly demonstrated by Taglia and Lederer*. To our knowledge, with the exception of model separations9, no attention has been paid to the use of capillary ITP (c-ITP) for the determination of CrV’, in spite of the fact that it is often present in anionic test mixtures used in this method (see, e.g., refs. 10 and 11). The aim of this work was to investigate the capabilities of c-ITP in the trace determination of Cr”, with emphasis on its determination in water. To achieve detection limits at the low ppb level with an adequate load capacity of the separation compartment and a high overall selectivity of the analysis, we employed a column coupling configuration of the separation unit l*, l 3 with photometric detection of the analyte. EXPERIMENTAL
Instrumentation A CS Isotachophoretic Analyser (VVZ PJT, SpiSsk6 Nova Ves, Czechoslovakia) was assembled in the column coupling configuration of the separation unit using modules provided by the manufacturer. The analyser was equipped with a UVD 01 photometric detector (VVZ PJT) in the analytical column. This fixed-wavelength detector can be used for detection at 254 and 405 nm and its detection cell is placed upstream of the conductivity sensor. Peak areas from the detector were evaluated with an HP 3390 A reporting integrator (Hewlett-Packard, Avondale, PA, U.S.A.). Chemicals Chemicals used for the preparation of the leading and terminating electrolytes were obtained from Lachema (Brno, Czechoslovakia), Reanal (Budapest, Hungary) and Sigma (St. Louis, MO, U.S.A.). Some of them were purified by conventional methods. Naphthalene-1,3,6-trisulphonic acid (NTS) was purchased from K & K Labs. (Plainview, N.Y., U.S.A.) and hydroxyethylcellulose 4000 (HEC) and methylhydroxyethylcellulose 30000 (MHEC) from Serva (Heidelberg, F.R.G.). Other chemicals were also obtained from the above suppliers. HEC, after purification on a mixed-bed ion exchanger, was used as an additive to the leading electrolyte at a concentration of 0.2% and MHEC served for coating of the walls of the separation compartment as described previously14. RESULTS AND DISCUSSION
Choice of working conditions As photometric detection can be employed for quantitation when the analytes are migrating in the interzonal boundary layers* 5’l 6, we preferred its use in this work. l
Throughout the article the American billion (log) is meant.
ITP OF CHROMIUM(V1)
AT LOW ppb CONCENTRATIONS
113
This approach to quantitation is very convenient when determinations at trace concentration levels are required. In our introductory experiments with practical water samples spiked with Cr”’ we found that c-ITP has potential for the trace determination of this constituent and both wavelengths of the detector (see Experimental) are suitable for this purpose. Although the absorptivity of Cr”’ at 405 nm is lower than that at 254 nm (see Fig. l), we preferred the former wavelength as in this way we could increase the selectivity of the analysis. The isotachopherograms shown in Fig. 2 clearly illustrate the advantages of this choice. To minimize the number of migrating constituents and thus increase the selectivity of the separation conditions, we preferred to work a’t a lower pH of the leading electrolyte while the ratio of the absorptivities at 405 nm (Fig. 1) favours the use of operational system No. 2 (Table I). The choice of suitable spacers for the determination of Cr”’ in the above mode was one of the tasks of this part of the work. As the choice was not straightforward (see below), possible alternatives are discussed and the final proposal is given in the section devoted to the analysis of practical samples. Disturbances in the ITP analysis of 0”’ and means for their elimination In the introductory experiments we observed losses of Cr”’ owing to its adsorption on the walls of the volumetric flasks made of Simax glass. Different substances were applied as additives to the stock solutions of chromate and to the solutions used for the measurement of the calibration graphs. Whereas some of them were effective for only a short time (e.g., citric acid and salts of phosphoric acid), sodium sulphate when added at a concentration of lop4 h4 prevented the adsorption for at least 48 h. With these precautions in the sample handling, the quantitation of Cr”’ was trouble-free using the operational system at pHL = 6.9 with the expected value of the detection limit (see below). As already stated, one of the aims of this work was to increase the selectivity of the ITP analysis by working at a low pH of the leading electrolyte and operational system No. 1 (Table I) was chosen for this purpose. Experimentss carried out with this system, however, indicated a high detection limit, the reproducibilities of the
0.6
Fig. 1. UV-visible absorption spectra of a 10m4M solution of K2Cr04 in the leading electrolyte solutions (see Table I) at (1) pH 3.5 and (2) pH 6.9. The spectra were measured relative to the electrolyte solutions without Cr” (HEC was not added to the solutions for the spectral measurement). The cell pathlength was 2 cm.
I. ZELENSK-?, V. ZELENSKA,
114 a
dr(Vf)
D. KANIANSKY
b
:f
Fig. 2. Isotachopherograms from the determination of Cr”’ with photometric detection at (a) 254 and (b) 405 nm. Operational system No. 1 (Table I) was used. The CrV’ concentration was IO” M and (1) sulphate, (2) NTS and (3) methanesulphonate were added to the sample at lw4 Mconcentrations; Picrate (4) present in the sample at cu. lD_’ Mconcentration served as a marker for the identification of the migration position of Cr”‘. The sample was injected by a valve (30 ~1) and the driving currents were 250 and 45 ,uA for the pre-separation and analytical columns, respectively. A = Increasing light absorption; f = increasing time.
determinations were lower than those with the system at pHL = 4.9 and CrV’ was detected in the analytical column even in experiments when none of the separands was allowed to enter this column after the preseparation stage (see Figs. 3 and 4). Obviously, the above precautions concerning the sample handling were followed in these experiments and good performance of the equipment was revealed by repeating some runs at pHL = 6.9. Isotachopherograms from the photometric detector for a series of runs at pHL = 3.5 (Fig. 3) illustrate this unexpected behaviour of CrV’. These findings could be explained by a pH-dependent adsorption of CrV’ on the walls of the separation compartment (with the exception of 2-3 mm long channel in the bifurcation block the sample in this compartment came into contact only with polytetrafluoroethylene and a copolymer of fluorinated ethylene and propylene) and, therefore, we carried out an experimental investigation along this line. Adsorption of the separands is known in zone electrophoresis on carriersl’ and in capillary zone electrophoresis l *#lg. Its disturbing role in c-ITP seems less obvious, as the time course of the field strength along the separation compartment TABLE I OPERATIONAL
SYSTEMS
Parameter
Sysrem No.
Solvent Leading anion Concentration (m&f) Counter ion Additive to the leading electrolyte pH of leading electrolyte Terminating anion l
1
2
Water Cl10 BALA* HEC (0.1%, w/v) 3.5 Acetate
Water Cl10 BTP* HEC (0.1%. w/v) 6.9 Citrate
BALA = B-Alanine; BTP = 1,3-bis[tris(hydroxymethyl)methylamino]propane.
ITP OF CHROMIUM(V1) a
‘;
I
c
b
1
d
~-j--L- -L1
1
A
r
115
AT LOW ppb CONCENTRATIONS
&
CrWl)
Fig, 3. Isotachopherograms from the photometric detector (405 nm) for the determination of Crv’ in operational system No. .l (Table I). A 30-,ul volume of a solution containing Cr”’ at 1OV M concentration was injected in all the runs. Ethylenediaminetetraacetate (EDTA), phosphate, citrate and succinate present in the sample solution at 10m4Mconcentrations served as spacers. Fen-EDTA complex (1) at a 2. 1OV h4 concentration served as an internal standard for the evaluation of Cr” present in the analytical column. (a) Complete train of the zones was introduced into the analytical column; (b) sulphate, Crv’ and half of the EDTA zone (the constituents migrated in this order) were led out the separation compartment after the pre-separation stage (no Crw should be present in the analytical column); (c) as in (b) and in addition also the remainder of the EDTA zone and half of the phosphate zone were led out of the separation compartment after the first stage; (d) all separands and part of the terminating zone were eliminated as in (b) and (c) (Fe”‘-EDTA was not present in the sample in this experiment). The amount of Crw found in run (d) was cu. 30% of the amount detexcted in (a). For the driving currents and the meaning of the other symbols, see Fig. 2.
and displacement effects of the less mobile separands act against it. This feature of ITP has found practical use in desorption isotachophoresiszO. Nevertheless, a strong memory effect of the walls of the separation compartment to additives acting against electroosmosis10J4,21 can be explained only in terms of their preferential adsorption, Adsorption must also be considered in the ITP separation of some proteinsz2. In our case anionic constituents which were expected to exhibit adsorption on the walls of the separation compartment were added to the samples containing constant concentrations of CrV’ (for details see the legend to Fig. 4) and ITP runs with such samples were evaluated by the photometric detector. When naphthalene-1,3,6trisulphonate (NTS) was present in the sample by far the highest recovery of Cr” was achieved. The isotachopherograms in Fig. 4 show the efficiency of this constituent migrating before Cr ” . It is also apparent that the amount of NTS injected was important. Cr ‘I lost in the separation compartment could be recovered when the required amount of NTS was injected with some time delay after the injection of CrV’. All these results provide strong evidence of reversible adsorption of Crv’ on the walls
i&
trcvo
&“I)
&VI,
Fig. 4. Role of naphthalene-1,3,6_trisulphonate in the determination of CrV’ at a low pH of the leading electrolyte. Operational system No. 1 (Table I) was used. The concentration of Cr”’ was 2 . 1OV M in all experiments. Sulphate (1) and methanesulphonate (2) were present in the sample solutions at 10e4 M concentrations; the concentrations of NTS were as follows: (a) 0; (b) 1V; (c) 2 . 1OV; (d) 5 10m5;(e) 1c4 A4. The peak areas for Cry’ in runs (d) and (e) agreed within 2%. The sample volume and the driving currents were as in Figs. 2 and 3. For the meaning of the other symbols, see Fig. 2.
I. ZELENSKY,
116
V. ZELENSKA,
D. KANIANSKY
of the separation compartment at a low pH of the leading electrolyte. It is also apparent that neither a high field strength in the terminating zone nor displacement effects of the less mobile separands were effective in this instance. Therefore, a detrimental effect of the adsorption was eliminated in subsequent experiments by the use of NTS. As this constituent did not exhibit a long-term memory effect, it was added to the samples at the required concentration (see Fig. 4 and below). Discrete spacers for the determination of CrV’ in operational system No. 1 were chosen according to their effective mobilities relative to the chromate zone among the constituents proposed for this system 23. For some spacers migrating behind the chromate zone, with effective mobilities close to that .of the analyte, we observed typical chromate tailinglo*’ l (see also Fig. 5b) also when NTS was used. The tailing was detected when the amount of Cr”’ injected was sufficient for the formation of its own zone. On the other hand, when the amount injected was such that CrV’ migrated in the interzonal layer its position in the train of separands indicated its lower effective mobility under otherwise identical conditions (see Fig. 5~). A possible explanation of this migration behaviour can be found in the chemical equilibria. At a pH of the leading electrolyte (pH 3.5) the following equilibria must be considered4: _ Cr,O$ - + Hz0 = 2 HCrOi
HCrO;
(1)
Y? H+ + CrOf-
(2) T
a
JQL
I
R i3 ;k&l
I
A
2 9
v
:
$--LL i3c (VI) 4
:
‘L
I--
/3 /2
1
/L
t Fig. 5. Migration behaviour of CrV’ in the operational system at pHL = 3.5. (a) Blank run with (1) sulphate, (2) NTS and (3) sulphamate injected at 1W M concentrations; (b) as in (a), but Cr”’ was present in the sample at 10e4 ii/l concentration [the sensitivity of the detector (405 nm) was five times lower in comparison with runs (a) and (c)l; (c) as (a), but the concentration of sulphamate (3) was doubled and Cr”’ was present in the sample at a 10m6M concentration. L and T = leading and terminating zones, respectively; R = increasing resistance in the conductivity detector. For the meaning of the other symbols and for further experimental parameters, see Fig. 2.
ITP OF CHROMIUM(VI)
AT LOW ppb CONCENTRATIONS
117
and three ionic forms of CrV1coexist under such conditions. At lower concentrations of CrV’ (at cu. 10e4 M or less), HCrO, is the main ionic form, whereas at higher concentrations dichromate will dominate. (ref. 4, p, 56). When we consider the ionic mobilities of the forms involved 24, the migration behaviour is understandable. When Cr” is injected in a smaller amount it is present at a lower concentration in the interzonal layer in the preferred HCrO; form (having lower ionic mobility), whereas when it is migrating in its own zone the form Cr 20 =- (having a higher ionic mobility) prevails at the ITP steady-state concentration. Hence the tailing of Cr” (its presence in the sulphamate zone in Fig. 5b) can be explained in the following way. When for any reason (diffusion, thermal convection, wall adsorption, etc.) CrzO$ - is entering the sulphamate zone, the equilibrium (1) prefers the HCrOi form. At the same time HCrO; species present in the chromate zone bleedz5 into the sulphamate zone. As both sulphamate and HCrOi behave as anionic constituent of strong electrolytes (for the pK values see, e.g., ref. 24) they must migrate in a normal modez6 under the separation conditions employed. Consequently, their migration order according to the ionic mobilities is decisive and HCrO; will accumulate behind the sulphamate zone (see Fig. 5~). The accumulation of HCrO; must lead to the formation of CrzO$- species which are forced to migrate into the chromate zone. Under the conditions of ITP steady-state an equilibrium in this CrV’ transport is established and the zone stacked between the migration positions of chromate will resemble a steady-state mixed zone 27,28. The trace of the photometric detector in Fig. 5b indicates labile kinetics of the equilibrium (l), otherwise a different photometric profile would be expected (see, e.g., ref. 29). For obvious reasons, discrete spacers behaving in this way were excluded in the quantitative work. In the analysis of practical samples, however, some constituents can act similarly and this possibility must be considered when a universal detector indicates the presence of constituent(s) with effective mobility(ies) intermediate between those of HCrO; and CrzO?-. Methanesulphonate was found to be a suitable spacer for the determination of Cr” as it migrates behind HCrOi with an effective mobility only slightly lower than that of sulphamate (see Figs. 3 and 4). Quantitative aspects of the determination of Cry’ and its applicability in the analysis of practical water samples
With the above findings in mind, we added to the samples in quantitative work at pHL = 3.5 the following constituents: (1) sulphate at 10m4Mconcentration (to prevent the adsorption of CrV’ on the walls of the volumetric flasks); (2) naphthalene-1,3,5-trisulphonate at 2 . 1OV M concentration (to eliminate the loss of Cr” in the separation compartment and to space chromate from the side of the leading zone); and (3) methanesulphonate at lop4 M concentration (a spacer from the side of the terminating zone). In the work at pHL = 6.9 the use of NTS was not necessary (see above) and sulphate served as a spacer from the side of the leading zone. Parameters of the regression lines describing the calibration graphs for the determination of CrV’ in both operational systems are given in Table II. The relative
I. ZELENSKY, V. ZELENSKA,
118
D. KANIANSKY
TABLE II REGRESSION EQUATIONS AND CORRELATION 1O-6 M CONCENTRATION RANGE Operational system No.
Regression equation*
1 1
yA = 70 308 + yu = 1.77 +
2 2
yA = 71575 y, = 2.00
633 240 17.88 + 1225 100 + 36.53
x x x x
COEFFICIENTS
FOR Crv’ IN THE IO-‘-5 .
Correlation coejicien t
Number of data points
0.9983 0.9982 0.9934 0.9993
14 14 15 15
l yA, yu = Peak area (counts of the integrator) and peak height (mm), respectively; x = concentration (pmole/l).
standard deviations were 3.4 and 2.6% for the determination of 10 and 250 ppb concentrations, respectively, in the operational system No. 1 (Table I) when the peak areas were measured. The peak-height measurements (using a ruler on the isotachopherogram) gave reprodueibilities of 0.52% for the concentration range studied a b
b
c
‘r
XVI)
1 A
,,Cr(VI) A
Fig. 6. Determination of Cr”’ in drinking water with photometric detection (405 nm). (a) Sample of tap water with (1) NTS and (2) methanesulphonate at low4 Mconcentrations [picrate (3) present at a 2 . 10m7 M concentration, served as a marker; see Fig. 21; (b) as (a), but the sample was spiked with Cr”’ at a 2 ’ lo-’ A4 concentration and the concentration of picrate was lo-’ M. The analysis was carried out in operational system No. 1 (Table I). For the driving currents and the meaning of the other symbols, see . Fig. 2. Fig. 7. Analysis of wastewater from a chromium electroplating process. The working conditions were identical with those in Fig. 6. (a) Wastewater sample diluted 1:10; (b) as (a), but with a dilution of 1:100; (c) analysis of sample (b) after addition of ascorbate. The concentration of picrate (3) was ca. lo-’ M.
ITP OF CHROMIUM(V1)
AT LOW ppb CONCENTRATIONS
119
(lo-250 ppb). The determinations using the operational system No. 2 were characterized by similar relative standard deviations. The detection limits (peaks three times higher than the short-term noise’l) were 5 and 4 ppb of CrV’ for operational systems Nos. 1 and 2, respectively. The injection volumes were 30 ~1 in both instances. When the detection limits provided by c-ITP are compared with those in the literature30 for the determination of Cr by other methods employed in the trace analysis of this element (5-7000 ppb), it is apparent that c-ITP is a suitable alternative at low ppb concentration levels. Ion chromatography (IC) combined with photometric detection at 365 nm provides a 10 ppb detection limit for CrV1 present in a 275 ,MIinjection volume3r. When the detection limits claimed for ITP and IC, the absorptivities at 365 and 405 nm (see Fig. 1) and the injection volumes are compared, the advantages of the concentrating power of ITP are apparent. In the analysis of practical water samples we found the operational system at pHL = 3.5 to be more suitable than the alternative system at pHL = 6.9 as it eliminates the migration of carbonate, which can be present in some water samples at high concentrations. The isotachopherograms in Fig. 6 were obtained in the determination of CrV’ in drinking water. The isotachopherogram in Fig. 6a is from the analysis of tap water and that in Fig. 6b was obtained for the same sample spiked at 10.4 ppb (2. lop7 M) with Cr “’ . An example of the use of c-ITP in the determination of CrV1 in waste water is shown in Fig. 7. In this instance CrV’ discharged into an industrial effluent was determined; a very dilute sample of the effluent could be reliably analysed (see the legend to Fig. 7). The identity of Cr” was confirmed by spiking the sample with this constituent. Disappearance of the CrV1 peak on the addition of ascorbate to the sample (Fig. 7c) also confirmed the identity of the analyte and at the same time showed the purity of the peak. CONCLUSIONS
This work shows the promising possibilities of c-ITP in the trace determination of Cr”‘. Photometric detection at 405 nm and the separation at low pH make the analysis sufhciently selective at low ppb concentrations. The limit of detection (4-5 ppb for 30-~1 samples volumes) depends on the pH of the leading electrolyte and can probably be decreased further by increasing the sample volume and/or by detection at a wavelength close to the absorption maximum (see Fig. 1). Such an improvement, however, must take into a consideration the limits of the load capacity32 and a decrease in the detection selectivity. Reliable analytical results were achieved when adsorption of the analyte on the walls of the sample-handling glassware and on the walls of the separation compartment (at pHL = 3.5) was prevented. The latter disturbance could be effectively eliminated by adding naphthalene-1,3,6-trisulphonate to the sample solution. The results of the analyses of practical water samples indicate the applicability of c-ITP to the determination of CrV’ in materials of environmental importance. REFERENCES 1 J. Marhold, PFehled PrCmyslod Toxikologie, Avicenum, Prague, 1980, p. 275. 2 H. E. Wise, Jr., and P. D. Fahrenthold, Environ. Sci. Technol., 15 (1981) 1292.
120
I1 12 13 14 15 16 17 18 19 20 21
I. ZELENSKY,
V. ZELENSKA,
D. KANIANSKY
A. K. Lavrukhina and L. V. Youkina, Analiticheskaya Kfiimya Khroma, Nauka, Moscow, 1979. W. J. Williams, Handbook of Anion Determination, Butterworths, London, 1979. J. Fries and H. Getrost, Organic lieagentsfor Trace Analysis, E. Merck, Darmstadt, 1977, p. 104. M. J. Fishman, D. E. Erdmann and J. R. Garbino, Anal. Chem., 55 (1983) 102R. J. R. Garbino, T. R. Steinhemer and H. E. Taylor, Anal. Chem., 57 (1985) 46R. V. Taglia and M. Lederer, J. Chromatogr., 77 (1973) 467. P. I. Bresler, I. A. Ivanova, 0. V. Oshurkova and G. A. Shtilerman, Zh. Anal. Khim., 36 (1981) 593. F. M. Everaerts, J. L. Beckers and Th. P. E. M. Verheggen, Isotachophoresis: Theory, Instrumentation and Applications, Elsevier, Amsterdam, Oxford, New York, 1976. P. HavaHi and D. Kaniansky, J. Chromatogr., 325 (1985) 137. F. M. Everaerts, Th. P. E. M. Verheggen and F. E. P. Mikkers, J. Chromatogr., 169 (1979) 21. D. Kaniansky, Thesis, Komensky University, Bratislava, 1983. D. Kaniansky, M. Koval and S. Stankoviansky, J. Chromatogr., 267 (1983) 67. L. Arlinger, J. Chromatogr., 91 (1974) 785. M. Svoboda and J. Vacik, J. Chromatogr., 119 (1976) 539. W. Ostrowski, in Z. Deyl (Editor), Electrophoresis: A Survey of Techniques and Apphcations, Part A, Elsevier, Amsterdam, Oxford, New York, 1979, p. 69. S. Hjerten, J. Chromatogr., 347 (1985) 191. H. H. Latter and D. McManigill, Trends Anal. Chem., 5 (1986) 11. V. Ka%ka and Z. Prusik, J. Chromatogr., 273 (1983) 117. J. C. Reijenga, G. V. Aben, Th. P. E. M. Verheggen and F. M. Everaerts, J. Chromatogr.. 260 (1983) 241.
22 Z. Prusik and V. KaSiEka, in P. Bo&k (Editor), Basic and Advanced Course of Isotachophoresis, ITP-84, September 24, 1984. Hradec Kralove, Czechoslovakia, VVZ PJT, SpiSskL Nova Ves, 1984, p. 49. 23 V. Madajova, D. Kaniansky and J. Marak, presented at the 5th International Symposium on Isotachophoresis, Maastricht, September 3-5, 1986. 24 T. Hirokawa, M. Nishino, N. Aoki, Y. Kiso, Y. Sawamoto, T. Yagi and J.-I. Akiyama, J. Chromatogr., 271 (1983) Dl. 25 P. Gebauer, P. BoEek, M. Deml and J. Janak, J. Chromatogr., 199 (1980) 81. 26 P. Gebauer and P. BoEek, J. Chromatogr., 267 (1983) 49. 27 J. P. M. Wielders, Thesis, University of Technology, Eindhoven, 1978. 28 J. P. M. Wielders and F. M. Everaerts, in B. J. Radola and D. Graesslin (Editors), Electrofocusing and Isotachophoresis, Walter de Gruyter, Berlin, New York, 1977, p. 527. 29 P. Gebauer and P. Bo&k, J. Chromatogr., 299 (1984) 321. 30 E. Wlnninen, in L. Niinistrii (Editor), EuroanaZysis IV, Reviews on Analytical Chemistry, Akademiai Kiadb, Budapest, 1982, p. 157. 31 Detection of Low ppb Levels of Chromiumj Vr) in Wastewater and Drinking Water, Application Note 51, Dionex, Sunnyvale, CA, 1985. 32 F. E. P. Mikkers, F. M. Everaerts and J. A. F. Peek, J. Chromatogr., 168 (1979) 293.