Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://analyt.chem.msu.ru/kinetics/papers/Sorpt-Cat%20Test%20of%20Cu%20Analyst.pdf Äàòà èçìåíåíèÿ: Thu Sep 25 14:13:56 2008 Äàòà èíäåêñèðîâàíèÿ: Mon Oct 1 20:16:58 2012 Êîäèðîâêà:

Sorption­catalytic testing of copper on a paper-based sorbent with attached alkylamino groups

Mikhail K. Beklemishev,*a Yuliya Yu. Petrovab and Inga F. Dolmanova
a

a

b

Department of Chemistry, Moscow State University, 119899 GSP-3, Moscow, Russia. E-mail: mkb@analyt.chem.msu.ru Medical Department, Surgut State University, Surgut, 626400, Russia. E-mail: chem@surgu.wsnet.ru

Received 17th May 1999, Accepted 20th July 1999

Copper(ii) was preconcentrated from aqueous solution on paper filters with chemically attached hexamethylenediamino (HMDA) groups and determined directly on the filters by its catalytic action in the oxidation of hydroquinone with H2O2 in the presence of 2,2A-dipyridyl; the reaction was monitored by measuring the absorbance of wet filters. The linear calibration graph of absorbance at 5 min versus log cCu permits the semiquantitative determination of CuII over the range 1 3 1026­0.1 mg l21 for a sample volume of 10 ml. Tolerance ratios for foreign ions are 3­4 orders of magnitude more favorable in the hybrid sorption­catalytic procedure than those for the catalytic determination of copper in solution without preconcentration. The procedure is simple and does not require any expensive instrumentation. Samples of tap and sea-water were analyzed. In order to improve the selectivity of catalytic techniques, separation and preconcentration methods are used.1 In cases when sorption is utilized, the determination is usually performed in aqueous solution after desorption of the analyte.2 Conducting a catalytic indicator reaction directly on the surface of the sorbent would simplify the procedure and improve its metrological characteristics by eliminating the superfluous stage of desorption. This approach (for which we use the term `sorption­catalytic determination') has not been developed until recently; we have demonstrated its feasibility by the determination of manganese after its preconcentration on filterpaper with aminocarboxylic groups.3 The purpose of the present studies was the further evaluation of the possibilities and the areas of applicability of the sorption­ catalytic principle. In this work we realized this approach in the determination of copper. For its preconcentration stage, efficient and selective sorbents are needed. Very efficient for the group preconcentration of transition metal ions are sorbents containing conformationally flexible amino and carboxylic groups chemically attached to paper cellulose4,5 (Scheme 1). A thorough investigation of the sorption properties of this series of sorbents6­8 has shown that aminocarboxylic diethylenetriaminetetraacetate (DETATA) sorbents are powerful chelating agents whereas diethylenetriamine (DETA) and hexamethylenediamine (HMDA) sorbents featuring alkylamino groups are not so efficient but are considered to be more selective towards copper(ii), which has a strong affinity for amine nitrogen. As the indicator reaction we selected the oxidation of hydroquinone with hydrogen peroxide catalyzed by CuII, which is one of the most sensitive, selective and well studied reactions for the catalytic determination of copper in solution.9­11 One of the main potential advantages of the new sorption­ catalytic method might be a substantial increase in selectivity in comparison with procedures performed in solution (owing to the preconcentration stage). The object of this work was to determine whether that increase in selectivity was achievable using the selected sorbent and the model catalytic indicator reaction as an example. For this purpose, it was necessary to optimize the conditions for the sorption­catalytic determination of copper(ii) after its preconcentration on the selected sorbent, to obtain the metrological characteristics of the determination and to estimate the selectivity parameters.

Experimental
Reagents, solutions and apparatus Hydroquinone `pure for analysis' (Reakhim, Moscow, Russia) was purified by sublimation. Other reagents were of analyticalreagent grade. Doubly distilled water (additionally de-ionized by pumping through a mixture of anion and cation exchangers) and rectified ethanol were used to prepare solutions. Buffer solutions for the pH range 1.8­12 were prepared from a mixture of acetic, boric and phosphoric acid (100 ml, 0.04 mol l21 of each acid) by adding appropriate amounts (0­100 ml) of 0.2 mol l21 NaOH. A stock standard copper solution containing 20 g l21 CuII in 0.01 mol l21 HNO3 was prepared; working standard solutions with lower CuII concentrations were prepared by dilution without additional acidification (solutions containing 1 mg l21 CuII or less were prepared daily). As the silica sorbent, plates for TLC were used (Sorbpolymer, Krasnodar, Russia). The paper-based sorbent obtained from Dr. G. I. Tsysin consisted of ashless filter-paper disks (2.5 cm in diameter) with chelating HMDA groups chemically attached to cellulose.4,5 After use the filters were regenerated by Analyst, 1999, 124, 1523­1527 1523

Scheme 1 Chelating groups attached to filter-paper: 1, hexamethylenediamine (HMDA); 2, diethylenetriamine (DETA); 3, diethylenetriaminetetraacetate (DETATA).

This journal is © The Royal Society of Chemistry 1999.


use of the following procedure: 30­40 filters were soaked for 5­10 min consecutively in 30 ml of distilled water, in 1 3 1024 mol l21 HCl and in a buffer solution (pH 6.8) (soaking in each solution was repeated three times). The filters were then rinsed with distilled water, dried in air and stored in a stoppered glass flask. The regeneration may be repeated at least four times. A KFK-3 spectrophotometer (ZOMZ, Zagorsk, Russia; precision 0.001 absorbance units) was used for measurements of the absorbance of solutions and filter-papers. All measurements were carried out in an air-conditioned laboratory at room temperature (23 ± 1 °C). Indicator reaction in solution To conduct the hydroquinone­CuII­2,2A-dipyridyl (dipy)­H2O2 indicator reaction, the following solutions are mixed in a 30 ml test-tube (the optimum concentrations are given): (1) 6.4 ml of buffer (pH 8.2); (2) 1 ml of copper solution; (3) 0.1 ml of 0.01 mol l21 dipy; (4) 1 ml of 5 mol l21 H2O2; and (5) 1.5 ml of 0.1 mol l21 hydroquinone (total volume 10 ml). Preliminary experiments showed that this mixing order results in the highest absorbance of the products. A stop-watch is started at the moment of adding of hydroquinone; the contents of the test-tube are shaken and transferred into a 2 cm spectrophotometer cell. After 7 min the absorbance of the oxidation product is measured at 490 nm against water (for research purposes, complete kinetic curves were obtained). Indicator reaction on filter-papers For the purpose of copper sorption, an HMDA filter is fixed in a polyethylene filter holder (Biospectr, St. Petersburgh, Russia) connected to a syringe by means of silicone-rubber tubing and 10 ml of the copper-containing solution is pumped through the filter at a rate of not higher than 20 ml min21. The filter is then dried and the indicator reaction is conducted on the filter. Coloration of the paper was most intense with the following mixing order: (1) CuII; (2) buffer; (3) activator (dipy); (4) H2O2; and (5) hydroquinone (other mixing orders led to a more diffuse colored spot). Two buffer solutions are used in this procedure: (a) to adjust the pH of the copper solution to be pumped through the filter (usually pH 5.6) and (b) to conduct the indicator reaction on the filter (pH 5.6 or 8.0). Buffer (a) is diluted 10-fold compared with buffer (b) to make sure that the pH on the filter will not be shifted from the desired value by buffer (a). The reaction on the filter and the measurements are performed as described in a previous paper:3 the absorbance of the reaction products is measured on the filter in the wet state against the absorbance of the same filter before the start of the reaction as described below. Aliquots of the reagent solutions are applied to the center of the filter by use of an Eppendorf-type pipette; upon adding each of the solutions (1­4), the filter is dried with a gentle flow of pressurized air until no traces of moisture are visible. Upon addition of peroxide solution (4), but before drying, the filter is placed between two glass plates fixed together with a clip and placed in the cell compartment of the spectrophotometer perpendicular to the light beam; the absorbance of this filter is taken as zero. The filter is then dried and the hydroquinone solution (5) is pipetted on to the filter; this moment is taken as the beginning of the reaction. The wet filter is then again placed between the glass plates for the measurements. The kinetic curves on the supports soon tend to bend so that the initial rate is difficult to measure. In this connection, the absorbance of the products at 5 min (A5) or the catalytic effect of copper at 5 min (DA5, which is the difference in A5 for the catalyzed and uncatalyzed reactions) is taken as the analytical signal. The detection limit is calculated as the concentration corresponding to the absorbance value of A5,blank + 3sa [A5,blank 1524 Analyst, 1999, 124, 1523­1527

= 0.019 and the standard deviation of the blank determined by direct experiments is sa = 0.002 (n = 3)].

Results and discussion
Evaluation of paper-based chelating sorbents for sorption­catalytic determination of copper The idea of the hybrid sorption­catalytic method involves preconcentration of CuII with its subsequent determination on the sorbent. One of the most powerful chelating sorbents of the studied series containing DETATA groups (Scheme 1) extracts copper(ii) at pH > 3.8 We also studied filter-papers with attached DETA and HMDA alkylamino groups. In order to examine the possibility of using these sorbents for the sorption­ catalytic determination of copper, we passed a coppercontaining solution through the filter-papers and then conducted the copper-catalyzed hydroquinone­H2O2 reaction directly on the filter-papers by pipetting reagent solutions on to the filter. The pH of the indicator reaction on the filters was kept constant while the pH of the copper solution to be pumped through the filter was varied. The results of this study (Table 1) show that the analytical signal on DETATA filters is close to that on unmodified filterpaper. Presumably, DETATA groups extract copper but exert an unfavorable effect on its catalytic activity in the indicator reaction, which is likely to be associated with a too high stability of the copper(ii) complex. Higher indicator reaction rates are obtained on HMDA and DETA filters containing only alkylamino groups that bind copper in complexes of lower stability. Further investigations were performed with HMDA filters that provided the highest analytical signal. Optimization of the indicator reaction conditions on HMDA filters Blank analytical signals on both HMDA filters and plain filterpaper are low [Fig. 1(a), curves 1 and 1A]; when copper solution is pumped through filter-paper, the signal is increased, which implies that some amount of metal is retained. The highest reaction rate is achieved with the filters with attached HMDA groups that provide copper preconcentration. Maximum signals are obtained at pH 4­6.5 and 8­9; further investigation was performed at pH 5.6. At pH > 9 a different product of the indicator reaction is formed (the pink product is rapidly converted into a yellowish brown compound), which makes it difficult to compare the reaction rates at different pH values. The study of the concentration effect of peroxide, hydroquinone and 2,2A-dipyridyl on the indicator reaction [Fig. 1(b)­ (d)] allowed us to select the optimum conditions of the indicator reaction on HMDA filters (8 3 1027 mol hydroquinone, 9 3 1026 mol H2O2 and 1 3 1027 mol dipy). The curves of reaction rate on HMDA filters versus concentration of the reagents
Table 1 Catalytic effect of copper on filter-paper and paper-based sorbents. DA5 is the difference in absorbances of the filters pumped with 10 ml of 10 mg ml21 CuII solution of the stated pH and with pure buffer solution, measured 5 min after the start of the indicator reaction conducted on the sorbents. Amounts of reagents: 9 mmol H2O2, 0.6 mmol hydroquinone, 0.05 mmol dipy. pH of indicator reaction: 8.0 DA5 on the sorbents pH of sorption 5.6 7.0 8.0 Filter-paper 0.036 0.025 0.022 DETATA 0.039 0.028 0.022 DETA 0.050 0.029 0.020 HMDA 0.059 0.031 0.058


feature either a maximum or a plateau similar to the main properties of the curves obtained in solution,4,5 which indicates that the reaction on HMDA filters involves the formation of a catalytically active intermediate complex of CuII with the reagents analogous to that proposed for this reaction in solution10,12 and probably including the nitrogen atoms of HMDA. The study of the effect of the pumped solution volume on the analytical signal showed that samples up to 100 ml may be used for the preconcentration of copper without any noticeable losses. The study of the pumping rate effect showed that the analytical signal remains constant over the rate range 8­20 ml min21, after which a slight decrease takes place. There is no need for strict control of the pumping rate within this range, which allows the procedure to be simplified by using a syringe rather than a peristaltic pump. Sensitivity and precision of sorption­catalytic determination of copper The calibration plots for the determination of copper on HMDA filters and in solution differ (Fig. 2): on HMDA filters the linearity range is very wide (1 3 1026­0.1 mg l21). Relative to the reaction in solution (Table 2), it is extended towards low concentrations of the metal. Correspondingly, the limit of detection in the case for the sorption­catalytic determination (1 3 1027 mg l21, or 0.001 ng for a sample volume of 10 ml) is several orders of magnitude lower than that in solution (50 ng for a 2 cm cell which holds 10 ml of solution), which indicates that the reaction rate is higher on HMDA filters than in solution (this phenomenon is not connected with the preconcentration of copper because we consider absolute amounts of the metal). A similar increase in the indicator reaction rate on a paper-based sorbent has also been observed in the KIO4­MnII­3,3A,5,5Atetramethylbenzidine reaction.3

The higher catalytic activity of copper on HMDA filters may be associated with the nature of the matrix itself (filter-paper) or with the nature of the attached groups. To reveal the role of the latter, the copper solution was pipetted on to the filters (in contrast to pumping, this technique guaranteed the presence of 100% of the introduced metal both on the filter-paper and on HMDA filters); then the indicator reaction was carried out as usual. Comparing the data for HMDA filters and filter-paper (Fig. 3, curves 1 and 1A), we can see that HMDA groups activate catalysis at pH 5.5­6 and inhibit it at pH 6­9. These effects are not very strong and cannot account for the order of magnitude difference in reaction rates and sensitivity for low copper concentrations on HMDA filters compared with solution. We conclude that the difference is primarily caused by the matrix (filter-paper). The described feature of the calibration curves (different slopes of the plots for solution and HMDA filters, Fig. 2) results in different precisions in the determination. The relatively low

Fig. 2 Appearance of the calibration plots for the determination of copper by the catalytic technique in solution (1) and sorption­catalytic technique on HMDA filters (2). The signals without copper are 0.103 ± 0.004 (1) and 0.019 ± 0.002 (2).

Fig. 1 Dependence of the rate of the hydroquinone (HQ)­CuII­H2O2­2,2A-dipyridyl (dipy) reaction on filter-paper with filter-paper (1A, 2A) (a) on the pH of the pumped CuII solution and on the amounts of (b) HQ, (c) hydrogen peroxide and curves 2 with copper. Conditions of sorption on the filters: pH 5.6 [for (a) the pH was varied], VCu = 10 ml, cCu = 10 mg (a) pH 8.0, (b)­(d) pH 5.6; amounts of reagents, (a) 6 3 1027 mol HQ, 9 3 1026 mol H2O2, 1 3 1027 mol dipy; (b) 9 dipy; (c) 8 3 1027 mol HQ, 5 3 1028 mol dipy; (d) 8 3 1027 mol HQ, 9 3 1026 mol H2O2.

attached HMDA groups (1, 2) and (d) dipy. Curves 1 without copper; l21; indicator reaction on the filter, 3 1026 mol H2O2, 5 3 1028 mol

Analyst, 1999, 124, 1523­1527

1525


slope of the graph on the support leads to high deviations in the results. It is notable that the reproducibility of the absorbance measurements on the filters is fairly high (Table 3): the precision is not impaired by the heterogeneous character of the process or irregularity in the filter-paper structure. However, the analytical signal (A5) is proportional to the logarithm of the copper concentration (Table 2), so that a small error in absorbance results in a higher error in the concentration (Table 3). For this reason, in spite of the high repeatability of the absorbance measurements, the procedure is semiquantitative. For example, around 1026 mg ml21 of copper, two metal concentrations will be confidently distinguished if they differ by 2 3 100.52 = 101.04 times, i.e., by an order of magnitude, and for the range 1025­1022 mg ml21 this difference will be 2 3 10(0.2­0.3) 3­4 times (Table 3). Interferences Most transition metal ions and iodide interfere with the determination of copper owing to acceleration of the indicator reaction. The criterion for interference was taken as a change of ±0.006 absorbance units for reaction in solution (0.01 mg l21 CuII) and ±0.002 absorbance units for reaction on HMDA filters (0.1 ng CuII). As demonstrated in Table 4, preconcentration of copper on the chelating sorbent improves the selectivity factors by 3­4 orders of magnitude (MnII ion is an exception). Analysis of water Samples were obtained from the Barents Sea and from water supply lines (before and after purification); a sample of tap water after its treatment by an AquaRoss (Surgut, Russia) household filter was also analyzed. An aliquot of the sample (5 ml; water samples Nos. 1 and 2 were diluted 10-fold) mixed with the buffer (pH 5.6, 5 ml) was pumped through an HMDA filter and the indicator reaction was performed as described under `Indicator reaction on filter-papers'. The reference technique involved preconcentration of copper on 8-hydrox-

yquinoline-modified activated carbon and determination by atomic absorption spectrometry (AAS) [Varian (Palo Alto, CA, USA) AA-1020 spectrometer, direct burning of the concentrate; for samples 1 and 2 no preconcentration was used]. As shown in Table 5, the agreement between the results of the sorption­ catalytic and AAS determinations is good (it should be taken into account that the proposed procedure is semiquantitative). Advantages of the sorption­catalytic approach The developed procedure demonstrates that a highly sensitive catalytic method may be successfully combined with selective sorption preconcentration of the analyte. The selectivity of the catalytic procedure is increased considerably in the hybrid sorption­catalytic technique, primarily owing to the separation stage. Elimination of the analyte desorption stage shortens the analysis time (12­14 min are required for an analysis from the beginning of pumping the sample). Conducting the indicator reaction directly on the sorbent broadens the linear range (from two orders of magnitude in solution to five on the support) and
Table 3 Precision of absorbance measurements and the corresponding deviations of log cCu in sorption­catalytic determination of copper (n = 3, P = 0.95). D(log cCu) = DA5/b = ts/bAn, where A5 is the absorbance of the filter measured 5 min after the start of the indicator reaction, b is the slope of the calibration graph, t is Student's t, n is the number of parallel runs and s is the standard deviation of A5 Log (cCu/mg ml21) 22 24 25 26 Blank RSD for A5 (%) 0.006 0.008 0.012 0.020 0.026 ±D(log cCu) 0.20 0.26 0.33 0.52 --

Table 4 Tolerance ratios for foreign ions (cion+cCu) in the determination of CuII by use of the hydroquinone­CuII­H2O2­2,2A-dipyridyl catalytic reaction in solution and on HMDA filters Tolerance ratio Reaction in solution (0.01 mg l21 Cu) 10 50 100 100 300 500 500 1000 1500 30 100 350 1000 Reaction on HMDA filters (0.1 ng Cu; preconcentration from 10 ml) 105 106 106 105 1000 106 106 106 1.5 3 10 105 5 3 105 2 3 106 2 3 106

Foreign ion Hg2+ Ni2+ Fe2+, Fe3+, Co Pb2+ Mn2+ Mg2+, Al3+ Ca2+ Na+, K+, Cd2+ Cl2 I2 CH3COO2 SO422 NO32

2+

6

Fig. 3 Dependence of the rate of the hydroquinone (HQ)­CuII­H2O2­ 2,2A-dipyridyl (dipy) reaction carried out by pipetting copper (1, 1A) and without copper (2, 2A) on HMDA filters (1, 2) and filter-paper (1A, 2A) on pH. Amounts of the reagents: 6 3 1027 mol HQ, 9 3 1026 mol H2O2, 5 3 1028 mol dipy; 0.1 mg Cu.

Table 2 Equations for the calibration graphsa and characteristics for the determination of CuII by use of the hydroquinone­CuII­H2O2­2,2A-dipyridyl catalytic reaction in solution and on filter-papers with attached HMDA groups (preconcentration of CuII from 10 ml of solution) Procedure for copper determination Linear range/ mg l21

a

sa

b

sb

r

LOD/mg l

21

In solution 0.06 0.018 4.1 0.44 0.97 0.01­0.07 5.0 3 1023 Sorption­catalytic 8.5 3 1022 9.5 3 1024 8.5 3 1023 2.6 3 1024 0.998 1 3 1026­0.1 1 3 1027 a For the equation y = a + bx, where x = 2log c Cu and y = A5 (absorbance measured 5 min after the start of the reaction; blank absorbance was not subtracted).

1526

Analyst, 1999, 124, 1523­1527


Table 5 Comparison of the results obtained by the proposed sorption­ catalytic and reference methods (n = 3) Copper found/mg l Sample Sea-water Water No. 1: Before treatment After treatment Water No. 2: Before treatment After treatment Tap water (after purification by a household filter) Sorption­catalytic method (4.5 ± 1.8) 3 10 0.16 ± 0.06 0.12 ± 0.05 0.41 ± 0.12 0.12 ± 0.05 (8 ± 4) 3 10
25 22 21

thank Dr. G. I. Tsysin for providing the chelating sorbents and for discussions and Surgutneftegaz Corporation and the Sanitary­Epidemiological Station of Surgut for the AAS measurements.

AAS (5.09 ± 0.03) 3 10 0.18 ± 0.01 0.13 ± 0.02 0.42 ± 0.01 0.13 ± 0.01 (9.0 ± 0.5) 3 10
25 22

References
1 H. MÝller, CRC Crit. Rev. Anal. Chem., 1982, 13, 313. 2 I. Ya. Kolotyrkina, L. K. Shpigun, Yu. A. Zolotov and G. I. Tsysin, Analyst, 1991, 116, 707. 3 M. K. Beklemishev, T. A. Stoyan and I. F. Dolmanova, Analyst, 1997, 122, 1161. 4 G. I. Tsysin, A. A. Formanovskii and I. V. Mikhura, USSR Pat., 1702659, 1989. 5 G. I. Tsysin, I. V. Mikhura, A. A. Formanovskii and Yu. A. Zolotov, Mikrochim. Acta, 1991, III, 53. 6 G. M. Varshal, T. K. Velyukhanova, V. I. Pavlutskaya, N. P. Starshinova, A. A. Formanovsky, I. F. Seregina, A. M. Shilnikov, G. I. Tsysin and Yu. A. Zolotov, Int. J. Environ. Anal. Chem., 1994, 57, 107. 7 G. I. Tsysin, G. I. Malofeeva, O. M. Petrukhin, G. A. Evtikova, V. P. Sokolov, I. N. Marov and Yu. A. Zolotov, Zh. Neorg. Khim., 1988, 33, 2617. 8 G. I. Tsysin, I. V. Mikhura, A. A. Formanovsky, G. A. Evtikova, V. P. Sokolov and I. N. Marov, Zh. Neorg. Khim., 1990, 35, 960. 9 I. F. Dolmanova, V. P. Poddubienko and V. M. Peshkova, Zh. Anal. Khim., 1973, 28, 592. 10 I. F. Dolmanova, O. I. Melnikova, G. I. Tsysin and T. N. Shekhovtsova, Zh. Anal. Khim., 1980, 35, 728. 11 S.-A. Cao, J.-C. Zhong, K. Hasebe and W.-Z. Hu, Anal. Chim. Acta, 1996, 331, 257. 12 I. N. Marov, V. K. Belyaeva, E. B. Smirnova and I. F. Dolmanova, Inorg. Chem., 1978, 17, 1667.

lowers the detection limit relative to the reaction in solution (in our case by a factor of 104), which is associated with the different properties of the calibration curves in solution and on the sorbent. We believe that the sorption­catalytic principle may be regarded as a general approach to the determination of metal ions, as it can be easily developed for numerous indicator reactions on various sorbents.

Acknowledgements
The research was supported by the Russian Foundation for Basic Research (grants 95-03-08854 and 97-03-33578a) and by the Program `The Universities of Russia­Basic Research'. We

Paper 9/03900G

Analyst, 1999, 124, 1523­1527

1527