Journal of Analytical Chemistry, Vol. 60, No. 3, 2005, pp. 247–251. Translated from Zhurnal Analiticheskoi Khimii, Vol. 60, No. 3, 2005, pp. 278–283. Original Russian Text Copyright © 2005 by Antonova, Vershinin, Dedkov.

ARTICLES

Use of Triphenylmethane Dyes for the Spectrophotometric Determination of Polymer Flocculants in Aqueous Solutions T. V. Antonova*, V. I. Vershinin*, and Yu. M. Dedkov** * Omsk State University, pr. Mira 55-A, Omsk, 644077 Russia ** Moscow State Regional University, ul. Radio 10a, Moscow, 105007 Russia Received May 6, 2004; in final form, July 15, 2004

Abstract—The interaction of anionic forms of triphenylmethane dyes (TPM) with cationic polymer flocculants (e.g., with polydiallyldimethylammonium chloride (PC)) yields inextricable ion pairs. This leads to a shift of absorption bands or a redistribution of their intensities in the TPM spectra and can be used for the spectrophotometric determination of microgram amounts of flocculants in aqueous solutions. Among 14 studied TPMs, erythrosine is the most promising dye. The corresponding procedure allows the determination of PC at a level of the maximum permissible concentration. To decrease the effect of impurities, the use of the standard addition method is recommended.

Water preparation and other industrial processes employ polymer flocculants, derivatives of quaternary ammonium salts, e.g., polydiallyldimethylammonium chloride (PC). The toxicity of PC requires the control of its residual concentration in tap water. As demonstrated in review [1], the problem of the rapid and reliable determination of polymer flocculants in waters is yet to be solved. A spectrophotometric procedure based on the interaction of PC with one of triphenylmethane dyes (TPM), namely, eosine is commonly used for this purpose. Previously the stoichiometry of the interaction of eosine with PC was determined and the conditions of the formation of the colored product (ion pair) were optimized [2, 3]. Unfortunately, even under optimum conditions, the sensitivity of the reaction with eosine is insufficient for the determination of PC at a level of the maximum permissible concentration (MPC). Systematic studies of the interaction of other TPM with polymer flocculants could lead to the development of more sensitive procedures for the determination of polymer flocculants, with was the aim of this work. For the study, we selected the following TPM: Pyrocatechol Violet (PCV), Bromophenol Blue (BPB), Bromocresol Green (BCG), Bromothymol Blue (BTB), Phenol Red (PR), Thymol Blue (TB), aluminon (AL), Bromocresol Purple (BCP), Xylenol Orange (XO), fluoresceine (FL), fluorexon (FLX), erythrosine (ER), Bengal Red (BR), and eosine (EO). The reagents were selected for the following reasons: The above reagents are well studied, are used as indicators, and are available in the pure form. Their dissolution in water (<10−3 M) yields true solutions stable on storage and intensely absorbing in the visible spectral region (logε = 3.5–5.0). All of the reagents have functional groups of the donor character and are potential ligands. Anionic forms of TPM interact with cationic forms of

monomeric quaternary ammonium salts yielding ion pairs [4]; hence, changes in absorption spectra are observed. It is expected that TPM will analogously react with polymer quaternary ammonium salts, e.g., with PC. However, the formation of ion pairs of polymer flocculants with dyes must have its specific features, which should be taken into account in the course of analysis. EXPERIMENTAL Stock 1 × 10 M solutions of dyes were prepared by dissolving weighed portions of chemically pure reagents. Stock solutions of PC (25 mg/L) were prepared in the day of use from a weighed portion of the VPK-402 flocculant (TU 6-05-2009-86; AO Kaustik, Sterlitamak, Russia) containing 39.66% PC and inert impurities. Conventional buffer solutions were used to adjust pH 4–10, and 2 M HCl was added to adjust pH < 4. The absorption spectra of mixtures of TPM and PC were recorded in the range 380–650 nm with a step of 10 nm on SF-26 and KFK-3 spectrophotometers in glass cells (3.0 cm) with reference to a blank solution with the same concentration of TPM and the same pH. PC alone does not absorb in the visible spectral region. All experiments were performed at 20–22°ë without thermostatting. Experiments were repeated 3–5 times. The statistical processing of the results was performed by conventional methods assuming the normal error distribution. The significance of changes in the absorbance of TPM in the presence of PC (hereafter ∆A) was checked by the Student’s test. The stoichiometry of the interaction of PC with dyes was determined by the Asmus and Bent–French methods [5]. The effect of ionic strength was checked by the addition of potassium salts (nitrates, sulfates, and chlorides). The extractability of products of the interaction between PC –3

1061-9348/05/6003-0247 © 2005 Pleiades Publishing, Inc.

248

ANTONOVA et al.

A 2

A 0.5

2 3

0.3

2

1

1 1

0.1 3

0 400

500

600

700 λ, nm

–0.1 400

500

600

700 λ, nm

Fig. 1. Absorption spectra of a 4.0 × 10–5 M solution of Bengal Red at pH 4.2; (1) without PC, (2) in the presence of 20 mg/L PC, and (3) ∆A; water was used as the reference solution.

Fig. 2. Absorption spectra of a 5.0 × 10–5 M solution of Bromocresol Green at pH 5.0; (1) without PC, (2) in the presence of 20 mg/L PC, and (3) ∆A; water was used as the reference solution.

and TPM was checked using tetrachloromethane and chloroform as extractants; the recovery was determined by the spectrophotometric method.

value of ∆A is the analytical signal of PC and the optimization parameter; it depends on the pH of the solution and on the ratio between the concentrations of PC and TPM. The pH dependence of ∆A was studied at constant concentrations of the reagent (cR = 4 × 10–5 M) and PC (cï = 1 mg/L) at the wavelengths that provide the largest value of ∆A in each case. For xanthene dyes, these wavelengths were 20–30 nm larger than λmax for the anionic form of the corresponding reagent. For the other TPM, the analytical wavelengths approximately coincided with λmax for the anionic forms. The pH dependence of ∆A for all reagents has the form of a curve with a maximum (Fig. 3), and the optimum values of pH (hereafter çopt) are different for different reagents. Xanthene dyes, for which this maximum is observed in weakly acidic solutions, in alkaline solutions exhibit one more weakly pronounced maximum. For explaining the revealed dependences, the state of TPM reagents in the solution should be taken into consideration. A suitable example is erythrosine, which occurs in aqueous solutions in three forms. According to the values of ionization constants [6], the molecular form H2R dominates in the pH range 0.5–4.0 and has λmax = 500 nm and ε = 1.14 × 104, the anionic form HR– dominates at pH 4.0–5.3 (λmax = 535 nm, ε = 2.3 × 104), and the second anionic form R2– dominates at pH 5.3– 12.0 (λmax = 525 nm, ε = 9.46 × 104). The maximum value of ∆A in the presence of PC was observed at pH 4.2, i.e., in the zone of the transition from H2R to HR–. For the other dyes, the maximum value of ∆A was observed in the same zone (at the dominance of çR–), and the second maximum appeared in the zone of the transition from çR– to R2–. Under the conditions when the molecular or the mostly deprotonated form dominates, the values of ∆A

RESULTS AND DISCUSSION Changes in the absorption spectra of reagents in the presence of PC. Preliminary experiments demonstrate that in the presence of PC (up to 1 mg/L) the absorbance of TPM solutions in the visible region is reliably changes only for FL, FLX, ER, BR, PCV, BTB, BPB, BCG, and EO. The interaction of the other five TPM with PC at the same concentration level does not lead to statistically reliable changes in absorption spectra; irrespective of the pH of the solution, ∆A ≈ 0. Subsequent studies were performed with only the above nine reagents. The character of changes in broad-band spectra of TPM on the addition of PC depends on the structure of the dye. Namely, eosine and its closest structural analogues, xanthene dyes (FL, FLX, ER, and BR) in the presence of PC exhibit a small (5–30 nm) bathochromic shift and an increase in the intensity of absorption bands (Fig. 1). In the absorption spectra of other triphenylmethane dyes (PCV, BPB, BTB, and BCG), the positions of both absorption bands commonly remain unchanged, but the redistribution of band intensities occurs (Fig. 2). Namely, in the presence of PC, the absorbance reliably increases (∆A > 0) in the more long-wave region of the spectrum and decreases (∆A < 0) in the relatively short-wave region. Changes in the spectra of the reagents of the second group in the presence of PC are analogous to the changes that would be observed without PC on some increase in pH (which was not actually observed). Optimization of the conditions of the interaction of PC with dyes. For the reagents of both groups, the

JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 60

No. 3

2005

USE OF TRIPHENYLMETHANE DYES

were nearly zero. Thus, the absorption spectra of BCG in the presence of PC exactly coincide with the spectra of pure BCG at pH 2, when H2R dominates, and at pH 10, when R2– dominated.

∆A 0.8

It is known that PC and other polymer quaternary ammonium salts in dilute aqueous solutions occur in the cationic form irrespective of pH [7]. Obviously, both the shift of bands and the redistribution of their intensities in the absorption spectra of TPM are due to the formation of ion pairs of the PC polycation and the anionic form of the reagent. A similar interpretation was proposed previously for changes in the spectra of TPM in the presence of low-molecular cationic surfactants [8]. Evidently, the formation of the ion pair of TPM with the polycation enhances the acidity of the dye (decreases K‡). Simultaneously, the optical properties of the anionic form of the TPM molecule are changed, which must lead to a change in the absorbance at pH ≈ pK‡, i.e., in the zone of the color change of the corresponding indicator. Evidently, the second mechanism is more typical for xanthene dyes, and the first mechanism is more typical for the other TPM.

0.6

At the constant concentration of PC and the optimum pH, the value of ∆A first increases with the concentration of the reagent, passes through a maximum at approximately 100-fold molar excess of the reagent (exact molar concentration of the PC polymer cannot be determined; however, it can be approximately estimated from the known average molar mass (3 × 105) and the concentration of PC in mg/L), and significantly decreases with the further increase in the excess of the reagent. The optimum concentrations of different TPM reagents are rather close to each other. At cï = 1 mg/L, the optimum concentrations are 1 × 10–5–5 × 10–5 M (Table 1). The study of the stoichiometry of the reaction demonstrated that similar products with the molar ratio 1 : 1 are formed in all of the studied systems, i.e., as in the case of eosine, there is one TPM anion for one elementary group of the polycation. It is expected that, at extremely high concentrations of the reagent, the product of another composition (with lower molar coefficient) is formed or another type of interaction, e.g., adsorption is observed.

1

2

0.22

0

2

4

6

8

10

12

14 pH

Fig. 3. Effect of pH on the value of ∆A for some TPM reagents; (1) BPB, (2) BTB, and (3) PCV; in all cases cR = 4 × 10–5 M and cï = 1 mg/L; wavelengths are given in Table 1.

For the stabilization of neutral ion pairs, a watermiscible polar solvent (ethanol or acetone) is sometimes added to the solution for photometry. However, in the case of PC, this addition produced an unexpected effect: a decrease in the analytical signal rather than an increase was observed. At low concentrations of ethanol or acetone (up to 5%), the values of ∆A decreased with increasing concentration of the nonaqueous solvent, and at 10–30% the values of ∆A were even undoubtedly negative, although small in absolute value and poorly reproducible. This effect can be explained by the following reasons. In aqueous–alcohol solutions, as distinct from aqueous solutions, the uncoiled conformation of PC dominates [10]. Under these conditions, Table 1. Optimum conditions for the determination of PC (1 mg/L) with the use of different triphenylmethane dyes

Effect of the solvent and the ionic strength of the solution. The extractability of TPM with nonpolar solvents like CH3Cl does not increase in the presence of PC as was observed in the case of the formation of neutral ion pairs [9]. Evidently, only some of the quaternary ammonium groups in the PC macromolecule interact with TPM anions. In this case, the resulting ion pairs of PC and TPM must be charged, which decreases their solubility in nonpolar organic solvents. Vol. 60

3

0.4

Under the optimized conditions, the value of ∆A increased with cï for all of the studied reagents. The obtained relationships can serve as calibration plots for the determination of PC (Table 2).

JOURNAL OF ANALYTICAL CHEMISTRY

249

No. 3

λmax, nm

pHopt

cR × 105 M

FLX

540

6.0–7.1

4.0

ER

550

2.8–5.0

4.0

BR

565

2.8–5.0

8.0

FL

505

5.2–6.8

1.0

EO

550

2.8–4.5

1.0

PCV

590

9.5–10.5

4.0

BPB

630

4.2–5.7

5.0

BCG

570

5.3–6.9

4.0

BTB

630

6.0–7.5

4.0

Reagent

2005

250

ANTONOVA et al.

Table 2. Characteristics of the spectrophotometric determination of PC with triphenylmethane dyes Equation of the calibration plot

Linearity range, mg/L

PCV

∆A = 0.71cx

0.10–1.00

ER

∆A = 0.64cx

0.05–1.00

BR

∆A = 0.27cx

0.10–4.00

BTB

∆A = 0.21cx

0.30–2.00

Reagent

the number of open quaternary ammonium groups increases. In addition, the introduction of an organic solvent decreases the dielectric constant of the solution, and the degree of dissociation of functional groups of the polyelectrolyte decreases. The cooperative action of both factors leads to the change in the character of the process: the polyelectrolyte begins to bind dye anions by the adsorption mechanism like on a large increases in the reagent excess rather than by the formation of ion pairs. Of course, this explanation is hypothetical, and the mechanism of the interaction of dyes with flocculants in nonaqueous solutions needs further investigation. However, whatever this mechanism may be, the presence of organic solvents like ethanol or acetone in the determination of polymer flocculants is undesirable, because it leads to a decrease in the analytical signal and impairs the reproducibility of the results of analysis. The effect of ionic strength was studied with the examples of PCV, BPB, BCG, and BTB. The introduction of small amounts of foreign electrolytes (I < 0.1) does not change the value of ∆A, and the further increase in ionic strength (0.1 < I < 1.0) decreases ∆A irrespective of the nature of the anion. Evidently, ion pairs of dyes with polymer flocculants have low stability, which is in agreement with the necessity of adding excess dye. The intensely colored and highly labile ion pair formed by Pyrocatechol Violet is especially sensitive to the presence of foreign electrolytes. Evidently, the presence of strong electrolytes in test solutions can lead to systematic errors in the determination of flocculants; therefore, the analysis should be performed at constant ionic strength. Sensitivity of the reaction and the selection of the reagent. All of the studied TPM reagents can be used for the photometric determination of trace PC. The analysis of model solutions was performed under the optimum conditions for each reagent. For example, when erythrosine was used, 20.0 mL of a test solution of PC or analyzed water, 20.0 mL of a 2 × 10–4 M solution of erythrosine, and 5 mL of a buffer solution with pH 4.25 and a high ionic strength were placed in a 50-mL volumetric flask, and the mixture was diluted with distilled water to the mark and stirred. The absorbance was measured immediately after mixing the reac-

tants in a cell with a layer thickness of 50 mm at 550 nm. The concentration of PC was determined by the equation of the calibration plot obtained previously (Table 2). In the region of low concentrations (0.1– 1.0 mg/mL), the value of ∆A is directly proportional to the concentration of PC for all of the reagents; linear correlation coefficients are above 0.9. The further increase in cï leads to a bend of the plot, and next ∆A goes to a plateau or even decreases (BPB). For xanthene dyes, the linear character of the relationship is observed in a wider interval (up to 10.0 mg/L). The sensitivities of procedures for the determination of PC with different TPM were compared by the value of the slope coefficient (k) in the equation ∆A = kc (µg/mL). The maximum sensitivity is attained with Pyrocatechol Violet (k = 0.71); however, the determination of PC with PCV exhibits the lowest accuracy. The sensitivity of the determination of PC with erythrosine is somewhat lower (k = 0.64). The other dyes exhibit even lower values of k, in particular, for eosine k = 0.40. Bromocresol Green exhibits the lowest sensitivity (k = 0.07). The relative standard deviations (RSD) for all of the reagents are rather close to each other and are no larger than 3%. The detection limit of PC with erythrosine calculated by the Kaiser method was lower than with eosine, which is commonly used for this purpose, and substantially lower than that set up as MPC in the Russian Federation, i.e., this reagent can be used for the determination of PC in potable water at a level of MPC. The only advantage of eosine is the lower value of absorbance in the blank experiment. The further decrease in the detection limit requires the change to basically different processes (e.g., the formation of ion pairs involving transition metal ions) or a different technique for measuring the analytical signal (spectrofluorimetry) [1]. The accuracy of the corresponding procedures was verified by the added–found method. In the determination of concentrations by a previously constructed calibration plot, commonly the systematic error of the determination (δ, %) is statistically significant and negative and its absolute value increases with the dilution of the solution (Table 3). At cï < MPC (0.1 mg/L), it begins to exceed 10%. Note that rather low accuracy is characteristic of all spectrophotometric procedures involving the formation of ion pairs; however, this trend is particularly pronounced in the determination of trace polymer flocculants. Most likely, the partial uncoiling of the globular PC macromolecule to the linear form leads to a change in the number of TPM molecules in the ion pair and, consequently, to a change in its optical properties. The completeness of this process and its rate can be affected by many factors including impurities in the analyzed water. Because the labile flocculant molecule can be hardly stabilized, it is desirable to use the standard addition method instead of the calibration plot for eliminating the effect of impurities and improving

JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 60

No. 3

2005

USE OF TRIPHENYLMETHANE DYES Table 3. Performance characteristics of procedures for the determination of PC in model solutions by the calibration plot (n = 3, P = 0.95) Reagent

Added, mg/L

Found, mg/L

0.08 0.15 0.30 0.08 0.15 0.50 0.30 0.50 0.30 0.50 0.40 0.80

0.06 ± 0.01 0.14 ± 0.01 0.29 ± 0.01 0.06 ± 0.01 0.14 ± 0.01 0.48 ± 0.01 0.28 ± 0.01 0.48 ± 0.01 0.28 ± 0.01 0.49 ± 0.01 0.33 ± 0.07 0.74 ± 0.02

EO

ER

BR BTB PCV

|δ|, % RSD, % 25 6.0 3.0 25 6.7 4.0 6.6 4.0 6.6 2.0 17.5 6.2

3 1 1 3 2 1 1 1 1 1 7 1

Found, mg/L

|δ|, %

RSD, %

0.10* 0.20* 0.10** 0.20**

0.09 ± 0.01 0.21 ± 0.01 0.10 ± 0.03 0.24 ± 0.07

11.1 4.8 0.0 17.8

4 2 14 11

1. Antonova, T.V., Vershinin, V.I., and Dedkov, Yu.M., Zavod. Lab., 2004, vol. 74, no. 1, p. 3. 2. Antonova, T.V., Achkasova, E.Yu., Baranova, S.V., and Dedkov, Yu.M., Vestn. Omsk. Univ., 2003, no. 1, p. 30. 3. Klyachko, Yu.A., Shnaider, M.A., Korshunova, M.L., et al., Izv. Vyssh. Uchebn. Zaved., Pishch. Tekhnol., 1984, no. 4, p. 95. 4. Savvin, S.B., Marov, I.N., Chernova, R.K., et al., Zh. Anal. Khim., 1981, vol. 36, no. 5, p. 850.

the accuracy of the analysis (Table 4). Both the computational version of this method (for one addition) and the graphical version with several additions of different amounts can be used.

Vol. 60

5. Bulatov, M.I. and Kalinkin, I.P., Prakticheskoe rukovodstvo po fotokolorimetricheskim i spektrofotometricheskim metodam analiza (A Manual of Photocolorimetric and Spectrophotometric Methods of Analysis), Leningrad: Khimiya, 1976. 6. Mchedlov-Petrosyan, N.O., Zh. Org. Khim., 1983, vol. 19, no. 4, p. 797. 7. Veitser, Yu.I. and Mints, D.M., Vysokomolekulyarnye flokulyanty v protsessakh ochistki prirodnykh i stochnykh vod (High-Molecular Flocculants in the Purification of Natural and Waste Waters), Moscow: Stroiizdat, 1984.

Note: Analysis was performed by the standard addition method (* graphical version, ** computational version).

JOURNAL OF ANALYTICAL CHEMISTRY

The selectivity of the spectrophotometric determination of cationic flocculants and the applicability of the corresponding procedures to the analysis of waters of different types will be considered elsewhere. REFERENCES

Table 4. Verification of the procedure for the determination of PC in tap water with erythrosine by the added–found method (n = 3, P = 0.95) Added, mg/L

251

8. Savvin, S.B., Chernova, R.A., Lobacheva, I.V., and Beloliptseva, G.M., Zh. Anal. Khim., 1981, vol. 36, no. 8, p. 1471. 9. Volkova, A.I., Shevchenko, T.L., and Pshinko, G.N., Khim. Tekhnol. Vody, 1994, vol. 16, no. 4, p. 372. 10. Shnaider, M.A., Kamenskaya, E.V., Korshunova, M.A., et al., Izv. Vyssh. Uchebn. Zaved., Pishch. Tekhnol., 1983, no. 6, p. 32.

No. 3

2005