A new spectrophotometric method for determination of residual polydiallyldimethylammonium chloride flocculant in treated water based on a diazotization-coupled ion pair

Polydiallyldimethylammonium chloride (polyDADMAC) is a water-soluble cationic polyelectrolyte used for water treatment. Its residues in treated water are contaminants as they react with chlorine to produce a carcinogenic compound. Commonly-used techniques for quantification of the polycation, such as colloidal and potentiometric titration as well as 1H NMR, have poor sensitivity and detection limits. This paper describes a more sensitive UV spectrophotometric method for quantitative determination of residual polyDADMAC in treated water, through formation of an ion pair with (4-hydroxy1-napthylazo) benzene-sulphonic acid (dye). The ion pair, which is a colloidal solid material, was characterised by FTIR, 13C NMR and 1H NMR techniques. The colloid materials formed with different concentrations of polyDADMAC were dissolved in either N,N-dimethylformamide or 1,4-dioxane, followed by determination with UV spectrophotometry. The wavelength of maximum absorption (λmax) was found to be dependent on the solvent used, with 1,4-dioxane showing a better linear range of 0.1–1.8 mg∙l-1 polyDADMAC. Varying the pH of the solutions had no significant effect on λmax.


Introduction
Domestic water is normally abstracted from rivers, dams and groundwater, in its raw state, and contains a substantial amount of suspended matter.It therefore must undergo treatment that will render it suitable for human consumption.The initial stage (clarification) involves the use of coagulants and flocculants which accelerate the settling of suspended particles.Organic cationic polyelectrolytes for the water treatment process are currently preferred over inorganic flocculants due to advantages in clarification of suspended particles.Examples of the most frequently applied flocculants are polydiallylydimethyl ammonium chloride (polyDADMAC) and epichlorohydrindimethylamine (epi-DMA), which are used in 75% of waterworks in South Africa (Majam et. al., 2004).The advantages of organic cationic polyelectrolytes over inorganic coagulants include an increase in the solid-liquid separation rate, a smaller sludge volume, non pH-dependence and low levels of dissolved salts in the final treated water.These quaternary ammonium polyelectrolytes have been found to solve the problem of slowsettling flocs in low-temperature coagulation and in the treatment of soft coloured waters, by improving settling of particles (Faust, 1983).The mechanism of water treatment is by inducing aggregation of molecules with opposite charges to form a new combination product containing colloidal particles (Mocchiutti and Zanuttini, 2007).
Quaternary ammonium cations are permanently charged (Xia et. al., 2007), and when they interact with the environ-, and when they interact with the environment while in active form (charged) they continue to form free primary radicals which are difficult to terminate (Assem et. al., 2007).Such free radicals initiate reactions that can form undesirable compounds.For instance, it has been realised from recent studies that organic cations react with chlorine to form N-nitrosodimethylamine (NDMA) during the chlorine water-disinfection process (Park et. al., 2009).NDMA is a human carcinogen which has been identified as a by-product of disinfection of water containing trace levels of polyDADMAC (Park et. al., 2009).There is therefore a need to determine the concentration of the polyelectrolyte cations that pass through the flocculation stage, before the chlorination process, with a view to ascertaining the safety of water to consumers.In addition, it is necessary to routinely monitor residual levels of polyDADMAC in the final treated water so as to advise on its removal.
It should be noted that polyDADMAC, like any other polyelectrolyte, when dissolved in water dissociates to form simple units of varying monomeric sizes (Gennes, 1979).The dissociated repeating units bear an electrolyte group, making the polymers charged, and, hence, vulnerable to further reaction.These properties of polyelectrolytes are thus similar to those of ionic salts, which gives rise to their being termed polysalts (Gennes, 1979).
Available methods for determining the concentrations of polyelectrolytes are unsatisfactory because polyelectrolytes have high detection limits and are therefore not suitable for trace-level analysis.For instance a detection limit of 10-40 mg•ℓ -1 was reported by Becker et. al. (2004), who estimated the level of polyDADMAC in water by fluorescent tagging.(Wanga et. al., 1989).These methods have severe limitations in terms of achieving good linearity, accuracy, precision, reproducibility and detection limits (Sakai., 2001).The other challenge is that the cationic poly- . The other challenge is that the cationic polymer is UV inactive and it is therefore not possible to employ UV-Vis spectrophotometry for its analysis (John et. al., 2002).However, a study by Parazak et. al. (1987) observed that complexation of the polyelectrolyte with an anionic dye forms an insoluble complex.Studies have been undertaken to determine the correlation between charge density and antimicrobial efficacy for quantification of quaternary ammonium salts attached on a surface, by Tiller et. al. (2001), Al-Shihry (2005), Kügler et. al. (2005) and Zander et. al. (2008), which showed that dyes bind to quaternary ammonium at a molar ratio of 1:1.
In view of the challenges faced in the determination of polyDADMAC levels in water, the study reported herein exploited the solubility property of the polyDADMAC-dye ion pair formed when polyDADMAC is reacted with (4-hydroxy-1-naphylazo) benzene-sulphonic acid dye.The ion pair (complex) is a coloured colloid which is soluble in selected non-aqueous solvents.The quantity of the ion pair in the nonaqueous solution was then determined by UV spectrophotometry.This approach was used to develop an efficient new method for the determination of trace levels of polyDADMAC residuals in treated water.

Chemicals and reagents
All of the solutions were prepared in double-distilled water.The reagents used were of analytical grade.Sodium nitrite, sodium hydroxide and sodium bicarbonate were supplied by Associated Chemical Supplies Ltd, Fluka Analytical (Buchs, Switzerland).Sulphanilic acid, N,N-dimethylformamide (DMF), 1,4-dioxane and 1-naphthol were supplied by BDH (London, UK) while hydrochloric acid and deuterated water were supplied by Sigma Aldrich (Johannesburg, South Africa).The polyDADMAC was supplied by Umgeni Water Treatment Works (Durban, South Africa), and had a density of 1.09 g•mℓ -1 and 35% w/v solution in water (diallyldimethyl ammonium chloride, DADMAC monomer has a molecular mass 161.66 g•mol -1 ).

Instrumentation
Fourier transform infrared (FT-IR) spectrophotometer, Perkin Elmer 100 (USA) with sampling accessory, was used to characterise the polyDADMAC-dye ion pair (Mukamel, 2000). 1 H NMR spectrometry was performed with a 400 MHz Bruker Avance spectrometer (Germany) using deuterium oxide as the lock solvent.Solid samples of the ion pair were characterised by solid-state carbon-13 nuclear magnetic resonance, 13 C NMR, using a Bruker 600 spectrometer with cross polarisation/magic angle spinning (CP/MAS).The 13 C NMR experiments were performed at a frequency of 10 MHz while cross-polarisation was calibrated with glycine at 10 KHz and spinning rate was set at 5 KHz.The contact time for CP/MAS was 1 ms and the delay time for acquisition was 5 s.Optical absorbance measurements were carried out on a Perkin Elmer LAMBDA 35 UV-Vis spectrophotometer (USA).The analysis was carried out using a wavelength range between 300 and 200 nm.

Diazotization of sulphanilic acid
A 6 mM solution of diazosulphanilic acid was prepared by dissolving 0.116 g of sulphanilic acid and 0.035 g sodium nitrite in approximately 500 mℓ of water.The solution was cooled to 0 o C on ice, and 2 mℓ of 2 M HC1 was added with vigorous stirring.The presence of excess nitrous acid was tested using a starch-iodide paper and the excess nitrous acid mopped up with a few drops of sulphanilic acid solution (Schoutissen, 1933).

Formation of polyDADMAC-dye ion pair
Composite diazotization-coupling reagent containing sulphanilic acid as the diazotisable aromatic amine and 1-naphthol as the coupling agent was used for the spectrophotometric determination of trace amounts of polyDADMAC.The coupled azo dye (5 mℓ) was reacted with several portions (1 ℓ) of varying concentrations (0.1-25 mg•ℓ -1 ) of polyDAD-MAC and the mixture was allowed to equilibrate for 2 h resulting in the formation of a colloid (Horita et. al., 1997).

Characterisation experiments
The polyDADMAC-dye ion pair colloid was filtered through a sintered glass crucible (number 4, porosity specification) and the resulting solid mass dissolved in either 10 mℓ of dimethylformamide (DMF) or 1,4-dioxane, separately (Horita et. al., 1997).The solutions were diluted to 1 ℓ and then analysed using a UV-Vis spectrophotometer.Each respective blank reference solution for this analysis was prepared by dissolving 10 mℓ of either DMF or 1,4-dioxane and diluting the resultant solution to 1 ℓ with water, similarly to the samples.
Samples for FTIR and 13 CNMR analyses were prepared by reacting solutions of known polyDADMAC concentration with the coupled dye.The resulting mixture was filtered through a sintered glass crucible.The residual solid colloidal material was dried at 60 o C and cooled in a desiccator before analysis.Samples for characterisation experiments with 1 HNMR were prepared by measuring 9 µℓ of polyDADMAC (1.09 g•ℓ -1 , 35% solution in water) into a vial and diluting this with 1 mℓ of deuterated water.Solutions for characterisation with UV-Vis spectrophotometry were prepared by reacting known volumes of dye solutions with known quantities of polyDADMAC and allowing the mixtures enough time to form colloids.The products were filtered through a sintered glass crucible and the resulting solid dissolved in either dimethylformamide or 1,4-dioxane.

Spectrophotometric determination of trace amounts of polyDADMAC
Experiments to quantify polyDADMAC by 1 H NMR technique were done by preparing the polyelectrolyte solution in deuterated water.Approximately 10 mg of polyDADMAC was placed in a vial and 1 mℓ of deuterated water was added.A series of dilutions with deuterated water were performed to obtain varying concentrations of the analyte.These solutions were transferred to clean dry NMR sample tubes and filled to a 5-cm length of the tube.The sample tubes were then introduced into the 400 MHz Bruker Avance spectrometer for analysis.
Quantification of polyDADMAC by UV-Vis spectrophotometric technique was done by adding 5 mℓ of coupled azo dye to 1 ℓ of varying concentrations (0.1-25 mg•ℓ -1 ) of polyDAD-MAC solution to form a colloidal material.The colloid was filtered through a sintered glass crucible (number 4, porosity specification) and the resulting solid mass dissolved in 10 mℓ of either dimethylformamide (DMF) or 1,4-dioxane.The resulting mixture was quantified on a UV-Vis spectrophotometer.Calibration curves for solutions prepared in both 1,4-dioxane and DMF were obtained for the determination of the content of the polyelectrolyte in water.
Water samples (4 samples of 3 ℓ each) were collected in polyethylene bottles on different dates (August 2010) from the Umgeni River in Durban.The samples were filtered through a sintered glass crucible and treated with polyDADMAC, immediately after collection.A similar procedure (as used for standards) was performed for the treatment of water samples where a volume of 5 mℓ of azo dye was reacted with 1.0 ℓ of the sample water.This was followed by filtration through the sintered glass crucible after equilibration, so as to obtain the solid ion pair for UV spectrophotometric analysis.
The ion pair formed was a dark red coloured solution due to the presence of azo dye.Several experiments were carried out to investigate the properties of the ion pair and the results are given in the respective sections.The ion pair has a molecular weight of 511.99 g•mol -1 (Horita et.al., 1997).

Characterisation of the polyDADMAC-dye ion pair
The ion pair was characterised by UV-Vis, FTIR, 13 C NMR and 1 H NMR techniques.These methods provide information about the composition and functional groups, as well as the structure of the ion pair.

FTIR characterisation
The FTIR-characterisation of polyDADMAC alone, and of the polyDADMAC-dye ion pair, resulted in the spectra shown in Fig. 1.
The spectra in Fig. 1 show the presence of functional groups found in both the dye and polyDADMAC.A band at 3 355.76cm -1 was observed which was assigned to aliphatic secondary amines (Coates, 2000).The band at 1 135.29 cm -1 was assigned to the C-N stretch of a secondary amine (Yang et. al., 1996;Khazaei et. al., 2007 ), whereas the band at 1 215.87 cm -1 was assigned to the organic sulphates.The organic sulphate band was absent in the spectrum of polyDADMAC but appears in the ion-pair spectrum due to the presence of the dye (Khazaei et. al., 2007 ).The band at 2 938.55 cm -1 was assigned to C-H of the aromatic ring from the dye and the band at 1 631.11cm -1 is characteristic of C-H in the conjugation of an aromatic ring.The latter confirms the presence of a delocalised π electron system of an aromatic ring (Khazaei et. al., 2007 ).The signal at 1 635.24cm -1 in (B) was also assigned to C-H stretch in the polyelectrolyte (Khazaei et. al., 2007).
The observed shoulder at 1 631.11cm -1 in Fig. 1(A), arises from  Coates (2000).The band at 1 473.56 cm -1 , appearing on both spectra, indicates a long carbon chain with a high degree of regularity for the linear backbone structure (Xamena et. al., 2004).This is pos-.This is possibly contributed by the polyDADMAC polyelectrolyte.It was observed that the signal intensity was enhanced upon the formation of the ion pair as compared to the polyelectrolyte alone, shown in Fig. 1(B), confirming formation of an ion pair.The increased signal intensity (Fig. 1(A)) may be a result of signal amplification due to the functional groups from both molecules of the ion pair.

Characterisation by CP/MAS 13 C NMR and 1 C NMR
A solid sample of polyDADMAC-dye ion pair was characterised by CP/MAS 13 C NMR.The results are given in Fig. 2.
From the spectrum (Fig. 2), the assignment of the peak of the ion pair corresponds to a linear polymer with 5-membered pyrollidinium ring.This peak assignment was identical to that described by Lancaster et. al (1976) and later confirmed by Assem et al. (2007) when using a trithiocarbonate RAFT (Reversible Addition Fragmentation Chain Transfer) agent.Signals (Fig. 2) contributed by the dye were observed at 127.4 δ and 138.3 δ but were very small owing to poor sensitivity of the technique.A sample solution of the polyDADMAC-dye solid ion pair previously analysed with CP/MAS 13 C NMR was prepared in deuterated water and also analysed with 1 H NMR. The results are given in Fig. 3.
The 1 H NMR spectrum of the polyDADMAC-dye ion pair showed peaks due to both the polyDADMAC and the dye.However, the peaks due to the aromatic hydrogen atoms from the dye were not easily observable because of the poor sensitivity of the method.This required one to perform repeated scanning to enable visualisation of the poorly-resolved tiny signals.The aromatic hydrogen atoms were observed in the range 7.5-8.0δ (see Fig. 3 inset, between 4 δ and 5 δ).By contrast, the hydrogen peaks in polyDADMAC, unlike those from the dye, were clearly detected on the 1 H NMR spectrum as shown in Fig. 4, which is similar to studies reported by Saveyna et. al. (2008).For the purpose of quantification, the band at 3.7 δ was considered and its intensity monitored at varying concentrations of polyDADMAC solutions in deuterated water.
Figure 4 shows an example of one of the results obtained when the concentration of the polyelectrolyte was 1.0 mg•ℓ -1 .The signal at 3.7 δ was visible, but only as a weak band, after the instrument's sensitivity was enhanced by a factor of 10 (compare the y-axes of Figs. 3 and 4).This implies that 1 H

UV-Vis characterisation
In the preliminary experiment, to establish if absorption in the visible region obeyed the Beer-Lambert law, solutions with different concentrations (0.4, 0.6, 0.8 and 1.0 mg•ℓ -1 of poly-DADMAC) of the ion pair were analysed by UV-Vis spectrophotometry (between 400 and 600 nm).The resulting spectra are shown in Fig. 5.
It was observed that the polyDADMAC-dye ion pair has an absorption peak in the visible region (as shown in Fig. 5(B) inset) at a wavelength of 520 nm.This agrees with the findings of other workers who analysed phenols (related organics) by spectrophotometric methods, after formation of derivatives with the same dye (Whitlock et. al., 1972).In addition, Oakes et. al (2001) reported similar observations of maximum absorb-reported similar observations of maximum absorbance at 520 nm for the oxidation of azo dyes in aqueous media.In this study, varying concentrations of standard solutions (0.4, 0.6, 0.8 and 1.0 mg•ℓ -1 ) showed a significant difference in absorbance values (obeying Beer-Lambert law) in the visible spectrum (at 520 nm).We also observed that the ion pair gave significantly higher intensities in the UV region than those in the visible spectrum (compare the spectra of Fig. 5(A), 235 nm with Fig. 5(B), 520 nm).The high intensities in the UV region could be attributed to the conjugation of the dye (Kotera et. al., 1955).Thus, in the UV region the Beer-Lambert law was obeyed up to 0.8 mg•ℓ -1 , and spectrophotometric determination of polyDADMAC was therefore carried out in this region.
The effect of pH on the wavelength of maximum absorbance (λ max ) was studied.Solutions of the ion pair were prepared at different pH values (1.50-12.00).The results obtained (spectra not shown) revealed that pH has no significant effect on λ max at 205 nm and 230 nm, for solutions prepared in 1,4-dioxane and DMF, respectively.This confirms observation by Al-Shihry (2005) that pH has no effect on the conjugation of the molecule.However, the effect of solvent type on λ max observed in our study was similar to observations reported by Dubas (2006).This solvatochromic effect was observed by Stamatovska and co-workers in 2006, in studies of the effect of solvents on electronic absorption spectra.They explained that the phenomenon was due to the free interaction of the chromophores in the molecules, whose absorption falls within a specific band arising from the delocalisation of the p-system, which spreads over the whole molecule.A study by Kotera  2006) also reported that the effects of sol-also reported that the effects of solvent polarity and hydrogen bonding on the absorption spectra were interpreted by the linear solvation energy relationships (LSER) concept developed by Kamlet and Taft (Bekárek, 1983;Kamlet et. al., 1983).

Calibration plots of polyDADMAC-dye ion pair by UV spectrophotometry
A calibration curve for the determination of the polyDADMAC ion pair was prepared with standard solutions in the range 0.1-25.0mg•ℓ -1 polyDADMAC (proportional to molar concentration range 6.18 x 10 -7 -1.55 x 10 -4 mol•ℓ -1 of DADMAC monomer, molar mass 161.66 g•mol -1 ).The dissociated units of polyDADMAC in water react with the dye to form ion-pair colloids with a molar mass of 511.99 g•mol -1 (De Gennes, 1979).On this basis, the resulting concentrations of the ion pair formed would be proportional to 1.95 x 10 -4 -4.88 x 10 -2 mol•ℓ -1 .Figure 6(A) shows the various overlaid spectra for standard solutions of polyDADMAC prepared in DMF (10 mℓ), showing the λ max at 230 nm, while the inset (B) is the corresponding calibration plot with concentrations in mol•ℓ -1 .Note that the molar absorptivity was calculated in mol•ℓ -1 •cm -1 .

Detection limit
The limit of detection (DL) was obtained graphically from the full calibration curve (prepared in 1,4-dioxane at λ max of 205 nm, with a range of 0.05-2.0mg•ℓ -1 of polyDADMAC which translates to 3.09 x10 -7 -1.24 x 10 -5 mol•ℓ -1 of ion-pair).The calculated DL value was 0.07 mg•ℓ -1 , which is proportional to a molar concentration of 1.37 x 10 -7 mol•ℓ -1 for the ion pair.On the basis of the DL value, this method would be suitable for the determination of the flocculant, polyDADMAC, at trace levels (concentrations < 0.1 mg•ℓ -1 ).

Analysis of environmental water samples
Water samples (collected from the Umgeni River in Durban) were treated with 0.045 mg•ℓ -1 polyDADMAC and filtered.Each of the filtrates (2 ℓ) was spiked with 2 different concentrations of polyDADMAC (100 mℓ of 0.5 mg•ℓ -1 and 1.2 mg•ℓ -1 ) and labelled as S1 and S2, respectively.The purpose of spiking the sample with the standards was to obtain relatively high concentration levels of polyDADMAC, detectable by the UV-Vis spectrophotometer.
The spiked samples were treated using a similar procedure to the standard solutions.The ion-pair formed was then dissolved in 10 mℓ of 1,4-dioxane and the absorbance readings recorded.It was observed that water sample S1 (spiked with 0.5 mg•ℓ -1 polyDADMAC) had an absorbance reading of 0.336 while S2 (spiked with 1.25 mg•ℓ -1 polyDADMAC) gave a reading of 0.193.Based on the amounts of polyDADMAC added to the water samples (after treatment and spiking), it is expected that the final concentrations would be 0.545 mg•ℓ -1 and 1.245 mg•ℓ -1 for the S1 and S2 samples, respectively, assuming that no coagulation took place.The experimental values obtained with the UV-Vis method were 0.546 mg•ℓ -1 and 1.245 mg•ℓ -1 for S1 and S2, respectively, or 1.01 x 10 -6 or 2. 43 x 10 -6 mol•ℓ -1 , (based on a molar mass 511.99 g•mol -1 for the ion pair formed).The results show an average difference of 0.045 mg•ℓ -1 , which is not significantly different (based on ANOVA at p = 0.05, n = 8) from the initial amount of 0.045 mg•ℓ -1 used in the treatment.This indicates that the UV method is reasonably accurate for the determination of polyDADMAC at these concentration levels.The implication of these results is that when polyDAD-MAC is applied in the treatment of uncontaminated water (as is the case in our study), the final product may possibly contain the unreacted polycation.The latter is known to form compounds of health concern when reacting with chlorinated by-products (Park et. al., 2009).

Conclusions
Quantitative determination by 1 H NMR spectroscopy of residual polyDADMAC in treated water at trace levels (<0.1 mg•ℓ -1 ) was not possible due to the poor sensitivity of the method.The UV spectrophotometric technique showed better sensitivity (in terms of the detection limit) for the quantification of polyDADMAC.Formation of stable UV-active colloids of a polyDADMAC-dye ion pair and their subsequent solvation in DMF or 1,4-dioxane made determination of the flocculant residue possible.The colloid formation was due to the high affinity of polyDADMAC for negatively-charged/neutral particles of the (4-hydroxy-1-napthylazo) benzene-sulphonic acid dye.Thus, the reaction between the polyelectrolyte and the dye enabled this UV-inactive compound (polyDADMAC) to be analysed by the spectrophotometric technique.The effect of pH on λ max for the polyDADMAC-dye complex was found to be insignificant for the two solvents systems studied.On the other hand, the nature of the solvent was found to play a significant role in determining the λ max value.The solvent chosen was 1,4-dioxane, because of the low concentration range investigated (linear range 0.1-1.8mg•ℓ -1 ), which gave a detection limit of 0.07 mg•ℓ -1 (70 ppb) of polyDADMAC, a value proportional to a molar concentration of 1.37 x 10 -7 mol•ℓ -1 of the ion pair.The developed method was found to be relatively easy for routine analysis, compared to the existing methods based on colloidal titration.The experimental results were found to be in close agreement with the expected values of the polyelectrolyte in the spiked water samples.The UV method developed in this study is suitable for trace analysis (concentrations < 0.1 mg•ℓ -1 ).The advantage of the developed method is that the UV spectrophotometric technique is simple, fast and does not require specialised skills.
In addition, the method is applicable for detection of polyDADMAC in the presence of interfering species in treated wastewater.This is because of the selective formation of the ion-pair with the dye which results in gravimetric precipitation of the ion-pair complex that can subsequently be determined spectrophotometrically.This selectivity is based on the fact that other substances in the water will not react with the dye and hence will remain in solution.Moreover in the case of co-precipitation of the analyte with any other species present in water, absorption of the complex will not take place at the same wavelength, as demonstrated in this study.