Introduction
Cationic surfactants are widely used in industrial, cleaning, disinfectant, cosmetic and pharmaceutical domains due to their anti-microbial, emulsifying, anticorrosion and softening properties.1 However, their final destiny is frequently disposal in wastewaters and consequently, the aquatic medium. This has in turn become a source of concern since, because they are not degraded to a great extent in anaerobic environments, they may pass almost unaltered through wastewater treatment plants.
Despite being biodegradable in aerobic conditions, they remain toxic even at low concentrations2 to the aquatic ecosystem,3 affecting organisms such as algae and invertebrate.4 Regarding human toxicity, they can affect the central nervous system, causing agitation, debility, ataxia and reduced consciousness when ingested in large quantities. They also promote the development of lesions in the mouth and the digestive tract and, when in contact with the eyes, they can cause serious damage to the cornea.5 A vigilant and continuous control of cationic surfactants in water sources thus becomes an obligation from the point of view of safeguarding public health.
Traditionally, cationic surfactant determinations have been based on liquid–liquid extraction of ionic pairs by organic solvents; most of such determinations are time-consuming, of low precision and include the need to consume and dispose of large volumes of solutions.
With the aim of circumventing these disadvantages, new analytical methods have been proposed.6–8 Flow injection (FI) systems have been developed based not only on solvent-
extraction methods9–11 but also on replacing the extraction and resorting to the effect of surfactants on dye absorption,12 emission intensity of chemiluminescence systems13 or fluorescent ion associates.14 The enhancement of colour intensity of binary metal complexes by surfactants has also been reported as an alternative method that avoids liquid–liquid extraction and the use of toxic solvents.4,15,16 Flow injection titration manifolds using ion selective electrodes17or micellar- enhanced fluorescence18have also been developed.
In spite of the broad use of FI methodologies, the process of automating the determination of cationic surfactants by sequential injection analysis (SIA), to our knowledge has only been considered recently19 in spite of its advantages for analytical control processes.20 FIA procedures usually require complex multiple-channel manifolds, with samples and reagents flowing continuously in parallel tubing, thus increasing the operational cost. SIA systems provide a good alternative to FIA systems due to the simplicity of the manifold design, the more versatile sample handling capabilities, decreased reagent consumption and minimum waste generation, all of which are important features regarding the need for greener chemistry.
Once designed, manifolds do not need to be physically reconfigured, as the parameters affecting the performance of the systems can be computationally modified.21
The purpose of this work was therefore to use the favorable characteristics of SIA already described to obtain an automatic system for monitoring cationic surfactants. Regarding the chemistry of the determination, the aim was to develop a sensitive and non-extractive method by selecting an improved aqueous spectrophotometric reaction based on the use of bromopyrogallol red and molybdenum(VI). Bromopyrogallol red (BPR) is often used as a chromogenic reagent for the determination of a large number of metals, particularly 2005 © The Japan Society for Analytical Chemistry
Application of Sequential Injection Analysis to the Determination of Cationic Surfactants Based on the Sensitized
Molybdenum–Bromopyrogallol Red Reaction
Marieta L. C. P
ASSOS, M. Lúcia M. F. S. S
ARAIVA,
†and José L. F. C. L
IMAREQUIMTE/Departamento de Química-Física, Faculdade de Farmácia, Universidade do Porto, Rua Aníbal Cunha 164, 4099-030 Porto, Portugal
The aim of this work was to develop a simple, automatic system for the evaluation of cationic surfactants by combining sequential injection analysis and the sensitized effect of cationic surfactants on the reaction between metal ions and chelating dyes. This reaction is based on the increase in absorbance of the complex formed among molybdenum, bromopyrogallol red and increasing concentrations of cationic surfactants. Under optimum conditions, two calibration plots were obtained for a concentration range between 2.50 ×10–7mol dm–3(detection limit) and 5.00 ×10–4mol dm–3of cetylpyridinium chloride, used as standard. Solubilization of water insoluble complexes formed for concentrations of cationic surfactants greater than 1.00 ×10–4mol dm–3were successfully achieved with Triton X-405. RSD values lower than 5.0% were obtained in all cases. The quality of the results obtained for 18 water samples were evaluated by comparison with conventional methods, with no statistically significant differences for a 95% confidence level.
(Received March 31, 2005; Accepted May 24, 2005)
†To whom correspondence should be addressed.
E-mail: lsaraiva@ff.up.pt
molybdenum(VI).22,23 The addition of surface-active substances improves the selectivity and sensitivity of the metal determinations due to the batho- and hyperchromic effects that can be observed.24 In the proposed determination, the effect of the cationic surfactants on the absorption spectra of the complex formed by molybdenum(VI) and bromopyrogallol red were used as the basis for establishing its quantitative determination.
Experimental
Reagents and chemicals
All solutions were prepared using analytical grade chemicals and high purity water (Milli Q) with a specific conductivity of less than 0.1 µS cm–1.
A 0.56 mol dm–3acetic/acetate buffer was prepared by mixing a 0.75 mol dm–3acetic acid solution with a required volume of sodium hydroxide solution of identical concentration to obtain a pH value of 4.5. The solution obtained was used as carrier.
Working standard solutions were prepared by appropriately diluting with water the 1.00 ×10–3mol dm–3and 1.00 ×10–4mol dm–3solutions, obtained from the stock solution of 1.00 ×10–2 mol dm–3 cetylpyridinium chloride (Sigma Chemical Co., St Louis, MO, USA), prepared in water from the solid.
A 2.5 ×10–4mol dm–3bromopyrogallol red (Sigma Chemical Co., St Louis, MO, USA) solution prepared in degasified water was used and a 25 mg dm–3 molybdenum(VI) solution was prepared from the 1000 mg dm–3 commercial solution (BDH- Spectrosol®refª 14050, Dorset, UK).
A 2.0 ×10–3mol dm–3Triton X-405 (Sigma Chemical Co., St Louis, MO, USA) solution was prepared from Triton X-405 70%.
Before analysis, the water samples were subjected to gravity filtration to remove the suspended material.
Apparatus
The SIA system (Fig. 1) comprised a Gilson Minipuls 3 (VilliersleBel, France) peristaltic pump, equipped with a 1.29 mm i.d. Gilson PVC pumping tube and a 10-port selection valve (Valco, Vici C25-3180EMH, Houston, USA). Manifold components were connected by 0.8 mm i.d. PTFE tubes, which were also used for the holding and reaction coils. The holding coil was 3 m in length, the reaction coil was 0.75 m in length, and both were figure eight-shaped in configuration.
A Nresearch 161 T031 solenoid valve (W. Caldwell, NJ, USA) and a synchronizer of the peristaltic pump rollers were introduced into the system to control the pump’s starting point, assuring flow reproducibility.25 Therefore, at the beginning of
each step of the analytical cycle, the carrier flow stayed in closed-circuit until it was ensured that the pump rollers were in a fixed position. Only then, and after one half revolution of the pump head, the solenoid valve was activated and the liquids were allowed to flow to the holding coil. These procedures minimized the effect of system inertia at the beginning of the pump’s rotation.
Another 3-way solenoid valve (Nresearch 161 T031, W.
Caldwell, NJ, USA) placed at one of the inlets of the selection valve was used to process the online addition of Triton X-405 and sample, when the concentration of cetylpyridinium in samples was higher than 1.00 ×10–4mol dm–3.
These devices were controlled by a home-made programme written in QuickBasic language, implemented in a microcomputer equipped with an interface card (Advantech Corp., PCL 711B, San Jose, CA).
Flow rate, flow direction and valve positions were determined by means of homemade software written in Microsoft Quick- Basic.
A Philips Pye Unicam® SP6-350 (Cambridge, UK) spectrophotometer was used as the detector with a Helma 178712-QS (Mullheim/Baden, Germany) flow cell of 18 mm3 inner volume. Analytical signals were recorded on a Kipp &
Zonen BD 111 (Delft, The Netherlands) strip chart recorder or viaa computer equipped with a convenient interface.
Procedure
For the determination of cationic surfactants (in terms of cetylpyridinium concentration) two analytical cycles were developed (Table 1): one (LowC) for samples with a surfactant concentration less than 1.00 × 10–4 mol dm–3 and the other (HighC) for higher concentrations.
With the exception of the sample routine (step 2), the two cycles had all other steps in common (1; 3; 4). Both analytical cycles initially included the aspiration of 65 mm3 of bromopyrogallol red solution to the holding coil (step 1), followed by the sample (step 2). Thereafter, 25 mm3 of molybdenum solution (step 3) were aspirated; finally, by flow reversal, the reaction zone was directed towards the detector, at a flow rate of 1.45 cm3min–1(step 4). For samples of lower cetylpyridinium concentrations, the sample routine (LowC–step 2) consisted merely of the aspiration of 70 mm3of sample.
Whenever concentrations of the analyte above 1.00 ×10–4mol dm–3were detected by the system, the other sample aspiration routine protocol was initiated. Following step 1, the Triton X- 405 and the sample solutions were alternately aspirated into the holding coil (HighC-steps 2.1 – 2.6) by binary sampling,26 Table 1 Analytical cycle used in the determination of cationic surfactants in waters
LowC, lower concentrations; HighC, higher concentrations
Flow rate/
cm3 min–1
Volume/
mm3 MV
position
Sol 2 position Steps
Event Time/s Direction
LowC HighC
1 1 BPR aspiration 1 — 5.0 Reverse 0.78 65
2 — Sample aspiration 2 — 5.4 Reverse 0.78 70
— 2.1 Triton aspiration 2 1 2.8 Reverse 0.34 15.9
— 2.2 Sample aspiration 2 0 1.3 Reverse 0.34 7.4
— 2.3 Triton aspiration 2 1 2.8 Reverse 0.34 15.9
— 2.4 Sample aspiration 2 0 1.3 Reverse 0.34 7.4
— 2.5 Triton aspiration 2 1 2.8 Reverse 0.34 15.9
— 2.6 Sample aspiration 2 0 1.3 Reverse 0.34 7.4
3 3 Molybdenum aspiration 3 — 1.9 Reverse 0.78 25
4 4 Propelling solutions to detector 4 — 10 Forward 1.45 2658
creating a string formed by sample slugs in tandem with Triton slugs (Fig. 1, BS). This was obtained by sequentially activating a solenoid valve (Fig. 1, Sol 2) between position 1 and 0 (on/off). This solenoid valve was connected at position 2 of the multi-port selection valve. Each solution insertion sequence consisted of 3 cycles of an intercalation time of 12.8 s for Triton X-405 and 1.3 s for sample, which implies the intercalation of 3 slugs of 15.9 mm3of Triton X-405 with 3 slugs of 7.4 mm3of sample, creating a well-homogenized zone of approximately 70 mm3. The cycle then proceeded with steps 3 and 4.
Each time that the sample was changed, an additional washing step was performed: 140 mm3of the new sample was aspirated to the holding coil and then flushed through the auxiliary waste (W2). When the mixture of a sample with Triton X-405 was intended, 100 mm3of air was aspirated through the sample tube, guaranteeing that when the mixture sequence was initiated, the tube length between the solenoid valve and the multi-port selection valve would be empty. Air that entered the holding coil was expelled through the auxiliary waste (W2) port.
Reference procedure
For a comparative evaluation, the samples analyzed by the developed SIA procedure were also analyzed using the conventional methodologies proposed by Rodier.27
The conventional methodology for the determination of cationic surfactants in superficial waters is based on the fact that the quaternary ammonium compounds in alkaline medium and in the presence of anionic dyes form a red compound which can be extracted by organic solvents. This compound was quantified spectrophotometrically at 508 nm. When the intention was to determine cationic surfactants in residual waters, the recommended method was based on the principle that quaternarium ammonium compounds in acidic media and in contact with sodium alizarin sulfonate form a yellow complex.
This complex was extracted by chloroform and measured spectrophotometrically at 400 nm.
Results and Discussion
Optimization
In this work, bromopyrogallol red (BPR) formed a red
complex with molybdenum(VI) (Mo). This complex showed a maximum absorption of 556 nm. In the presence of cetylpyridinium (CP) standards, a shift of the absorption maximum to 615 nm occurred along with an increase in sensitivity. This increase was to a certain extent proportional to the concentration of cationic surfactant used and therefore provided the basis for the proposed spectrophotometric SIA system.
The SIA system optimization was made taking into account the best performance characteristics of the method developed;
these included sensitivity, accuracy, precision, detection limits, linear range and sample frequency.
The reaction product formed by the mixture of the BPR, Mo and CP solutions was dependent on pH. To guarantee a constant pH in the reaction zone, we decided to use the buffer solution as carrier instead of water, because the aspiration of buffer zones along with the other three solution zones would significantly extend the analysis time. The pH values between 1.3 and 6 were studied for this carrier solution, with 4.5 being the value chosen; it was provided by an acetic/acetate buffer solution. This value yielded the lowest blank signal, the highest sensitivity and the widest linear range.
Studies were also made to evaluate the buffer capacity. To this end, acetic acid/acetate solutions with concentrations of 0.26, 0.35, 0.56 and 0.74 mol dm–3were tested as carrier. A sensitivity improvement was observed till 0.56 mol dm–3, while higher concentrations had no influence on reaction development. To evaluate if these results were a consequence of insufficient buffering of the sample zone, we made two consecutive calibration curves for each buffer concentration:
one with water standards and the other with standards prepared in the acetic/acetate solution used as carrier. When comparing the two calibration curves obtained with buffer concentrations below 0.56 mol dm–3, we observed a difference of about 10%, meaning that the sample region was not effectively buffered.
However for the concentrations 0.56 mol dm–3 and 0.74 mol dm–3, the calibration curves obtained with aqueous or buffered standards were similar; thus 0.56 mol dm–3was adopted for the acetic/acetate solution concentration for further studies.
Since electrostatic interactions are the bases of ternary complex formation,28 the effect on signals of different ionic strengths was also tested by the addition of NaCl to the different buffer solutions tested. These studies were then completed by comparing two consecutive calibration curves made with buffer concentrations ranging from 0.26 to 0.74 mol dm–3without and with NaCl, which was added to level the ionic strength to a final value of 0.74 mol dm–3. It was concluded that its effect was negligible, since only a slight increase in the absorbance with ionic strength was observed for buffer concentrations lower than 0.37 mol dm–3.
Having established the composition of the buffer solution, we next evaluated the optimum concentrations and volumes of other solutions used. The highest sensitivity was obtained with a solution of 2.5 × 10–4 mol dm–3 BPR and a 25.0 mg dm–3 solution of molybdenum(VI). When the volumes of reagents and sample were studied, the objectives were to reduce the consumption of reagents and to obtain the maximum sensitivity and reproducibility in CP determinations. The volumes chosen were 65 mm3for the BPR solution, 70 mm3for the sample and 25 mm3for the molybdenum(VI) solution. In each case, when greater volumes were tested, a decrease in absorbance was observed, due to the dilution effect caused by the higher volume solutions.
Another important factor that influences the kinetics of the reaction and the physical dispersion29is the aspiration order of Fig. 1 Sequential injection manifold for the determination of
cationic surfactants in waters. Pump, peristaltic pump; Carrier, acetate buffer pH = 4.5; Sol 1, Sol 2, solenoid valves; HC, holding coil (3 m figure-eight; 0.8 mm i. d.); RC, reaction coil (0.75 m figure- eight; 0.8 mm i. d.); MV, multi-port selection valve; D, detector; W, waste; W2, secondary waste; BS, binary string formed by sample slugs (S) in tandem with Triton slugs (T) in the holding coil, in steps 2.1 – 2.6 of the analytical cycle.
the reagents. This fact is even more pronounced here as the overlap of three zones is considered.30 All possible combinations were then tested. The results demonstrated that the sample slug should be aspirated between the BPR and molybdenum(VI) solutions, confirming the observations of Gubeli et al.31 After interchanging the order of the two reagents, a wide linear range and a higher sensitivity was shown when BPR was aspirated first. This could be justified by the effect of cationic surfactant on BPR hydration, which favors the coordination of the metal with the dye.24
During optimization, it was verified that the ternary BPR-CP- Mo complex tended to precipitate when cetylpyridinium concentration was above 1.00 ×10–4mol dm–3, which hindered its determination for higher concentrations. To solve this problem, we considered two hypotheses: the on-line dilution of the samples with water or the recourse to a non-ionic surfactant, Triton X-405, usually used to promote the solubility of similar complexes.28 The on-line dilution required, independently of being implemented by using a dilution coil, a zone sampling approach or a mixing chamber32,33would significantly enlarge the analytical cycle time. With Triton X-405, it was possible to avoid the precipitation of the complex formed and to maintain the optimized cycle time. The addition of this solution to the samples was then made on-line to avoid batch pretreatment of the samples with higher cationic surfactant contents. To accomplish this objective we placed a solenoid valve at position 2 of the multi-port selection valve connected to the two solutions, samples and the Triton X-405. The insertion of a sequence composed of just one aliquot of sample and one aliquot of Triton, would compromise its mixture and hinder further reaction development, due to the limited contact between the two zones during aspiration. Therefore, a binary sampling approach26 was selected (Fig. 1, BS) enabling the loading of very small aliquots of the two solutions in an alternating way directly into the holding coil. This approach promoted an almost immediate homogeneous mixture and favored the color reaction development. The optimized insertion sequence then consisted of 3 intercalation cycles of three 15.9 mm3Triton X-405 slugs with three 7.4 mm3sample slugs, creating a well-homogenized zone of approximately 70 mm3, the optimum sample volume having already been optimized. To ensure the reproducible aspiration of these low volumes, we used the aspiration flow rate of 0.34 cm3 min–1 instead of the rate of 0.78 cm3 min–1 established for the other solutions. The Triton X-405 concentration was also studied in the range between 4.0 ×10–4 and 2.8 ×10–3 mol dm–3and the 2.0 ×10–3 mol dm–3concentration was chosen, since this was the lowest concentration that yielded a high sensitivity and a low detection limit without any precipitate formation. In this sense, the non-ionic surfactant was a solubilizing agent as well as a sensitizing agent. We next checked whether this enhancement would be advantageous for the full concentration range of cetylpyridinium studied. To this end, a calibration
curve was made with all the standards intercalated with Triton X-405. It was not possible to obtain a linear calibration curve for all concentrations in the study with the non-ionic surfactant.
The analytical signals obtained with low CP concentrations were smaller than the signal obtained without the intercalation of a Triton solution in samples, indicating that the dilution effect superimposed on the Triton enhancement. Therefore, it was necessary to establish two calibration curves: one for cetylpyiridinium concentrations below 1.00 ×10–4mol dm–3and the other that was activated for concentrations above 1.00 ×10–4 mol dm–3. When the former was used, the samples were not intercalated with Triton X-405 solution, while with the latter intercalation was necessary.
The size of the holding coil must allow minimal dispersion, but must be of sufficient volume to prevent the different
“tandem” zones reaching the propulsion unit;34this balance was obtained with a 3 m tube length of 0.8 mm internal diameter.
The geometry of the holding coil was also investigated and a figure eight-shape was used. A linear geometry, in spite of the higher analytical signals obtained, gave rise to lower repeatability. The interpenetration of zones continues within the reaction coil when they are propelled to the detector.
Therefore, its length and shape was also evaluated in order to obtain the best reaction conditions. The analytical signals showed an increase of 9% and 272% for the cetylpyridinium concentrations of 1.00 × 10–5 and 5.00 × 10–4 mol dm–3, respectively, with the decrease in length from the 2.25 m to 0.25 m coil lengths studied. However as the blank signal also increased and the precision deteriorated, the 0.75 m value was selected as a compromise between the obtained sensitivity, precision and sampling rate. The figure eight-shaped configuration was selected, since by intensifying the radial transport and limiting the axial dispersion along the conduits,29 it permitted a compromise between detection limit, sensitivity and calibration curve linearity. The effect of propulsion flow- rate was also investigated between 0.67 cm3min–1and 2.23 cm3 min–1. A value of 1.45 cm3 min–1 was selected since this corresponded with the attainment of an enhanced sensitivity and sample frequency. For lower flow rates, a prevalence of dispersion was noted and for higher values the reaction time was insufficient for color development. Since the aspiration period extends the measurement cycle,29it is desirable to use the greatest flow rate that enables an acceptable precision in the volumes drawn up. After evaluating the flow rates between 0.78 and 1.45 cm3min–1, a value of 0.78 cm3min–1was selected as a compromise between sampling time and precision, since at higher flow rates the backpressure became too high.
Analytical figures of merit
The proposed sequential injection system was evaluated under optimum conditions (Table 2) regarding response concentration ranges, typical calibration curves, detection limits (LOD), quantification limits (LOQ), precision and determination Table 2 Figures of merit of the SIA system developed
Typical calibration LODa/ mol dm–3
LOQb/ mol dm–3
RSD, % (conc./
mol dm–3)
Determination rate/h–1 Linear range
8.00 × 10–6 – 1.00 × 10–4 AU = 2289(±52) Conc. + 0.0029(±0.0030) 2.20 × 10–6 7.34 × 10–6 4.57 (2.55 × 10–5) 29
mol dm–3 R = 0.999 3.27 (7.48 × 10–5)
1.00 × 10–4 – 5.00 × 10–4 AU = 2068(±46) Conc. + 0.045(±0.014) 2.96 × 10–5 9.88 × 10–5 4.93 (2.55 × 10–4) 28
mol dm–3 R = 0.999 2.70 (3.41 × 10–4)
a. Detection limit. b. Quantification limit.
frequency. Two calibration curves were obtained, one for lower concentrations of cetylpyridinium and the other for higher concentrations. The detection and quantification limits were obtained with a blank solution as three times and ten times the signal blank/slope (n= 30), while precision was evaluated by 10 repeated runs of a number of samples.35
Application to water samples
The proposed SIA procedure was applied to the evaluation of 18 water samples of superficial and residual origin that were analyzed with one of the analytical cycles optimized (LowC or HighC) depending on the concentration level of cationic surfactants.
Common cationic species in water such as Na, K, Ca, Mg, Cr(VI), do not react with BPR. Only the group of those easily hydrolyzed elements, in which molybdenum is included, react:
namely, Sn(IV), Ge(IV), Pb(VI), Al, Ni, Cu.24 However, no significant interference is expected from these ions, as Mo(VI) is used in the determination in a concentration of 25 mg dm–3 and, even when this value was doubled to 50 mg dm–3 (in the optimization studies), there was no change in the absorbance values obtained.
The accuracy of the results obtained was evaluated by comparison with the cationic surfactant concentration values obtained by conventional methods.27 Concentration values obtained by the proposed and the conventional methods for 18 samples, and the relative deviations, are shown in Table 3.
The significance of the correlation was also evaluated using the t-test, carried out as a bilateral coupled test.36 The tvalue tabulated (2.11) when compared with the t value calculated (1.20) shows the absence of statistical differences for the results obtained by the methodologies at the 95% confidence level.
The agreement between both methods was also confirmed for a linear relationship obtained: CSIA(mol dm–3) = (0.995 ± 0.024) CCONV (mol dm–3) + (3 × 10–7 ± 6.7 ×10–6), (where CSIA and CCONV are the sequential injection and the conventional procedure results, respectively) with 95% confidence limits for the intercept and the slope. The values obtained show that intercepts are close to zero and the slopes close to unity, which confirms yet again the agreement between the two methods.
Conclusions
The developed methodology was shown to be a valuable automatic alternative for the routine determination of cationic surfactants in terms of cetylpyridinium in waters over a wide concentration range. The methodology is simple, inexpensive and reliable.
When compared to the conventional methods, as it is a non- extractive method, this methodology avoids the use of organic solvents and is less laborious and more rapid. Regarding flow methodologies and while only considering those that do not involve organic solvents, it allows reagent saving between 300 and 10000% in comparison to the non-continuous working of SIA, a very important feature concerning environmentally friendly chemistry. It enables a working range of approximately two decades of concentration that is comparable or better to that obtained with flow methodologies.
Due to its configuration and mode of operation it can be adapted for applications with other matrixes and even other concentrations of cationic surfactants, as different routines of intercalation of samples with Triton X-405 can be implemented simply by accessing the keyboard of the computer.
References
1. J. Cross and E. J. Singer, in “Cationic Surfactants- Analytical and Biological Evaluation”, 1994, M. Dekker, New York, 3.
2. M. J. Scott and M. N. Jones, Biochim. Biophys. Acta, 2000, 1508, 235.
3. G. Nalecz-Jawecki, E. Grabinska-Sota, and P. Narkiewicz, Ecotoxicol. Environ. Saf., 2003, 54, 87.
4. K. Agrawal, G. Agnihotri, K. Shrivas, G. L. Mundhara, K.
S. Patel, and P. Hoffmann, Microchim. Acta, 2004, 147, 273.
5. P. F. González, in “Manual de Intoxicaciones en Pediatría, Intoxicaciones por Detergents”, 2003, Ergon, Madrid, 161.
6. H. Klotz, Tenside, Surfactants, Deterg., 1987, 24, 370.
7. T. Masadome, Anal. Lett., 2004, 37, 499.
8. A. I. Kulapin and T. V. Arinushkina, Anal. Chem., 2000, 55, 1096.
9. J. Kawase, Anal. Chem., 1980, 52, 2124.
10. T. Sakai, Analyst, 1992, 117, 211.
11. J. Y. Guo, Z. M. Zhang, and Y. P. Wang, Fenxi Huaxue, 2000, 28,124.
12. K. Yamamoto and S. Motomizu, Anal. Chim. Acta, 1991, 246, 333.
13. A. Safavi and M. Karimi, Anal. Chim. Acta, 2002, 468, 53.
14. T. Masadome, Anal. Lett., 1998, 31, 1071.
15. R. Patel and K. S. Patel, Talanta, 1999, 48, 923.
16. I. Nqmcová, V. Tománková, and P. Rychlovsky, Talanta, 2000, 52, 111.
17. C. J. Dowle, B. J. Cooksey, J. M. Ottaway, and W. C.
Campbell, Analyst, 1988, 113, 117.
18. C. A. Lucy and J. S. W. Tsang, Talanta, 2000, 50, 1283.
19. S. Feng, X. Chen, J. Fan, and Z. Hu, Intern. J. Environ.
Anal. Chem., 2005, 85, 63.
20. J. F. van Staden, Anal. Chim. Acta, 2002, 467, 61.
21. A. Ivaska and J. Ruzicka, Analyst, 1993, 118, 885.
22. J. H. Mendez, B. M. Cordero, and L. G. Davila, Analyst, 1987, 112, 1507.
23. M. M. Tananaiko and L. I. Gorenshtein, J. Anal. Chem.
USSR, 1988, 43, 239.
Table 3 Results obtained in the determination of cationic surfactants in water samples, by SIA and by conventional methodologies
Sample SIA/mol dm–3 Conventional method/mol dm–3
Relative error, % 1 (6.97 ± 0.13) × 10–5 (6.68 ± 0.05) × 10–5 + 4.3 2 (4.07 ± 0.04) × 10–5 (3.96 ± 0.06) × 10–5 + 2.8 3 (5.40 ± 0.09) × 10–5 (5.57 ± 0.07) × 10–5 – 3.0 4 (8.51 ± 0.02) × 10–5 (8.23 ± 0.04) × 10–5 + 3.4 5 (8.13 ± 0.14) × 10–5 (7.75 ± 0.07) × 10–5 + 4.9 6 (5.00 ± 0.01) × 10–5 (4.90 ± 0.04) × 10–5 + 2.0 7 (2.14 ± 0.06) × 10–4 (2.13 ± 0.25) × 10–4 + 0.5 8 (3.11 ± 0.05) × 10–4 (3.14 ± 0.24) × 10–4 – 1.0 9 (3.97 ± 0.06) × 10–4 (3.95 ± 0.14) × 10–4 + 0.5 10 (3.71 ± 0.06) × 10–4 (3.80 ± 0.04) × 10–4 – 2.4 11 (1.98 ± 0.02) × 10–4 (2.04 ± 0.02) × 10–4 – 2.9 12 (2.80 ± 0.03) × 10–4 (2.91 ± 0.10) × 10–4 – 3.8 13 (3.84 ± 0.07) × 10–4 (3.95 ± 0.02) × 10–4 – 2.8 14 (1.98 ± 0.01) × 10–4 (2.00 ± 0.07) × 10–4 – 1.0 15 (2.89 ± 0.00) × 10–4 (2.97 ± 0.05) × 10–4 – 2.7 16 (3.23 ± 0.02) × 10–4 (3.29 ± 0.12) × 10–4 – 1.8 17 (4.49 ± 0.04) × 10–4 (4.58 ± 0.09) × 10–4 – 2.0 18 (4.98 ± 0.07) × 10–4 (4.81 ± 0.00) × 10–4 + 3.5
24. S. B. Savvin, R. K. Chernova, and G. M. Beloliptseva, J.
Anal. Chem. USSR, 1980, 35, 757.
25. A. N. Araújo, R. C. C. Costa, J. L. F. C. Lima, and B. F.
Reis, Anal. Chim. Acta, 1998, 358, 111.
26. B. F. Reis, M. F. Giné, E. A. G. Zagatto, J. L. F. C. Lima, and R. A. Lapa, Anal. Chim. Acta, 1994, 293, 129.
27. J. Rodier, “L’analyse de l’eau”, 1996, Dunod, Paris, 434.
28. E. Pramauro and E. Pelizzetti, in “Surfactants in Analytical Chemistry: Application of Organized Amphiphilic Media.
Comprehensive Analytical Chemistry”, ed. S. G. Weber, 1996, Elsevier, Amsterdam, 203.
29. G. D. Marshall and J. F. van Staden, Process Control Quality, 1992, 3, 251.
30. P. C. A. G. Pinto, M. L. M. F. S. Saraiva, S. Reis, and J. L.
F. C. Lima, Anal. Chim. Acta, 2005, 531, 25.
31. T. Gubeli, G. D. Christian, and J. Ruzicka, Anal. Chem., 1991, 63, 2407.
32. J. F. van Staden and R. E. Taljaard, Fresenius J. Anal.
Chem., 1997, 357, 577.
33. J. C. Masini, P. J. Baxter, K. R. Detwiler, and G. D.
Christian, Analyst, 1995, 120, 1583.
34. R. E. Taljaard and J. F. van Staden, Lab. Robotics Automat., 1998, 10, 325.
35. International Union of Pure and Applied Chemistry, Anal.
Chem., 1976, 55, 712A.
36. J. C. Miller and J. N. Miller, “Estadística para Química Analítica”, 2nd ed., 1993, Addison-Wesley Iberoamerican, S.A., Wilmington, USA, 44.
