Sensitive measurement of methylene blue active substances by attenuated total reflection spectrometry with a trimethylsilane-modified glass slab optical waveguide

Abstract

Attenuated total reflection spectrometry with a slab optical waveguide (SOWG) was explored for the simple, rapid and sensitive measurement of total anionic surfactants by the methylene blue active substance (MBAS) method. A fused-silica sheet used as a guiding layer was modified with trimethylsilane (TMS) to extract and concentrate the MBASs on the SOWG surface. Based on preliminary studies of the adsorption behavior and visible ATR spectrum of MB on the modified silica surface, a detection wavelength of 600 nm was chosen for the sensitive measurement of anionic surfactants. When the concentration of MB was set at 10 μM in the final measurement solution, the calibration curve for a typical anionic surfactant (sodium dodecylbenzenesulfonate) was linear up to 0.6 μM and the detection limit was 0.07 μM. The proposed method was applied to the determination of total anionic surfactants in river water.

---

Introduction

Anionic surfactants are widely used in household cleaners, industrial detergents and cosmetic formulations. The surfactants expelled to natural water reservoirs as municipal and industrial wastes are well known to have adverse effects on aquatic organisms, hence the monitoring of surfactants in environmental samples is of great importance.

Various types of analytical methods have been proposed with the increasing demand for simple, rapid and sensitive determinations of anionic surfactants.^1–15 For the measurement of total surfactant concentration, two-phase titration methods have been extensively explored.^1–6 Several ion-selective electrodes sensitive to anionic surfactants are now commercially available and are widely exploited as end-point indicators in potentiometric titrations. Another major analytical method for total anionic surfactants is the extraction–spectrophotometric method.^7–12 This method is based on an ion-pair extraction reaction with a cationic dye and the ion associates extracted into organic solvents are subsequently determined with a visible-range spectrophotometer. Several cationic dyes such as methylene blue, ethyl violet and Rhodamine 6G have been tested to improve the sensitivity,^7–11 while the solid-phase extraction technique has been introduced to reduce the use of toxic organic solvents.¹²

Most of these methods, however, require troublesome sample pre-treatments such as preconcentration and purification owing to the low concentration of the analyte and the complexity of the sample matrix. In order to simplify handling, much research activity has been aimed at automating the manually operated stages. As a result, several promising flow injection analysis (FIA) systems including on-line sample pre-treatment and flow-through detection have been constructed.^13–15 In order to increase the sensitivity in such FIA systems, some researchers have employed the liquid core waveguide as an optical cell with a long pathlength,^16–18 and others have applied solid-phase spectrophotometry,^19,20 in which the absorbance of analyte species adsorbed and enriched on a solid support placed inside the light path of the flow cell is directly measured.

We have been applying attenuated total reflection (ATR) spectrometry with a slab optical waveguide (SOWG) to chemical sensing, e.g., trace detection of dyes, iron(II) and silicic acid.^21–23 This SOWG technique in practice can detect only species adsorbed on the SOWG surface because the evanescent wave of propagated light is used for the excitation of the analytes. Moreover, it has a highly sensitive detection ability because of its multi-reflection process; its sensitivity can be two to four orders of magnitude higher than that of conventional absorption spectrometry.^24–26 For these reasons, this technique can be a simple but powerful tool for chemical sensing.

In the present study, a successive visible ATR spectrum measurement system was constructed and applied to the sensitive determination of anionic surfactants based on the methylene blue active substance (MBAS) method. For highly sensitive measurement, a thin sheet of glass (50 μm thick) was employed as a guiding layer and the glass surface was modified with trimethylsilane (TMS) to facilitate the deposition of MBASs on the SOWG through hydrophobic interaction. As with solid-phase extractions, this surface extraction method requires little use of toxic organic solvents.

Experimental

Reagents

Methylene blue (MB) was of analytical-reagent grade from Junsei Chemicals (Tokyo, Japan) and was used without further purification. Trimethylchlorosilane (TMCS), used as a silylating reagent, was purchased from Shinetsu Chemicals (Tokyo, Japan). Toluene, ethanol and methanol were obtained from Wako (Osaka, Japan). Anionic surfactants tested as analytes were sodium dodecylbenzenesulfonate (DBS) from Kanto Chemicals (Tokyo, Japan) and sodium dodecylsulfate (SDS) from Wako. These reagents were of highest commercially available purity and used as received. Water was purified with a Milli-Q system (Millipore, Bedford, MA, USA).

Apparatus

Fig. 1 shows a schematic diagram of a visible ATR spectrum measurement system. The system used was basically the same as that in our previous paper.²⁷ In this study, a 150 W xenon arc lamp (Hamamatsu Photonics, Japan) was employed as a light source. The source light was collimated with an optical fiber collimator and was coupled into the SOWG with a coupler prism (La-SF08, n_D = 1.8785) (Kogakugiken, Japan). The SOWG was mounted on a 360° rotational stage with X–Y–Z translation to regulate the incident angle (θ_i) of the source light. In this experiment, the angle was set at 35°. The guided light was out-coupled through another coupler prism. The out-coupled light passed through an aperture, a lens and a polarizer and was finally detected by a multichannel charge-coupled device (CCD) detector (PMA-11) (Hamamatsu Photonics). The signal was processed by a personal computer to obtain visible ATR spectra. The polarizer was used to measure the out-coupled light in the transverse magnetic (TM) and/or electric (TE) mode; the TM mode was chosen for the MBAS determination.

Fig. 1 Schematic diagram of visible ATR spectrum measurement system with SOWG. 1, Xe arc lamp; 2, convex lens; 3, optical fiber; 4, glass slide; 5, flow cell; 6, coupling prism; 7, guiding layer (fused-silica sheet, 50 μm thick, n_D = 1.459); 8, poly(tetrafluoroethylene–co-hexafluoropropylene) (FEP) film (25 μm thick, n_D = 1.338); 9, sample inlet; 10, sample outlet; 11, aperture; 12, polarizer; 13, multichannel CCD detector; 14, personal computer; θ_i, incident angle of source light.

A flow cell was made from a polytetrafluoroethylene (PTFE) block and a PTFE spacer (0.55 mm thick) and was placed on the SOWG. The cell length was 10 mm and the cell volume was 16.5 μl (3 mm wide × 10 mm long × 0.55 mm high). Sample solutions were introduced manually into the cell with a syringe. After the measurement, the cell was washed with 0.1 M HCl, water, ethanol and water to desorb the ion associates from the modified silica surface.

Surface modification of fused-silica sheet

A reflux method was employed for the modification of silica surface with TMS.²⁸ At the outset of the modification, a fused-silica sheet (50 μm thick, n_D = 1.459) (Musashino Fine Glass, Japan) was immersed in concentrated nitric acid for 24 h and was washed with distilled water. After drying at 100 °C for 1 h, the sheet was refluxed in toluene containing 5 vol.% of TMCS for 24 h. The modified silica sheet was then rinsed with toluene, ethanol, methanol and distilled water and dried at 120 °C for 2 h.

Sample preparation

Methylene blue and anionic surfactants were dissolved in water to obtain 5.0 and 2.0 mM stock standard solutions, respectively. These solutions were diluted when necessary and were mixed with each other before the SOWG measurements. The solution pH was conditioned with phosphoric acid or sodium hydrogenphosphate.

River water samples from central Japan (Gunma) were collected in 100 ml polyethylene bottles, as prescribed in the literature.²⁹ The samples filtered with a 0.45 μm membrane filter were stored in a refrigerator at 4 °C.

Results and discussion

Visible ATR spectra of ion associates of MB with DBS

Visible ATR spectra of ion associates of MB with DBS were examined to choose the optimum detection wavelength for anionic surfactants. Fig. 2 shows the visible ATR spectra of 0.1 mM MB solutions containing DBS at concentrations of 0 (blank), 10 and 100 μM. Two absorption peaks were observed at 600 and 670 nm and the main absorption band was shifted from 670 to 600 nm on addition of DBS. The two peaks may be related to different forms of MB adsorbed on the TMS-modified silica surface. According to the literature,^30,31 the maximum absorptions at 600 and 670 nm can be attributed to the light absorption of the dimer and monomer of MB, respectively. In this case, the ion associate of MB with DBS may give a similar ATR spectrum to that of the dimer. Almost the same results were also obtained for ion associates of MB with SDS. As can be seen in Fig. 2, the contribution of the blank absorbance at 600 nm was small, whereas the analyte absorbance increased with increase in the DBS concentration. In the following experiments, the light absorption at 600 nm was used for the determination of anionic surfactants.

Fig. 2 Visible ATR spectra of 0.1 mM MB solutions containing DBS at concentrations of (A) 0, (B) 10 and (C) 100 μM.

The SOWG surface was also modified with octylsilane (OS) and octadecylsilane (ODS) besides TMS and the modified SOWGs were compared in terms of the sensitive measurement of MBASs. Among these, TMS-modified SOWG gave the minimum blank absorbance at 600 nm (derived from MB monomer), whereas there was little difference in the analyte absorbance (derived from the ion associate of MB with DBS). Thus, TMS–SOWG was chosen for further investigation. Moreover, the TM mode of measurement was found to reduce the blank absorbance, while no significant differences between the TM and TE modes were observed in the analyte absorbance. These phenomena were explained by the difference in the orientation of MB monomer and MB in the ion associates.³² Hence the use of the TM mode is advantageous for lowering the detection limit in this study. The adsorption behavior of MB on these modified SOWGs will be discussed in detail in another paper.

The relationship between the absorbance response of the DBS solution and the standing time was also examined. Fig. 3 shows the changes in absorbance response after the introduction of the sample into the flow cell, where the net absorbance was obtained by subtracting the absorption signal of the blank solution (0.1 mM MB solution) from that of the analyte solution. It was found that net-absorbance increased and reached a limiting value within 3.0 min.

Fig. 3 Relationship between absorbance response and standing time. Concentration of DBS: ▲, 1; ◆, 10; and ▼, 100 μM in 0.1 mM MB solution.

Effect of analyte solution pH on SOWG extraction–spectrophotometric detection

The effect of pH on the surface extraction of ion associates was examined in the range 1–8. Fig. 4 shows the results for several analyte solutions with different pH values. The absorption signals increased with increase in pH, but a signal increase of the blank solution (0.1 mM MB solution) was also observed at pH >4. This may be because MB with a positive charge is attached electrostatically to the negatively charged residual silanol on the SOWG surface. As seen in Fig. 4, the net absorbance was maximum and almost constant in the pH range 4–6. Therefore, this pH range was found to be appropriate for the determination and subsequent experiments were carried out at pH 5.0.

Fig. 4 Effect of analyte solution pH on the SOWG extraction–spectrometric detection of 0.1 mM DBS. ▲, 0.1 mM MB with 0.1 mM DBS; △, 0.1 mM MB (reagent blank); and ◆, net absorbance.

Effect of MB concentration on SOWG extraction–spectrophotometric detection

The concentration of MB used as the reacting reagent may have a significant influence on the extraction of anionic surfactants, hence the effect of MB concentration was examined. The results obtained by using 0.01 mM DBS solutions containing MB at concentrations up to 0.5 mM are shown in Fig. 5. It was noted that DBS provided a maximum and almost constant net absorbance by mixing with a 10-fold excess of MB.

Fig. 5 Effect of MB concentration on absorbance response.▲ , 0.02 mM DBS solution; △, reagent blank; and ◆, net absorbance.

On the other hand, the detection limit of anionic surfactants may also be dependent on the MB concentration. The detection limits were therefore calculated as the concentration of DBS corresponding to 3σ (σ = standard deviation) of the blank signal. The values obtained were 0.65 and 0.07 μM for 0.1 and 0.01 mM MB solutions, respectively. It was found that the detection limit could be enhanced by decreasing the concentration of the reacting reagent. It is concluded that the concentration of MB should be selected according to the concentration level in real samples.

Influence of sample matrices on the spectrophotometric detection of anionic surfactants

The absorption signals of ion associates may be severely affected by the presence of salts owing to the salting-out effect. Thus, the influence of co-existing ions on the absorption responses was evaluated. Table 1 shows the absorption responses of 0.1 mM MB solutions containing 0.1 mM DBS and 1 mM of each electrolyte. The relative standard deviation (RSD) for five replicate measurements was <7% under each experimental condition. As can be seen in Table 1, doubly charged cations were prone to cause negative errors in the absorption response.

Table 1 Effect of co-existing electrolytes on the absorbance^a of DBS

Co-existing electrolyte	Absorbance of blank solution^b	Absorbance of analyte solution^c	Net absorbance
a Average value of three measurements. b Solution containing 0.1 mM MB and 1 mM of the electrolyte. c Solution containing 0.1 mM MB, 1 mM of the electrolyte and 0.1 mM DBS.
None	0.118	0.260	0.142
1 mM NaCl	0.109	0.252	0.143
1 mM CaCl2	0.111	0.244	0.133
1 mM MgSO4	0.113	0.245	0.132

---

Application to the determination of anionic surfactants in river water

This SOWG detection method was applied to the determination of total anionic surfactants in river water. Just before the SOWG measurement, 5 ml of the river water sample was mixed with 5 ml of MB solution and 1 ml of the mixture was introduced manually into the flow cell (16.5 μl). In this measurement, the concentration of MB was set at 0.01 mM (in the final solution) to determine trace levels of surfactants of concentration <1 μM. Moreover, a calibration curve constructed with DBS standards prepared by diluting DBS with artificial river water was employed to eliminate the negative error from co-existing ions. The artificial river water was composed of 1 mM each of NaCl, CaCl₂, MgSO₄ and NaHCO₃. A good linear relationship was found up to 0.6 μM, as described by the linear regression equation y = 0.0113x (r² = 0.997), where x is the DBS concentration (μM) and y is the absorption signal at 600 nm. On the other hand, the result for the calibration curve constructed with DBS standards diluted with Milli-Q-purified water was y = 0.0203x (r² = 0.984). The slopes were completely different from each other. It was clearly confirmed that the preparation of proper standard solutions was of importance for accurate and precise determinations. The detection limit was 0.071 μM at a signal-to-noise ratio of 3.

Table 2 gives analytical results for real samples. Only one sample lying near the source (industrial drain) was found to be contaminated with anionic surfactants beyond the permissible limit (0.2 mg l⁻¹), while anionic surfactants were not detected in other samples obtained from a semi-urban area. In order to evaluate the validity and applicability of the proposed method, the recovery tests were carried out by adding known amounts of DBS to the river water samples. The results are also represented in Table 2. The recovery values were found to be almost 100% in the linear dynamic range and the RSDs for three replicate measurements were <5%. These results demonstrate that the proposed method can be employed satisfactorily for the determination of anionic surfactants in surface waters.

Table 2 Analytical results for real and spiked samples DBS/μM

Sample	DBS/μM		Recovery (%)^a
	Added	Found
a Average value of three measurements; the RSDs were <5%. b 20-fold diluted sample was analyzed using 0.1 mM MB reagent solution. c Not detected.
Watarase river (industrial)	0	54.5^b	—
Watarase river (semi-urban)	0	n.d.^c	—
Kiryu river (semi-urban)	0	n.d	—
Watarase river (rural)	0	n.d.	—
Watarase river (rural)	0.1	0.1	100.9
Watarase river (rural)	0.3	0.3	100.4
Watarase river (rural)	1.0	1.1	109.6

---

Conclusion

Visible ATR spectrum measurement system with a SOWG provided a promising analytical technique for MBAS. Preconcentration and detection were integrated in the same system, thus avoiding the use of conventional solvent extraction flow units. As a result, only small amounts of organic solvent are needed and toxicity hazards for the determination are decreased. It is also stressed that further improvements of the sensitivity and analytical performance can be achieved with the use of an automated flow system and an integrated SOWG with an end coupler or a grating coupler which has a smoother and more homogeneously modified surface.^23,33,34 Such studies are in progress and will be reported in subsequent papers.

Acknowledgement

This work was partly supported by a grant-in-aid for Scientific Research (No. 11450320) from the Ministry of Education, Science, Sports and Culture, Japan.

References

1. M. Gerlache, Z. Senturk, J. C. Vire and J. M. Kauffmann, Anal. Chim. Acta, 1997, 349, 59.
2. S. Alegret, J. Alonso, J. Bartroli, J. Baro-Roma, J. Sanchez and M. del Valle, Analyst, 1994, 119, 2319.
3. W. Frenzel, Analyst, 1988, 113, 1039.
4. S. Martinez-Barrachina, J. Alonso, L. Matia, R. Prats and M. del Valle, Anal. Chem., 1999, 71, 3684.
5. J. Alonso, J. Baro, J. Bartroli, J. Sanchez and M. del Valle, Anal. Chim. Acta, 1995, 308, 115.
6. M. del Valle, J. Alonso, J. Bartroli and I. Marti, Analyst, 1988, 113, 1677.
7. J. Longwell and W. D. Maniece, Analyst, 1955, 80, 167.
8. S. Motomizu, S. Fujiwara, A. Fujiwara and K. Toei, Anal. Chem., 1982, 54, 392.
9. M. Kamaya, Y. Tomizawa and K. Nagashima, Anal. Chim. Acta, 1998, 362, 157.
10. S. R. Barroso, V. R. Gamonal and L. M. Polo Diez, Anal. Chim. Acta, 1988, 206, 351.
11. K. Higuchi, Y. Shimoishi, H. Miyata, K. Toei and T. Hayama, Analyst, 1980, 105, 768.
12. O. A. Zaporozhets, O. Y. Nadzhafova, V. V. Verba, S. A. Dolenko, T. Y. Keda and V. V. Sukhan, Analyst, 1998, 123, 1583.
13. J. Kawase, A. Nakae and M. Yamanaka, Anal. Chem., 1979, 51, 1640.
14. M. del Valle, J. Alonso, J. Bartroli and I. Marti, Analyst, 1982, 54, 392.
15. M. Agudo, A. Rios and M. Valcarcel, Analyst, 1994, 119, 2097.
16. K. Fuwa, L. Wei and K. Fujiwara, Anal. Chem., 1984, 56, 1640.
17. K. Tsunoda, A. Nomura, J. Yamada and S. Nishi, Appl. Spectrosc., 1990, 44, 163.
18. R. D. Waterbury, W. Yao and R. H. Byrne, Anal. Chim. Acta, 1997, 357, 99.
19. K. Yoshimura, Analyst, 1988, 113, 471.
20. R. J. Cassella, L. S. G. Teixeira, S. Garrigues, A. C. S Costa, R. E. Santelli and M. de la Guardia, Analyst, 2000, 125, 1835.
21. K. Tsunoda, H. Itabashi and H. Akaiwa, Anal. Chim. Acta, 1993, 276, 133.
22. K. Tsunoda, H. Itabashi and H. Akaiwa, Chem. Lett., 1996, 919.
23. K. Tsunoda, H. Itabashi and H. Akaiwa, Anal. Chim. Acta, 1995, 299, 327.
24. N. Matsuda, A. Takatsu and K. Kato, Chem. Lett., 1996, 105.
25. P. L. Edmiston, J. E. Lee, L. L. Wood and S. S. Saavedra, J. Phys. Chem., 1996, 100, 775.
26. D. A. Stephens and P. W. Vohn, Anal. Chem., 1989, 61, 386.
27. K. Tsunoda, H. Itabashi and H. Akaiwa, Chem. Lett., 1995, 935.
28. D. G. I. Kingston and B. B. Gerhart, J. Chromatogr., 1976, 116, 182.
29. R. Patel and K. S. Patel, Analyst, 1998, 123, 1691.
30. D. A. Higgins, S. K. Byerly, M. B. Abrams and R. M. Corn, J. Phys. Chem., 1991, 95, 6984.
31. T. Hinoue, Y. Yokoyama and T. Ozeki, Bunseki Kagaku, 1994, 43, 443.
32. S. S. Saavedra and W. M. Reichert, Langmuir, 1991, 7, 995.
33. S. S. Saavedra and W. M. Reichert, Anal. Chem., 1990, 62, 2251.
34. W. Lukosz, C. Stamm, H. R. Moser, R. Ryf and J. Dubendorfer, Sens. Actuators B, 1997, 39, 316.

---