Anal. Chem. 1998, 70, 1255-1261
Molecular Design, Characterization, and Application of Multiinformation Dyes for Multidimensional Optical Chemical Sensings. 2. Preparation of the Optical Sensing Membranes for the Simultaneous Measurements of pH and Water Content in Organic Media Hideaki Hisamoto, Yukiko Manabe, Hiroko Yanai, Hajime Tohma, Toshiki Yamada, and Koji Suzuki*
Department of Applied Chemistry, Keio University, 3-14-1 Hiyoshi, Kohoku-ku, Yokohama 223, Japan
Optical chemical sensing of pH and water content in organic solvents is proposed, using multiinformation dyes (MIDs) based on the support matrixes for the dyes. In this investigation, four kinds of merocyanine-type dyes having a polymerizable olefin unit as the MIDs were synthesized. These dyes were copolymerized with hydrophilic monomer molecules to obtain dye-immobilized optical chemical sensor (optode) membranes. In this case, selection of the monomer molecule gave optode membranes having different color change properties, because different monomer molecules provided different chemical environments around the immobilized dye. These optode membranes were used for the measurement of pH and water content in organic solvents. These membranes offered two-dimensional sensing information in one spectrum when they were employed for water content sensing in organic solvents, in which the maximum wavelength represents the water content and the absorbance at this wavelength represents the pH of the water present. These polymer membranes have a long lifetime, which can be adequate for practical use. The practical importance of functional dyes used in optical chemical sensor (optode) development has received increased attention in recent years. Recently, we proposed molecular design concepts for a multiinformation dye (MID) of the merocyanine type for application to optodes, and the fundamental spectral characteristics were fully investigated and reported.1 The aim in the second stage of our MID development is to make optode devices utilizing MIDs that can be effectively used in practical measurements. The concept of the molecular design for the MIDs includes not only the spectral change of the dyes but also an effective applicability of the dyes as a sensing molecule or a component in a sensing device. For a sensing device, an MID (1) Hisamoto, H.; Tohma, H.; Yamada, T.; Yamauchi, K.; Siswanta, D.; Yoshioka, N.; Suzuki, K. Molecular Design, Characterization, and Application of the Multi-Information Dyes (MIDs) for Multi-Dimensional Optical Chemical Sensings. 1. Molecular Design Concepts of the Dyes and Their Fundamental Spectral Characteristics, submitted for publication. S0003-2700(97)00637-9 CCC: $15.00 Published on Web 02/20/1998
© 1998 American Chemical Society
molecule having an immobilization site is required that can be easily utilized for the application of MIDs to chemical sensor devices such as a membrane-based optode. Membrane preparation techniques with support matrixes and functional molecules play an important role in the development of optode devices for practical use. Several types of immobilization techniques have been reported, such as mechanical immobilization, electrostatic immobilization, and covalent immobilization.2 Recently reported optodes utilize these techniques, such as trapping functional dyes in a sol-gel-derived waveguide film,3 covalently immobilizing a dye having a reactive group on a cellulose support,4 and electrostatically immobilizing a cationic dye onto an anionic polymer surface.5 Among these techniques, covalent immobilization of the sensing elements on the polymeric support is the most effective technique for obtaining a sensing probe having a reproducible response and long lifetime. Therefore, we introduced an olefin unit into the MID molecules as one of the immobilization sites, in which this unit can be used as the copolymerization site with the polymer membrane matrix (monomer). Generally, an optical sensing device based on a polymer membrane is convenient and easy to prepare for practical measurements.6 Here we report the preparation of optode membranes with the MIDs and other monomer molecules as a useful example of the use of MIDs for chemical sensor application. In this case, we used four kinds of synthesized merocyanine-type dyes possessing an olefin unit that can be used to easily prepare an optode polymer membrane or film. A preliminary investigation of the spectral characteristics demonstrates that the MIDs prepared here can offer two-dimensional sensing information in one spectrum by utilizing both a shift in maximum wavelength and a change in the absorbance as the detection signals.1 (2) Wolfbeis, O. S., Ed. Fiber Optic Chemical Sensors and Biosensors; CRC Press: Boca Raton, FL, 1991; Vol.1, p 336. (3) Yang, L.; Saavedra, S. S. Anal. Chem. 1995, 67, 1307. (4) Mohr, G. J.; Wolfbeis, O. S. Anal. Chim. Acta 1994, 292, 41. (5) Igarashi, S.; Kuwae, K.; Yotsuyanagi, T. Anal. Sci. 1994, 10, 821. (6) Peterson, J. I.; Goldstein, S. R.; Fitzgerald, R. V.; Buckhold, D. K. Anal. Chem. 1980, 52, 864.
Analytical Chemistry, Vol. 70, No. 7, April 1, 1998 1255
These functional dye molecules were copolymerized with hydrophilic monomer matrixes such as 2-hydroxyethyl methacrylate (HEMA), diethyleneglycol monomethyl ether methacrylate (DMM), or their mixtures. In this case, we found that the selection of the chemical structure of the polymerizable monomer is an important factor for controlling the color change characteristics of the dye-immobilized polymeric membrane. Because different copolymerized monomer molecules produce different chemical environments around the dyes, the spectral or color change characteristics of functional dyes can be subsequently controlled by the chemical structures of the monomer molecules. This phenomenon can be used for controlling the color change characteristic of the dyes for the practical application of the optode device based on the MIDs. In this study, the optode membrane copolymerizations were performed by a UV irradiation technique, and these membranes were applied as detection devices for pH and for water content in various organic media. pH optodes have been widely investigated since the early 1980s.6,7 The application of pH optodes was recently widely expanded, e.g., for intracellular pH measurements using a submicrometer fiber sensor,8 for wide dynamic range pH determination using a polyaniline-coated optical fiber,9 and for a polymer swelling-based pH sensor.10 A water content optode was recently reported by Seitz et al.,11 and the response principle of this optode was similar to that of the polymer swelling-based pH optode. Different from the response mechanism of these optodes, the optode membranes prepared here gave simultaneous sensing information in one spectrum, in which the λmax represents the water content in organic media and the absorbance at this wavelength represents the pH value of the water present. Thus, multichemical detection was achieved by the optode using the MIDs. In this report, the preparation and characteristics of the MIDs in the “site-controlled” polymer membranes and their application to pH sensing and water content sensing in organic media are discussed. Finally, these optode membranes were covalently attached to a small glass plate to make an easy and handy sensing device. EXPERIMENTAL SECTION Reagents. 2-Hydroxyethyl methacrylate (HEMA) and diethyleneglycol monomethyl ether methacrylate (DMM) as membrane matrixes and benzoin methyl ether as a polymerization initiator were purchased from the Tokyo Chemical Industry Co. (Tokyo, Japan). The synthesis and chemical structures of the MIDs (1(5-hexenyl)-4-[(3′,5′-dibromo-4′-oxocyclohexa-2′,5′-dienylidene)ethylidene]-1,4-dihydropyridine (KD-M1), 1-(5-hexenyl)-4-[(4′-oxocyclohexa-2′,5′-dienylidene)ethylidene]-1,4-dihydropyridine (KD-M2), 1-(5-hexenyl)-4-[(3′,5′-di-tert-butyl-4′-oxocyclohexa-2′,5′dienylidene)ethylidene]-1,4-dihydropyridine (KD-M3), and 1-(5hexenyl)-4-[(3′,5′-dimethoxy- 4′-oxocyclohexa-2′,5′-dienylidene)ethylidene]-1,4-dihydropyridine (KD-M4)) used here were described (7) Seitz, R. Anal. Chem. 1984, 56, 16A. (8) Tan, W.; Shi, Z. U.; Smith, S.; Birnbaum, D.; Kopelman, R. Science 1992, 258, 778. (9) Ge, Z.; Brown, C. W.; Sun, L.; Yang, S. C. Anal. Chem. 1993, 65, 2335. (10) Shakhsher, Z.; Seitz, W. R.; Legg, K. D. Anal. Chem. 1994, 66, 1731. (11) Bai, M.; Seitz, W. R. Talanta 1994, 41, 993.
1256 Analytical Chemistry, Vol. 70, No. 7, April 1, 1998
in a previous paper.1 The prepared buffer solutions were as follows; pH 0.6-3.5, 0.2 M HCl-KCl buffer; pH 4-12, universal buffer (0.0286 M citrate, phosphate, borate, and 0.2 M NaOH were mixed and used). All organic solvents used in this work were distilled and stored over molecular sieves (4 Å) to eliminate any water residue. For absorbance measurements of the membrane in organic solvent-water mixtures, distilled and deionized water having resistivity values greater than 1.5 × 107 Ω‚cm at 25 °C was used as the “water” in the case where the pH of the water was not stated (pH ≈ 6-7). All other reagents of the highest grade commercially available were used. Preparation of the Dye-Immbilized Polymer Membrane. The functional dye-immobilized polymer membranes were prepared using the following procedures. The monomer, dye, and benzoin methyl ether (96:1:3 in weight ratio) were mixed and cast onto a dust-free poly(tetrafluoroethylene) (PTFE) plate. In this case, HEMA, DMM, or their mixtures was used as the monomer. When the HEMA-DMM mixture was used, the mixing ratio was varied with weight percent. To obtain thinner membranes, normal cover-glass plates having a 0.12-mm thickness were placed over the droplets, and UV radiation (type XX-15S UV-lamp, Funakoshi, Tokyo, Japan) was directed from 10 cm onto the membrane matrixes for ∼3 h in a nitrogen atmosphere. After the UV irradiation period, the functional dye covalently immobilized polymer membranes were completely washed with water and methanol to remove any unreacted species using a supersonic cleaner until no leaching of the dyes was observed. Thickness of the resulting membranes were ∼50 µm (wet state) unless otherwise stated. Preparation of the Chemical Sensing Plate. Conventional glass plates (38.5 × 9.5 (mm2) ×10 plates) were immersed in 30 mL of methanol, and 385 mg of 3-(trimethoxysilyl)propyl methacrylate was added in the solution. This reaction mixture was refluxed for 20 h. After the reaction period, the methacryl-groupmodified glass plates were washed with methanol to remove any unreacted species. The sensing plates were prepared using the following procedures. Monomer HEMA, dye, and benzoin methyl ether (96:1:3 in weight ratio) were mixed and cast onto a dust-free PTFE plate. Surface-modified glass plates were placed over the droplets, and UV radiation (type XX-15S UV lamp, Funakoshi) was directed in the same manner as previously described. Final treatment of the prepared chemical sensing plate is also the same as the membrane preparation procedure. Consequently, the prepared polymer membrane was covalently attached to the glass surface. Absorbance Measurements. A double-beam spectrophotometer (U-2000, Hitachi Co., Ltd., Tokyo, Japan) and a normal conventional glass vessel (1-cm length with ∼4-mL capacity standard optical cell) were used for the absorbance measurements. The optode membrane was fixed with a PTFE plate having a 9-mm-diameter hole and immersed into a 10- × 10- × 50-mm glass vessel (a normal absorption measurement cell). The reference cell (the same glass vessel as that for the sample) of the spectrophotometer was filled with the same sample solution in order to compensate for the background absorbance in the measurement spectrum.
Figure 1. Solvatochromic properties of the MID-immobilized poly-HEMA membranes, (a) KD-M1, (b) KD-M2, (c) KD-M3, and (d) KD-M4.
Table 1. Absorption Maxima and pKa Values of the MID-Immobilized Poly-HEMA Membranes dye
KD-M1
KD-M2
KD-M3
KD-M4
pKa λmax (nm)
5.6 468
9.1 477
10.0 581
9.2 524
RESULTS AND DISCUSSION pH Response Characteristics of the MID-Immobilized Polymer Membranes. We have prepared a pH-indicator-immobilized hydrophilic polymer membrane in which 2-hydroxyethyl methacrylate (HEMA) was previously used as the polymer matrix.12 The optode membrane based on this polymer matrix showed excellent properties, such as good response reproducibility and long-term stability. Based on this result, the four kinds of MIDs (KD-M1-KD-M4) were copolymerized with HEMA using the UV irradiation polymerization technique. The typical absorption spectra based on the dye-immobilized poly-HEMA membranes showed a clear isosbestic point in each case. This indicates that these MIDs can be used as a normal pH indicator. The pKa and λmax values of these membranes are summarized in Table 1. Based on the effect of the electron-donating or -accepting property of the substituting group in the MID molecule, pH-sensing optode membranes with the ability to measure different pH ranges were successfully prepared using the four kinds of MIDs (KD-M1KD-M4). Their short-time reproducibility was tested using the KD-M3-based poly-HEMA optode membrane. This optode membrane exhibited good reproducibility, with a relative standard deviation of (0.21% (n ) 10), when sample solutions of pH 12.2 and 7.0 were alternately measured. Response time for each measurement was within 5 min. When these poly-HEMA membranes were stored with water in a dark place, a reversible color (12) Hisamoto, H.; Tsubuku, M.; Enomoto, T.; Watanabe, K.; Kawaguchi, H.; Suzuki, K. Anal. Chem. 1996, 68, 3871.
change was obtained after 2 years. Thus poly-HEMA membranes were prepared that had an excellent color change durability after long-term storage. Spectral Characteristics of the MID-Immobilized Polymer Membranes in Organic Solvents. The spectral characteristics of the dye-immobilized poly-HEMA optode membranes in various organic solvents are shown in Figure 1. The absorbance spectra in relation to the protonated form of the MIDs were obtained for all optode membranes except the KD-M1-based membrane in all the six kinds of tested organic solvents, except pyridine. The appearance of a higher wavelength band in Figure 1a was attributed to the difference in the pKa values of the KD-M1immobilized poly-HEMA membrane and those of the other dyeimmobilized membranes. Among the four kinds of MIDs, KDM1 has the lowest pKa value (see Table 1). Consequently, the degree of protonation at the quinoid position of the KD-M1 molecule, which is caused by the contained water or hydroxyl group of the base polymer chain, was relatively lower than that with the membranes based on other dyes (KD-M2-KD-M4). Furthermore, as shown in Figure 1b-d, the membranes based on KD-M2-KD-M4 in pyridine showed interesting spectral properties; the appearance of a higher wavelength band is attributed due to the basic property of pyridine. Thus, in consideration of the pKa values shown in Table 1, the order of the band appearance can be expected to be KD-M2 > KD-M4 > KD-M3. However, the results demonstrated in Figure 1b-d showed the inverse order of that expected. Taking into account the steric structure around the quinoid ring of these molecules, KD-M3 has a much bulkier tert-butyl subunit relative to the other dyes that have a methoxy or hydrogen. Subsequently, the “protonation effect” from the hydroxyl group of the base polymer chain or water molecule contained in the KD-M3-based membrane Analytical Chemistry, Vol. 70, No. 7, April 1, 1998
1257
Figure 2. Solvatochromic properties of the MID-immobilized poly-HEMA-DMM (50:50 w/w) membranes, (a) KD-M1, (b) KD-M2, (c) KD-M3, and (d) KD-M4.
can be relatively less than that in KD-M2- or KD-M4-based membranes. In relation to this absorption measurements using the MIDbased poly-HEMA membranes, different permeabilities of the organic solvents into the membrane were observed. The polyHEMA membranes quickly swelled when they were immersed in a hydrophilic and protic solvent such as water or methanol. However, a hydrophobic and nonprotic solvent molecule such as chloroform, ethyl acetate, or hexane does not swell the polyHEMA membranes. In this case, the optode membranes turned white or shrank when they were immersed in hydrophobic and organic nonprotic solvents. These results indicated that the color change properties of the MIDs are controlled by the selection of the polymer matrix. Therefore, our attention then focused on the control of the chemical environment around the copolymerized dye and the permeability of organic solvents into the polymer membranes by the selection of the monomer molecule required for the copolymerization of the optode membrane. One of the monomer molecules we selected, diethyleneglycol monomethyl ether methacrylate (DMM), has a chemical structure similar to that of the HEMA monomer. DMM is a nonprotic monomer molecule which does not possess a hydrophilic hydroxyl group. Figure 2 shows the absorbance spectra of the MID-immobilized polymer membranes in various organic solvents, in which an HEMA-DMM mixture (50:50 w/w) was used as the membrane matrix. As expected, higher wavelength bands in the 550-650nm region appeared in various organic solvents, compared with those obtained with the MID-immobilized membranes with the poly-HEMA matrix (see Figure 1). The noteworthy fact is the appearance of the higher wavelength bands in the nonpolar solvents such as hexane, chloroform, and toluene, which hardly swelled when a poly-HEMA membrane was used. Especially, the KD-M4-based membrane shown in Figure 2d exhibited this property, while the KD-M2-based membrane did not exhibit the higher wavelength band in nonpolar solvents, as shown in Figure 2b. This fact can also be explained by the steric structure of the dye molecule, as in the case when the poly-HEMA membrane was used. KD-M2 does not have a bulky subunit to prevent the protonation effect of the contained water molecule or base polymer chain, despite its high pKa value. Subsequently, a further increase in the DMM content in the optode membrane could produce a responsive membrane for nonpolar and aprotic solvents. However, overdoping of DMM in the polymer membrane resulted in a fragile membrane; thus, the membrane containing DMM at ∼60% is the limit for practical measurement with the HEMA1258 Analytical Chemistry, Vol. 70, No. 7, April 1, 1998
Figure 3. DMM content dependence of the λmax of the MIDimmobilized poly-HEMA membranes. Plots for the KD-M3-based membranes in the low DMM content region could not be measured due to the disappearance of the higher wavelength band because of the protonation effect in the membranes (see Figure 1c).
DMM matrix. When a poly-HEMA-DMM (50:50) membrane was used, this membrane retained its transparency during our measuring procedure. These results suggested that permeability control of the organic solvents into the membrane phase can be achieved by controlling the membrane composition. This methodology can also be applied to the polarity measurement in the organic solvents. Among the spectral data shown in Figure 2, the KD-M3-based membrane showed a relatively small hypsochromic shift in wavelength (see Figure 6c). Especially, even in the protic solvents such as methanol or ethanol, a drastic absorbance change rather than a spectral shift was observed. The spectral characteristics of this membrane in organic solvents were similar to the pH response characteristics. This spectral change behavior of KDM3 is different from those of other dyes, such as KD-M1, -M2, and -M4. For comparison, the DMM content dependence of the absorption maxima of the KD-M1- and KD-M3-immobilized polyHEMA-DMM membranes in various organic solvents are shown in Figure 3. In the case of the KD-M1-based membranes, a certain λmax shift with increasing DMM content in the membrane or with solvent polarity was observed. In contrast to this phenomenon, the KD-M3-based membrane showed only a slight λmax shift and exhibited an absorbance decrease with increasing solvent polarity (see also Figure 2c). This indicates that the oxygen in the quinoid ring of the KD-M3 molecule has poor interaction with the sol-
Figure 4. Spectral characteristics of the KD-M1-immobilized polyHEMA-DMM (50:50 w/w) membranes in THF-water mixture. (a) Absorption spectra in THF-water mixture (pH 5.6); (b) absorption spectra in THF-water mixture (pH 11.7); (c) water content dependence of the λmax shift for different pH values.
Figure 5. Absorption spectra of the KD-M1-immobilized polyHEMA-DMM (50:50) membranes in THF-water mixture (80:20) when pH of mixing water was varied.
vent molecule, because of the steric hindrance from the bulkiness of the tert-butyl units which are situated very near the oxygen atom. Though the sensing based on a simple solvatochromic λmax shift is already used with several dye molecules dissolved in different solvent media,13-17 our optode membranes based on synthesized dyes of merocyanine derivatives with a two-monomer mixed-membrane system are more widely and reversibly applicable to pH and solvent molecule sensings. From an analytical point of view, spectral changes such as a λmax shift (e.g., KD-M1based membrane) or absorbance change (e.g., KD-M3-based membrane) which can be basically induced by solvent polarity are useful for practical applications (application examples of these membranes are described in the next section). Based on these results concerning the spectral characteristics of the MIDs, (13) Brooker, L. G. S.; Keyes, G. H.; Heseltine, D. W. J. Am. Chem. Soc. 1951, 73, 5350. (14) Langhals, H. Anal. Lett. 1990, 23, 2243. (15) Kumoi, S.; Oyama, K.; Yano, T.; Kobayashi, H.; Ueno, K. Talanta 1970, 17, 319. (16) Dimroth, K.; Reichardt, C.; Siepmann, T.; Bohlmann, F. Liebigs Ann. Chem. 1963, 661, 1. (17) Reichardt, C. Angew. Chem., Int. Ed. Engl. 1979, 18, 98.
increasing the DMM content in the poly-HEMA membrane can decrease the amount of water molecules existing in the membrane phase. Thus, controlling the site (OH group in this case) of the base polymer chain can control both the color change characteristics of the dye and the permeation property of the organic solvent in the membrane. In other words, the matrix-dye interaction and solvent permeabilty are simultaneously controlled by the “site control” of the base polymer chain. Water Content Sensing in Organic Solvents. For analytical purposes, the membranes can be applied to the polarity measurement of the solvent mixtures, and the most suitable system for a practical analytical example is the measurement of an organic solvent-water mixture. Hence, methodologies for water content sensing in organic media with these membranes were further investigated. For the determination of water content in various organic solvents using MID-immobilized polymer membranes, the following two methods are useful, as were described in a previous report:1 (1) measuring the polarity change of the organic solvent and (2) measuring the change in the “pseudo-pH” value in organic solvents caused by the presence of water. The required characteristics of the MIDs for these respective measurements are as follows: (1) the dye has a lower pKa value and larger λmax shift with varying solvent polarity, and (2) the dye has a higher pKa value and smaller λmax shift with varying solvent polarity. The reasons for these requirements have already been fully discussed in a previous report.1 In consideration of these requirements, the first method, which is based on the measurement of the λmax change, was carried out with a KD-M1-based polymer membrane for its suitability regarding requirement 1. In this case, an attempt was made to clarify the relationship between the λmax shift and the pH value of the water in the organic solvents. Figure 4a,b shows the water content dependence of the absorption spectra of the KD-M1-immobilized polymer membrane (HEMADMM 50:50) in a THF-water mixture with different pH values of pH 5.6 and 11.7. As shown in Figure 4a,b, absorbance values for the same composition sample were different due to the protonation effect. However, as shown in Figure 4c, the λmax shift for the KD-M1-immobilized membrane is essentially independent of the pH of water. Figure 5 shows the absorption spectra of this membrane when the pH of water was varied in the THF-water mixture (80:20 v/v). In this case, a thin membrane (∼10 µm) was used. As a result, no obvious shifts in wavelength of the absorption maximum with varying pH were observed. These facts lead to the conclusion that this MID (KD-M1)-immobilized polymer membrane provides two-dimensional sensing information in one spectrum in which the maximum wavelength represents the water content in the organic solvent, and the absorbance at this wavelength represents the pH value of the mixed water. In other words, two-dimensional sensing information was integrated in one spectrum. The second method, which is based on the measurement of the absorbance change at a higher wavelength band, was carried out with KD-M3-based polymer membranes in which KD-M3 has the most suitable properties mentioned in requirement 2. Figure 6 shows the absorption spectra of the KD-M3-based membranes with different membrane composition in a THF-water mixture system. As expected on the basis of the results shown in Figure Analytical Chemistry, Vol. 70, No. 7, April 1, 1998
1259
Figure 8. Spectral characteristics of the KD-M1-immobilized polyHEMA plate in ethanol-water mixture. (a) Absorption spectra in ethanol-water mixture; (b) water content dependence of the λmax shift.
Figure 6. Absorption spectra of the KD-M3-immobilized polymer membranes in THF-water mixture with different polymer compositions in weight mixing ratio: (a) HEMA-DMM ) 100:0; (b) HEMADMM ) 67:33; (c) HEMA-DMM ) 50:50; (d) HEMA-DMM ) 33: 67; and (e) HEMA-DMM ) 0:100.
Figure 7. Schematic illustration of the preparation of the membranemodified conventional glass plate (for details, see Experimental Section).
2c, the higher wavelength band decreased with increasing water content. Furthermore, when the DMM content in the polymer membrane was increased, the absorption band was increased due to the elimination of the protonation by the contained water molecule or the OH residue in the polymer matrix. Thus, 1260 Analytical Chemistry, Vol. 70, No. 7, April 1, 1998
Figure 9. Spectral characteristics of the KD-M3-immobilized polyHEMA plate in pyridine-water mixture. (a) Absorption spectra in pyridine-water mixture; (b) water content dependence of the absorbance response.
controlling the site (OH group) in the membrane matrix successfully controlled the sensitivity of the water content sensing in organic solvents. Preparation of the Chemical Sensing Plates Using MIDImmobilized Optode Membranes. Recently, dry reagent, solid thin-film, and surface modification techniques have provided several useful analytical devices which have a wide use for the simple and rapid determination of analytes.18-20 In consideration of this fact, the conventional sensing plate using MIDs was prepared, based on the surface modification technique for a glass plate.21 Figure 7 is a schematic representation of the process of surface modification of the glass plate using MIDs. Surface silanization with the monomer group and the subsequent copolymerization with the MID and HEMA (monomer) can easily prepare the optode membrane covalently attached to a glass plate that is an easy handled sensing device. The sensing plate was applied to the water content sensing in organic solvents. Figure (18) Walter, B. Anal. Chem. 1983, 55, 498A. (19) Kaneko, E.; Tanno, H.; Yotsuyanagi, T. Mikrochim. Acta 1988, XX, 333. (20) Hisamoto, H.; Miyashita, N.; Watanabe, K.; Nakagawa, E.; Yamamoto, N.; Suzuki, K. Sens. Actuators B 1995, 29, 378. (21) Plueddemann, E. P. Silane Coupling Agents; Plenum Press: New York, 1982.
8 shows the spectral characteristics of the KD-M1-based polyHEMA plate when the water content in ethanol was varied. In this case, increasing the water content in ethanol gave a hypsochromic shift to the higher wavelength band. As shown in Figure 8a, the absorbance value decreased with increasing water content due to the protonation. However, as expected, disapperance of the higher wavelength band was not observed when utilizing the dye having a low pKa value, such as KD-M1. Thus, the water content sensing using the λmax shift was successfully performed with the sensing plate. Figure 9a shows the absorption spectra of the KD-M3-based sensing plate that were obtained when the water content in pyridine was varied. As expected, the polarity and pH-dependent absorption spectra were obtained. The response curve of this plate for water content sensing is shown in Figure 9b. In this case, the water content determination by visible observation can also be possible; visible colors of this sensing plate are blue at 0%, green at 5-10%, and yellow at 20% water in pyridine. The short-time reproducibility of the sensing plate based on KD-M3 was tested by measuring alternately using pyridine (100%) and pyridine-water mixture (5% water) samples. As shown in Figure 9c, the relative standard deviations of 0.33% (n ) 5) and 0.25% (n ) 5) were obtained for these sample solutions, respectively. In this case, the response time was within 5 min for each measurement. CONCLUSIONS In this paper, four kinds of MIDs possessing a copolymerization unit were synthesized, and optode membranes having
different spectral characteristics were prepared using the MIDs. The spectral features of these membranes were successfully controlled by the membrane composition in which a matrix-dye interaction and organic solvent-dye interaction are simultaneously controlled by the selection of the dye and membrane composition. For example, using a KD-M1 dye with the HEMADMM mixture membrane system, pH sensing and water content sensing in organic solvents were successfully achieved with the membrane. In the case of water content sensing in organic solvents, two-dimensional sensing information was obtained with one spectrum. In our opinion, the molecular design of multiinformation dyes, which have multiple color change mechanisms and permit coloration control with membrane matrixes, as demonstrated in the present report, will play an important role in the development of multipurpose optodes that are desired in the future.
ACKNOWLEDGMENT Partial support of this investigation by The Kawakami Memorial Foundation, Iketani Science and Technology Foundation, and the Ministry of Education is acknowledged.
Received for review June 18, 1997. Accepted January 7, 1998. AC970637+
Analytical Chemistry, Vol. 70, No. 7, April 1, 1998
1261
