Quartz Crystal Microbalance and Infrared Reflection Absorption

A survey of the 2001 to 2005 quartz crystal microbalance biosensor literature: applications of acoustic physics to the analysis of biomolecular intera...
0 downloads 0 Views 231KB Size
Download PDF
Anal. Chem. 2004, 76, 788-795

Quartz Crystal Microbalance and Infrared Reflection Absorption Spectroscopy Characterization of Bisphenol A Absorption in the Poly(acrylate) Thin Films Guifeng Li,† Shigeaki Morita,‡ Shen Ye,*,†,‡, § Masaru Tanaka,| and Masatoshi Osawa†,§

Graduate School of Environmental Earth Science, Hokkaido University, Sapporo 060-0810, Japan, Japan Science and Technology Agency (JST), Japan, Catalysis Research Center, Hokkaido University, Sapporo 001-0021, Japan, and Research Institute for Electronic Science, Hokkaido University, Sapporo 060-0812, Japan

The absorption process of bisphenol A (BPA) in a number of poly(acrylate) thin films, such as poly(2-methoxyethyl acrylate) (PMEA), poly(ethyl acrylate) (PEA), poly(n-butyl methacrylate) (PBMA), and poly(methyl methacrylate) (PMMA), has been investigated by quartz crystal microbalance (QCM) and infrared reflection absorption spectroscopy (IRRAS) measurements. Both QCM and IRRAS measurements show that the BPA molecules absorb in PMEA, PEA, and PBMA thin films but not in PMMA thin film. The differences in the BPA absorption behavior are mainly attributed to the difference in the glass transition temperature (Tg) between these polymers. This absorption behavior also depends on the BPA concentration and polymer film thickness. Furthermore, IRRAS characterization demonstrates that the hydrogen bonding is formed between the hydroxyl group in BPA and the carbonyl group in the poly(acrylate) thin films. BPA molecule absorbed in these polymer thin films can be removed by ethanol rinse treatment. By optimizing experimental conditions for the QCM electrode modified by PMEA thin film, detection limitation of ∼1 ppb for BPA can be realized by the in situ QCM measurement. This method is expected to be a sensitive in situ detection way for trace BPA in the environmental study. Bisphenol A (BPA) is widely used as a monomer for the synthesis of raw materials for engineering plastic materials. It has been recently reported that BPA can be dissolved from polycarbonate products such as food packaging, and a trace amount of BPA can disturb the hormone balance in a living body; i.e., it is considered as an endocrine disrupter (an environmental hormone).1-3 Endocrine disrupters are attracting extensive attention in the field of environmental science, and the quantitative detection * Corresponding author. E-mail: [email protected]. † Graduate School of Environmental Earth Science, Hokkaido University. ‡ PRESTO, Japan Science and Technology Agency. § Catalysis Research Center, Hokkaido University. | Research Institute for Electronic Science, Hokkaido University. (1) Wright, A. N.; Fischli, A. E. Pure Appl. Chem. 1998, 70, 1617-1865. (2) Hunt, P. A.; Koehler, K. E.; Susiarjo, M.; Hodges, C. A.; Ilagan, A.; Voigt, R. C.; Thomas, S.; Thomas, B. F.; Hassold, T. J. Curr. Biol. 2003, 13, 546553. (3) Staples, C. A.; Dorn, P. B.; Klecka, G. M.; O’Block, S. T.; Harris, L. R. Chemosphere 1997, 36, 2149-2173.

788 Analytical Chemistry, Vol. 76, No. 3, February 1, 2004

of BPA molecules with a high sensitivity and selectivity becomes an extremely important issue. The quantitative detection of BPA has been carried out by using gas chromatography coupled with mass spectrometry (GC/MS),4-6 liquid chromatography with electrochemical detection (LC-ED)7 and liquid chromatography coupled with mass spectrometry (LC-MS).5,8 The detection limit of BPA by LC/MS is about 0.5 ppb.8 However, these analysis systems are rather complicated and expensive, and need a long time for pretreatment before measurement, and are hardly employed on site. Recently, the enzyme-linked immunosorbent assay (ELISA) method has also been employed to detect BPA with a detection limit of 0.1 ppb in water8,9 but is still hard to be employed under in situ condition. It is really desirable to develop a simple and sensitive analytical method for the in situ detection of a trace amount of BPA in environment. The quartz crystal microbalance (QCM) is an effective method to detect a small mass change on the solid surface in a mass level of nanograms but with a poor selectivity.10-13 QCM electrodes modified by functional organic thin films have been used to improve its selectivity in the detection of biomolecules.10-19 Ha et al. investigated the interaction of indolicidin with a lipid bilayer (4) Meesters, R. J. W.; Schroder, H. F. Anal. Chem. 2002, 74, 3566-3574. (5) Petrovic, M.; Eljarrat, E.; Lopez de Alda, M. J.; Barcelo, D. J. Chromatogr., A 2002, 974, 23-51. (6) del Olmo, M.; Gonzalez-Casado, A.; Navas, N. A.; Vilchez, J. L. Anal. Chim. Acta 1997, 346, 87-92. (7) D’Antuono, A.; Dall′Orto, V. C.; Balbo, A. L.; Sobral, S.; Rezzano, I. J. Agric. Food Chem. 2001, 49, 1098-1101. (8) Inoue, K.; Wada, M.; Higuchi, T.; Oshio, S.; Umeda, T.; Yoshimura, Y.; Nakazawa, H. J. Chromatogr., B 2002, 773, 97-102. (9) Zhao, M. P.; Li, Y. Z.; Guo, Z. Q.; Zhang, X. X.; Chang, W. B. Talanta 2002, 57, 1205-1210. (10) Buttry, D. A.; Ward, M. D. Chem. Rev. 1992, 92, 1355-1379. (11) Oyama, N.; Ohsaka, T. Prog. Polym. Sci. 1995, 20, 761-818. (12) O’Sullivan, C. K.; Guilbault, G. G. Biosens. Bioelectron. 1999, 14, 663-670. (13) Janshoff, A.; Steinem, C. Sens. Update 2001, 9, 313-354. (14) Sota, H.; Yoshimine, H.; Whittier, R. F.; Gotoh, M.; Shinohara, Y.; Hasegawa, Y.; Okahata, Y. Anal. Chem. 2002, 74, 3592-3598. (15) Ha, T. H.; Kim, C. H.; Park, J. S.; Kim, K. Langmuir 2000, 16, 871-875. (16) Pope, L. H.; Allen, S.; Davies, M. C.; Roberts, C. J.; Tendler, S. J. B.; Williams, P. M. Langmuir 2001, 17, 8300-8304. (17) Takada, K.; Naal, Z.; Park, J.-H.; Shapleigh, J. P.; Bernhard, S.; Batt, C. A.; Abruna, H. D. Langmuir 2002, 18, 4892-4897. (18) Tanaka, M.; Mochizuki, A.; Motomura, T.; Shimura, K.; Onishi, M.; Okahata, Y. Colloid Surf., A 2001, 193, 145-152. (19) Tanaka, M.; Mochizuki, A.; Shiroya, T.; Motomura, T.; Shimura, K.; Onishi, M.; Okahata, Y. Colloid Surf., A 2002, 203, 195-204. 10.1021/ac0348874 CCC: $27.50

© 2004 American Chemical Society Published on Web 12/30/2003

Table 1. Physical Properties of the Polymers polymer PMEA Mw Tg (°C)

Figure 1. Molecular structures of various polymers used in the study.

membrane of L-R-dipalmitoylphosphatidic acid deposited on a QCM electrode modified by an octadecanethiol self-assembled monolayer (SAM).15 Pope et al. investigated the formation of a DNA duplex and the adsorption of nogalamycin molecules on the surface of oligonucleotides fixed on a hexadecanethiol SAMmodified QCM electrode.16 On the other hand, the functional polymer thin film directly deposited on the QCM electrode has been employed to increase its selectivity and sensitivity. Takada et al. investigated the interaction between maltose binding protein-nitroreductase fusion and electropolymerized films by the QCM measurement.17 Tanaka and Okahata carried out in situ QCM detection of protein adsorption behaviors of bovine serum albumin and human fibrinogen on the poly(2-methoxyethyl acrylate) (PMEA) thin film,18,19 which shows excellent blood compatibility and is expected to be one of the promising biomaterials.20 Recently, we have reported that the BPA molecules strongly interact with the PMEA thin film using sum frequency generation and infrared reflection absorption spectroscopy (IRRAS).21 In the present study, BPA absorption behaviors on a number of poly(acrylate) thin films, such as PMEA, poly(ethyl acrylate) (PEA), poly(n-butyl methacrylate) (PBMA), and poly(methyl methacrylate) (PMMA), have been systematically and quantitatively investigated by QCM and IRRAS measurements, including the dependences of the BPA concentration and polymer film thickness. Molecular structure changes of poly(acrylate) thin films induced by the BPA absorption were discussed in detail based on the IRRAS results. Reasons for the selective absorption of BPA in PMEA thin film are discussed. These experimental results show that there is a possibility to develop the PMEA thin film as a BPA sensor with a high sensitivity. EXPERIMENTAL SECTION Materials. The chemical structures of four poly(acrylate)s (PMEA, PEA, PBMA,- PMMA) and polystyrene (PS) used in this study were shown in Figure 1. PMEA and PEA were synthesized (20) Tanaka, M.; Motomura, T.; Kawada, M.; Anzai, T.; Kasori, Y.; Shiroya, T.; Shimura, K.; Onishi, M.; Mochizuki, A. Biomaterials 2000, 21, 1471-1481. (21) Ye, S.; Morita, S.; Li, G.; Noda, H.; Tanaka, M.; Uosaki, K.; Osawa, M. Macromolecules 2003, 36, 5694-5703.

8.5 × -50

104

PEA 8.5 × -24

104

PBMA 1.8 × 20

105

PMMA 1.2 × 105

105

PS 4.5 × 104 100

by the radical polymerization method according to a previous paper.20 PBMA, PMMA, and PS were purchased from Aldrich. The physical properties for these polymers, such as weightaverage molecular weight (Mw) and glass translation temperature (Tg), are shown in Table 1. BPA and other organic solvents were Suprapure reagents from Wako Pure Chemicals. All chemicals were used without further purification. Milli-Q water was used to prepare aqueous solutions in the present experiment. QCM Measurement. A 5-MHz AT-cut quartz plate (Maxtek Inc.) with gold electrode was used as the QCM resonator. The gold electrode with a thickness of 200 nm was prepared by vacuum evaporation. The QCM electrode was cleaned with ethanol and acetone and was then dried by purified N2 gas. Polymer thin films were deposited on the surface of the gold electrode by spin coating or casting methods.21 The QCM electrodes modified by polymer thin films were stored in a vacuum box for at least 12 h before use. Details about the QCM systems have been reported elsewhere.22 A universal frequency counter (Agilent 53132A, 225 MHz) was used to monitor the frequency change (∆f ) of the QCM resonator. ∆f was monitored and transferred to a computer every second. The stability of QCM system in present work is ∼0.1 Hz. All experiments were carried out under in situ conditions and room temperature. The aqueous solution was stirred by a small bar magnet. BPA molecules were introduced into the QCM cell (volume, 225 mL) by a microsyringe from its ethanol stock solution (0.1 µM-0.1 M). The mass change (∆m) from ∆f of a AT-cut quartz crystal can be estimated by the Sauerbrey equation as23

∆m ) -

AxFd ∆f 2f o2

(1)

where f 0 is the fundamental resonant frequency (5 MHz) of the QCM resonator and F and d are density (2.65 g/cm3) and shear modulus (2.95 × 1011 dyn/cm2) of quartz, respectively. A is the effective area of the QCM electrode. However, it is not always possible to correlate ∆f to ∆m directly. It is well known that, as a mass sensor, the frequency shift of a QCM resonator is influenced by many factors, such as viscoelastic effect, solvent uptake, surface roughness, surface mass loadings and distribution, surface stress, and interfacial slippage.10-13 Only when the polymer thin film is rigidly coupled to the quartz surface and the influences of its viscoelasticity and other effects are negligible, can ∆f be related to the ∆m exactly. However, such conditions are not satisfied by all polymers, especially for those with low Tg since the adhesive rigidity of the such a polymer film to the QCM (22) Ye, S.; Haba, T.; Sato, Y.; Shimazu, K.; Uosaki, K. Phys. Chem. Chem. Phys. 1999, 1, 3653-3659. (23) Sauerbrey, G. Z. Phys. 1959, 155, 206-222.

Analytical Chemistry, Vol. 76, No. 3, February 1, 2004

789

Figure 3. ∆f of QCM electrode modified by (a) 83-nm-thick PMEA and (b) PS film after introduction of ethanol, water, phenol, and BPA. The relative positions of the results were shifted vertically for an easy reading.

Figure 2. (a) AFM image (100 µm × 100 µm) of the PMEA film on the gold substrate; (b) thickness profiles of (i) PMEA + Au and (ii) Au on the surface of glass along the dotted line region in (a).

electrode surface is poor when the temperature is higher than its Tg. For example, Tg of the PMEA is ∼-50 °C (Table 1) and the ideally rigid adhesion to the QCM electrode is not good enough under room temperature. To improve this situation, a PS thin film (Tg ) 100 °C, Table 1) with a thickness of ∼300 nm was sandwiched between PMEA and the gold surface.21 Furthermore, as will be described later in detail, ex situ IRRAS measurements were also carried out to calibrate the mass changes induced by the BPA absorption and to elucidate the interaction between BPA molecules and the poly(acrylate) thin films. IRRAS Measurement. IRRA spectra were recorded by a BioRad FTS 575C FT-IR spectrometer equipped with a liquid N2cooled MCT detector and a Harrick Reflector grazing angle reflectance unit (incident angle, 75°).24 Each IRRA spectrum was obtained by coadding 32 interferrograms. The samples for IRRAS measurements were prepared by deposition of polymer thin film on a slide glass (2 × 2 cm2) covered by a 200-nm-thick gold film. The IRRA spectrum of a bare gold film was used as the reference. The spectral resolution used was 4 cm-1. All the measurements were carried out in ex situ conditions. Morphology and Thickness Characterization of the Thin Polymer Films. Figure 2a shows an AFM image (100 µm × 100 µm) of a PMEA film on a 200-nm-thick gold film evaporated on a slide glass substrate using the tapping mode (Nanopics 1000, Seiko instruments Inc.). A straight step observed in the center of the image was made by lifting the PMEA/Au film with an adhesive tape from the glass substrate. The morphology of the PMEA thin film (left top) is relatively smooth in wide range and its root-meansquare roughness was estimated to be ∼1.6 nm, which is (24) Harrick, N. J.; Milosevic, M. Appl. Spectrosc. 1990, 44, 519-522.

790 Analytical Chemistry, Vol. 76, No. 3, February 1, 2004

comparable to that of gold film (∼1.0 nm). Similar results have been observed on the other polymer thin films. The profile along the dotted line in Figure 2a is shown in Figure 2b. The thickness of the polymer film can be estimated from the step height shown in Figure 2 by subtracting the thickness of the gold film. As shown in Figure 2b, from the cross sections of (i) PMEA + Au film and (ii) Au film, the thickness of the PMEA film was determined to be ∼83 nm. Furthermore, the thicknesses of the polymer films have also been confirmed by an ellipsometry measurement with a SOPRA GESP-5 spectrometer using a 30-W Xe lamp as a light source, and its details have been given elsewhere.21 The thicknesses of other PMEA films were estimated from the IR absorbance of PMEA thin film by assuming a linear relation between the IR absorbance and film thickness. The thicknesses of other polymer thin films were evaluated in the same way. RESULTS AND DISCUSSION QCM Detection of BPA Absorption in Poly(acrylate) Thin Films. Figure 3a shows the frequency responses (open squares) of a QCM electrode modified by a 83-nm-thick PMEA film when ethanol, water, BPA, and phenol were introduced into the aqueous solution in sequence. No frequency change was observed when 200 µL of ethanol or 200 µL of water was introduced. However, introduction of 200 µL of 1 mM BPA ethanol solution (final concentration in the QCM cell was 0.2 ppm, the same notation will be used below) yielded a frequency decrease of ∼0.5 Hz. This frequency decrease (∼0.5 Hz) reproduced well when the same amount of BPA was introduced into the solution (Figure 3a). However, no frequency change was observed when a 10 ppm phenol solution, which has a phenol group similar to that in BPA molecule, was introduced into the solution. As a comparison, the same in situ QCM observation was also carried out on a QCM electrodes modified by a 300-nm-thick PS thin film. As shown in Figure 3b (filled squares), no frequency change was observed when the same chemicals as those in a PMEA-covered QCM electrode were introduced into the solution (Figure 3a), indicating that the frequency decreases observed in Figure 3a should be attributed to BPA absorption in the PMEA thin film.

Figure 4. Frequency changes of QCM electrode modified by (a) PMEA (83 nm), (b) PEA (74 nm), (c) PBMA (96 nm), (d) PMMA (81 nm) after introduction of 10 ppm BPA. Table 2. QCM Results of Various Polymers after 10 ppm BPA Introduction polymer

d (nm) ∆f max (Hz) ∆f max/d (Hz/nm) τ (s)

PMEA

PEA

PBMA

PMMA

83 ( 5 105 ( 3 1.27 215 ( 5

74 ( 4 37 ( 2 0.50 160 ( 3

96 ( 6 10 ( 1 0.10 130 ( 3

81 ( 5

Figure 4 shows the frequency response of the QCM electrodes modified by different 80-100-nm-thick poly(acrylate) films, such as (a) PMEA (open stars), (b) PEA (open triangles), (c) PBMA (open squares), and (d) PMMA (open circles), as a function of time after 10 ppm BPA was introduced into the solution. The resonant frequencies of QCM electrodes covered by (a) PMEA, (b) PEA, and (c) PBMA thin films immediately decreased after the BPA molecules were introduced into the solution while almost no frequency change was observed for that covered by PMMA thin film. The frequencies of QCM electrodes modified by PMEA, PEA, and PBMA quickly decreased in the initial stage after the BPA introduction and then changed slowly and finally became almost constant after a certain period. The time profile of the frequency decrease (∆f ) can be expressed by a simple exponential function (eq 2),

∆f ) ∆f max (1 - e-(t/τ))

(2)

where τ is the relaxation time to reach the maximum frequency change (∆f max). The observed QCM responses were fitted by eq 2. The fitting results are shown as solid lines in Figure 4, and the fitting parameters are summarized in Table 2. To compare the frequency response of different polymers, ∆f max was normalized by the respective thickness (d) as shown in Table 2. Both ∆f max/ d and τ decreased in the sequence PMEA > PEA > PBMA . PMMA, demonstrating clearly that the BPA molecules generate larger frequency decreases in the PMEA thin film with a longer time period than those in the PEA and PBMA thin films. In comparison with the QCM responses for introduction of 0.2 (Figure 3a) and 10 ppm BPA (Figure 4a) on the PMEA film with the same thickness, it is clear that the values of ∆f max and τ

significantly depend on BPA concentration in the solution. The higher the BPA concentration, the larger the ∆f max, and the longer the time to reach the equilibrium. IRRAS Confirmation of BPA Absorption in Poly(acrylate) Thin Films. As described in the Experimental Section, a number of factors can give rise to the frequency change of a QCM electrode modified by the thin polymer films.10-13 To check whether the frequency decreases observed in Figures 3 and 4 quantitatively reflect the mass increases due to the BPA absorption in the poly(acrylate) thin films, ex situ IRRAS measurements were also carried out after the samples were immersed into the BPA solution. Figures 5a and 6a show IRRA spectra of a 83-nm-thick PMEA thin film with a PS intermediate layer on the gold-covered slide glass substrate after immersion in 10 and 340 ppm BPA solutions for various periods, respectively, in the frequency region of 37003100 and 1800-1500 cm-1. The spectra between 3700 and 3100 cm-1 in Figure 5 were multiplied by a factor of 5 for comparison. As shown in the IRRA spectra of the original PMEA/PS thin film (top curve in Figures 5a and 6a), an intense peak at 1740 cm-1 corresponding to the CdO stretching of the free carbonyl group in the PMEA layer and several peaks in the region of 1600-1500 cm-1 due to the CsC ring stretching modes of monosubstituted benzene in the PS intermediate layer were observed.21,25 A small absorption, probably by the residue water in the PMEA film, could be found around 3450 cm-1. Detailed assignments about the main IR peaks for the PMEA/PS thin films were given in the Table 3. A number of new spectral features can be observed after the immersion of the PMEA/PS sample in the BPA solution as indicated by the dotted lines (Figures 5 and 6). To show the changes clearly, the IRRA spectra in Figures 5a and 6a were normalized by the IRRA spectrum of the original PMEA/PS thin films before immersion in BPA solution, and the new IRRA spectra were showed in Figures 5b and 6b, respectively. In these normalized IRRA spectra, upward and downward peaks correspond to the species formed and disappeared during the immersion process in BPA solution, respectively. When the PMEA thin film was immersed into 10 ppm BPA solution (Figure 5b), two upward peaks at 1514 and 3400 cm-1 and a bipolar peak (upward one at 1715 cm-1 and downward one at 1740 cm-1) could be clearly observed. These spectral changes become virtually bigger and two upward peaks were also undoubtedly found at 1616 and 1603 cm-1 after the immersion in a concentrated (340 ppm) BPA solution (Figure 6b). Obviously, the upward peaks observed at 1514, 1603, and 1616 cm-1 can be attributed to typical ring C-C stretching modes of phenyl ring of BPA (Table 3).21,25 The broad peak around 3400 cm-1 can be assigned to the O-H stretching of BPA molecules.21,25 The intensities of these IR peaks increased with immersion time and BPA concentration. These results suggest that BPA molecules were absorbed in the PMEA thin film during the immersion process. The appearance of the bipolar peak (1740/1715 cm-1) can be attributed to the environmental changes in the CdO stretching of free carbonyl group in the PMEA after the BPA absorption. Since the carbonyl group in PMEA is a strong electron donor, it is expected to form strong hydrogen bonds with the OH group in the BPA molecules. The upward peak that appeared at 1715 (25) Colthup, N. B.; Daly, L. H.; Wiberley, S. E. Introduction to Infrared and Raman Spectroscopy; 3rd ed.; Academic Press: San Diego, CA, 1990.

Analytical Chemistry, Vol. 76, No. 3, February 1, 2004

791

Figure 6. (a) IRRA spectra and (b) difference spectra of an 83nm-thick PMEA film at different times after 340 ppm BPA injection. The IRRA difference spectra were normalized by the IRRA spectrum before BPA introduction. Figure 5. (a) IRRA spectra and (b) difference spectra of an 83nm-thick PMEA film at different time after introduction of 10 ppm BPA. The IRRA difference spectra were normalized by the IRRA spectrum before BPA introduction.

cm-1 is attributed to the carbonyl groups, which are bonded with the OH group in the BPA via hydrogen-bonding interaction, and the downward peak at 1740 cm-1 can be attributed to the decrease of the free carbonyl group.21 Similar peak shift-induced hydrogen bonding was also observed in the blends of PMMA and poly(4vinyl phenol).26-28 (26) Li, D.; Brisson, J. Polymer 1998, 39, 793-800. (27) Li, D.; Brisson, J. Polymer 1998, 39, 801-810. (28) Dong, J.; Ozaki, Y. Macromolecules 1997, 30, 286-292.

792 Analytical Chemistry, Vol. 76, No. 3, February 1, 2004

Table 3. Assignments of IRRAS Spectra PMEA

PS

BPA 1514 1603 1616

1600-1500 1715 1740

∼3400

assignment ring ν19a ring ν8b ring ν8a ring CsC stretching bands CdO‚‚‚HsO hydrogenbonding interaction CdO stretching OsH stretching

These spectral features due to BPA absorption completely disappeared after the PMEA thin film was rinsed with ethanol for a short time (bottom curve in Figure 5). No change was observed when the polymer surface was only rinsed with pure

Figure 8. IR intensities of the peak at 1514 cm-1 for (a) PMEA (open stars), (b) PEA (open triangles), (c) PBMA (open squares), and (d) PMMA (open circles) thin films after different immersion periods (0-720 s) in 10 ppm BPA solution.

Figure 7. Time-dependent IR intensities of peaks at 1740, 1715, and 1514 cm-1 on the 83-nm-thick PMEA thin film after (a) 10 and (b) 340 ppm BPA introduction.

water. The possibility of hydrolysis of the ester group in the PMEA and other poly(acrylates) thin films seems to be low under the present experimental condition from the effect of ethanol rinsing. These results suggest that a weak chemical bond between the BPA and poly(acrylate) was formed. Figure 7 shows the intensities of IR peaks at 1740, 1715, and 1514 cm-1 as a function of immersion time in (a) 10 and (b) 340 ppm BPA solutions, respectively. The absorbed amount of BPA, which is proportional to the intensity of the IR peak at 1514 cm-1, increases with immersion time and BPA concentration. The amount of free carbonyl group (1740 cm-1) in the PMEA thin film decreases while that of hydrogen-bonded carbonyl group (1715 cm-1) increases with BPA absorption. The BPA absorption process seems to be quicker and absorption reaches equilibrium after a shorter period in the concentrated BPA solution. On the other hand, PMEA thin films were also prepared from PMEA solution containing different amounts of BPA on gold substrates by the casting method, and a calibration curve obtained between the amounts of BPA in the film and the intensity of the IR peak at 1514 cm-1 was used to estimate the amounts of BPA absorbed in polymer thin films during the immersion process. Figure 8 shows IR peak intensities at 1514 cm-1 for (a) PMEA (open stars), (b) PEA (open triangles), (c) PBMA (open squares), and (d) PMMA (open circles) thin films after different immersion periods (0-720s) in 10 ppm BPA solution. The IR peak intensities of BPA in the PMEA, PEA, and PBMA thin films increased with the immersion process and became almost constant after ∼400

Figure 9. Relationship between mass changes estimated from QCM experiments (Figure 4) and those obtained from IRRAS measurements for (a) PMEA (open stars), (b) PEA (open triangles), and (c) PBMA (open squares) thin films. The dotted line corresponds to an ideal QCM response following the Sauerbrey equation.

s. The amount of BPA absorbed decreases in a sequence of PMEA > PEA > PBMA . PMMA. The intensity profiles of the IR peak for these polymers look similar to those of frequency decreases observed by QCM (Figure 4), implying that at least part of these frequency changes can be attributed to the mass changes following BPA absorption in the poly(acrylate) films. Since the QCM responses are controlled by a number of factors besides mass change,10-13 it is very important to know the contribution of the mass changes in the frequency changes observed. Figure 9 shows the relationship between mass changes estimated from QCM experiments (Figure 4) and the mass changes in the films measured by IRRAS measurements. The dotted line in Figure 9 shows an ideal case when the conditions for the Sauerbrey equation are satisfied. When the absorbed amount of BPA in the films was low, mass changes estimated from the QCM were relatively close to those obtained by IRRAS measurement (Figure 9). When the absorbed amount of BPA in PMEA or PEA increased, however, the deviation between the QCM and IRRAS measurements became larger; i.e., QCM measurements gave larger mass changes compared with those obtained from the IRRAS measurements. As already reported in our previous paper, Analytical Chemistry, Vol. 76, No. 3, February 1, 2004

793

Figure 10. ∆f max and IR intensity of the peak at 1514 cm-1 on an 83-nm-thick PMEA film as a function of BPA concentration.

swelling and morphology changes of the PMEA film were observed clearly after the BPA absorption.21 It is known that the swelling and morphology changes can generate larger changes in the film viscoelasticity, which in turn can result in larger frequency changes in the polymer films. Although detailed QC impedance measurement10-13 is still necessary to determine the exact contribution of the film viscoelasticity in the QCM responses, the present IRRAS measurement can quantitatively evaluate the contribution of the viscoelasticity changes after BPA absorption on the QCM electrode modified by poly(acrylate) thin films. Therefore, ∆f obtained from the present QCM measurement can be partly related to the mass increases due to BPA absorption in the bulk of the poly(acrylate) films and can be used as in situ determination of BPA concentration, but a careful calibration is needed for its real mass changes in the films by IRRAS measurements. Optimum Condition for BPA Detection. (1) Concentration Dependence. The absorption behavior of BPA in the PMEA thin film has been investigated by QCM and IRRAS measurements in a wide concentration range of BPA. Figure 10 shows ∆f max of a 83-nm-thick PMEA film (circle) and intensity of IR peak at 1514 cm-1 (star) as a function of BPA concentration (0-360 ppm). With increase of BPA concentration in solution, ∆f and absorbance quickly increased in the lower concentration region and the slope of the curve decreased slightly when the BPA concentration became higher. The similar BPA concentration dependence of QCM and IRRAS responses further confirmed that ∆f correlates the mass change in the PMEA film due to BPA absorption. (2) Thickness Dependence. Figure 11 shows the QCM frequency response of the QCM resonator modified by (a) 83-, (b) 435-, and (c) 600-nm-thick PMEA films after 10 ppm BPA was introduced into the QCM cell. It is clear that a thicker film has a higher frequency response but takes a longer time to reach equilibrium condition. ∆f max for (a) 83-, (b) 435-, and (c) 600-nmthick PMEA films were 105, 364, and 608 Hz, respectively. The thickness dependence of QCM and IRRAS responses is compared in Figure 12. The thicker the PMEA film, the higher is the IR peak intensity at 1514 cm-1, and therefore, the larger is the BPA absorption. This result demonstrates that QCM response, i.e., BPA absorption, in the PMEA thin film is strongly dependent on the PMEA thickness. Therefore, a highly sensitive detection of BPA can be realized by optimizing the film thickness and other 794 Analytical Chemistry, Vol. 76, No. 3, February 1, 2004

Figure 11. Frequency responses of QCM electrodes modified by PMEA with thickness of (a) 83, (b) 435, and (c) 600 nm after 10 ppm BPA introduction.

Figure 12. Frequency change and IR intensity of the peak at 1514 cm-1 as a function of PMEA film thickness after 10 ppm BPA introduction.

experimental conditions. Figure 13 shows the QCM response of 600- and 83-nm-thick PMEA thin films during 1 ppb BPA injection. Although no response could be observed on the 83-nm-thick PMEA film, an obvious frequency decrease of ∼5 Hz could be observed on the 600-nm-thick PMEA film after 1 ppb introduction. As described above, QCM modified by PMEA thin film is an excellent way to detect BPA. Even higher sensitivity and stability by in situ QCM measurement will be realized by further optimization for other experimental conditions in the future. Possible Factors Affecting BPA Absorption in the Poly(acrylate) Thin Films. As observed by IRRAS characterization, hydrogen bonding between the OH group of BPA and carbonyl group of poly(acrylate) films seems to play an important role in the BPA absorption process. However, only a hydrogen-bonding effect cannot explain all experimental results. For example, although PMMA has the similar carbonyl group in its side chain, as shown in Figures 4 and 8, no BPA absorption could be observed. Relative values of the Tg of these polymers also have a significant effect on BPA absorption behavior in the polymer bulks. When the temperature is higher than Tg, the internal mobility of the polymer main chain is higher, and the free volume of the polymer is larger. The penetration of BPA into the polymer bulk can be affected by the micro-Browian motion of the polymer main

is necessary for direct correlation between the frequency change and mass change. As shown in Figure 9, mass changes estimated from the Sauerbrey equation were different from those obtained from IRRAS measurement. The discrepancies for PMEA (Tg ) -50 °C) and PEA (Tg ) -24°C) seem to be larger than that for PBMA (Tg ) 20 °C). The rigidity of the PMEA and PEA on the QCM electrode is poor than that of PBMA under room temperature. As discussed above, changes in film viscoelasticity by BPA absorption will also contribute to the frequency changes in the QCM measurement. A careful calibration by IRRAS measurement is necessary to determine the adsorbed amount of BPA in these polymer films. Furthermore, as shown in Figure 3, although phenol has a fragmental structure of the phenyl ring similar to that in the BPA molecule, its absorption in PMEA and other poly(acrylate) films was not observed by both QCM and IRRAS measurements. This behavior can be related to the higher solubility of phenol in water (77.5 g/L) than that of BPA in water (∼0.12 g/L).3 Compared with BPA, phenol will be readily dissolved in water due to its high solubility. Figure 13. Frequency change of 83- and 600-nm PMEA thin film after 1 ppb BPA introduction. The relative positions of the results were shifted vertically for an easy reading.

chain and its free volume, which are related to Tg. The structure of a polymer with a higher Tg will be more rigid and other molecules will not easily penetrate. The lower the Tg, the higher is the mobility of absorbed molecules in the polymer bulk, the easier is the structure change, and therefore, the higher is the absorbed amount. Compared with that of the PEA, PBMA, and PMMA, PMEA has the lowest Tg (-50 °C, Table 1). It is expected that BPA molecules can easily penetrate into PMEA film and lead to larger ∆f and higher IR intensity due to BPA absorption, in comparison with those of PEA and PBMA thin films. The Tg of PMMA (105 °C, Table 1) is much higher and leads to a rigid structure under room temperature. Thus, it is difficult for BPA to penetrate into PMMA thin film. In fact, the BPA absorption in PMMA bulk was clearly observed after immersion in boiling water containing BPA, indicating that Tg plays an important role in BPA absorption in poly(acrylate) films.29 On the other hand, it should be mentioned that the polymer with lower Tg may make it hard to satisfy the ideal condition of the Sauerbrey equation in QCM measurement where a rigid film (29) Morita, S.; Li, G.; Ye, S.; Osawa, M. Vib. Spectrosc., in press.

CONCLUSION In the present study, the absorption behaviors of BPA in PMEA, PEA, PBMA, and PMMA thin films have been investigated by QCM and IRRAS measurements. Compared with PMMA, PBMA, and PEA films, more BPA molecules were adsorbed in PMEA thin film. This behavior can be attributed to the lower Tg of the PMEA film. Hydrogen bonding and solubility also show important effects on BPA absorption behavior in the poly(acrylate) thin films. A 1 ppb detection limit of BPA has been realized by in situ QCM in the present work. This method is expected to be developed as a sensitive in situ detection method for trace BPA in environmental study after further optimization of experimental conditions in the near future. ACKNOWLEDGMENT S.Y. greatly acknowledges the support from PRESTO, Japan Science and Technology Agency (JST). S.Y. also acknowledges support from the Shiseido Fund for Science and Technology and 2001 Corning Research Grants.

Received for review July 31, 2003. Accepted November 17, 2003. AC0348874

Analytical Chemistry, Vol. 76, No. 3, February 1, 2004

795