Adsorption Behavior of Glucose Oxidase on a Dipalmitoylphosphatic

Luiz C. Salay , Marystela Ferreira , Osvaldo N. Oliveira , Clovis R. Nakaie , Shirley Schreier. Colloids and Surfaces B: Biointerfaces 2012 100, 95-10...
0 downloads 0 Views 2MB Size
Download PDF
2042

Langmuir 2007, 23, 2042-2051

Adsorption Behavior of Glucose Oxidase on a Dipalmitoylphosphatic Acid Monolayer and the Characteristics of the Mixed Monolayer at Air/Liquid Interfaces Yuh-Lang Lee,* Jing-Yi Lin, and Shin Lee Department of Chemical Engineering, National Cheng Kung UniVersity, Tainan 70101, Taiwan ReceiVed August 24, 2006. In Final Form: October 27, 2006 A dipalmitoylphosphatic acid (DPPA) monolayer at the air/liquid interface is used as a binding layer to incorporate glucose oxidase (GOx) from the subphase. The effects of the adsorption time of GOx on the behavior of the mixed DPPA/GOx monolayer and the relevant structure of the mixed LB film were studied using the characteristics of the pressure-area (π-A) isotherm, Brewster angle microscopy (BAM), and atomic force microscopy (AFM). The experimental results show that two equilibrium states of GOx adsorption exist in the presence of a DPPA monolayer. The first equilibrium stage occurs at tens of minutes after spreading of DPPA, and a surface pressure of ca. 7.5 mN/m is obtained. The second equilibrium stage approaches slowly, and a higher equilibrium surface pressure (ca. 16 mN/m) was obtained at ca. 8 h after the first stage. The BAM and AFM images show that, after the second equilibrium stage is reached, a more condensed phase and rough morphology are obtained on the mixed DPPA/GOx monolayer, indicating a higher amount of GOx incorporated into the mixed film. For the first equilibrium stage of GOx adsorption, DPPA molecules can still pack regularly and closely under compression, suggesting that GOx molecules are mainly located beneath the DPPA monolayer at the compressed state. A more uniform phase was detected on a film prepared after the first equilibrium stage was reached. The present result indicates that distinct structures and properties of mixed DPPA/GOx films can be prepared from the various stages of GOx adsorption.

1. Introduction

* To whom correspondence should be addressed. Phone: (+) 886-62757575 ext 62693. Fax: (+) 886-6-2344496. E-mail: yllee@ mail.ncku.edu.tw.

with precisely controlled thickness and molecular orientation, which should give a biosensor with a very short response time.8,13 However, poor reproducibility of the calibration curves was reported for these biosensors, attributing to the events occurring during the preparation and deposition of the mixed films.12,13 The nature of the amphiphile molecules was also found to be important to the sensitivity and response time of a biosensor. A densely packed and poorly permeable monolayer was found to prevent the diffusion of ions and molecules and thus decreases the activity and sensitivity of a mixed film.14 The lipid/GOx interaction plays an important role in determining the adsorption of GOx in a lipid monolayer and the related characteristics of the mixed enzyme/lipid films. Various amphiphile compounds have been used to study their interaction with GOx at the air/liquid interface including octadecylamine,15,16 stearic acid,17 behenic acid,13,18 dibehenoylphosphatidylcholine (DBPC),16,19,20 dimyristoylphosphtidylcholine (DMPC),20 and dipalmitoylphosphatidylcholine (DPPC).20 The lipid chain length,20,21 headgroup charge,16,17 and enzyme concentration15,18 are considered to be important factors in determining the binding behavior of GOx to the lipid monolayer. It is generally accepted that protein will adsorb spontaneously from the subphase to the air/liquid interface, and the presence

(1) Sackmann, E. Science 1996, 271, 43-48. (2) Lukes, P. J.; Petty, M. C.; Yarwood, J. Langmuir 1992, 8, 3043. (3) Nakahara, H.; Nakamura, S.; Nakamura, K.; Inagaki, K.; Aso, M.; Higuchi, R.; Shibata, O. Colloids Surf., B 2005, 42, 157-174. (4) Kodama, M.; Shibata, O.; Nakamura, S.; Lee, S.; Sugihara, G. Colloids Surf., B 2004, 33, 211-226. (5) Feng, S.-S.; Gong, K.; Chew, J. Langmuir 2002, 18, 4061-4070. (6) Yasuzawa, M.; Hashimoto, M.; Fujii, S.; Kunugi, A.; Nakaya, T. Sens. Actuators, B 2000, 65, 241-243. (7) Singhal, R.; Takashima, W.; Kaneto, K.; Samanta, S. B.; Annapoorni, S.; Malhotra, B. D. Sens. Actuators, B 2002, 86, 42-48. (8) Okahata, Y.; Tsuruta, T.; Ijiro, K.; Ariga, K. Langmuir 1988, 4, 13731375. (9) Bartlett, P. N.; Whitaker, R. G. J. Electroanal. Chem. 1987, 224, 37. (10) Tse, P. H. S.; Gough, D. A. Anal. Chem. 1987, 59, 2339. (11) Sun, S.; Ho-Si, P.; Jed Harrison, D. Langmuir 1991, 7, 727-737. (12) Sriyudthsak, M.; Yamagishi, H.; Moriizumi, T. Thin Solid Films 1988, 160, 463.

(13) Fiol, C.; Valleton, J. M.; Delpire, N.; Barbey, G.; Barraud, A.; RuaudelTeixier, A. Thin Solid Films 1992, 210/211, 489. (14) Porter, M. D.; Bright, T. B.; Allara, D. L.; Chidsey, C. E. D. J. Am. Chem. Soc. 1987, 109, 3559. (15) Chovelon, J. M.; Wan, K.; Jaffrezic-Renault, N. Langmuir 2000, 16, 6223-6227. (16) Li, J.; Rosilio, V.; Boissonnade, M.-M.; Baszkin, A. Colloids Surf., B 2003, 29, 13-20. (17) Zaitsev, S. Y. Colloids Surf., A 1993, 75, 211-216. (18) Dubreuil, N.; Alexandre, S.; Fiol, C.; Valleton, J. M. J. Colloid Interface Sci. 1996, 181, 393-398. (19) Rosilio, V.; Boissonnade, M.-M.; Zhang, J.; Jiang, L.; Baszkin, A. Langmuir 1997, 13, 4669-4675. (20) Zhang, Z.; Rosilio, V.; Goldmann, M.; Boissonnade, M.-M.; Baszkin, A. Langmuir 2000, 16, 1226-1232. (21) Du, Y.-K.; An, J.-Y.; Tang, J.; Li, Y.; Jiang, L. Colloids Surf., A 1996, 7, 129-133.

Lipid monolayers and Langmuir-Blodgett (LB) films in a well-defined structure have been studied for a long time due to their structure similarity to that of biomembranes and their practical applications in biofunctionalization of solid surfaces.1-5 The incorporation of biomolecules, such as enzymes, into the monolayers or LB films of lipids provides them with biological activity and biospecific recognition properties. Such films can be used to fabricate artificial systems with biological functions and are of great interest in the field of biosensors.6-8 Among various biosensors studied in the literature, the glucose biosensor is a widely studied system.6-8 Glucose biosensors are operated on the basis of the catalytic transformation of glucose into gluconic acid by glucose oxidase (GOx). To immobilize the GOx onto an electrode surface, various methods have been employed, including binding the enzyme directly to the electrode, entrapping the enzyme in an inert polymer matrix,9,10 and LB film immobilization.11,12 The advantage of the LB technique over other methods is its capability in preparing ultrathin films

10.1021/la062501p CCC: $37.00 © 2007 American Chemical Society Published on Web 12/29/2006

Adsorption BehaVior of GOx on a DPPA Monolayer

of a monolayer will help the adsorption of protein to the interface. Thus, for a mixed protein/lipid film prepared by the LB technique, the incorporation of protein will be determined by the initial concentration of lipid spread on the subphase and the time allowed to perform the adsorption before the compression of the mixed monolayer. Surprisingly, there is little research devoted to the effects of the adsorption state of the protein on the behavior of the mixed monolayers. In a previous work, the dipalmitoylphosphatic acid (DPPA) monolayer was found to be stable at the air/liquid interface and is able to deposit as Y-type multilayer LB films.22 The DPPA monolayer is used in this work to investigate its binding characteristic to GOx. Several techniques including the pressurearea (π-A) isotherm, Brewster angle microscopy (BAM), and atomic force microscopy (AFM) have been applied to characterize the adsorption behaviors of GOx and the properties of mixed DPPA/GOx monolayers. BAM is a powerful tool for in situ observation of the phase evolution of a monolayer at the air/ liquid interface.23-25 AFM provides the possibility of imaging the surface morphology down to the molecular level and has been applied to image films of GOx and protein.26-28 Two equilibrium stages of GOx adsorption are found on the DPPA monolayer. The monolayer behaviors and structure of the mixed LB films obtained for the two equilibrium stages are extensively studied in this work. 2. Experimental Section 2.1. Materials. DPPA (>99% pure) was purchased from Avanti Polar Lipids (Alabama). GOx (130000 units/g, isoelectric point 4.2) was supplied from Fluka. These chemicals were used as received. DPPA was dissolved in chloroform/methanol (9:1 by volume) mixed solvent to prepare a stock solution with a concentration of 0.5 mg/ mL. Water purified by means of a Milli-Q plus water purification system (Millipore) with a resistivity of 18.2 MΩ·cm was used. The subphase was prepared by dissolving GOx in pure water, and the concentration of GOx was controlled at 3.3 µg/mL. 2.2. Methods. Surface pressure-area per molecule (π-A) isotherm measurements and LB film deposition were performed on a film balance system with a deposition apparatus (model KSV 2000) constructed by KSV Instruments Ltd., Finland. The Teflon trough (600 × 75 mm2 in size) was mounted on an aluminum base plate with built-in water channels for subphase temperature control. All experiments were performed at 25 °C. The Wilhelmy plate technique was used for surface pressure measurements. For the π-A isotherm measurements, stock solution containing DPPA was spread on the subphase containing GOx by a microsyringe (Hamilton Co.). Prior to the spreading of the DPPA solution, the possible GOx adsorbed at the air/liquid interface was removed by sweeping the subphase surface with a capillary pipet connected to a vacuum pump. After a zero surface pressure was reached, the lipid solution was spread immediately. To allow both the evaporation of organic solvent and the adsorption of GOx, the monolayer compression was started after a waiting period. Two waiting periods (30 min or 8 h) were adopted in this work to study the effect of the adsorption state of GOx on the monolayer behavior. The monolayer was compressed symmetrically at a rate of 4.6 Å2/(molecule·min) by two barriers to obtain a π-A isotherm. (22) Lee, Y.-L.; Lin, J.-Y.; Chang, C.-H. J. Colloid Interface Sci. 2006, 296, 647-654. (23) Tanaka, H.; Akatsuka, T.; Ohe, T.; Ogoma, Y.; Abe, K.; Kondo, Y. Polym. AdV. Technol. 1998, 9, 150-154. (24) Lee, Y.-L.; Yang, Y.-C.; Shen, Y.-J. J. Phys. Chem. B 2005, 109, 46624667. (25) Lee, Y.-L.; Liu, K.-L. Langmuir 2004, 20, 3180-3187. (26) Losic, D.; Shapter, J. G.; Gooding, J. J. Langmuir 2001, 17, 3307-3316. (27) Losic, D.; Shapter, J. G.; Gooding, J. J. Langmuir 2002, 18, 5422-5428. (28) Hou, Y.; Jaffrezic-Renault, N.; Martelet, C.; Tlili, C.; Zhang, A.; Pernollet, J.-C.; Briand, L.; Gomila, G.; Errachid, A.; Samitier, J.; Salvagnac, L.; Torbiero, B.; Temple-Boyer, P. Langmuir 2005, 21, 4058-4065.

Langmuir, Vol. 23, No. 4, 2007 2043

Figure 1. Variation of the surface pressure with the adsorption time of GOx. The GOx adsorption was carried out on an air/liquid interface in the absence (a) or in the presence (b-d) of a DPPA monolayer. Various DPPA densities with a molecular area of 336 (b), 168 (c), or 112(d) Å2/molecule were used to study the density effect on the GOx adorption. The concentration of GOx in the subphase was controlled at 3.3 µg/ mL. For monolayer relaxation experiments, a monolayer was compressed to a specific surface pressure (30 mN/m) and the surface pressure was held constant throughout the remainder of the experiment by automatically adjusting the interfacial area using barriers. The change in surface area was then recorded as a function of time. To prepare LB films for the analysis of AFM, freshly cleaved mica plates (Alfa Aesar) were used as substrates. The substrate was immersed into the subphase first, and then a monolayer was formed at the interface. After the monolayer was compressed to the target surface pressure, the substrate was withdrawn with a rate of 1.5 mm/min to deposit the LB film. During the deposition, the surface pressure was kept constant by adjusting the barriers automatically. Before the AFM analysis, the LB films were dried in a desiccator at room temperature for about 5 days to decrease the humidity of the samples. The morphology of the films was then examined with an atomic force microscope (Digital Instruments, NanoScope IIIa) via the tapping mode at a scan rate of 0.5 Hz. A silicon tip on a cantilever of 125 µm length (noncontact silicon cantilever, model NCH-50, Germany) was used. The resonance frequency and force constant of the tip are 320 kHz and 42 N/m, respectively. The engagement set point was controlled at 0.85 V. BAM was used to observe the morphological characteristics of the monolayer at the air/liquid interface. For this experiment, a larger trough (70 × 7 cm2) constructed by Nima Technlogy Ltd., England (model 601 BAM) was used. The Brewster angle microscope used was designed by Nanofilm Technology (NFT), Go¨ttingen, Germany (model BAM2 plus). The p-polarized light from an NdYAG laser, with a wavelength of 532 nm, was incident at the Brewster angle (53.1°) to the air/water interface. The images were visualized on a CCD camera and were recorded by means of a video recorder. The lateral resolution of BAM was about 2 µm. The BAM images were obtained during the compression process and recorded along the isotherm.

3. Results and Discussion 3.1. Adsorption of GOx at the Air/Liquid Interface. Proteins and enzyme are known to adsorb spontaneously onto an air/ liquid interface.11 The adsorption behaviors of GOx with and without the presence of a DPPA monolayer were studied first by measuring the variation of surface pressure with adsorption time, and the results are shown in Figure 1. In the absence of a DPPA monolayer (curve a), the surface pressure increases slowly with adsorption time and approaches a constant value (ca. 7.5 mN/m) after about 120 min. With a further increase of adsorption time, above 24 h, no significant change in surface

2044 Langmuir, Vol. 23, No. 4, 2007

Lee et al.

Figure 2. BAM images of the GOx adsorption monolayer at the air/water interface. The adsorption time and surface pressure of the images were (a) 45 min, π ) 3.5 mN/m, (b) 80 min, π ) 6.2 mN/m, and (c) 180 min, π ) 7.5 mN/m. The length bar corresponds to 50 µm.

Figure 3. Evolution of BAM images along the adsorption process of GOx in the presence of a DPPA monolayer. The molecular area of DPPA was controlled at 112 Å2/molecule. The adsorption time and surface pressure of the images were (a) 20 min, π ) 7 mN/m, (b) 40 min, π ) 7.5 mN/m, (c) 2 h, π ) 8.0 mN/m, (d) 3.5 h, π ) 10 mN/m, (e) 5 h, π ) 14 mN/m, and (f) 8 h, π ) 16.5 mN/m. The length bar corresponds to 50 µm.

Adsorption BehaVior of GOx on a DPPA Monolayer

pressure was observed. Similar results are obtained in a pH 3 subphase adjusted by HCl and in a pH 7 phosphate buffer solution. The result obtained here is contrary to that reported by Rosilio et al.19 In their work, the adsorption of GOx was performed by injection of the GOx solution into a subphase (pH 7 buffer solution) without stirring. No significant elevation of surface pressure was reported for GOx concentrations below 4 µg/mL, probably due to the slow diffusion rate of GOx in the subphase without proper mixing. Different adsorption behaviors are expected for the different experimental conditions. In the presence of a DPPA monolayer, the adsorption rate of GOx is dependent on the DPPA density at the air/liquid interface. Zhang et al. studied the lipid density dependence of GOx adsorption by compressing the lipid monolayer first to a target surface pressure followed by an adsorption process.20 In their method, the lipid density effect was not accounted for in the evaporation and compression periods where GOx adsorption may be significant. To take into consideration the GOx adsorption in the solvent evaporation period, various amounts of DPPA stock solutions (25, 50, or 75 µL) were spread at the air/liquid interface to prepare lipid monolayers with various initial densities. The molecular areas equivalent to the spreading volumes mentioned above are 336, 168, or 112 Å2/molecule, respectively. In the presence of a DPPA monolayer, the adsorption rate of GOx is greatly enhanced and the adsorption rate is proportional to the DPPA density. The surface pressure increases steadily with the adsorption time and approaches a constant value (6-8 mN/m) in tens of minutes. This constant-pressure state is termed the first equilibrium stage of GOx adsorption, and the time required to reach this state increases with decreasing density of the monolayer. The first equilibrium stage lasts for several hours (2-6 h), and the surface pressure increases again when the adsorption time is further prolonged. Finally, a second equilibrium surface pressure (ca. 14-18 mN/m) is approached. Apparently, the interaction between DPPA and GOx leads to a fast adsorption rate and a higher adsorption amount of GOx. However, an energy barrier existed for the second adsorption stage to proceed. Such phenomena will be discussed in a later section. It is noteworthy that a slight deviation of the equilibrium surface pressure, as well as the time required to reach the equilibrium stage, may be found between different runs. Such deviation is more obvious in the second equilibrium stage where the experimental time is long. However, the reproducibility of the experiment is sufficient enough to distinguish the effect of the DPPA density on the adsorption rate of GOx. The evolution of the surface morphology at the air/liquid interface along the GOx adsorption process was imaged using Brewster angle microscopy. In the absence of a DPPA monolayer, the surface morphology does not change significantly in the early adsorption stage. No specific surface structure can be identified in the image shown in Figure 2a, which was acquired after 45 min (π ≈ 3.5 mN/m). With increasing surface pressure, the image becomes progressively whiter as demonstrated by parts b and c of Figure 2, imaged after 80 min (π ≈ 6.2 mN/m) and 3 h (π ≈ 7.5 mN/m), respectively. The variation of the BAM images indicates an increasing amount of GOx adsorbed at the air/liquid interface. However, the small gray level (GL) increment, from 19 (Figure 2a) to 23 (Figure 2c), implies a limited amount of GOx adsorption in the absence of DPPA. When DPPA molecules are presented with a molecular area of 112 Å2/molecule (75 µL of stock solution was spread), clear adsorption phases can be observed in the early adsorption stage. A typical image obtained near the first equilibrium stage (20 min, π ≈ 7 mN/m, GL ) 20) is shown in Figure 3a. A cloudy-

Langmuir, Vol. 23, No. 4, 2007 2045

Figure 4. Pressure-area isotherms of the GOx adsorption monolayer (a), pure DPPA monolayer (b), and mixed DPPA/GOx monolayers compressed after the first (c) or second (d) equilibrium stage of GOx adsorption was reached.

Figure 5. Area relaxation of the GOx adsorption monolayer (a), pure DPPA monolayer (b), and mixed DPPA/GOx monolayers compressed after the first (c) or second (d) equilibrium stage of GOx adsorption was reached. The relaxation was performed at π ) 30 mN/m.

like morphology appears at about 40 min (Figure 3b, π ≈ 7.5 mN/m, GL ) 22) and gradually changes by increasing the coverage and density of the adsorption phase. Figure 3c shows a typical image obtained at the end of the first equilibrium stage (2h, π ≈ 8.0 mN/m, GL ) 24). During the first equilibrium stage, the slight increases in surface pressure and GL value of the BAM image indicate that the GOx adsorption performs very slowly in this period. After the first equilibrium stage, a more condensed phase gradually forms accompanied by a moderate increase in surface pressure. Parts d (GL ) 46) to f (GL ) 71) of Figure 3 demonstrate the evolution of the monolayer morphology between the first and second equilibrium stages. The significant increase of the gray level indicates the formation of a highly condensed monolayer, attributable to the higher incorporation of GOx in this period. According to the results mentioned above, the adsorption enhancement of GOx onto the air/liquid interface by the presence of a DPPA monolayer is initially confirmed by the rapid increase of the surface pressure in the initial adsorption stage. By prolonging the adsorption process, an additional amount of GOx may incorporate into the DPPA monolayer, which leads to an increase of the surface pressure at the later stage. Apparently, there are two equilibrium states of GOx adsorption when a DPPA

2046 Langmuir, Vol. 23, No. 4, 2007

Lee et al.

Figure 6. Hysteresis curves show the compression-reexpansion behaviors of the pure DPPA monolayer (a), pure GOx adsorption monolayer (b), and mixed DPPA/GOx monolayers compressed after the first (c) or second (d) equilibrium stage of GOx adsorption was reached.

monolayer is present. The effect of the two states on the behavior of mixed DPPA/GOx monolayers was also studied in this work. In the following experiments, a DPPA monolayer with an initial molecular area of 112 Å2/molecule is controlled. The monolayer compression starts at 30 min and 8 h, respectively, after the spreading of DPPA to control the state of the first and second equilibrium stages. 3.2. Surface Pressure-Area per Molecule Isotherm. The π-A isotherms of a DPPA monolayer in an aqueous solution with or without the presence of GOx are demonstrated in Figure 4. For the pure DPPA monolayer, a steep rise of the isotherm suggests that a condensed phase was formed. The lift-off area and collapse pressure of the pure DPPA monolayer are about 56 Å2/molecule and 53 mN/m, respectively. In the presence of GOx, the initial surface pressures of the monolayer are about 7 and 16 mN/m, respectively, for the first and second equilibrium stages. The lift-off area of the isotherm extends to about 83 Å2/molecule for the first equilibrium stage, and the surface pressure increases slowly in the later compression stage. For the second stage, the surface pressure elevates at the initial compression stage. Obviously, the monolayer becomes more extended and compressible when GOx is incorporated into the monolayer. The adsorption monolayer of GOx, in the absence of DPPA, was also compressed, and the variation of the surface pressure was recorded as a function of the trough area. For comparison with the isotherms in the presence of DPPA, the relative area is used for the adsorption monolayer, and the isotherm is also shown in Figure 4. The surface pressure of the GOx monolayer is nearly constant before a 50% area decrease and increases slowly when the surface is further compressed. The surface pressure of the GOx monolayer increases to about 40 mN/m when the surface area is compressed to 10% of its initial area.

For the first equilibrium stage, the collapse pressure of the DPPA/GOx mixed monolayer is about 51 mN/m and the collapse occurs at an area (40 Å2/molecule) smaller than the limiting area of the DPPA monolayer (52 Å2/molecule). For the second stage, no obvious collapse point was found. These phenomena indicate that the regularity of DPPA molecules decreases when GOx molecules are incorporated. It is inferred that there are GOx molecules at the air/liquid interface, especially for the second equilibrium stage, which prevent the intimate contact of DPPA molecules during compression. For the first equilibrium stage (curve c), the plateau region after collapse is similar to the characteristic of a pure DPPA monolayer, implying similar collapse structures between the two monolayers. This result is consistent with the previous interpretation that fewer GOx molecules are adsorbed in this stage. On the other hand, the plateau region did not appear on the isotherm of the second equilibrium stage, a characteristic similar to that for the GOx adsorption monolayer (curve a). This result supports the previous inference that there is a significant amount of GOx at the air/ liquid interface, which prevents the close packing of DPPA molecules during compression. 3.3. Relaxation and Hysteresis. The stability of the monolayer was evaluated by the area relaxation of the monolayer at constant surface pressure. The results shown in Figure 5 demonstrate that the area loss of the DPPA monolayer after 3 h of relaxation is less than 5%, being the most stable of the monolayers studied here. On the contrary, the area loss of the pure GOx monolayer after 3 h of relaxation is about 20%, attributable mainly to the desorption of GOx under the applied pressure. For the mixed DPPA/GOx monolayers, the area losses are about 6-7%, ranging between those of DPPA and GOx monolayers. Apparently, the

Adsorption BehaVior of GOx on a DPPA Monolayer

binding effect between DPPA and GOx can enhance the stability of the GOx in the monolayer. The hysteresis characteristics of the monolayers are shown in Figure 6. For the DPPA monolayer (Figure 6a), the expansion curve closely follows the compression curve and the curves are nearly identical for three successive hysteresis cycles. Apparently, the DPPA monolayer has a high reexpansion characteristic with very small hysteresis. For the pure GOx monolayer shown in Figure 6b, the expansion curve is located close to the compression curves when the pressure decreases from 30 mN/m to the equilibrium pressure before compression (7 mN/m). However, the surface pressure keeps decreasing with a further increase of the surface area and approaches a zero surface pressure when the area expands to the initial value. In the second cycle, the surface pressure elevates at the initial compression stage and approaches a plateau region when the equilibrium surface pressure is obtained. The following compression-expansion curves of the second cycle are nearly identical to that of the first cycle, and so is that of the third one. In the first compression cycle of the pure GOx monolayer (Figure 6b), the compression causes an increase of the surface concentration of GOx, and therefore, GOx will desorb simultaneously to maintain the equilibrium surface pressure. Since the compression rate of the barrier is constant (10 mm/min), the area decreasing ratio (defined as ∆A/A) increases with a decrease of the surface area. That is, the concentration of GOx increases slower in the initial compression stage, and the equilibrium state can be maintained by the desorption of GOx. At the later compression stage, the GOx concentration increases more quickly and the desorption rate of GOx is not fast enough to compensate for the density increase of GOx. Consequently, the surface pressure increases with compression. The closeness of the curves of compression and expansion in the early expansion stage (from π ) 30 mN/m to π ) 7 mN/m) indicates that the desorbed amount of GOx in this period is insignificant. However, the GOx loss in the entire compression period is meaningful as indicated by the steady decrease of the surface pressure below the equilibrium value in the expansion stage. In the later stage of the first expansion cycle, GOx cannot adsorb fast enough to compensate the desorption loss during compression; therefore, the surface pressure decreases with increasing surface area, approaching a value as low as zero. In the second and third compression cycles, the initial increase of the surface pressure is a result of the density increase of the preexisting GOx molecules due to compression and the simultaneous adsorption of GOx since the surface pressure is lower than the equilibrium value. Once equilibrium is approached, the compression and desorption behavior is similar to that of the first cycle. The hysteresis behavior of mixed DPPA/GOx monolayers is demonstrated in parts c and d of Figure 6 for, respectively, the first and second equilibrium stages of GOx adsorption. For the first equilibrium stage, obvious hysteresis occurs when the monolayer expands from π ) 40 mN/m to π ) 7 mN/m. The high hysteresis may be attributed to the desorbed GOx (and/or the GOx/DPPA complex) in the compression stage, which cannot readsorb quickly in the early expansion stage where the surface pressure is high. In the later expansion stage, the surface pressure approaches the equilibrium value and does not decrease further with surface expansion. This phenomenon is caused by the fast adsorption of GOx in this period due to the presence of DPPA. For the threee successive cycles studied here, it is interesting to find that the curves slightly shift right cycle by cycle, implying that more GOx molecules are incorporated into the monolayer in the later cycle. This result also indicates that the GOx adsorption

Langmuir, Vol. 23, No. 4, 2007 2047

Figure 7. BAM photographs of the pure DPPA monolayer imaged during the compression of the π-A isotherm. The images were taken at (a) A ) 92.1 Å2/molecule (π ) 0 mN/m), (b) π ) 30 mN/m, and (c) π ) 52 mN/m. The length bar corresponds to 50 µm.

proceeds steadily during the hysteresis experiment since the second equilibrium stage has not yet been reached. When the hysteresis experiment was performed after the second equilibrium stage of GOx adsorption was reached, a significant hysteresis of the mixed monolayer was also found (Figure 6d). Compared with the result of the first equilibrium stage, two different phenomena were observed for the second equilibrium stage (Figure 6d). The first is that, after a hysteresis cycle, the surface pressure cannot approach the initial value before compression, and the second is the left shift of the successive hysteresis curves. Both facts are attributed to the loss of GOx (and/or the DPPA/GOx complex) from the monolayer due to the desorption effect in the compression-expansion cycle. The adsorption experiments shown in Figure 1 indicate that a long time is required to reach the second equilibrium stage, which also implies a slow adsorption rate in this period. That is, the

2048 Langmuir, Vol. 23, No. 4, 2007

Lee et al.

Figure 8. BAM images of the mixed DPPA/GOx monolayer compressed after the first equilibrium stage of GOx adsorption was reached. The images were taken just before compression at π ) 7 (a) and after compression to various surface pressures of 8 (b), 16 (c), 22 (d), 43 (e), and 50 (f) mN/m . The length bar corresponds to 50 µm.

desorbed GOx molecules in the compression stage cannot readsorb fast enough to retain the second equilibrium state before compression. 3.4. Monolayers Observed using Brewster Angle Microscopy. The BAM images of a pure DPPA monolayer along the π-A isotherm are shown in Figure 7. At the states before the lift-off point of the isotherm, coexisting gas and liquid phases were observed, as shown in Figure 7a (92.1 Å2/molecule). Upon compression, the gas phase disappears gradually and a homogeneously condensed phase forms at a surface pressure of ca. 5-10 mN/m. This phase does not change in the following compression process before the collapse of the monolayer, and a typical image taken at π ) 30 mN/m is shown in Figure 7b. When the surface pressure increases near the collapse point, striped structures appear as demonstrated in Figure 7c. Such collapse structures are always found on a monolayer with a rigid characteristic,24 and their appearance implies condensed and rigid properties of the DPPA monolayer. Figure 8 shows the phase evolution of a mixed DPPA/GOx monolayer compressed after the first equilibrium stage of GOx

adsorption is reached. Before compression of the monolayer, dotlike aggregations of low contrast were observed in the monolayer. Such structures are identified as the phase of the DPPA/GOx complex since it was not found on both pure DPPA and GOx monolayers. In the initial compression stage, the dotlike domains coalesce gradually and bright zones of larger size and low contrast appear (Figure 8b, π ≈ 8 mN/m). The brightness of the image (the GL) increases significantly after a significant elevation of surface pressure, and a typical image acquired at π ≈ 16 mN/m is shown in Figure 8c. At least two kinds of domains can be distinguished in this image. The first one consists of densely distributed bright dots (the central region in Figure 8c), and the second exhibits a more uniform morphology without an obvious aggregative structure. The different phases are caused by heterogeneity of DPPA monolayers in the initial spreading stage. Since gas and liquid expanded phases coexist before compression (as shown in Figure 7a), the heterogeneity will lead to different adsorption rates and compositions between various phases and is responsible for the nonuniform adsorption of GOx on the DPPA monolayer.

Adsorption BehaVior of GOx on a DPPA Monolayer

Langmuir, Vol. 23, No. 4, 2007 2049

Figure 9. BAM images of the mixed DPPA/GOx monolayer compressed after the second equilibrium stage of GOx adsorption was reached. The images were taken after compression to surface pressures of 18.5 (a), 22 (b), 25 (c), 33 (d), 40 (e), and (f) 45 mN/m. The length bar corresponds to 50 µm.

After further compression, the dotlike aggregations disappear and, instead, uniform phases with different brightnesses appear as shown in Figure 8d. The contrast between various phases decreases gradually with a steady increase of the surface pressure, and a uniformly smooth phase appears at about 30 mN/m (not shown here). With a further elevation of surface pressure, the smooth surface becomes rougher and protruded blotches appear as shown on the left side of Figure 8e. Occasionally, striped aggregates can also be seen. The significant collapse phase of the mixed monolayer appears at about 50 mN/m and appears as striped structures (Figure 8f). The appearance of this collapse phase, similar to that of DPPA, indicates that DPPA molecules are still packed closely, probably with GOx adsorbed beneath the DPPA monolayer. A different characteristic of the mixed DPPA/GOx collapse structure, compared with that of pure DPPA, is that the bright stripe is lumpy and seems to be composed of successive dots. The adsorption of GOx to the DPPA monolayer is also responsible for the formation of such a collapse structure.

For the second equilibrium stage of GOx adsorption, the BAM images did not change a lot in the initial compression stage. The image acquired at π ) 18.5 mN/m (Figure 9a) shows the existence of two distinct phases. The first phase consists of densely distributed fine dots, and the other exhibits a more smooth morphology. Apparently, the distinct phases are a result of different compositions caused by the heterogeneous state of GOx adsorption. Upon compression, a more uniform phase forms due to the shrinkage of area between aggregates (Figure 9b). A fully extended condensed phase appears at about π ) 25 mN/m (Figure 9c) on which no separated dot is observed. The following compression leads to the reappearance of morphological heterogeneity. The aggregative structures grow rapidly in this period, and the evolution is shown in Figure 9d-f. Compared with the result obtained after the first equilibrium adsorption of GOx, a more heterogeneous and rough monolayer is formed due to the incorporation of more GOx molecules. 3.5. AFM Investigation. BAM can only observe the morphology on a micrometer scale. To examine the finer structure

2050 Langmuir, Vol. 23, No. 4, 2007

Lee et al.

Figure 10. AFM images (1 × 1 µm2) of mixed GOx/DPPA LB films transferred onto mica. The LB films were prepared from monolayers compressed after the first (a, b) and second (c, d) equilibrium stages of GOx adsorption were reached. The LB films were transferred at molecular areas of 80 (a, c) and 58 (b, d) Å2/DPPA molecule.

of the mixed monolayer, the monolayer was transferred onto a mica substrate and analyzed by AFM. The effect of various adsorption stages was studied by comparing the mixed LB films transferred at the same area (80 or 58 Å2 per DPPA molecule), and the results are shown in Figure 10. For an area of 80 Å2/ DPPA molecule, the AFM image obtained for the first equilibrium stage (Figure 10a) has a uniform phase with outstanding dots positioned on the LB film. It is inferred that outstanding aggregations are attributed to the GOx molecules that have a larger molecular volume compared with DPPA. At a low surface pressure (π ) 11 mN/m), the GOx density is low as inferred from the lower population of distributed dots shown in Figure 10a. When the monolayer is compressed to π ) 30 mN/m (58 Å2/DPPA molecule), the image shown in Figure 10b demonstrates that the monolayer is still uniform but its morphology becomes rougher (the rms roughness increases from 0.34 to 0.42 nm) due to the increase of the GOx density. For the second equilibrium stage of GOx adsorption, the AFM morphology of the LB film, transferred at 80 Å2/DPPA molecule (π ) 30 mN/m, Figure 10c), demonstrates the existence of two distinct phases. The first one exhibits a uniform and highly condensed structure with an rms roughness of 0.41 nm. The second phase has a rougher and expanded structure formed by the aggregation of separated dots, and the rms roughness is estimated to be 0.69 nm. When the pressure elevates to 45 mN/ m, the expanded phase cannot transfer to the condensed one and both phases become rougher (Figure 10d). The rms roughnesses for the condensed and expanded phases are 1.15 and 1.06 nm, respectively. The overlap of the condensed phases due to overcompression, as well as the desorpion of GOx at high surface pressure, is responsible for the formation of the coarser

morphology. The evolution of the two phases during compression also indicates that the two phases have different compositions. The more condensed phase is considered to be composed of a higher composition of DPPA since DPPA can help to incorporate GOx into the mixed monolayer and form a condensed phase upon compression. On the contrary, the expanded phase should be DPPA-poor or DPPA-free. Due to the large molecular volume of GOx and the desorption effect at a high surface pressure, a condensed phase is not easily reached without the presence of a sufficient amount of DPPA. 3.6. Proposed Model for the DPPA/GOx Mixed Monolayer. According to the results described above, a model illustrating the evolution of the mixed DPPA/GOx monolayer at various compressed states is proposed and shown in Figure 11. Since the presence of DPPA can enhance the adsorption of GOx, GOx molecules should be mainly beneath the DPPA monolayer in the initial adsorption stage. It is possible that GOx has a chance to move to the air/liquid interface in the first equilibrium stage (Figure 11a-1). However, this movement should have a highenergy barrier because the dispersion interaction between proteins and the air/water interface is always considered to be repulsive29 and may be why a long time is required to approach the second equilibrium stage. Upon compression, the desorption of GOx does not occur at a low surface pressure, but some GOx present at the air/liquid interface may be pushed back beneath the interface (Figure 11a-2). A uniform morphology (Figure 11a-3) is obtained at about π ) 30 mN/m as deduced from the BAM image. As the area decreases further, the mixed monolayer is also compressible because the molecular area of DPPA at π ) 30 (29) Sengupta, T.; Damodaran, S. Langmuir 1998, 14, 6457-6469.

Adsorption BehaVior of GOx on a DPPA Monolayer

Langmuir, Vol. 23, No. 4, 2007 2051

Figure 11. Schematic model illustrating the evolution of the mixed DPPA/GOx monolayer at different compression states. The monolayer was compressed after the first (a) or second (b) equilibrium stage of GOx adsorption was reached.

mN/m (ca. 58 Å2) is still larger than the occupied area of a pure DPPA molecule in a closely packed arrangement (ca. 48-50 Å2). However, GOx may be expelled from the monolayer or aggregates to form a heterogeneous phase. The state where the DPPA molecules are closely packed is illustrated in Figure 11a4, which is equivalent to the BAM image shown in Figure 8e. Further compression causes the collapse of the DPPA monolayer. Due to the rigid characteristic of the DPPA monolayer, the monolayer was pushed upward, leading to a structure of a folded stripe. The aggregation and desorption of GOx may occur simultaneously after the collapse of the monolayer (Figure 11a5). For the second equilibrium stage, more GOx molecules are adsorbed and have the chance to transfer up to the air/liquid interface. Two routes are proposed for such a movement. The first is the transferring of the adsorbed GOx beneath the DPPA monolayer, and the second is a direct adsorption without the help of DPPA. As shown in Figure 11b-1, the two mechanisms lead to DPPA-rich (left region) and DPPA-poor (right region) phases, respectively. In the initial compression stage, both GOx and DPPA concentrations are increased. Because the DPPA-rich domain is more concentrated in both DPPA and GOx, the compression will cause a fast increase in the monolayer thickness, leading to a higher heterogeneity and a rougher morphology of the monolayer (left region in Figure 11b-2). On the other hand, since the incorporation effect of DPPA to GOx is weaker in the DPPA-poor region, such a domain should contain a lower amount of GOx, and thus, a thinner and smoother morphology will be observed (right region in Figure 11b-2). The two domains are equivalent to the two phases observed in BAM images (Figure 9a,b). When the surface pressure increases further to 30 mN/m, a more condensed and homogeneous phase is formed (Figure

11b-3). At the later compression stage, both aggregation and desorption of GOx perform simultaneously and a monolayer with higher heterogeneity results (Figure 11b-4). The arrangement of DPPA molecules is irregular in the monolayer of the second equilibrium stage, and thus, no apparent collapse point can be found from the isotherm.

Conclusion The present study demonstrates that the adsorption rate of GOx onto an air/liquid interface is significantly enhanced by the presence of a DPPA monolayer. An equilibrium adsorption stage with a surface pressure of ca. 7.5 mN/m can be obtained within 20-120 min regardless of the presence of DPPA. However, a second equilibrium stage with a higher surface pressure (ca. 16 mN/m) can further be reached after several hours only when the DPPA monolayer is present. BAM and AFM analysis indicate that, after the second adsorption stage is reached, a larger amount of GOx molecules are incorporated into the monolayer, which is also responsible for the increase of the surface pressure at this period. Due to the different adsorption amounts of GOx in the two equilibrium stages, the mixed DPPA/GOx monolayers compressed after the two adsorption stages are reached exhibit distinct characteristics in relaxation, hysteresis, BAM image, and LB film morphology. Acknowledgment. The support of this research by the National Science Council of Taiwan through Grant No. NSC 94-2214-E-006-022 is gratefully acknowledged. LA062501P