Selectively Erasable Multilayer Thin Film by Photoinduced

Mar 12, 2010 - A multilayer film consisting of poly(acrylic acid) (PAA) and an azobenzene-containing surfactant (AzoTEA) was fabricated via a layer-by...
3 downloads 0 Views 2MB Size
Download PDF
pubs.acs.org/Langmuir © 2010 American Chemical Society

Selectively Erasable Multilayer Thin Film by Photoinduced Disassembly Yapei Wang, Peng Han, Guanglu Wu, Huaping Xu, Zhiqiang Wang, and Xi Zhang* Key Lab of Organic Optoelectronics & Molecular Engineering, Department of Chemistry, Tsinghua University, Beijing 100084, PR China Received January 31, 2010. Revised Manuscript Received March 3, 2010 A multilayer film consisting of poly(acrylic acid) (PAA) and an azobenzene-containing surfactant (AzoTEA) was fabricated via a layer-by-layer assembly technique. The micellar structure favored by AzoTEA while in solution results in a bilayer structure when deposited on a substrate surface. This aggregation conversion behavior favors the deposition of AzoTEA and PAA in an alternating pattern to form a photoresponsive multilayer film. The molecular amphiphilicity of AzoTEA can be tuned by photoisomerization of the azobenzene moiety, which affects the aggregation behavior of AzoTEA in both films and solutions. The disassembly of AzoTEA aggregates caused by photoirradiation can induce the disassembly of the whole multilayer film. The AzoTEA-PAA multilayer films were found to be stable in pH 4 acetic acid (AcOH) solution, unless treated with UV radiation. On the basis of the different stability of multilayers with or without photoirradiation, when the multilayer films are selectively irradiated with UV light, the regions exposed to UV radiation disassembled after being immersed in pH 4 AcOH solution for 10 min but the regions not exposed to photoradiation are maintained on the substrate. Moreover, the plausible mechanism for the assembly and disassembly of these multilayer films and the confirmation of erasable films by atomic force microscopy are discussed.

Introduction Layer-by-layer (LbL) assembly is a well-established technique for preparing multilayer thin films on the surface of a wide variety of substrates. This versatile technique does not require any elaborate instrumentation and is independent of substrate shape, making it promising for the fabrication of thin films with integrated functions. Although Decher pioneered the technique by alternately depositing a pair of oppositely charged polymers on a substrate,1 several other systems have been explored for multilayer fabrication on the basis of different driving forces (e.g., hydrogen bonding,2 host-guest,3 and charge-transfer4 interactions have been exploited for LbL deposition on substrates5). Lacking strong covalent bonds, films formed by LbL assembly can be disassembled by interrupting the weaker supramolecular interactions between the layers. Such an interruption has been considered to be applied to selectively erase multilayer LbL films in a controlled fashion. This family of erasable thin films is becoming increasingly more popular because of its potential applications in new functional surface coatings. Granick gave the first example of an erasable multilayer film by disrupting intercalated hydrogen bonds through the manipulation of pH or ionic strength.6 We developed a similar approach, using a base to extract the acidic component selectively from a multilayer film.7 These base-induced porous films have acted as ink reservoirs for

the surface patterning of proteins8 and are expected to be applied in gas separation, selective ion-permeation membranes, and so on. Other than interrupting the interlayer interactions, the approach by using external stimuli to switch the multilayer LbL films has extensively become attractive. For example, LbL films consisting of degradable polyelectrolytes were found to disassemble gradually along with the degradation of the polyelectrolytes.9 Beyond these ways to adjust multilayer films, the wide development of “smart” stimuli-responsive materials has inspired the development of multilayer films that are responsive to external stimuli. For example, by using a thermally responsive polymer as a building block, multilayer films were developed that responded to changes in temperature.10 Hammond et al. fabricated an electroactive multilayer film that was disassembled by applied voltage, allowing the controlled release of loaded cargo.11 The use of light to control the disassembly of multilayer films is desirable because the process does not produce additional chemicals and the reactions are typically clean and rapid. Etching of electrostatic-based multilayer films by light irradiation is attracting interest because of the capability of regioselective erasing and partial decomposition. This type of erasable multilayer films is anticipated for use as either a carrier for controllable cargo release or as a template to make patterned arrays or porous films for selective separation and molecular adsorption. Selectively photocuring multilayer films can add stability while allowing the un-cross-linked electrostatic-based parts to be removed.12

*Corresponding author. E-mail: [email protected]. (1) (a) Decher, G.; Hong, J. D. Makromol. Chem., Macromol. Symp. 1991, 46, 321. (b) Decher, G. Science 1997, 277, 1232. (2) (a) Wang, L. Y.; Wang, Z. Q.; Zhang, X.; Shen, J. C. Macromol. Rapid Commun. 1997, 18, 509. (b) Stockton, W. B.; Rubner, M. F. Macromolecules 1997, 30, 2717. (3) (a) Suzuki, I.; Egawa, Y.; Mizukawa, Y.; Hoshi, T. Chem. Commun. 2002, 164. (b) Crespo-Biel, O.; Dordi, B.; Reinhoudt, D. N.; Huskens, J. J. Am. Chem. Soc. 2005, 127, 7594. (4) Shimazaki, Y.; Mitsuishi, M.; Ito, S.; Yamamoto, M. Langmuir 1997, 13, 1385. (5) Zhang, X.; Chen, H.; Zhang, H. Chem. Commun. 2007, 1395. (6) (a) Sukhishvili, S. A.; Granick, S. J. Am. Chem. Soc. 2000, 122, 9550. (b) Sukhishvili, S. A.; Granick, S. Macromolecules 2002, 35, 301. (7) (a) Fu, Y.; Bai, S. L.; Cui, S. X.; Qiu, D. L.; Wang, Z. Q.; Zhang, X. Macromolecules 2002, 35, 9451. (b) Zhang, H. Y.; Fu, Y.; Wang, D.; Wang, L. Y.; Wang, Z. Q.; Zhang, X. Langmuir 2003, 19, 8497.

9736 DOI: 10.1021/la1004648

(8) (a) Xu, H.; Gomez-Casado, A.; Liu, Z.; Reinhoudt, D. N.; Lammertink, R. G. H.; Huskens, J. Langmuir 2009, 25, 13972. (b) Wu, C.; Xu, H.; Otto, C.; Reinhoudt, D. N.; Lammertink, R. G. H.; Huskens, J.; Subramaniam, V.; Velders, A. H. J. Am. Chem. Soc. 2009, 131, 7526. (9) (a) Vazquez, E.; Dewitt, D. M.; Hammond, P. T.; Lynn, D. M. J. Am. Chem. Soc. 2002, 124, 13992. (b) Smith, R. C.; Riollano, M.; Leung, A.; Hammond, P. T. Angew. Chem., Int. Ed. 2009, 48, 8974. (10) (a) Kharlampieva, E.; Kozovskaya, V.; Tyutina, J.; Sukhishvili, S. A. Macromoelcules 2005, 38, 10523. (b) Zhu, Z.; Sukhishvili, S. A. ASC Nano 2009, 3, 3595. (11) Wood, K. C.; Zacharia, N. S.; Schmidt, D. J.; Wrightman, S. N.; Andaya, B. J.; Hammond, P. T. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 2280. (12) (a) Shi, F.; Dong, B.; Qiu, D.; Sun, J.; Wu, T.; Zhang, X. Adv. Mater. 2002, 14, 805. (b) Shi, F.; Wang, Z. Q.; Zhao, N.; Zhang, X. Langmuir 2005, 21, 1599. (c) Wu, G.; Shi, F.; Wang, Z.; Liu, Z.; Zhang, X. Langmuir 2009, 25, 2949.

Published on Web 03/12/2010

Langmuir 2010, 26(12), 9736–9741

Wang et al.

Article

Figure 1. Synthesis routes for AzoTEA.

Recently, micellar or vesicle-like aggregates formed by amphiphilic molecules have emerged as new candidates for LbL film fabrication because of their utility as nanocontainers in encapsulating functional cargo.13 Considering that the self-assembly and disassembly of some amphiphiles can be tuned by changing the molecular amphiphilicity through external stimuli,14 a multilayer film composed of amphiphilic molecules can be decomposed by a change in molecular amphiphilicity. Herein, we report the development of a new concept combining the controlled selfassembly of amphiphiles within LbL thin films that are selectively erased via photoinduced disassembly. To this end, we designed and synthesized a photoresponsive azobenzene-containing surfactant, the aggregate of which is used as a polyvalent, positively charged building block to deposit alternatively with negatively charged poly(acrylic acid) on a solid surface. Similarly, in aqueous solution, the surfactant-formed aggregates in these multilayer films became unstable with a photoinduced amphiphilicity change. The photoinduced disassembly of the aggregates further led to the destruction of the multilayer film because the electrostatic equilibrium between the aggregates and the polyelectrolytes was broken. By permitting the light to pass through a mask for selective irradiation of the film, the multilayer film was found to be regioselectively erasable in providing a desired pattern.

Experimental Section Materials. Poly(sodium styrenesulfonate) (PSS, Mw = 70 000), poly(acrylic acid) (PAA, Mw = 2000), and 3-amimopropyltrimethoxysilane (APTS) were purchased from Acros Organics. 18-Crown-6 was obtained from Aldrich. All other reagents and solvents were purchased from Beijing Chemical Reagent Company, China. Triethylamine was dried with NaOH before use. Synthesis. 4,40 -Dihydroxylazobenzene was synthesized according to the literature.15 The photoresponsive azobenzene-containing surfactant (AzoTEA) was synthesized according to the synthesis route shown in Figure 1. 4,40 -Dihydroxylazobenzene (I) (0.500 g, 2.34 mmol) and n-butylbromide (0.317 g, 2.33 mmol) were dissolved in 150 mL of acetone. Potassium carbonate (1.293 g, 9.37 mmol) was subsequently added as the weak base in the presence of the phase-transfer reagent, 18-crown-6. After being refluxed for 48 h under N2, one hydroxyl group of 4,40 -dihydroxylazobenzene was alkylated with (13) (a) Ma, N.; Zhang, H. Y.; Song, B.; Wang, Z. Q.; Zhang, X. Chem. Mater. 2005, 17, 5065. (b) Ma, N.; Wang, Y.; Wang, Z.; Zhang, X. Langmuir 2006, 22, 3906. (c) Ma, N.; Wang, Y.; Wang, B.; Wang, Z.; Zhang, X.; Wang, G.; Zhao, Y. Langmuir 2007, 23, 2874. (d) Jiang, J. Q.; Tong, X.; Morris, D.; Zhao, Y. Macromolecules 2006, 39, 4633. (e) Cho, J.; Hong, J.; Char, K.; Caruso, F. J. Am. Chem. Soc. 2006, 128, 9935. (f) Michel, M.; Vautier, D.; Voegel, J.-C.; Schaaf, P.; Ball, V. Langmuir 2004, 20, 4835. (g) Liu, X.; Zhou, L.; Geng, W.; Sun, J. Langmuir 2008, 24, 12986. (h) Manna, U.; Patil, S. J. Phys. Chem. B 2008, 112, 13258. (i) Zhao, Y.; Bertrand, J.; Tong, X.; Zhao, Y. Langmuir 2009, 25, 13151. (14) (a) Wang, Y.; Xu, H.; Zhang, X. Adv. Mater. 2009, 21, 2849. (b) Wang, Y.; Han, P.; Xu, H.; Wang, Z.; Zhang, X.; Kabanov, A. V. Langmuir 2010, 26, 709. (15) Wei, W.; Tomohiro, T.; Kodaka, M.; Okuno, H. J. Org. Chem. 2000, 65, 8979.

Langmuir 2010, 26(12), 9736–9741

an n-butyl chain. Then excess 1,4-dibromobutane (2 mL, 16.74 mmol) and another 0.600 g of potassium carbonate were added to convert the other hydroxyl group to a bromobutyloxyl chain, yielding II. Excessive triethylamine (0.5 mL) was added to a II (0.130 g, 0.32 mmol) tetrahydrofuran/acetonitrile (1:1 v/v) solution. The mixture continued to stir for 12 h at 60 °C under Ar. After evaporation of the solvent, the yellow solid was dissolved in tetrahydrofuran. The concentrated solution was added dropwise to a large volume of stirring petroleum ether, and the resulting yellow precipitate was filtered and dried in a vacuum oven. The characterization of AzoTEA was as follows: 1H NMR (300 MHz, CDCl3): δ 7.92 (d, 4H, J = 8.94 Hz); 7.01 (d, 4H, J = 8.58 Hz); 4.17-4.03 (m, 4H); 3.55 (m, 8H); 2.10 (m, 8H); 1.81-1.39 (m, 9H); 1.02 (t, 3H, J = 7.40 Hz). ESI-MS: (except bromide anion) calculated 426.61, found 426.33. Instrumentation. 1H NMR spectra were recorded using a JEOL JNM-ECA300 spectrometer. Electrospray ionization mass spectrometry (ESI-MS) was performed using a PE SciexAPI 3000 spectrometer. UV/vis spectra were obtained using a Hitachi U-3010. A high-pressure mercury lamp with an optical fiber and an intensity of 900 mW/cm2 was used as the radiation light source for the photoisomerization of azobenzene. Band-pass filters with wavelengths of 365 ( 10 and 450 ( 10 nm were used to produce UV light (365 nm) and visible light (450 nm). To avoid the heat generated by light irradiation, the optical fiber was kept 3 cm away from the sample. To irradiate the surface of the films selectively, UV light was permitted to pass through a mask that blocked parts of the film surface when irradiated. Transmission electron microscopy (TEM) was performed on a JEMO 2010 electron microscope, operating at an acceleration voltage of 110 kV. The samples were prepared by drop-coating the aqueous solution onto a carbon-coated copper grid and staining with 0.2% phosphotungstic acid hydrate before TEM observation. The size distribution of the aggregates in aqueous solution was confirmed by a Zetasizer 3000HS measurement system. The light wavelength was set at 633 nm, which is much longer than the absorption band of the trans- or cis-azobenzene group. Atomic force microscopy (AFM) images were obtained on a Multimode Nanoscope IV in tapping mode using silicon cantilevers. Critical Micelle Concentration. The dependence of solution conductivity on surfactant concentration was used to determine the critical micelle concentration (cmc) of AzoTEA. Typically, the slope of the change in conductivity versus the concentration below cmc is steeper than the slope above the cmc. Therefore, the junction of the conductivity-concentration plot represents the cmc value. To measure the cmc of cis-AzoTEA, the AzoTEA solution at each concentration was irradiated with UV light to keep AzoTEA in the cis form. Substrate Treatment. Quartz slides (0.1  1  3 cm3) were immersed in fresh piranha solution (30% H2O2/98% H2SO4 1:3 v/v) and heated until no bubbles were observered. Caution! Piranha solution is very corrosive, and appropriate safety precautions should be utilized, including the use of acid-resistant gloves and adequate shielding. The slides were rinsed with deionized water and dried for 12 h at 100 °C in an oven. The substrates were then DOI: 10.1021/la1004648

9737

Article

Wang et al.

Figure 2. Critical micelle concentration determination of trans-AzoTEA and cis-AzoTEA by the dependence of conductivity on concentration. immersed in 1  10-5 M APTS in toluene for 12 h to form a self-assembled monolayer terminated with amine groups at the exposed surface. The substrate was subjected to toluene, tetrahydrofuran, acetone, and water in turn for a short time to remove the physically adsorbed APTS.

LbL Assembly and Photoresponse of Multilayer Films. The substrates coated with amine were immersed in PAA (1 mg/ mL, pH 4) for 20 min. After washing with water and drying with N2, they were immersed in AzoTEA solution (9.5  10-5 M, which is between the cmc’s of trans-AzoTEA and cis-AzoTEA) for another 20 min, followed by cleaning and drying. After the first bilayer adsorption, these substrates were alternately immersed in PAA and AzoTEA for 10 and 20 min, respectively. The multilayer structure was fabricated by repeating these two steps in a cyclic fashion. The deposition of AzoTEA in each layer was indicated by the UV-vis absorbance at 360 nm that is attributed to the π-π* and n-π* transitions of the azobenzene moiety. The coverage of AzoTEA in the multilayer film can be deduced by n/S = A/2ε, where n, S, A, and ε refer to the molar quantity, the substrate area, the UV absorbance at 360 nm, and the molar extinction coefficient of AzoTEA, respectively. Unless stated otherwise, the final multilayer film is composed of a (PAA/ AzoTEA)8PAA 8.5 bilayer where the outermost layer is PAA. In addition, all of the LbL films were kept in the dark before use. For irradiation, the substrates coated with multilayer films were kept in glass bottles. UV or visible light was administered through the bottle opening to irradiate the multilayer film. Film Stability Test. Multilayer films that were or were not subjected to UV light irradiation were immersed in a pH 4 AcOH aqueous solution to test their stability. After exposing each film to the acidic solution for 5 min, the film was taken out and the solution residue was blown away by N2 gas. Then, the film with trans-AzoTEA was tested by UV/vis spectroscopy to monitor the UV absorbance change at 360 nm. Films with cis-AzoTEA were irradiated with visible light for 900 s to invert cis-AzoTEA to trans-AzoTEA before UV/vis measurement.

Result and Discussion Self-Assembly Behavior of AzoTEA in Aqueous Solution. The photoisomerization of azobenzene changes the molecular symmetry and polarity, significantly affecting the amphiphilicity. The change in the dipole between trans-azobenzene and cisazobenzene is estimated to be 4.4 D and is responsible for the polarity change between trans-azobenzene and cis-azobenzene.16 The evidence for the amphiphilicity change of AzoTEA in aqueous solution is that the cmc of AzoTEA increases with conversion of the azobenzene moiety from the trans to cis form. As shown in Figure 2, according to the conductivity dependence on concentration, the cmc of trans-AzoTEA is determined to be 8.0 10-5 M and the cmc of cis-AzoTEA is 1.1 10-4 M. (16) Tong, X.; Wang, G.; Soldera, A.; Zhao, Y. J. Phys. Chem. B 2005, 109, 20281.

9738 DOI: 10.1021/la1004648

Figure 3. TEM images of (a) AzoTEA with a concentration of 9.5 10-5 M, (b) 5 mL of AzoTEA (9.5 10-5 M) mixed with 100 μL of 1 mg/mL PAA (pH 4 AcOH solution), (c) the solution in b after UV irradiation for 600 s, and (d) size distribution of solution b by DLS.

The difference in the two cmc values means that AzoTEA at a concentration between the two values will be responsive to UV irradiation, self-assembling when in the trans form but disassembling when in the cis form. The self-assembly behavior of AzoTEA in aqueous solution was studied by TEM and DLS. As shown in Figure 3, at a concentration above the cmc AzoTEA self-assembled to form micelles. Upon addition of PAA, the spontaneous formation of an electrostatic complex between AzoTEA and PAA yielded large aggregates. The size distribution, determined by dynamic light scattering (DLS), of this kind of large aggregate ranged from 220 to 480 nm, in which the main size was 331 nm with a polydispersity of 0.01 (Figure 3d). Upon UV irradiation, the aggregates formed by AzoTEA as well as the electrostatic complex of AzoTEA with PAA were completely destroyed (Figure 3c), agreeing well with the above cmc shift that indicates that the self-assembly of AzoTEA in aqueous solution can be well controlled by photoirradiation. Formation of Photoresponsive Multilayer Film of AzoTEA. Because AzoTEA can self-assemble to form aggregates above its cmc, it is plausible that polyvalent and positively charged AzoTEA aggregates can alternately deposit with a negatively charged polyelectrolyte on a substrate to form a multilayer thin film. To verify this possibility, both a strong polyelectrolyte (PSS) and a weak polyelectrolyte (PAA) were employed to deposit AzoTEA aggregates for multilayer fabrication. However, AzoTEA in the Langmuir 2010, 26(12), 9736–9741

Wang et al.

Article

Figure 4. Layer-by-layer assembly of PAA (1 mg/mL, dissolved in pH 4 AcOH solution) with (a) AzoTEA (9.5 10-5 M), (b) AzoTEA (1.9 10-5 M), and (c) AzoTEA (2.2 10-4 M).

Figure 5. Photoisomerization of AzoTEA in the 8.5 bilayer multilayer film.

first adsorbed bilayer mostly fell off of the substrate upon exposure to PSS for the deposition of the second bilayer, terminating the LbL assembly. The weaker polyelectrolyte, PAA, can evidently assemble AzoTEA on a substrate in an even-layer deposition, although most of the layers were lost when exposed to PAA aqueous solution. We attribute the unsuccessful LbL assembly to the extraction of AzoTEA into the solution by the polyelectrolyte. PAA is less able to extract AzoTEA because it is less absolutely ionized in comparison to PSS. We attempted to decrease the extraction of AzoTEA by free PAA in the solution through reducing the degree of ionization of PAA. To address this issue, PAA (1 mg/mL) was dissolved in a pH 4 AcOH aqueous solution in which only 10% carboxyl acid group are estimated to be ionized.17 For the case of 10 min of PAA adsorption and 20 min of AzoTEA deposition, the alternating LbL film was well assembled on the substrate. By evaluating the absorbance at 360 nm of the 8.5 bilayer multilayer film shown in Figure 4a, the coverage of AzoTEA in the film is estimated to be 4.6 10-9 mol/cm2. Additionally, the linear growth of even layers indicates the controlled fabrication of a multilayer film. It should be noted that AzoTEA at a concentration of 1.9 10-5 M, which is lower than the cmc of trans-AzoTEA, cannot be assembled into a multilayer film in the same fashion as discussed above (Figure 4b). AzoTEA at a concentration of 2.2 10-4 M, which is indeed above the cmc of cis-AzoTEA, can be assembled well (Figure 4c). Both control experiments suggest that the aggregation of AzoTEA in solution is a prerequisite for multilayer formation. Although embedded in the multilayer film, the azobenzene moiety of AzoTEA has photoisomerization behavior that was well maintained. As shown in Figure 5, upon irradiation with UV light of 365 nm for 300 s, the absorption band at around 360 nm decreased remarkably. Concomitantly, the absorption band around 450 nm increased. These absorption bands at 360 and 450 nm are ascribed to π-π* and n-π* transitions of azobenzene, (17) Choi, J; Rubner, M. F. Macromolecules 2005, 38, 116.

Langmuir 2010, 26(12), 9736–9741

Figure 6. Stability of multilayer films with and without UV irradiation. The absorbance of the multilayer film at 360 nm depending on the time immersed in AcOH solution implies the quantity of trans-AzoTEA left in the film. To convert cis-AzoTEA to transAzoTEA in the photoirradiated film, the film was irradiated with visible light for 900 s after immersion in AcOH solution and irradiated with UV for the other 600 s for the remaining immersion.

respectively. Therefore, the change in the absorption bands is indicative of the photoisomerization of AzoTEA from the trans to cis form, induced by UV irradiation. The opposite transition was observed upon irradiation of the cis form by visible light of 450 nm for 900 s. In this case, the π-π* absorption increased while the n-π* absorption decreased, suggesting that AzoTEA inverted from the cis to trans form. Controlled Disassembly of Multilayer Film. The unique property of AzoTEA in which the trans form is more stable than the cis form in the self-assembled structure is applied to tune the controllable disassembly of multilayer films. UV irradiation can decompose the multilayer film because the aggregates of AzoTEA are not stable when AzoTEA inverts to the cis form. To distinguish the stability of trans-AzoTEA-based films and from that of cis-AzoTEA-based films, electrostatic interactions between AzoTEA and PAA must be maintained. Otherwise, the multilayer film will not be disassembled in a controlled way, degrading even without UV irradiation when the interlayer electrostatic interaction weakens. To amplify the disparity of stability between photoirradiated and nonphotoirraditated multilayer films, film stabilities were examined with exposure to three aqueous solutions: sodium chloride, pure water, and pH 4 AcOH. Because the electrostatic interaction is sensitive to ionic strength, the nonphotoirradiated multilayer film was observed to be nearly completely disassembled in 1 M sodium chloride solution after 5 min of immersion. Whereas the multilayer film shows greater stability in pure water, it remained difficult to determine an optimized immersion time in which nonphotoirradiated films maintained their stability while photoirradiated films were disassembled. Considering that the multilayer film was fabricated in pH 4 AcOH solution, the nonphotoirradiated multilayer film was expected to display DOI: 10.1021/la1004648

9739

Article

Wang et al.

Figure 7. AFM image of (a) the surface of (PAA/AzoTEA)8 and (b) the surface of (PAA/AzoTEA)8/PAA.

Figure 8. Selectively erased multilayer film. The multilayer film is irradiated with UV light for 600 s under the cover of the mask. Furthermore, the film is immersed in AcOH solution for 10 min.

relatively good stability against pH 4 AcOH immersion. As shown in Figure 6, only 5% of AzoTEA in the nonphotoirradiated film decomposed after immersion in pH 4 AcOH solution for 5 min. However, the photoirradiated film lost 72% of its AzoTEA after 5 min of exposure to pH 4 AcOH. This AcOHresistance comparison shows that the photoisomerization of AzoTEA can significantly influence its stability in multilayer films. Additionally, after another 10 min of immersion, the photoirradiated film remained the same as that after 5 min of immersion, indicating that a 5 min immersion in AcOH was sufficient to disassemble the photoirradiated film. It should be noted that the residual AzoTEA resulted from the adsorption of AzoTEA onto the PAA monolayer, which always remains on the positively charged substrate. Plausible Mechanism for the Self-Assembly and Disassembly of Multilayer Film. In the case of micelle aggregation, the AzoTEA surfactant can act as a polyvalent building block to assemble with PAA. However, the aggregates formed by surfactants mostly become unstable upon dehydration and usually adopt a bilayer structure at the interface between solid and air. To determine if this feature also existed when AzoTEA was assembled in a multilayer film, the morphologies of even and odd outer layers of multilayer films were observed by AFM. As shown in Figure 7a, the surface of the (PAA/AzoTEA)8 film contained domains with a thickness of 3.4 nm after subtracting 9740 DOI: 10.1021/la1004648

the inner neighboring layer. The thickness was ascribed to an intercalated bilayer of AzoTEA because the single AzoTEA molecule is estimated to be as long as 2.6 nm. The AFM observation indicates that the micelles spread on the surface to form a bilayer structure when AzoTEA was assembled from aqueous solution onto a multilayer film. However, when the outer layer was PAA, as shown in Figure 7b, AzoTEA was assembled with deposited PAA to form larger aggregates, correlating with the conversion of AzoTEA from a micelle structure to sphere aggregates by TEM observation. Evidently, through the aggregation conversion, the surface charge of the multilayer film inverted from positive to negative, which can be subsequently applied to combine the next layer of AzoTEA for multilayer growth. The disassembly of multilayer films under UV irradiation is proposed to result from the breaking of AzoTEA-based aggregates. Similar to the behavior of aggregated AzoTEA in aqueous solution, after UV irradiation, cis-AzoTEA’s constructed in the multilayer film were assumed to become unstable against AcOH immersion because they were repulsed by each other. Therefore, after the AzoTEA release, the multilayer film could no longer be maintained because free PAA, lacking electrostatic pairing with AzoTEA, also departed from the film. Selectively Erasable Multilayer Film. In the case of the disassembly of multilayer films under exposure to UV irradiation, we explored the possibility of selectively erasing sections of the Langmuir 2010, 26(12), 9736–9741

Wang et al.

Article

multilayer film; the degree of loading was 4 times that of the film assembled by AzoTEA at the lower concentration described above. However, when the approach based on disassembly was employed to erase the multilayer film selectively, the photoirradiated regions were not adequately washed away, as shown in Figure 9. A reasonable explanation is that a large amount of AzoTEA assembled in the multilayer film made the disruption less sensitive and their aggregates exhibited sufficient stability to maintain the photoirradiated film against complete disassembly after 10 min of AcOH immersion. Accordingly, in the aqueous solution, AzoTEA at 2.2  10-4M in comparison to AzoTEA at a lower concentration displayed poor sensitivity to UV light because AzoTEA maintained its aggregation after the photoisomerziation of the azobenzene moieties. This comparison provides us a convenient profile for evaluating the optimized conditions for selectively erasable multilayer films by photoirradiation. Figure 9. AFM image of a multilayer film formulated from 2.2  10-4 M AzoTEA and PAA with selective photoirradiation and further AcOH immersion (10 min); the scale is 40  40 μm2.

multilayer film through regioselective photoirradiation. A mask was used to cover the multilayer film, permitting the light to irradiate the surface in a patterned fashion. As stressed above, these multilayer films display different stability in pH 4 AcOH solution before and after photoirradiation. Therefore, the multilayer film exposed to UV light in a patterned fashion should undergo regioselective disassembly with exposure to pH 4 AcOH. When the multilayer film covered by a mask was irradiated with UV light for 600 s and subsequently immersed in AcOH solution for 10 min, the regions exposed to UV irradiation were erased from the multilayer film (Figure 8). Concomitantly, the regions that were not photoirradiated remained on the substrate to form stripes, which were determined by AFM to be about 17 nm higher than the erased regions. We also examined multilayer films assembled by AzoTEA at a higher concentration (2.2  10-4M), which was not only higher than the cmc of trans-AzoTEA but also higher than that of cis-AzoTEA for selective erasability. For (PAA/AzoTEA)8/PAA multilayer films, a great deal of AzoTEA was assembled in the

Langmuir 2010, 26(12), 9736–9741

Conclusions A photoresponsive azobenzene-containing surfactant has been successfully employed as a building block to fabricate layer-bylayer films with a polyelectrolyte. On the basis of the photoisomerization of azobenzene, the disassembly of bilayered surfactants induces the decomposition of these multilayer films. By employing a mask to irradiate the multilayer film selectively, only the photoirradiated regions are observed to be disassembled against AcOH immersion. Considering that the thickness of the multilayer film can be well controlled on the nano/microscale via tuning the number of layer, this new concept of controllable disassembly provides a framework for the facile fabrication of precisely patterned surfaces for protein or cell adhesion as well as controllable cargo delivery. Acknowledgment. This work was financially supported by the National Basic Research Program of China (2007CB808000), the National Natural Science Foundation of China (20974059, 50973051), the Tsinghua University Initiative Scientific Research Program (2009THZ02-2), and the joint project between NSFC and DFG (TRR61). We acknowledge the careful editing by Timothy J. Merkel, University of North Carolina, Chapel Hill.

DOI: 10.1021/la1004648

9741