Layer-by-Layer Technique as a New Approach to Produce

Langmuir 2009, 25, 2331-2338

2331

Layer-by-Layer Technique as a New Approach to Produce Nanostructured Films Containing Phospholipids as Transducers in Sensing Applications P. H. B. Aoki,† D. Volpati,† A. Riul, Jr.,‡ W. Caetano,§ and C. J. L. Constantino*,† Faculdade de Cieˆncias e Tecnologia, UNESP, Presidente Prudente/SP, 19060-900 Brazil, UniVersidade Federal de Sa˜o Carlos, campus Sorocaba/SP, 18052-780 Brazil, and UniVersidade Estadual de Maringa´, Maringa´/PR, 87020-900 Brazil ReceiVed August 18, 2008. ReVised Manuscript ReceiVed December 10, 2008 Phospholipids are widely used as mimetic systems to exploit interactions involving biological membranes and pharmacological drugs. In this work, the layer-by-layer (LbL) technique was used as a new approach to produce multilayered thin films containing biological phospholipids applied as transducers onto Pt interdigitated electrodes forming sensing units of an electronic tongue system. Low concentrations (nM level) of a phenothiazine compound were detected through impedance spectroscopy. Both negative 1,2-dipalmitoyl-sn-3-glycero-[phosphor-rac-(1-glycerol)] (DPPG) and zwitterionic L-R-1,2-dipalmitoyl-sn-3-glycero-phosphatidylcholine (DPPC) phospholipids were used to produce the LbL films, whose molecular architecture was monitored combining spectroscopy and microscopy at micro and nanoscales. The sensor array was complemented by Langmuir-Blodgett (LB) monolayers of DPPG and DPPC deposited onto Pt interdigitated electrodes as well. It was found that the distinct molecular architecture presented by both LbL and LB films plays a key role on the sensitivity of the sensor array with the importance of the LbL films being demonstrated by principal component analysis (PCA).

1. Introduction Detection and quantification of organic compounds such as phenothiazine pharmaceuticals through a selective method envisaging clinical analysis have been highly keen in the last decades1-3 due to their great therapeutic importance in pharmacology. Therefore, the immobilization and stabilization of biological molecules, such as phospholipids, onto solid substrates might have a substantial importance in biosensing development.4-9 They are mainly based on interactions of immobilized biomolecules with drugs or proteins that might be incorporated in artificial membranes10 or through impedance measurements inhighlydilutedsystems.11-14 Consequently,theLangmuir-Blodgett * Corresponding author. E-mail: [email protected]. † Faculdade de Cieˆncias e Tecnologia, UNESP. ‡ Universidade Federal de Sa˜o Carlos. § Universidade Estadual de Maringa´.

(1) Basavaiah, K.; Krishnamurthy, G. Anal. Sci. 1999, 15, 67–71. (2) Choi, H. N.; Cho, S. H.; Park, Y. J.; Lee, D. W.; Lee, W. Y. Anal. Chim. Acta 2005, 541, 49–56. (3) Bouchta, D.; Izaoumen, N.; Zejli, H.; El Kaoutit, M.; Temsamani, K. R. Biosens. Biolectron. 2005, 20, 2228–2235. (4) Nikolelis, D. P.; Hianik, T.; Krull, U. J. Electroanalysis 1999, 11, 7–15. (5) Fiol, C.; Valleton, J. M.; Delpire, N.; Barbey, G.; Barraud, A.; RuaudelTeixier, A. Thin Solid Films 1992, 210, 489–491. (6) Okahata, Y.; Tsuruta, Y.; Ijiro, K.; Atiga, K. Langmuir 1988, 4, 1373– 1375. (7) Singhal, R.; Takashima, W.; Kaneto, K.; Samanta, S. B.; Annapoorni, S.; Malhotra, B. D. Sens. Actuators, B 2002, 86, 42–48. (8) Yasuzawa, M.; Hashimoto, M.; Fujii, S.; Kunugi, A.; Nakaya, T. Sens. Actuators, B 2000, 65, 241–243. (9) Tamm, L. K. C.; Bohm, J. Y.; Shao, Z.; Hwang, J.; Edidin, M.; Betzig, E. Thin Solid Films 1996, 284, 813–816. (10) Lavan, D. A.; Mcguire, T.; Langer, R. Nat. Biotechnol. 2003, 21, 1184– 1181. (11) Shi, H.; Yanga, Y.; Huanga, J.; Zhaoa, Z.; Xu, X.; Anzai, J.; Osac, T.; Chen, Q. Talanta 2006, 70, 852–858. (12) Zucolotto, V.; Pinto, A. P. A.; Tumolo, T.; Moraes, M. L.; Baptista, M. S.; Riul, A., Jr.; Araujo, A. P. U.; Oliveira, O. N., Jr Biosens. Bioelectron. 2006, 21, 1320–1326. (13) Zucolotto, V.; Daghastanli, K. R. P.; Hayasaka, C. O., Jr.; Ciancaglini, P., Jr Anal. Chem. 2007, 79, 2163–2167. (14) Kukol, A.; Li, P.; Estrela, P.; Ko-Ferrigno, P.; Migliorato, P. Anal. Biochem. 2008, 374, 143–153.

(LB) and layer-by-layer (LbL) techniques have been well explored in the immobilization of phospholipids onto solid substrates,15-19 allowing the fabrication of supramolecular architectures with molecular level control. In this work, LB and LbL20 techniques were applied to two phospholipids, L-R-1,2-dipalmitoyl-sn-3-glycero-phosphatidylcholine (DPPC) and 1,2-dipalmitoyl-sn-3-glycero-[phosphorrac-(1-glycerol)] (DPPG), aiming to detect trace levels of a phenothiazine compound (methylene blue, MB) in water.21 It is worth mentioning that it is the first time that a systematic study is being reported on the formation of multilayered LbL films containing DPPG and DPPC, in opposition to the LB technique that restricts the phospholipid films to mono- or bilayers (ref 21 and references therein). Three major pieces of information were acquired in this study: (i) the optimized growth conditions of DPPC and DPPG LbL films; (ii) the quality of the LbL films in terms of thickness control and surface morphology; and (iii) their efficiencies as transducers when applied as sensing units in an e-tongue system (sensor array). The growth of the films was monitored by ultraviolet-visible absorption (UV-vis) spectroscopy, and their morphologies were investigated at the microscopic level via micro-Raman measurements, combining chemical and morphological information through the coupling of an optical microscope to a Raman spectrograph, while morphological information at the nanoscale was investigated by atomic force microscopy (AFM). Comple(15) Pavinatto, F. J.; Caseli, L.; Pavinatto, A., Jr.; Nobre, T. M.; Zaniquelli, M. E. D.; Silva, H. S.; Miranda, P. B., Jr Langmuir 2007, 23, 7666–1640. (16) Caseli, L.; Masui, D. C.; Furriel, R. P. M.; Leone, F. A.; Zaniquelli, M. E. D. Thin Solid Films 2007, 515, 4801–4807. (17) Lee, Y. L.; Lin, J. Y.; Chang, C. H. J. Colloid Interface Sci. 2006, 296, 647–654. (18) Wang, L. Y.; Scho¨nhoff, M.; Mo¨hwald, H. J. Phys. Chem. B 2002, 106, 9135–9142. (19) Moraes, M. L.; Batista, M. S.; Itri, R.; Zucolotto, V., Jr Mater. Sci. Eng., C 2008, 28, 467–471. (20) Paterno, L. G.; Mattoso, L. H. C., Jr Quı´m. NoVa 2001, 24, 228–235. (21) Aoki, P. H. B.; Caetano, W.; Volpati, D., Jr.; Constantino, C. J. L. J. Nanosci. Nanotechnol. 2008, 8, 4341–4348.

10.1021/la802696j CCC: $40.75  2009 American Chemical Society Published on Web 01/22/2009

2332 Langmuir, Vol. 25, No. 4, 2009

Figure 1. Molecular structures of MB, DPPG, DPPC, and PAH.

mentary, Fourier transform infrared absorption (FTIR) spectroscopy was performed to check possible interactions between the MB and the phospholipids. Additionally, impedance spectroscopy measurements carried out for LB monolayers and 5-bilayer LbL films of DPPC and DPPG deposited onto Pt interdigitated electrodes were quite successful in the detection of low levels of MB in aqueous solutions.

2. Materials and Methods 2.1. Reagents. The anionic DPPG [1,2-dipalmitoyl-sn-3-glycero[phosphor-rac-(1-glycerol), purity >99%]] and zwitterionic DPPC [L-R-1,2-dipalmitoyl-sn-3-glycero-phosphatidylcholine, purity >99%] phospholipids were purchased from Avanti Polar Lipids Inc. Phenothiazine methylene blue (MB) and poly(allylamine hydrochloride) (PAH) were acquired from Sigma-Aldrich Co. The molecular weights of DPPC, DPPG, and MB are 734, 745, and 319 g/mol, respectively. All chemicals (see Figure 1) were used without further purification. Ultrapure water (18.2 MΩ cm and pH 5.6) acquired from a Milli-Q system, model Simplicity, was used to prepare both the LbL films and the MB solutions. 2.2. Solutions and LbL and LB Films. The LbL films were fabricated using DPPC or DPPG and PAH aqueous solutions at pH ) 5.6. The following solution concentrations were used: DPPC ) 0.73 mg/mL (1.0 mM), DPPG ) 0.74 mg/mL (1.0 mM), and PAH ) 0.80 mg/mL. All aqueous solutions were prepared without any special procedure: the powder was simply added to ultrapure water, and the solution was gently vortexed. In 1.0 mM aqueous solutions, DPPC and DPPG self-assemble as multilamellar vesicles (MLV), whereas sonication or extrusion yields formation of unilamellar vesicles.19,22,23 The DPPG/PAH LbL film was fabricated by immersions of the substrate into distinct solutions according to the following sequence: PAH solution f ultrapure water gently stirred to remove excess of adsorbed PAH f DPPG solution f ultrapure water to remove excess of adsorbed DPPG. After that, the first DPPG/ PAH LbL bilayer is formed and the multilayered DPPG/PAH LbL film is grown repeating the “four-step sequence”. The growth of multilayered DPPC/PAH LbL films follows the same procedure. The number of deposited LbL bilayers and the type of substrate were chosen according to the characterization technique. The immersion of the substrates into distinct solutions for a specific time must be previously determined by a growth kinetic study. Briefly, it was carried out by following the growth of a phospholipid/PAH LbL bilayer through UV-vis spectroscopy monitoring the increase of the absorbance according to the immersion time of the substrate (1, 3, 5, and 7 min). The appropriate immersion time of the substrate is determined by a plateau in the absorbance curve indicating that a maximum adsorption of material is reached and, consequently, a layer is formed. The LB films were fabricated using a Langmuir KSV 2000 trough with the water subphase at 23 °C and a compression speed of 10 (22) Vivares, E.; Ramos, L. Langmuir 2005, 21, 2185–2191. (23) Nassar, P. M.; Almeida, L. E.; Tabak, M. Langmuir 1998, 14, 6811–6817.

Aoki et al. mm/min. The target surface pressure during the compression was 30 mN/m, which corresponds to the condensed phase of the phospholipids as indicated by the mean molecular area-surface pressure isotherms.21 When the Langmuir monolayer reaches a stable state, it is then transferred from the air-water interface to Pt interdigitated electrodes (50 pairs, with 10 µm width, 0.5 mm length, 100 nm height being 10 µm apart from each other) using a dipper speed of 10 mm/min. The LB monolayer of each phospholipid was also transferred under these conditions while keeping the transfer ratio close to one. 2.3. Characterization Techniques. UV-vis absorption spectroscopy was carried out for LbL films deposited onto quartz substrates using a Varian spectrophotometer (model Cary 50) from 190 to 1100 nm. Raman analysis and optical microscopy were obtained using a micro-Raman Renishaw spectrograph model in-Via equipped with a Leica microscope, whose 50× objective allows collection of spectra with ∼1 µm2 spatial resolution, CCD detector, lasers at 514.5, 633, and 785 nm, 1200 and 1800 grooves/mm gratings with additional notch filters, and a computer-controlled three-axis-encoded (XYZ) motorized stage to take Raman images with a minimum step of 0.1 µm. In addition, FTIR measurements were carried out for LbL films deposited onto Ge substrate using a Bruker spectrometer (model Vector 22). Both Raman and FTIR spectra were taken from LbL films containing 21 deposited bilayers. AFM images were collected using a Digital Instrument microscope (model Nanoscope IV) with a tip of silicon nitride, using the contact mode. The images were characterized by using the software WsXM 4.0 Develop 12.0 to analyze the root mean square (RMS) roughness and average height of the topographic images. Impedance spectroscopy measurements were carried out with a Solartron analyzer (model 1260A). The whole sensor array comprises five sensing units: a bare Pt electrode and Pt electrodes coated with 5-bilayer LbL films of DPPC/PAH and DPPG/PAH and LB monolayers of DPPC and DPPG as well. All sensing units were dipped in ultrapure water, used as a reference, and different MB aqueous solutions from 0.01 to 10.0 nM. The measurements were recorded at the frequency range from 1 Hz to 1 MHz using 50 mV of amplitude as the input signal. After immersion, the sensing units were left soaking for 20 min prior to data acquisition to enable a stable reading. It is important to stress that between each set of measurements the electrodes were immersed in ultrapure water under moderate stirring to avoid possible contaminations by particles aggregated in the structure of the thin films.

3. Results and Discussion 3.1. Growth of DPPG/PAH and DPPC/PAH LbL Films. Initially, a study of the growth kinetics of the DPPG/PAH and DPPC/PAH LbL films was performed using UV-vis absorption spectroscopy. Figures 2 and 3 show, respectively, the UV-vis spectra for both DPPG/PAH and DPPC/PAH LbL films indicating that the optimized immersion time for the LbL fabrication (inset A in Figures 2 and 3) is 3 min. A linear increase in the absorbance was measured as a function of the deposited number of LbL bilayers, as shown in inset B in Figures 2 and 3, suggesting that, for each phospholipid, the same amount of material was adsorbed at each deposition step. However, the DPPC/PAH LbL film, when compared to the DPPG/PAH LbL film, possesses smaller absorbance values for the same number of deposited bilayers, besides the maximum deposition of the 13 bilayers. The differences in the absorbance for both LbL films are related to the negative charge of DPPG while DPPC is electrically neutral. Since PAH is a cationic polyelectrolyte, the deposition of DPPG/ PAH is facilitated by electrostatic interactions between the groups NH3+ of PAH and PO4- of DPPG, leading to LbL films with higher absorbance values. A scheme representing a possible structure of the phospholipids in the LbL film is shown in inset C in Figure 3.

LbL Technique to Produce Nanostructured Films

Langmuir, Vol. 25, No. 4, 2009 2333

Figure 2. UV-vis absorption spectra for LbL film of DPPG/PAH containing different numbers of bilayers. Inset A: kinetic growth for LbL film of DPPG/PAH. Inset B: linear dependence of the absorbance at 200 nm as a function of DPPG/PAH bilayers in the LbL film.

Figure 3. UV-vis absorption spectra for LbL film of DPPC/PAH containing different numbers of bilayers. Inset A: kinetic growth for LbL film of DPPC/PAH. Inset B: linear dependence of the absorbance at 200 nm as a function of DPPC/PAH bilayers in the LbL film. Inset C: scheme of the LbL films containing PAH/phospholipids where DPPC or DPPG is in the form of a multilamellar vesicle (MLV).

3.2. Morphology of DPPG/PAH and DPPC/PAH LbL Films. The morphology of DPPG/PAH and DPPC/PAH LbL films was investigated at microscopic and nanoscopic scales through optical microscopy and AFM. Figure 4 shows the optical microscopy and AFM images, respectively, for 21- and 10-bilayer DPPG/PAH LbL films, while Figure 5 shows the same measurements accomplished, respectively, for 13- and 10-bilayer DPPC/ PAH LbL films. In agreement with the optical microscopy results exhibited in Figures 4a and 5a, in two and three dimensions, for the DPPG/PAH and DPPC/PAH LbL films, it was verified that both possess a homogeneous morphology at the micrometric scale. The AFM images, exhibited in Figures 4b and 5b, show the presence of domains (molecular aggregates), whose height distributions are given by the histograms in the insets. These domains can be assigned to the phospholipid multilamellar organization (MLV), similar to the results reported for vesicles by Fang et al., Lunelli et al., You et al., and Tarasova et al.24-27 The presence of MLVs is supported by the coalescing domains (24) You, H. X.; Qi, X. Y.; Grabowski, G. A.; Yu, L. Biophys. J. 2003, 84, 2043–2057. (25) Lunelli, L.; Pasquardini, L.; Pederzolli, C.; Vanzetti, L.; Anderle, M. Langmuir 2005, 21, 8338–8343. (26) Tarasova, A.; Griesser, H. J.; Meagher, L. Langmuir 2008, 24, 7371– 7377. (27) Fang, Y.; Yang, J. Biochim. Biophys. Acta 1997, 1324, 309–319.

and globular protuberances noted in the AFM images (Figures 4b and 5b) as well, whose globular-like structures present dimensions (“diameter”) between 1000 and 3000 Å for DPPG and smaller than 1000 Å for DPPC. Volodkin et al.28 reported the immobilization of DPPC and DPPG in the form of unilamellar vesicles (extrusion) onto polyelectrolytes using the LbL technique. In the latter, AFM images similar to ours revealed that the vesicles kept their shape upon LbL immobilization. Moreover, AFM results were analyzed through RMS roughness and average height. The RMS roughness is given by the standard deviation following the equation

Rrms )



N

∑ (Zn - Zj )2

n)1

N-1

j is the mean of the Z values within a given area, Zn is where Z the height of the nth pixel, and N is the number of pixels considered within a given area, while the average height is simply the arithmetic mean of the measured heights. Measurements were taken for 5.0 µm × 5.0 µm, 2.0 µm × 2.0 µm, and 1.0 µm × (28) Volodkin, D.; Arntz, Y.; Schaaf, P.; Moehwald, H.; Voegel, J. C.; Ball, V. Soft Matter 2008, 4, 122–130.

2334 Langmuir, Vol. 25, No. 4, 2009

Aoki et al.

Figure 4. (a) Optical images for 21-bilayer LbL film of DPPG/PAH in two and three dimensions and (b) AFM topographic images for 10-bilayer LbL film of DPPG/PAH in two and three dimensions. The inset shows a histogram with a distribution of height recorded along the area of 2 µm × 2 µm.

Figure 5. (a) Optical images for 13-bilayer LbL film of DPPC/PAH in two and three dimensions and (b) AFM topographic images for 10-bilayer LbL film of DPPC/PAH in two and three dimensions. The inset shows a histogram with a distribution of height recorded along the area of 2 µm × 2 µm.

1.0 µm areas, and the obtained values are given in Table 1. Both DPPC/PAH and DPPG/PAH LbL films presented RMS roughness and average height values varying with the scanned area size, which is expected since the distribution of aggregates is not homogeneous either along the film surface or in terms of size. It can be observed that the RMS roughness for the DPPG/PAH LbL film at different areas is higher than that for the DPPC/PAH

LbL film. This result is related to the difference in the average height of the analyzed films, where greater values were observed for the DPPG/PAH LbL film. This fact might be associated with the intense interaction between DPPG and PAH, leading to the formation of molecular aggregates. Besides, the fact that the DPPG/PAH LbL film possesses greater average height than the DPPC/PAH LbL film may suggest that a larger amount of material

LbL Technique to Produce Nanostructured Films

Langmuir, Vol. 25, No. 4, 2009 2335

Table 1. RMS Roughness and Average Height for DPPC/PAH and DPPG/PAH LbL Films sample

area/µm2

RMS roughness/nm

average height/nm

DPPC/PAH

5×5 2×2 1×1 5×5 2×2 1×1

5.26 1.14 0.89 10.07 5.51 5.03

8.65 3.61 2.03 21.51 16.69 11.34

DPPG/PAH

is present there. This result is in good agreement with the UV-vis measurements, which presented absorbance values for the DPPG/ PAH LbL films of 1 order of magnitude higher than those for the DPPC/PAH LbL films. 3.3. Sensor Application. Before using the phospholipids as sensing units in an e-tongue system, several measurements were performed with the bare Pt electrodes immersed in ultrapure water to guarantee similar electrical responses. Therefore, any minor changes observed in the impedance data, after the deposition of the LbL and LB films, are in fact due to the presence of MB in solution instead of subtle differences attributed to the design of the electrodes.21 Moreover, a sequence of measurements at every 5 min was made with the bare Pt electrodes dipped in ultrapure water to check the temporal stabilization of the electric signal due to the double-layer formation at the electrode/electrolyte interface29 (results not shown). The electrical signals start to overcome each other ∼15 min after the immersion of the electrodes in ultrapure water; therefore, the electrodes were left soaking for 20 min in the solutions before data acquisition to better ensure a stable reading of the data. After that, capacitance measurements were performed on Pt electrodes covered with the LB and LbL films of DPPC and DPPG, dipped again in ultrapure water and MB solutions at 0.01, 0.1, 1.0, and 10.0 nM. In addition, after each set of measurements, the electrodes were carefully washed with ultrapure water to remove possible MB molecules entrapped within the films. Five independent measurements made from the same stock solutions were taken for each liquid sample, apart from measurements in ultrapure water before and after dipping the electrodes in MB solutions, as seen in Figure 6, to better check for possible contamination of the electrodes. It is clear that even at low molar concentrations the sensor was able to differentiate ultrapure water from MB solutions, with a downward trend in the measured capacitance with increasing MB concentration in solution at lower frequencies (