4198
Langmuir 2006, 22, 4198-4207
Controlled, Simultaneous Assembly of Polyethylenimine onto Nanoparticle Silica Colloids Joseph F. Bringley,*,† Andrew Wunder,† Andrew M. Howe,‡ Robin D. Wesley,‡ Tiecheng A. Qiao,† Nancy B. Liebert,† Brian Kelley,† John Minter,† Brian Antalek,† and James M. Hewitt† Research and DeVelopment, Eastman Kodak Company, Rochester, New York 14650-02002, and Kodak European R&D, Harrow, Middlesex, HA1 4TY, United Kingdom ReceiVed December 16, 2005. In Final Form: February 24, 2006 A novel precision-assembly methodology is described on the basis of the controlled, simultaneous assembly (CSA) of a core nanoparticle substrate and polyelectrolyte solutions. The method is capable of assembly rates at least as fast as 1016 core particles s-1 L-1 and affords concentrated suspensions of stable colloids with an adsorbed polyelectrolyte. The resulting dispersions are highly homogeneous, have a low viscosity and narrow particle-size distribution, and are stable colloids, even at solid concentrations of at least 33 wt %. The adsorption isotherm and the saturation adsorption for polyethylenimine (PEI) assemblies onto a 15 nm silica colloid have been evaluated with 1H NMR spectroscopy. The saturation adsorption is highly dependent upon the pH at assembly and is given by the equation PEIa (µmol m-2) ) 1.73pH - 1.89, R2 ) 0.986, where micromoles refers to the concentration of the EI monomer. The saturation concentration increases from 6.8 µmol m-2 at pH 5.0 to 13.7 µmol m-2 at pH 9.0. The adsorbed polyelectrolyte may be cross-linked and thereby permanently fixed to the colloid surface to prepare nanoparticle-polyelectrolyte colloidal assemblies having enhanced colloid stability, high homogeneity, and a high fraction (>80%) of permanently adsorbed polyelectrolyte. These assemblies are stable at physiological pH and ionic strength and may represent ideal substrates for bioconjugation and, ultimately, the design of nanocarriers for in vivo applications.
Introduction The ordered assembly of nanoscale and molecular components has promise to create assemblages capable of mimicking biological function, to create electronic devices on a molecular scale, and to create assemblages capable of interacting with living cells and cellular components. Many techniques for creating nanoscale assemblies are being developed and include LangmuirBlodgett,1 polyelectrolyte assembly,2 soft lithography,3 coreshell assemblies,4 micellular synthesis,5 and heterogeneous precipitation.6 However, a significant challenge lies in creating methods for assembling or fashioning nanoparticles or molecules into “materials” capable of being fabricated into free-standing, stable, working “devices”. While much progress has been made regarding two-dimensional molecular assemblies on planar surfaces, discrete three-dimensional assemblies of nanoparticles or molecules are generally more difficult to “fabricate” precisely and far fewer examples and methodologies have been reported. Three-dimensional nanoscale assemblies often suffer from instabilities and resist integration into working systems. A simple example involves the integration of nanoscale assemblies into living organisms. Successful integration requires assemblies that * To whom correspondence should be addressed. E-mail:
[email protected]. Telephone: +1-585-722-7678. Fax: +1-585722-4771. † Eastman Kodak Company. ‡ Kodak European R&D. (1) Stine, K. J.; Moore, B. G. Nano-Surface Chemistry, 1st ed.; Marcel Dekker: New York, 2001. (2) Decher, G. Science 1997, 277, 1232. (b) Caruso, F. AdV. Mater. 2001, 13, 11. (c) Caruso, F.; Caruso, R. A.; Mohwald, H. Science 1998, 282, 1111. (3) Gates, B. D.; Whitesides, G. M. J. Am. Chem. Soc. 2003, 125, 14986. (4) Liz-Marzan, L. M.; Correa-Duarte, M. A.; Pastoriza-Santos, I.; Mulvaney, P.; Ung, T.; Giersig, M.; Kotov, N. A. In Handbook of Surfaces and Interfaces of Materials; Nalwa, H. S. Ed.; Academic Press: NewYork, 2001; Vol. 3, p 189. (5) Thurmand, K. B.; Kowalewski, T.; Wooley, K. L. J. Am. Chem. Soc. 1996, 118, 7239. (6) Bringley, J. F.; Marchetti, A. P.; Eachus, R. S. J. Phys. Chem. B 2002, 106, 5346.
are colloidally stable under highly specific conditions (physiological pH and ionic strength), are compatible with blood components, are capable of avoiding detection by the immune system, and can survive the multiple-filtration and waste-removal systems inherent to living organisms.7 Precise methods of assembly are necessary for building ordered nanoscale assemblies capable of performing under such stringent conditions. One method that has received much attention in the past decade is the layer-by-layer (LbL) adsorption of polyelectrolytes onto charged surfaces.2,8 In this process, soluble polyelectrolytes are “assembled” onto surfaces having an opposite charge via Coulomb attraction, and they are held in place through the combination of Coulombic forces, hydrogen-bonding and van der Waals forces, and entropic forces.9 The combination of binding forces leads to an overcompensation of charge and, hence, a reversal of the surface charge and allows for the stepwise assembly of consecutive polyelectrolyte layers having alternating charges. Decher at al.10 first demonstrated this general approach for thinfilm assembly onto two-dimensional surfaces, and since that time, various authors have demonstrated multilayer assemblies of polymers,2,8 proteins,11 nanoparticles,12 dye molecules,13 and hollow microcapsules14 and a multitude of applications have been envisioned. Sukhorukov et al.15 described the assembly of (7) Moghimi, S. M.; Hunter, A. C.; Murray, J. C. Pharmacol. ReV. 2001, 53, 283. (8) Sukhorukov, G. B.; Donath, E.; Lichtenfeld, H.; Knippel, E.; Knippel, M.; Budde, A.; Mohwald, H. Colloid Surf., A 1998, 137, 253. (b) Burke, S. E.; Barrett, C. J. Langmuir 2003, 19, 3297. (c) Shiratori, S. S.; Rubner, M. F. Macromolecules 2000, 33, 4213. (d) Yoo, D.; Shiratori, S. S.; Rubner, M. F. Macromolecules 1998, 31, 4309. (9) Fleer, G. J.; Cohen Stuart, M. A.; Scheutjens, J. M. H. M.; Cosgrove, T.; Vincent, B. Polymers at Interfaces; Chapman and Hall: London, U.K., 1993. (10) Decher, G.; Hong, H.-G. Makromol. Chem. Macromol. Symp. 1991, 46, 321. (11) Cassier, T.; Lowack, K.; Decher, G. Supramel. Sci. 1998, 5, 309. (b) Caruso, F.; Schuler, C. Langmuir 2000, 16, 9595. (c) Caruso, F.; Kiikura, K.; Furlong, D. N.; Okahata, Y. Langmuir 1997, 13, 3427. (12) Caruso, F.; Mohwald, H. Langmuir 1999, 15, 8276.
10.1021/la0534118 CCC: $33.50 © 2006 American Chemical Society Published on Web 03/25/2006
CSA of PEI onto Nanoparticle Silica Colloids
polyelectrolyte multilayers onto 640 nm polystyrene latexes and showed that polyelectrolyte desorption and colloid aggregation could be a problem. Gittins and Caruso16 showed the assembly of polyelectrolyte multilayers onto gold nanoparticles and briefly discussed the role of polyelectrolyte size (i.e., Mw) and salt concentrations in affecting colloid stability. Kato et al.17 showed the assembly of weak polyelectrolytes onto latex particles and examined the role of pH and salt concentration in determining the extent of polymer adsorption. More recently, Schneider and Decher18 showed the assembly of as many as 20 polyelectrolyte layers onto gold nanoparticles and demonstrated the general colloidal stability of such particles. While the assembly of polyelectrolytes onto two-dimensional surfaces is complex but fairly well-understood, the assembly onto spherical colloids, particularly as they approach the nanosize range, is very complex and less well-understood. Messina et al.19 have suggested that, in the absence of strong van der Waals forces, polyelectrolyte multilayers should not be stable on spherical colloids. Furthermore, the method of producing polyelectrolyte multilayers on colloids is extremely tedious, involving mixing the colloid substrate in an excess of oppositely charged polyelectrolyte, allowing for adsorption, followed by repeated centrifugations, and washing of the surface-modified colloid. These steps must be repeated for each successive layer assembly. The resulting colloids can only be prepared at low concentrations, about 0.1-1.0% solids, and may not yield homogeneous, narrow particle-size distributions of colloids, and the temporal stability of the resulting colloids are rarely noted. The equilibrium conditions of polyelectrolyte adsorption are not well-understood, and rarely are attempts made to measure the amount of unadsorbed polymer free in solution. Finally, the constraints upon the system become acute, as the particle diameter of the substrate colloid is reduced below 100 nm and, particularly, below 25 nm. This is a direct result of the increase in surface area per unit mass of nanoparticles and the decrease in particle separation as the particle size is reduced at a given volume concentration. Therefore, it is important to understand the size dependence of the adsorption process, especially for substrate particles less than 100 nm and even more particularly less than 25 nm. It is also of importance to understand the dependence of polyelectrolyte adsorption on the average molecular weight of the polymer, particularly as the size of the polymer becomes appreciable with respect to the substrate particle. Methods are required for determining quantitatively the extent of polymer adsorption onto nanoparticle colloids, the equilibrium between adsorbed and free polyelectrolytes, and the influence of pH and salt concentration upon the equilibria. It is important to understand processes leading to desorption of the polyelectrolyte, especially when subjected to changes in pH and ionic strength, and in the presence of competing molecules or polymers. The solution to the above problems will facilitate the application of polyelectrolyte-modified colloids and could allow for the (13) Cooper, T. M.; Campbell, A. L.; Crane, R. L. Langmuir 1995, 11, 2713. (b) Araki, K.; Wagner, M. J.; Wrighton, M. S. Langmuir 1996, 12, 5393. (c) Ariga, K.; Lvov, Y.; Kunitake, T. J. Am. Chem. Soc. 1997, 119, 2224. (d) Yoo, D.; Wu, A. P.; Lee, J.; Rubner, M. F. Synth. Met. 1997, 85, 1425. (e) Tedeschi, C.; Caruso, F.; Mohwald, H.; Kirstein, S. J. Am. Chem. Soc. 2000, 122, 5841. (14) Donath, E.; Sukhorukov, G. B.; Caruso, F.; Davis, S. A.; Mohwald, M. F. Angew. Chem. Int. Ed. 1998, 37, 2202. (15) Sukhorukov, G. B.; Donath, E.; Davis, S. A.; Lichtenfeld, H.; Caruso, F.; Popov, V. I.; Mohwald, M. F. Polym. AdV. Technol. 1998, 9, 759. (16) Gittins, D. I.; Caruso, F. J. Phys. Chem. B 2001, 105, 6846. (17) Kato, N.; Schuetz, P.; Fery, A.; Caruso, F. Macromolecules 2002, 35, 9780. (18) Schneider, G.; Decher, G. Nano Lett. 2004, 4, 1833. (19) Messina, R.; Holm, C.; Kremer, K. Langmuir 2003, 19, 4473.
Langmuir, Vol. 22, No. 9, 2006 4199 Table 1. Composition of Silica Suspensions Used in This Study
material
mean diameter (nm)
concentration (%, w/w)
supply pH
specific surface area (m2 g-1)
Nalco 1130 Nalco 1140 Nalco 1050 Nalco 2329
8 15 20 90
30 40 50 40
10 9.7 9 10.8
375 200 150 40
successful integration of surface-modified polyelectrolyte colloids in biological and other highly complex systems. In particular, the aforementioned information is necessary for the successful integration of nanoparticle polyelectrolyte assemblies for in vivo applications. One of the primary goals of our research is to explore methods to prepare and develop surface-modified nanoparticulate materials that are capable of carrying biological, pharmaceutical, or diagnostic components. The components, which might include drugs, therapeutics, diagnostics, and targeting moieties, could potentially be delivered via the nanocarrier and directed to diseased tissue, tumors, or bones. This approach has promised to improve significantly the treatment of cancers and other lifethreatening diseases and may revolutionize their clinical diagnosis and treatment.20 The “components” that may be carried by the nanoparticles can be attached to the nanoparticle by well-known bioconjugation techniques, the most common of which involves conjugation or linking to an amine functionality.21 Thus, polyelectrolyte colloidal assemblies might be ideal candidates as nanoparticle carriers. However, many challenges remain regarding their integration in in vivo applications. Simply regarding conjugation of useful biological molecules, the colloidal assemblies must be robust and stable under a variety of pH conditions and salt concentrations, particularly under physiological conditions, and avoid detection by the immune system, and finally, the charge and surface characteristics of the carrier must be precisely controlled to prevent nonspecific adsorption when introduced into a biological system. Experimental Section Materials. Aqueous suspensions of silica were purchased from Nalco Chemical Company. The mean particle diameters range from 8 to 90 nm, and details are given in Table 1. Polyethylenimines (PEIs) were branched polymers purchased from Aldrich Chemicals and were of average molecular weight Mn ) 1800 (herefter referred to as 2 kDa), 10 000, and 60 000 Da, with corresponding degrees of polymerization of 46.5, 233, and 1395 monomers/mol, respectively. The molecular weight of an “average” monomer of PEI was taken to be 43.0 Da. The units of PEI concentration (µmol m-2, with micromoles referring to the EI monomer concentration and m-2 referring to the silica surface area) were chosen to keep the ratios independent of the polymer molecular weight, core particle size, and core particle concentration. In the text, the term “shelling ratio” is simply the nominal (or beginning) amount of PEI added (in µmol of monomer m-2) to the reaction mixture; it does not describe the adsorbed amount of PEI under any given conditions. Bisethene,1,1′-[methylenebis(sulfonyl)] (BVSM) was obtained from Eastman Kodak Company. A phosphate buffer system (PBS) was prepared in our laboratory and consisted of an aqueous solution of 0.137 M NaCl, 0.0027 M KCl, 0.010 M Na2HPO4, and 0.002 M KH2PO4 at pH 7.4. Sample Preparation. Procedure for Controlled, Simultaneous Assembly (CSA). The CSA procedure involves the simultaneous (20) Weisslener, R.; Ntziachristos, V. Nat. Med. 2003, 9, 123. (b) Mahmood, U.; Weissleder, R. Mol. Cancer Ther. 2003, 2, 489. (21) Hermanson, G. T. In Bioconjugate Techniques; Academic Press: San Diego, CA, 1996.
4200 Langmuir, Vol. 22, No. 9, 2006 addition at controlled rates of two or more fluids into a region of controlled geometry and with constant mixing.22 Methods for preparing PEI-silica composites based on core silica particles with a surface area of 30 m2 g-1, with a PEI amount of 20 µmol m-2, and with a final concentration of the composite of 33.1% (w/w) are described below. Similar procedures were followed for other shelling ratios, other PEI molecular weights, and other silica surface areas. Into a 2 L container containing 200 mL of distilled water, which was stirred with a prop-like stirrer at a rate of about 2000 rpm, was simultaneously added (i) 1548.0 g of a 40% (w/w) silica colloid core particle (Nalco 2329, 92 nm) at a rate of 40.0 mL min-1; (ii) 158.8 g of a 10% (w/w) solution of PEI (Mn ) 2 kDa, adjusted to pH 5.0 with nitric acid) at a rate of 5.3 mL min-1, each for 30 min; and (iii) a 1.0 N solution of nitric acid, added at a rate necessary to keep the pH pinned at or near pH 5.0. The addition rates were controlled using calibrated peristaltic pumps. The rates were set to keep the ratio of PEI/silica constant throughout this process. CSA at Varying pH. The above procedure was varied in experiments to determine the effect of pH upon the assembly process. Reagent concentrations were adjusted to achieve a final concentration of solids of about 15.0%, and the initial PEI solution (10%, w/w) was not preadjusted to pH 5.0 but was left at its nominal pH of 10.8. HNO3 (1.0 N) was added during the controlled simultaneous addition at a rate necessary to keep the pH pinned at the values given in the Results. Effect of Colloid Size and Polymer Molecular Weight. The particle size (d ) 8-90 nm) of the silica substrate was varied, as well as the PEI molecular weight (2-60 kDa), to determine the facility of the CSA method. In this series of experiments, the starting materials (PEI and silica) were diluted to 10% (w/w) with distilled water and the pH of each was adjusted to 5 with 1 N HNO3. For each experiment, 200 g of distilled water was placed into a 1 L container. A total of 200 g of 10 wt % colloidal silica at pH 5 was measured out and placed into a separate container. The calculated amount of 10 wt % PEI at pH 5 was also measured out into a graduated cylinder and was added at a rate corresponding to shelling ratios of 10, 20, and 30 µmol m-2 of the silica surface area. The system was assembled using the CSA method. During CSA, the pH of the solution was monitored with a handheld pH-meter, and the pH was pinned at pH 5 with the controlled addition of 1 N HNO3. The colloidal suspensions produced were approximately 5% solids by weight. Particle-size analysis was completed for each sample 1, 7, and 14 days after assembly. While aging, suspensions were kept at rest in plastic containers at room temperature, and before each analysis, the solutions were shaken lightly by hand. Effect of Cross-Linking. Polyelectrolyte assembly onto 15 nm silica particles was performed as described above. The resulting polyelectrolyte colloidal assembly was cross-linked by the following general procedure. Into a 1.0 L container containing 200 mL of distilled water, which was stirred with a prop-like stirrer at a rate of about 2000 rpm, was simultaneously added (i) 200 g of a PEI colloidal assembly [PEI (Mn ) 2 kDa) assembled onto a 15 nm silica colloid at 20 µmol m-2], at a rate of 20.0 mL min-1 and to cross-link the PEI formed on the particles, and (ii) 59.7 g of a 0.45% solution of BVSM cross-linking reagent at 6 mL min-1 each for 10 min. The rates were set to keep the ratio of BVSM/PEI (polymer) constant at 3:1 (mol/mol). Methods. Particle Sizing. Particle-size distributions of colloids were measured by a dynamic light-scattering technique using a Microtrac Ultrafine Particle Analyzer (UPA) 150 from Leeds and Northrop. The light-scattering measurements were performed using a 780 nm laser diode in a backscattering geometry (θ ) 180°). The backscattering geometry was employed to reduce the effects of multiple scattering. The measured signal was processed using a frequency-based power spectrum. Samples were measured at a solids concentration of about 1-2 wt %, at the indicated pH. A dilution series of a single sample to concentrations as low as 0.05 wt % did (22) Bringley, J. F. U.S. Patent application 10/622354 A1, 2005. (b) Bringley, J. F.; Qiao, T.; Patton, D. U.S. Patent pending, 2005. (a) Bringley, J. F., Broadus, K.; Shaw-Klein, L.; Barber, G. N. U.S. Patent application 10/622230 A1, 2005.
Bringley et al. not significantly affect the measured particle-size distribution. Corrections for background scattering because of unadsorbed PEI were not performed, but it was determined that unadsorbed PEI contributed negligbly to the overall scattering. The volume-weighted mean particle diameter (d) was calculated from the area distribution of the particle size, as described in the instrument manual. The standard deviation (σd) describes the width of the particle-size distribution. Electrophoresis. Electrophoretic mobility measurements were made using a Malvern zetasizer. Suspensions of the colloidal materials (0.010%, w/v) were prepared using deionized distilled water. The pH was adjusted using 0.1 M HCl or 0.1 M KOH. The ζ potential (ζ) of the particles was computed using the Henry equation UE )(ζ/1.5η)(f(Ka))
(1)
where η ) the viscosity of the medium, is the dielectric constant, UE is the electrophoretic mobility of the particle, and f(Ka) is Henry’s function. The experiments were performed on samples as prepared, having salt concentrations between about 0.01 and 0.001 M, but later repeated at a constant ionic strength (0.135 M NaCl). Centrifugation. An ultracentrifuge was used to separate the colloidpolymer assemblies from the supernatant to measure the composition of the aqueous phase. This was achieved with a L8-60 M ultracentrifuge (Beckman Coulter, Inc., Fullerton, CA) operating at 25 000 rpm for 15 h. Cryogenic Transmission Electron Microscopy (Cryo-TEM). Samples were prepared in a controlled environment vitrification system. The chamber temperature was 25 °C, and the relative humidity was kept close to saturation to prevent evaporation from the sample during preparation. A 10 µL sample drop was placed on a carboncoated holey film supported by a copper grid and gently blotted with filter paper to obtain a thin biconcave liquid film (∼20-400 nm) spanning the holes in the grid. The grid was rapidly plunged into liquid ethane at -180 °C and transferred into liquid nitrogen (-196 °C). The vitrified specimens were stored in liquid nitrogen and transferred into a Philips CM20 TWIN microscope equipped with cryoblades (Gatan 651) mounted in a cryotransfer holder (Gatan 626) from its workstation. The acceleration voltage was 120 or 200 kV, and the working temperature was kept below -175 °C. Images were recorded digitally with a CCD camera (Gatan SSC 694) under low-dose conditions (
