pubs.acs.org/Langmuir © 2010 American Chemical Society
Biphasic Synthesis of Au@SiO2 Core-Shell Particles with Stepwise Ligand Exchange Mathias Schulzendorf,† Christian Cavelius,† Philip Born, Eoin Murray, and Tobias Kraus* Leibniz Institute for New Materials (INM), Structure Formation group and Nano Cell Interaction group, Campus D2 2, 66123 Saarbruecken, Germany. †These authors contributed equally to this work Received June 30, 2010. Revised Manuscript Received November 12, 2010 We report the synthesis of well-dispersed core-shell Au@SiO2 nanoparticles with minimal extraneous silica particle growth. Agglomeration was suppressed through consecutive exchange of the stabilizing ligands on the gold cores from citrate to L-arginine and finally (3-mercaptopropyl)triethoxysilane. The result was a vitreophilic, stable gold suspension that could be coated with silica in a biphasic mixture through controlled hydrolysis of tetraethoxysilane under L-arginine catalysis. Unwanted condensation of silica particles without gold cores was limited by slowing the transfer across the liquid-liquid interface and reducing the concentration of the L-arginine catalyst. In-situ dynamic light scattering and optical transmission spectroscopy revealed the growth and dispersion states during synthesis. The resulting core-shell particles were characterized via dynamic light scattering, optical spectroscopy, and electron microscopy. Their cores were typically 19 nm in diameter, with a narrow size distribution, and could be coated with a silica shell in multiple steps to yield core-shell particles with diameters up to 40 nm. The approach was sufficiently controllable to allow us to target a shell thickness by choosing appropriate precursor concentrations.
Introduction Silica shells render gold nanoparticles virtually inert and protect the soft, high-energy metal surface.1,2 The optical activity of the core particle is retained3 and can be complemented by dyes incorporated in the shell.4,5 Molecules adsorbed on the gold surface cannot desorb through the shell and are protected from the solvent.5 If interaction with solutes is desired, the silica shell provides ample opportunities for the introduction of linker molecules.6,7 Silica is a weak acid and has a Hamaker constant below that of gold,8 thus forming colloids that are more easily stabilized than gold colloids. Au@SiO2 core-shell particles are thus a robust and versatile alternative to pure gold. However, gold nanoparticle syntheses are reliable and yield particles having low dispersity using very simple procedures,9-11 a critical advantage that core-shell syntheses should retain. Gold nanoparticles are often synthesized in aqueous solvents by the reduction of chloroauric acid with sodium borohydride9 or citrate10 in the presence of a stabilizer.9,10 The surfaces of the resulting particles are charged by weakly adsorbed molecules and are easily modified by strongly adsorbing molecules such as thiols, more easily so than the hydrophobic particles synthesized using the method developed by Brust et al.9 Silica, however, will not adsorb readily on either of these particles as there is little
interaction between silica and citric acid or gold: the particle surfaces are vitreophobic. Liz-Marzan et al.3 were the first to replace the original stabilizing ligands with more vitreophilic molecules (in their case, (3-aminopropyl)trimethoxysilane, APTS), and several groups have followed their lead, using dye-labeled APTS12 or poly(vinylpyrrolidone) (PVP)13 to increase the interaction between the gold surface and the reactive silica precursors. The kinetics of ligand exchange reactions on gold have previously been monitored using NMR,14 EPR,15 SERS,16 and UVvis spectroscopy of the surface plasmon resonance.17 From these techniques it has been shown that ligand exchange on gold nanoparticles is generally initially rapid and slows to give an overall second-order reaction.18 This process can be mechanistically compared to ligand exchange reactions on inorganic complexes, in that it shows both associative and dissociative processes.19 Recently, it has been suggested that there is also some degree of diffusion control in the process.20 The apparent complexity of the exchange reaction, especially on smaller particles, is mainly due to the nonuniform morphology (facets, edges, curvature, etc.) and, hence, reactivity of the particle surface. A previous study on the effect of ligand chain length on affinity showed that gold nanoparticles follow the HSAB (hard-soft acid-base) principle.17 In the current study we exploit this principle, for example to replace citric acid on gold (a weak acid) by L-alanine.
*To whom correspondence should be addressed. E-mail: tobias.kraus@ inm-gmbh.de. (1) Giersig, M.; Ung, T.; Liz-Marzan, L. M.; Mulvaney, P. Adv. Mater. 1997, 9, 570. (2) Chen, M. M. Y.; Katz, A. Langmuir 2002, 18, 8566. (3) Liz-Marzan, L. M.; Giersig, M.; Mulvaney, P. Langmuir 1996, 12, 4329. (4) Aslan, K.; Wu, M.; Lakowicz, J. R.; Geddes, C. D. J. Am. Chem. Soc. 2007, 129, 1524. (5) Ow, H.; Larson, D. R.; Srivastava, M.; Baird, B. A.; Webb, W. W.; Wiesner, U. Nano Lett. 2005, 5, 113. (6) Philipse, A. P.; van Bruggen, M. P. B.; Pathmamanoharam, C. Langmuir 1994, 10, 92. (7) Badley, R.; Ford, W.; McEnroe, F.; Assink, R. Langmuir 1994, 6, 792. (8) Visser, J. Adv. Colloid Interface Sci. 1972, 3, 331. (9) Frens, G. Nat. Phys. Sci. 1973, 241, 20. (10) Brust, M.; Walker, M.; Bethell, D.; Schiffrin, D.; Whyman, W. Chem. Commun. 1994, 801. (11) Zheng, N.; Fan, J.; Stucky, G. D. J. Am. Chem. Soc. 2006, 128, 6550.
Langmuir 2011, 27(2), 727–732
(12) Van Blaaderen, A.; Vrij, A. Langmuir 1992, 8, 2921. (13) Graf, C.; Vossen, D. L. J.; Imhof, A.; van Blaaderen, A. Langmuir 2003, 19, 6693. (14) Woehrle, G. H.; Brown, L. O.; Hutchison, J. E. J. Am. Chem. Soc. 2005, 127, 2172. (15) Lucarini, M.; Franchi, P.; Pedulli, G. F.; et al. J. Am. Chem. Soc. 2004, 126, 9326. (16) Feng, Y.; Xing, S.; Xu, J.; et al. Dalton Trans. 2010, 39, 349. (17) Ghosh, S. K.; Nath, S.; Kundu, S.; Esum,i, K.; Pal, T. J. Phys. Chem. B 2004, 108, 13963. (18) Guo, R.; Song, Y.; Wang, G.; Murray, R. W. J. Am. Chem. Soc. 2005, 127, 2752. (19) Caragheorgheopol, A.; Chechik, V. Phys. Chem. Chem. Phys. 2008, 10, 5029. (20) Kassam, A.; Bremner, G.; Clark, B.; Ulibarri, G.; Lennox, R. B. J. Am. Chem. Soc. 2006, 128, 3476.
Published on Web 12/13/2010
DOI: 10.1021/la102630y
727
Article
Schulzendorf et al.
Figure 1. Biphasic synthesis route for the preparation of Au@SiO2 nanoparticles.
Figure 3. UV-vis spectra recorded for gold nanoparticles incubated for 24 h in 0.5 mM L-arginine solution after the addition of sufficient MPTS to form 0.5, 1, 2, and 4 monolayers.
Figure 2. UV-vis spectra recorded for gold nanoparticles incubated for 24 h in ammonia (a) and L-arginine solutions at concentrations between 0.5 and 10 mM (b). The asterisk and double asterisk in (a) indicate the evolution of additional plasmon peaks attributed to the agglomeration of the gold nanoparticles at the given concentrations.
Nanoparticle coating requires the mixing of the particles into a reactive solution while preserving their stability and integrity. 728 DOI: 10.1021/la102630y
Ligands that are vitreophilic are not necessarily efficient in stabilizing the particles in these mixtures. This problem is exacerbated if the solvent has to be adapted to the silica precursor. Typically, tetraethoxysilane (TEOS) is used as a precursor in the classic St€ober process for the synthesis of pure silica nanoparticles.21 TEOS is not soluble in water and requires the addition of ethanol (EtOH). To avoid agglomeration, one can first coat the gold particles with a thin silicate layer from a sodium silicate solution3 and only then transfer to EtOH. It is also possible to grow shells exclusively from silicate, but it is difficult to obtain thick, uniform shells without using TEOS.2 Even obtaining a thin, but reproducible, silica shell from silicate solutions is challenging.13 PVP can also stabilize gold particles in EtOH and is reasonably vitreophilic13 but remains in the core-shell particle as a thick and poorly defined interface layer. The approach presented in this paper minimizes agglomeration by successively replacing ligands such that particles remain as stable as possible and by entirely avoiding the use of (agglomeration-prone) EtOH. Instead, silanes are added to a (21) St€ober, W.; Fink, A.; Bohn, E. J. Colloid Interface Sci. 1968, 26, 62.
Langmuir 2011, 27(2), 727–732
Schulzendorf et al.
Article
Figure 4. Schematic drawing of the ligand exchange at the surface of the gold nanoparticles prior to shell formation.
Figure 5. Normalized UV-vis spectra recorded for gold nanoparticles before synthesis and Au@SiO2 core-shell particles under synthesis conditions at an L-arginine concentration of 0.5 mM. The shift of the surface plasmon peak is due to silica shell growth after addition of TEOS to a MPTS-containing gold nanoparticle solution.
second, unpolar phase in a biphasic solution with the aqueous particle solution and then transferred into the aqueous suspension, following a strategy originally devised by Hartlen et al. for the synthesis of pure silica particles22 (Figure 1). Gold nanoparticles prepared by the method described by Frens et al.7 (reduction of HAuCl4 with an excess of citrate) are exposed to L-arginine, which displaces the citrate on their surface and makes the particles stable under the slightly basic conditions required in the next step. Cyclohexane is layered on top of the suspension, and a mercaptosilane ((3-mercaptopropyl)triethoxysilane, MPTS) is added. The silane is slowly hydrolyzed under alkaline conditions and replaces the L-arginine, thereby producing a stable, vitreophilic gold colloid. Finally, TEOS is added to the organic phase. The silane is hydrolyzed, enters the aqueous phase, and rapidly binds to the silanol groups on the gold particles. Depending on the TEOS concentration, silica shells of variable thickness are formed with a well-defined MPTS interlayer.
Materials and Methods Chemicals. All solvents and chemicals in this work were used without further purification unless otherwise stated. Ultrapure Milli-Q water (R > 18 MΩ) was used for all aqueous preparations. (22) Hartlen, K. D.; Athanasopoulos, A. P. T.; Kitaev, V. Langmuir 2008, 24, 1714.
Langmuir 2011, 27(2), 727–732
(3-Mercaptopropyl)triethoxysilane (purity >92%) (MPTS) was purchased from ABCR, Germany. Cyclohexane (puriss., purity >99.5%) and L-arginine (reagent grade, purity g98%) were purchased from Sigma-Aldrich, Germany. Gold(III) chloride hydrate was prepared by dissolving a 1 g of gold ingot (purity >99.99%) in 20 mL of freshly prepared aqua regia and subsequent evaporation of the liquid at boiling temperature. Tetraethoxysilane (puriss., purity>99%) (TEOS) was purchased from Fluka, Germany. Trisodiumcitrate dihydrate (pro analysi, purity >99%) was purchased from Merck, Germany. Purification of particles by dialysis was performed using cellulose hydrate membranes (NADIR, Carl Roth, Germany) with a molecular weight (MW) cutoff of 10-20 kDa, equivalent to a mean pore size of 25-30 A˚. Synthesis of Gold Nanoparticles. Gold nanoparticles with a mean diameter of 20 nm were prepared using the method introduced by Frens.2 In a typical synthesis, 26 mg (76.5 μmol) of HAuCl4 were dissolved in 250 mL of water and heated. Upon boiling, 5 mL of a 34.3 mM solution of trisodium citrate tetrahydrate were added to induce particle formation. During prolonged heating under reflux for 10 min, the color of the solution turned from yellow to wine red. After filtration through a 0.22 μm membrane and dialysis against ultrapure water, 20 nm diameter citrate-stabilized gold nanoparticles had formed. Synthesis of Core-Shell Nanoparticles. 0.5 mL of a 20.7 mM L-arginine solution was added to 20 mL of the freshly prepared gold colloid to achieve an overall amino acid concentration of 0.5 mM. The solution was transferred into a 100 mL three-neck round-bottom flask via a 0.22 μm cellulose triacetate filter and covered with a layer of 5 mL cyclohexane to yield a biphasic system. While stirring at 90 rpm, 6.2 μL (32.8 μmol) of MPTS were added to form a MPTS monolayer on the nanoparticle surface. After stirring at room temperature overnight, the solution was heated to 60 °C, and TEOS was added to the cyclohexane layer to form a silica shell. This procedure was repeated several times to yield particles with silica shells of defined size. The color of the solution changed from red to red-violet during shell growth.
Characterization. Transmission Electron Microscopy (TEM). A CM 200 (200 keV) electron microscope (Philips, Netherlands) was used to take TEM images. Samples were prepared by dipping a 200 mesh carbon-coated copper grid into the nanoparticle suspension and allowing the grid to dry in air. Images were analyzed for particle size distributions using the ImageJ software from the National Institutes of Health. A representative number of between 20 and 100 particles per TEM image were counted and their sizes measured. The histograms were fitted to Gaussian distributions to obtain mean and distribution of the particle sizes. Dynamic Light Scattering (DLS). A Dyna Pro Titan instrument (Wyatt Technology, Wyatt Technology Europe GmbH, Germany) with a laser wavelength of 831 nm was used to determine particle size distributions via dynamic light scattering. Approximately 0.6 mL of each sample was filtered through a DOI: 10.1021/la102630y
729
Article
Schulzendorf et al.
Table 1. Core-Shell Particles Obtained from Gold Cores (First Line) after Four Growth Steps with 0.5 mM L-Arginine
sample gold core first shell second shell third shell fourth shell
TEOS conca [mM]
hydrodynamic radiusb (DLS) [nm]
TEM diameter [nm]
shell thickness from TEMc [nm]
plasmon peak position [nm]
4 15
23.2 ( 3.2 25.6 ( 3.9 28.4 ( 6.1
18.9 ( 2.6 21.0 ( 0.4 26.2 ( 3.4
1.07 ( 0.02 3.7 ( 0.5
521 524 528
60
33.2 ( 4.2
32.6 ( 2.9
6.9 ( 0.6
532
98
56.4 ( 13.8
38.9 ( 1.0
10.0 ( 0.3
534
a The TEOS concentration in the system was adjusted by addition of pure TEOS to the cyclohexane phase. b The hydrodynamic radius was estimated from DLS results. c Shell thicknesses were calculated by subtracting the core diameter (which remains unchanged after shell growth) from the total diameter obtained by TEM.
0.22 μm membrane filter prior to measurement. Wyatts’ Dynamics software, which uses a proprietary non-negative least-squares algorithm based on an inverse Laplace transformation, was used to fit all autocorrelation data. UV-vis Characterization. A Cary 5000 photospectrometer (Varian Inc.) was used to record spectra in the wavelength range from 300 to 800 nm. Approximately 2 mL of the respective sample was measured in quartz cuvettes with a beam path 1 cm in length.
Inductive Coupled Plasma-Optical Emission Spectroscopy. An ICP OES Ultima 2 (Horiba Jobin Yvon, Horiba Jobin Yvon Gmbh, Germany) was used to measure the gold content of the solutions.
Results and Discussion Gold Nanoparticles. The gold nanoparticles obtained by citrate reduction exhibit a narrow size and shape distribution. A mean diameter of 19 ( 3 nm was calculated by image analysis of TEM images (Figure 8), whereas a hydrodynamic diameter of 23 ( 3 nm was obtained by DLS measurements. The particles are negatively charged with a zeta-potential of -40 mV due to the citrate capping of the gold nanoparticles. Colloidal Stability under Synthetic Conditions. The simplest catalyst for the formation of silica particles from silane is ammonia.21 Ammonia is, however, known to destabilize citratecapped gold nanoparticles suspensions due to its high ionic strength at the concentrations necessary for silane hydrolysis.16 An alternative are amino acids: several researchers (including Hartlen) describe the growth of silica particles at small concentrations of L-arginine between 0.5 and 2 mM.22-24 We investigated the stability of our gold seeds under different silica-shell growth conditions by recording UV-vis spectra of the seeds incubated in ammonia and L-arginine solutions for 24 h using concentrations between 0.5 and 10 mM (Figure 2). As expected, citrate-capped gold seeds show a low stability against agglomeration even at low ammonia concentrations. The absorption maximum caused by the gold particles’ surface plasmon resonance, originally at around 522 nm, decreases upon the addition of ammonia at concentrations between 0.5 and 10 mM. Additional bands (marked with asterisk and double asterisk in Figure 2) appear in the red and near-IR region of the spectrum, (23) Yokoi, T.; Sakamoto, Y.; Terasaki, O.; et al. J. Am. Chem. Soc. 2006, 128, 13664. (24) Davis, T. M.; Snyder, M. A.; Krohn, J. A.; Tsapatsis, M. Chem. Mater. 2006, 18, 5814. (25) Mulvaney, P. Langmuir 1996, 12, 788.
730 DOI: 10.1021/la102630y
Figure 6. Grown shell volume over original TEOS concentration, indicating a constant growth rate that enables targeting core-shell nanoparticles with an adjustable shell thickness. The L-arginine concentration was fixed at 0.5 mM. Numbers above the data points indicate the evolution of shell thickness with increasing TEOS concentration.
indicating the formation of particle agglomerates.26,27 In contrast, gold nanoparticles incubated with L-arginine did not agglomerate at L-arginine concentrations up to 2 mM. At higher concentrations, a broadening of the surface plasmon resonance peak and a shift of the center toward longer wavelengths was observed. First Ligand Exchange. L-Arginine is not only a catalyst for silica formation but also a ligand capable of stabilizing gold colloids.28 With its isoelectric point at 10.76, L-arginine should be a more suitable stabilizer for gold nanoparticles during silane hydrolysis.29 We replaced the citrate-containing original solvent of the particles with a 0.5 mM L-arginine solution. The UV-vis traces shown in Figure 3 indicate a slight red shift of the single particle surface plasmon peak, likely caused by the adsorption of 25,30 L-arginine. Thus, a nanoparticle suspension was formed that could withstand silane hydrolysis. Second Ligand Exchange. L-Arginine is a catalyst for silica formation,22 but not vitreophilic itself;3 we replaced it with a second ligand in the subsequent step but retained the amino acid in the solvent. The necessary concentration of the vitreophilic ligand MPTS was calculated employing the approach of Liz-Marzan,3 which used (3-aminopropyl)triethoxysilane (APTS) as stabilizer to coat citrate-capped gold nanoparticles with a thin vitreophilic silica shell. Gold content and particle size were estimated by ICP-PES and TEM. The diameter was used to calculate the mass of one gold nanoparticle, and the particle concentration was obtained by dividing the gold content by the single particle mass. Shell growth experiments were conducted with a fixed particle concentration equivalent to 0.3 nM of gold. The amount of MPTS used was that calculated to form a monolayer on gold particles at slightly basic pH. No particle agglomeration was observed after this final ligand exchange, even (26) Fujiwara, H. F.; Yanagida, S.; Kamat, P. V. J. Phys. Chem. B 1999, 103, 2589. (27) Thomas, K. G.; Kamat, P. V. J. Am. Chem. Soc. 2000, 122, 2655. (28) Bhargawa, S. K.; Booth, J. M.; Agrawal, S.; Coloe, P.; Kar, G. Langmuir 2005, 21, 5949. (29) Russel, W. B.; Saville, D. A.; Schowalter, W. R. Colloidal Dispersions; Cambridge University Press: Cambridge, 1989; Chapter 8. (30) Selvakannan, P. R.; Mandal, S.; Phadtare, S.; Pasricha, R.; Sastry, M. Langmuir 2003, 19, 3545.
Langmuir 2011, 27(2), 727–732
Schulzendorf et al.
Article
Figure 7. TEM micrographs of Au@SiO2 core-shell nanoparticles (a-d) obtained after different subsequent growth steps. Image (e) shows particles obtained after a 3-fold scale-up of the reaction which yielded increased nanoparticle numbers with otherwise unchanged properties, while (f) shows the effect of excessively high initial TEOS concentrations which lead to the formation of empty silica nanoparticles. Scale bars are 100 nm; the insets show high-resolution micrographs.
when enough silane was added to coat the particles with 0.5-1 monolayer of MPTS at L-arginine concentrations between 0.5 and 2 mM. This is in agreement with the results of Marzan et al. for APTS (Figure 3). In two exchange steps, the gold surface was consecutively modified with ligands having an increasing affinity to gold in the order of sodium citrate < L-arginine < (3-mercaptopropyl)triethoxysilane (Figure 4).31 Silica Shell Formation. The ligand-exchanged nanoparticle suspension is sufficiently stable and vitreophilic to grow a silica layer on the particle surfaces. Shell formation was induced by addition of TEOS to the MPTS-capped gold (Figure 5, Table 1). We chose a biphasic mixture to ensure the solubility of TEOS, limit its rate of hydrolysis and prevent homogeneous nucleation of silica from excess silicic acid.22 Unpolar TEOS was added to the cyclohexane phase and was slowly hydrolyzed at the cyclohexane/ water interface by L-arginine from the aqueous phase. A constant supply of water-soluble, active silicic acid is thus produced and transferred to the aqueous phase. The silica shell is slowly formed on the gold particles, avoiding the growth of additional silica particles due to excess silicate and homogeneous silica shell growth. The flux of the reactive silica source is tunable by variation of the L-arginine concentration, which governs the hydrolysis rate of TEOS precursor. We calculated the amount of TEOS that was required to form a shell from the data obtained from the synthesis of pure silica nanoparticles of the same volume after Hartlen. Careful subsequent additions of TEOS yielded
(31) Pearson, R. G. J. Am. Chem. Soc. 1963, 85, 3533.
Langmuir 2011, 27(2), 727–732
shells with thicknesses of 1-10 nm. In four growth steps, a final particle size of 40 nm could be obtained (Table 1). Silica shells red shift the surface plasmon peak of gold particles,3,32 an effect we exploited by measuring UV-vis spectra in situ as the shell diameter increased. We observed a shift in peak position from an initial 523 nm at a particle diameter of 18.9 nm (MPTS-capped gold) to 534 nm at a particle diameter of 38.9 nm (Figure 5). The TEM radii given in Table 1 allow the calculation of the average volume of the shell. A growth rate was interpolated from the volume, and the volume-growth per step was calculated. We assumed an original volume for the gold core of 3503 ( 384 nm3. A plot of the grown volume (ΔV) against the added amount of precursor per volume cTEOS (proportional to the number of particles) indicates a linear growth (Figure 6). However, simply adding the amount of TEOS theoretically required to grow a 50 nm shell in a single step did not succeed. The initial TEOS concentration from such an injection (cTEOS = 530 mM) likely is too high, resulting in the formation of pure silica particles (Figure 7f). Particle Geometry and Dispersity. Figure 7a-f shows TEM images of Au@SiO2 nanoparticles obtained from samples prepared by the described method. The gold seeds obtained by citrate reduction show an acceptable size and shape distribution. The TEM diameter of 18.9 ( 2.6 nm was in good agreement with the hydrodynamic diameter from DLS of 23.2 ( 3.2 nm. The content of silica particle without gold cores was below the detection limits of TEM and DLS. (32) Lu, Y.; Yin, Y.; Li, Z. Y.; Xia, Y. Nano Lett. 2002, 2, 785.
DOI: 10.1021/la102630y
731
Article
Schulzendorf et al.
Figure 8. Schematic drawing of the ligand exchange at the surface of the gold nanoparticles prior to shell formation.
Overall, the hydrolysis of the ethoxy groups in silane triols is catalyzed by L-arginine, which reacts as a base in aqueous solutions due to the guanidine groups present in the molecule.19 As intended, L-arginine also promotes the formation of a negatively charged monolayer of silane triols from MPTS and acts as a stabilizer prior to the ligand exchange due to its high affinity to gold surfaces (Figure 8). Thus, rather well-dispersed core-shell particles are formed. While the problem of stability in the formed particles has been overcome, it remains important to limit the concentration of reactive TEOS in the system. Even in a biphasic mixture with its limited hydrolysis rate, excessive silicic acid rapidly forms silica particles.
Conclusions and Outlook Consecutive ligand exchange is a gentle and controllable route to well-dispersed core-shell particles. As the process steps are consecutive, the synthesis can be optimized stepwise, limiting the complexity of each task. When combined with a relatively slow
732 DOI: 10.1021/la102630y
two-phase system for silica formation, this leads to a synthetic route that is less sensitive to the details of gold particle seed, exact timing, or details of mixing than one-step approaches. We have shown that it allows the preparation of well-controlled shells, and it remains possible to pinpoint problems in the synthesis (such as excessive TEOS concentration). This strategy should be compatible both with continuous synthesis (for example, in capillarybased microreactors) and with other core particles as long as suitable ligands can be found. Acknowledgment. The authors thank Annette Kraegeloh for her support of this project, Anika Weber for titration series to analyze colloidal stability, and Eduard Arzt for his continuing support. Funding from the Deutsche Forschungsgemeinschaft (DFG) is gratefully acknowledged. Supporting Information Available: Figures 1 and 2. This material is available free of charge via the Internet at http:// pubs.acs.org.
Langmuir 2011, 27(2), 727–732
