Langmuir 2005, 21, 4277-4280
4277
Sto1 ber Synthesis of Monodispersed Luminescent Silica Nanoparticles for Bioanalytical Assays Liane M. Rossi,†,‡ Lifang Shi,† Frank H. Quina,‡ and Zeev Rosenzweig*,† Department of Chemistry and the Advanced Material Research Institute (AMRI), University of New Orleans, New Orleans, Louisiana 70148, and Institute of Chemistry, University of Sa˜ o Paulo, Sa˜ o Paulo, SP 05508-900 Brazil Received February 15, 2005. In Final Form: March 7, 2005 We have developed a simple method to prepare bright and photostable luminescent silica nanoparticles of different sizes and narrow size distribution in high yield. The method is based on the use of Sto¨ber synthesis in the presence of a fluorophore to form bright silica nanoparticles. Unlike micro-emulsion-based methods often used to prepare luminescent silica particles, the Sto¨ber method is a one-pot synthesis that is carried out at room temperature under alkaline conditions in ethanol:water mixtures and avoids the use of potentially toxic organic solvents and surfactants. Our luminescent particles contained the transition metal complex tris(1,10-phenanthroline) ruthenium(II) chloride, [Ru(phen)3]Cl2. They showed higher photostability and a longer fluorescence lifetime compared to free Ru(phen)3 solutions. Leakage of dye molecules from the silica particles was negligible, which was attributed to strong electrostatic attractions between the positively charged ruthenium complex and the negatively charged silica. To demonstrate the utility of the highly luminescent silica nanoparticles in bioassays, we further modified their surface with streptavidin and demonstrated their binding to biotinylated glass slides. The study showed that digital counting of the luminescent nanoparticles could be used as an attractive alternative to detection techniques involving analogue luminescence detection in bioanalytical assays.
Introduction Luminescent nanoparticles have been used in recent years as an alternative to organic fluorophores as luminescent labels in biological systems. They offer significant advantages compared to free dyes.1,2 For example, the encapsulation of luminescent molecules in nanoparticles often increases their photostability and emission quantum yield due to their isolation from possible quenchers such as molecular oxygen and water.3,4 Silica nanoparticles have emerged as a particularly attractive host matrix due to the relative ease of functionalization of the particles’ surface. This enabled their conjugation to bioactive molecules for targeting purposes.1-10 When organic fluorophores were attached to the particles covalently, their synthesis involved pre-modification of the dyes with organoalkoxysilanes.5,11 Recently, Trau and co-workers described the synthesis of porous functionalized silica particles using the Sto¨ber method.12 Fluorescent dyes were incorporated into the particles during the synthesis through the addition of 3-aminopropyl trimethoxysilane which was coupled to isothiocyanate-modified dyes. How† ‡
University of New Orleans. University of Sa˜o Paulo.
(1) Bagwe, R. P.; Zhao, X.; Tan, W. J. Dispersion Sci Technol. 2003, 24, 453. (2) Schuetz, W.; Caruso, F. Chem. Mater. 2002, 14, 4509. (3) Santra, S.; Zhang, P.; Wang, K. M.; Tapec, R.; Tan, W. H. Anal. Chem. 2001, 73, 4988. (4) Santra, S.; Wang, K.; Tapec, R.; Tan, W. H. J. Biomed. Opt. 2001, 6, 160. (5) Zhao, X.; Tapec-Dytioco, R.; Tan, W. J. Am. Chem. Soc. 2003, 125, 11474. (6) Tapec, R.; Zhao, X. J.; Tan, W. J. Nanosci. Nanotechnol. 2002, 2, 405. (7) Schardl, W. Adv. Mater. 2000, 12, 1899. (8) He, X.; Wang, K.; Tan W.; Liu, B.; Lin, X.; He, C.; Li, D.; Huang, S.; Li, J. J. Am. Chem. Soc. 2003, 125, 7168. (9) Zhao, X.; Bagwe, R. P.; Tan, W. Adv. Mater. 2004, 16, 173. (10) Azioune, A.; Slimane, A. B.; Hamou, L. A.; Pleuvy, A.; Chehimi, M. M.; Perruchot, C.; Armes, S. P. Langmuir 2004, 20, 3350. (11) Collinson, M. M. Trends Anal. Chem. 2002, 21, 30. (12) Johnston, A. P. R.; Battersby, B. J.; Lawrie, G. A.; Trau, M. Chem. Commun. 2005, 848.
ever, most dye molecules are not easily modified without affecting their luminescence properties. Organic fluorophores were also encapsulated in silica particles. This proved to be difficult due to the high hydrophilicity of the silica matrix and the relatively high hydrophobicity of most dye molecules. Microemulsion-based techniques were also used to prepare luminescent silica particles.1,3-5,9 However, these techniques require the use of large amounts of surfactants and organic solvents. Therefore, scaling-up these procedures proved to be difficult. The Sto¨ber synthesis is a widely used method to prepare silica particles.13 This simple one-step synthesis protocol involves the condensation of tetraethyl orthosilicate (TEOS) in ethanol:water mixtures under alkaline conditions at room temperature. Using the Sto¨ber method it is possible to achieve excellent control of size, narrow size distribution, and smooth spherical morphology of the resulting silica particles. Our study focused on the use of the Sto¨ber synthesis method to prepare silica nanoparticles in which ruthenium diimine complexes were physically encapsulated and their application in streptavidin-biotin affinity assays. Experimental Methods Materials and Reagents. Dichloro tris(1,10-phenanathroline)ruthenium(II) hydrate, TEOS, and 3-mercaptopropyltrimethoxysilane (MPTMS) were purchased from Aldrich Chemicals, Inc. Streptavidin-maleimide and biotin-maleimide were purchased from Sigma. All reagents were used as received unless mentioned otherwise. Deionized water was prepared to a specific resistivity of at least 18 MΩ‚cm and was used to prepare all buffer solutions. Preparation of Fluorescent Silica Spheres. Silica particles were prepared by adding a premixed ethanol solution (25 mL) containing ammonium hydroxide and tris(1,10-phenanthroline) ruthenium(II) chloride {[Ru(phen)3]Cl2} aqueous solution to a TEOS solution in ethanol (5 mL) under stirring. The amounts of reagents are given in Table 1. The mixture was stirred for 1 (13) Sto¨ber, W.; Fink, A.; Bohn, E. J. Colloid Interface Sci. 1968, 26, 62.
10.1021/la0504098 CCC: $30.25 © 2005 American Chemical Society Published on Web 04/07/2005
4278
Langmuir, Vol. 21, No. 10, 2005
Letters
Table 1. Reagents and Quantities Used in the Preparation of Fluorescent Silica Nanospheres
TEOS
NH4OH [Ru(phen)3]Cl2 product (30%) (0.5 mg/mL) ethanol weight
1.3 mL 3 mL 1.3 mL 7 mL 2 mL 1.5 mL
3 mL 3 mL 0.5 mL
mean diameter (from TEM images)
22.7 mL 0.30 g 800 ( 20 nm 18.7 mL 0.30 g 440 ( 18 nm 25 mL 0.50 g 65 ( 8 nm
h and further sonicated for 10 min. The luminescent nanospheres were isolated by centrifugation (5000 rpm, 10 min) and washed three times with ethanol. The samples were dried under reduced pressure and then at 100 °C for 1 h. The yield for this preparation was about 80%. Preparation of Thiol-Modified Luminescent Silica Particles. A total of 2 mg of luminescent silica particles was dispersed in 10 mL of ethanol under sonication. A total of 100 µL of MPTMS was then added to this solution. The mixture was stirred overnight in a sealed vessel. The particles were separated by centrifugation (5000 rpm, 10 min) and washed three times with ethanol. The particles were then re-dispersed in 4 mL of phosphate buffer at pH 7.4 and stored at 4 °C until use. Preparation of Streptavidin-Modified Luminescent Silica Particles. The thiol-modified luminescent silica particles (1 mL of stock solution) were mixed with 5 mL of phosphate buffer solution at pH 7.4 that contained 0.25 mg of maleimidelabeled streptavidin. The mixture was incubated at room temperature under gentle stirring for 2 h. The streptavidinlabeled particles were separated by centrifugation (5000 rpm, 10 min) and washed three times with a phosphate buffer solution. The particles were then re-dispersed in 4 mL of phosphate buffer at pH 7.4 and immediately used for biotin-streptavidin assays. Preparation of Biotin-Modified Glass Slides. Microscope glass slides were treated with 1 M HNO3 aqueous solution, then water and then ethanol by sonication for 15 min each, to remove deposited organic material. The pre-cleaned glass slides were blow-dried with a stream of dry nitrogen and incubated with MPTMS in toluene solution (5%, v/v) for 4 h at room temperature. The slides were washed with toluene and ethanol and dried under a stream of nitrogen followed by thermal treatment at 100 °C for 16 h. Then, the thiol-modified glass slides were placed in flasks containing 0.5 mg of maleimide-labeled biotin dissolved in 10 mL of phosphate buffer solution at pH 7.4. Following 2 h of incubation at room temperature, the biotin-labeled glass slides were washed with the phosphate buffer and then used for biotinstreptavidin assays. Binding of Streptavidin-Modified Luminescent Silica Particles to Biotin-Modified Glass Slides. A total of 100 µL of streptavidin-modified luminescent silica particle solutions (concentration range 0-200 µg/mL) was placed on the surface of a biotin-modified glass slide. The slides were incubated for 30 min at room temperature, washed three times with water, and then dried under a stream of nitrogen. The fluorescent particles attached to the glass slides were counted using our digital imaging microscopy system. Characterization of the Fluorescent Silica Particles. Transmission electron microscopy (TEM) images of the particles were taken using a JEOL-2010 electronic microscope operating at an accelerating voltage of 200 kV. Samples for TEM were prepared by placing a drop of an aqueous particle solution on a copper grid. The particle size distribution was estimated from the measurement of about 200 particles found in an arbitrary chosen area in enlarged images. Fluorescence emission measurements were performed in a quartz cuvette using a PTI Quanta Master fluorescence spectrometer equipped with a 75-W xenon short-arc lamp as a light source. Fluorescence lifetimes were determined with an Edinburgh Analytical Instruments LP900 laser flash photolysis system operated in the emission mode (monitoring lamp shutter closed). Aliquots (3.00 mL) of solutions of the ruthenium dye or of suspensions of the ruthenium-doped silica particles were placed in a 1-cm-path-length quartz fluorescence cuvette. The samples were excited with the third harmonic (355 nm) of a Surelite I-10 Nd:YAG laser. When required, samples were deoxygenated by bubbling for 15 min with nitrogen. All samples were stirred continuously during the
Figure 1. TEM image of [Ru(phen)3]Cl2-doped silica particles of 440 ( 18 nm (left) and a digital fluorescence image of the particles taken through a 40× microscope objective using a 460 ( 10 nm band-pass excitation filter, a 505-nm dichroic mirror, and a 515-nm long pass emission filter (right). measurements and monitored for laser-induced decomposition by conventional UV-vis absorption spectroscopy (HewlettPackard 8452A diode array spectrometer). The emission decay curves at 590 nm were collected by averaging a total of 10 excitation laser shots and analyzed with the standard exponential decay routines of the Edinburgh Analytical Instruments LP900 system software to obtain the lifetimes of the excited species. Digital Fluorescence Imaging Microscopy System. Fluorescence images of the fluorescent particles attached to glass slides through biotin-streptavidin interactions were taken using a digital fluorescence imaging microscopy system that consisted of an inverted fluorescence microscope (Olympus IX70) equipped with a 100-W mercury lamp as a light source. The fluorescence images were collected through a 20× microscope objective using a 460 ( 10 nm band-pass excitation filter, a 505 nm dichroic mirror, and a 515-nm long pass emission filter. A highperformance ICCD camera (Roper Scientific, model BH2RFLT3) was employed for digital imaging.
Results and Discussion This paper describes the use of the Sto¨ber synthesis to prepare highly luminescent monodispersed silica nanoparticles that encapsulate the inorganic dye [Ru(phen)3]Cl2 with an average yield of 80%. Unlike microemulsionbased methods often used to prepare luminescent silica particles this technique completely avoids the use of potentially toxic organic solvents and surfactants. Further conjugation of the particles to biomolecules is easier because there is no need to wash the particles off surfactant molecules, which often requires multiple washing steps when microemulsion techniques are used to prepare nanoparticles. TEM and digital fluorescence microscopy images of the luminescent silica particles are shown in Figure 1. Slight modifications of the ammonia and water content in the reaction mixture resulted in monodispersed silica spheres of different diameters (see Table 1). The dye loading efficiency was determined by subtracting the fluorescence intensity of the supernatant from the fluorescence intensity of the original Ru(phen)3 solution. Under our experimental conditions the loading efficiency was estimated at about 10%. The dye concentration in the preparation solution was optimized to yield silica nanoparticles with the highest possible emission quantum yield. The emission quantum yield of the luminescent particles was estimated on the basis of absorption and emission measurements of the fluorescent silica particles and free Ru(phen)3 solutions. For similar absorption signals at 460 nm the emission signal of the particle suspension was two-fold lower that the emission of the free dye solution. On the basis of these measurements
Letters
Langmuir, Vol. 21, No. 10, 2005 4279
Figure 2. Time-resolved intensity decay and fit for [Ru(phen)3]Cl2 in aerated solutions (left) and in nitrogenated solutions (right). Free dye (curve a), dye adsorbed in silica nanospheres (curve b), and dye-doped silica nanospheres prepared by Sto¨ber synthesis (curve c).
and previous studies14 that reported the emission quantum yield of Ru(phen)3 to be around 0.1 we estimated the emission quantum yield of the luminescent nanoparticles to be around 0.05. TEM images of the particles showed that the particles were highly monodispersed and exhibited smooth morphology and narrow size distribution. The ability to control the dimensions of the particles is very useful because different particle sizes are required for different applications. It should be noted that the visualization of individual particles using conventional fluorescence microscopy instrumentation was easier for particles larger than 400 nm. In Figure 1 (right), the particles appear bright with a signal-to-background ratio of about 100. The Ru(phen)3 encapsulating particles emitted intense red light (emission 590 nm) when excited at 460 nm. Chemical stability measurements indicated that the leakage rate of the encapsulated molecules was minimal and could not be observed using our instrumental capabilities. The minimal or no leakage was attributed to strong electrostatic attraction forces between the positively charged Ru(phen)3 molecule and the negatively charged silica. Comparative photostability measurements were conducted by illuminating samples of free dye and luminescent nanoparticles using a continuous 1.5-KW xenon lamp. The luminescent nanoparticles showed a twofold higher photostability than the free dye under identical irradiation conditions. The increased photostability of the newly prepared particles was attributed to the poor permeability of molecular oxygen into the silica particles. Molecular oxygen is often required for dye photobleaching because it is a precursor for singlet oxygen, which triggers photodecomposition reactions. Ruthenium diimine complexes such as Ru(phen)3 were used previously in oxygensensing applications.15,16 The luminescence intensity of these complexes increases in an oxygen-free environment as a result of decreasing oxygen quenching efficiency. However, when encapsulated in our silica particles the dye did only minimally respond to changes in external oxygen levels. The luminescence intensity of the Ru(phen)3containing silica particles that were prepared using the (14) Carraway, E. R.; Demas, J. N.; DeGraff, B. A.; Bacon, J. R. Anal. Chem. 1991, 63, 337. (15) Zhang, P.; Guo, J.; Wang, Y.; Pang, W. Mater. Lett. 2002, 53, 400. (16) Tang, Y.; Tehan, E. C.; Tao, Z.; Bright, F. V. Anal. Chem. 2003, 75, 2407.
Sto¨ber method decreased by only 30% in oxygenated solution compared to the luminescence of the same particle suspension saturated with nitrogen. In contrast, the luminescence intensity decreased by 65% for luminescent silica particles that were prepared by dye loading into pre-prepared silica particles (physical adsorption) and by about 80% for free dye solutions. Luminescence lifetime measurements were carried out to determine the effect of encapsulating Ru(phen)3 in the silica particles on the dye luminescence properties. See details in the experimental section. Figure 2 describes the luminescence lifetime measurements of (a) free Ru(phen)3 in aqueous solution, (b) luminescent silica particles prepared by loading Ru(phen)3 into pre-prepared silica particles, and (c) luminescent particles prepared using the Sto¨ber method in aerated solutions (left) and nitrogenated solutions (right). In the aerated solutions the luminescence lifetimes of the aerated solutions were 530 ( 12 ns for the free dye, 990 ( 45 ns for the luminescent particles prepared by dye loading, and 1357 ( 86 ns for the luminescent particles prepared using the Sto¨ber method. In the nitrogenated solutions the luminescence lifetime of the free dye solutions increased to 998 ( 11 ns (curve a). The luminescence lifetime of particles prepared by dye loading was 1273 ( 85 (curve b), which was about 25% higher than the luminescence lifetime of the same particle suspension under aerated conditions. This indicated that a relatively large number of molecules were bound to the surface of the particles rather than encapsulated in the particles’ bulk. The luminescence of these molecules was affected by changes of oxygen concentrations in the sample. In contrast the fluorescence lifetime of Ru(phen)3-containing silica particles that were prepared using the Sto¨ber method only increased by about 6% to 1445 ( 108 ns (curve c) compared to the luminescence lifetime of the same particles under aerated conditions. It should be noted that the stated lifetimes were approximated values because the decay curves, particularly for dye-doped particles, exhibited significant variation from first-order exponential decay. These variations were attributed to the varying accessibility of oxygen to dye molecules that were entrapped in the silica particles. Nevertheless, it is fair to conclude that dye molecules were better protected from photobleaching and oxygen quenching when prepared using the Sto¨ber method rather than by dye loading into pre-prepared particles.
4280
Langmuir, Vol. 21, No. 10, 2005
Letters
Figure 3. Digital fluorescence images of streptadivin-modified luminescent silica particles that are immobilized to a biotinmodified glass slide. (a) A control experiment: 125 µg/mL thiol-modified luminescent silica particles adsorb nonspecifically to a biotinylated glass slide. Minimal nonspecific adsorption is observed. (b) 12.5 µg/mL and (c) 125 µg/mL of streptadivin-coated luminescent silica particles. All glass slides were washed thoroughly with a phosphate buffer at pH 7.4 to remove nonspecifically bound particles. The images were taken through a 20× microscope objective using a 460 ( 10 nm band-pass excitation filter, a 505-nm dichroic mirror, and a 515-nm long pass emission filter.
As previously mentioned we also successfully modified the surface of the luminescent silica particles with thiol groups through hydrolysis and condensation with MPTMS17 to enable covalent coupling of maleimidemodified biomolecules to the silica particles. As an example, we attached maleimide-labeled streptavidin to the thiol-modified surface of the luminescent silica particles. The presence of streptavidin on the particles surface was verified through streptavidin-biotin interactions using biotin-modified glass slides. Biotin-modified glass slides were incubated for 1/2 h with streptavidinmodified luminescent silica particles in a phosphate buffer solution at pH 7.4. The slides were washed with a buffer solution and water to remove all nonspecifically adsorbed particles. The glass slides were dried under nitrogen and analyzed using digital fluorescence imaging microscopy (see Figure 3). The images confirm the presence of luminescent particles attached to the glass slide through streptavidinbiotin interactions. The number of luminescent silica spheres on the glass slides increased linearly with increasing concentration of luminescent particles in the sample solutions from 0 to 150 µg/mL. Aggregation of the silica particles was only observed at higher concentrations than 150 µg/mL. A luminescence image resulting from a negative control experiment (Figure 3a) shows negligible nonspecific adsorption of luminescent silica particles to the glass surface. The well-known avidin-biotin interactions18 could be used to bind other biomolecules such as antibodies, proteins, and nucleic acids to the luminescent silica particles to facilitate their use as signal amplifiers in immunoassays and DNA hybridization assays. However, variation in the silanization of the glass surface remained a problem. Even though the silanization of glass surfaces with trialkoxysilanes is relatively simple and well-known, the silane density is difficult to control. We observed a standard deviation of 25% for the thiol group (17) Halliwell, C. M.; Cass, A. E. G. Anal. Chem. 2001, 73, 2476. (18) Diamandis, E. P.; Christopoulos T. K. Clin. Chem. 1991, 37 (5), 625.
density on glass slide surfaces prepared by liquid-phase silanization. At this stage of our studies this variation precludes the application of the luminescent particles in highly quantitative assays on glass slides. However, the digital counting approach could prove very useful in application requiring qualitative analysis of pathogens. In summary, we have developed a simple method to prepare bright and photostable luminescent silica nanoparticles of different sizes and narrow size distribution. The method is based on the use of Sto¨ber synthesis in the presence of Ru(phen)3, a ruthenium diimine complex, to form bright silica particles. While the Sto¨ber method is well-known, its use for the synthesis of highly luminescent silica particles in which fluorescent dyes are physically encapsulated was not previously reported. It is a one-pot synthetic method that avoids microemulsion formulations and the use of surfactants. In addition, the method results in luminescent nanoparticles that are characterized by higher photostability and longer fluorescence lifetime compared to free Ru(phen)3 solutions. To demonstrate the utility of the highly luminescent silica nanoparticles in bioassays, we further modified their surface with streptavidin. Digital imaging microscopy measurements showed that the streptavidin-labeled particles attached readily to biotinylated glass slides. The utility of these luminescent nanoparticles as signal amplifiers in digital counting assays is currently studied in our laboratory. Digital counting of nanoparticles as a detection mode in bioanalytical assays is an attractive alternative to detection techniques involving analogue luminescence detection. Solving the observed silanization problem could enhance the sensitivity of bioanalytical assays down to the singlemolecule limit. It could also be used in field applications requiring qualitative analysis to indicate the presence or absence of a targeted analyte. Acknowledgment. This work was supported by DARPA Grant HR0011-04-C-0068 and by NSF Grant CHE-0314027. LA0504098
