2526
Langmuir 2005, 21, 2526-2536
Synthesis, Characterization of Dihydrolipoic Acid Capped Gold Nanoparticles, and Functionalization by the Electroluminescent Luminol Ste´phane Roux,*,† Bruno Garcia,†,‡ Jean-Luc Bridot,† Murielle Salome´,§ Christophe Marquette,| Laurence Lemelle,‡ Phillipe Gillet,‡ Loı¨c Blum,| Pascal Perriat,⊥ and Olivier Tillement† Laboratoire de Physico-Chimie des Mate´ riaux Luminescents, UMR CNRS 5620, Universite´ Claude Bernard Lyon 1, Domaine Scientifique de la Doua, 69622 Villeurbanne Cedex, France, Laboratoire de Sciences de la Terre, UMR CNRS 5570, Ecole Normale Supe´ rieure de Lyon, 46 alle´ e d’Italie, 69007 Lyon, France, ID21 X-ray Microscopy Beamline, European Synchrotron Radiation Facility (ESRF), 6 rue J. Horowitz, BP 220, 38043 Grenoble Cedex, France, Laboratoire de Ge´ nie Enzymatique et Biomole´ culaire, UMR CNRS 5013, Universite´ Claude Bernard Lyon 1, Domaine Scientifique de la Doua, 69622 Villeurbanne Cedex, France, and Groupe d’Etude de Me´ tallurgie Physique et de Physique des Mate´ riaux, UMR CNRS 5510, Institut National des Sciences Applique´ es de Lyon, Domaine Scientifique de la Doua, 69621 Villeurbanne Cedex, France Received July 29, 2004. In Final Form: October 30, 2004 The use of gold nanoparticles as biological probes requires the improvement of colloidal stability. Dihydrolipoic acid (DHLA), a dithiol obtained by the reduction of thioctic acid, appears therefore very attractive for the stabilization and the further functionalization of gold nanoparticles because DHLA is characterized by a carboxylic acid group and two thiol functions. The ionizable carboxylic acid groups ensure, for pH g 8, the water solubility of DHLA-capped gold (Au@DHLA) nanoparticles, prepared by the Brust protocol, and the stability of the resulting colloid by electrostatic repulsions. Moreover almost all DHLA, adsorbed onto gold, adopts a conformation allowing their immobilization by both sulfur ends. It is proved by sulfur K-edge X-ray absorption near edge structure spectroscopy, which appears as an appropriate tool for determining the chemical form of sulfur atoms present in the organic monolayer. Such a grafting renders the DHLA monolayers more resistant to displacement by dithiothreitol than mercaptoundecanoic acid monolayers. The presence of DHLA on gold particles allows their functionalization by the electroluminescent luminol through amine coupling reactions assisted by 1-(3-dimethylaminopropyl)3-ethylcarbodiimide hydrochloride and N-hydroxysuccinimide. As a luminol-functionalized particle is nine times as bright as a single luminol molecule, the use of the particles as a biological probe with a lower threshold of detection is envisaged.
Introduction Since the mid 1990s, many works emphasized the crucial role of bioconjugated gold nanoparticles for production of nanostructured materials,1 biomolecule detection, and biological diagnostics.2 Bioconjugation involves the use either of alkylthiol-terminated biomolecules or of thiolated organic species which act as anchorage sites for biomolecules. Although the nature of the bonding of the thiol group on the gold nanoparticle surface is not yet fully understood, the interaction between gold and sulfur atoms is sufficiently strong to allow immobilization of thiolated * To whom correspondence may be addressed. Tel: +33 472 43 28 80. Fax: +33 472 43 12 33. E-mail:
[email protected]. † Laboratoire de Physico-Chimie des Mate ´ riaux Luminescents, UMR CNRS 5620, Universite´ Claude Bernard Lyon 1. ‡ Laboratoire de Sciences de la Terre, UMR CNRS 5570, Ecole Normale Supe´rieure de Lyon. § ID21 X-ray Microscopy Beamline, European Synchrotron Radiation Facility (ESRF). | Laboratoire de Ge ´ nie Enzymatique et Biomole´culaire, UMR CNRS 5013, Universite´ Claude Bernard Lyon 1. ⊥ Groupe d’Etude de Me ´ tallurgie Physique et de Physique des Mate´riaux, UMR CNRS 5510, Institut National des Sciences Applique´es de Lyon. (1) (a) Storhoff, J. J.; Mirkin, C. A. Chem. Rev. 1999, 99, 1849. (b) Niemeyer, C. Angew. Chem., Int. Ed. 2001, 40, 4128. (c) Stevens, M. M.; Flynn, N. T.; Wang, C.; Tirell, D. A.; Langer, R. Adv. Mater. 2004, 16, 915. (d) Li, M.; Mann, S. J. Mater. Chem. 2004, 14, 2260.
species.3 Their presence on the particles significantly limits the aggregation and improves the stability of the colloidal suspension of gold nanoparticles.4 Unfortunately, a partial desorption of thiolated species5 and/or a partial replacement by ligand exchange6 are commonly observed during the aging or in biological media. This phenomenon, if it is not controlled, is detrimental for the colloid stability and consequently for biological applications.7 Moreover (2) (a) Mirkin, C. A.; Letsinger, R. L.; Mucic, R. C.; Storhoff, J. Nature 1996, 382, 607. (b) Jin, R.; Wu, G.; Li, Z.; Mirkin, C. A.; Schatz, G. C. J. Am. Chem. Soc. 2003, 125, 1643. (c) Maxwell, D. J.; Taylor, J. R.; Nie, S. J. Am. Chem. Soc. 2002, 124, 9606. (d) Dubertret, B.; Calame, M.; Libchaber, A. J. Nat. Biotechnol. 2001, 19, 365. (e) Sato, K.; Hosokawa, K.; Maeda, M. J. Am. Chem. Soc. 2003, 125, 8102. (f) Tkachenko, A. G.; Xie, H.; Coleman, D.; Glomm, W.; Ryan, J.; Anderson, M. F.; Franzen, S.; Feldheim, D. L. J. Am. Chem. Soc. 2003, 125, 4700. (g) Sokolov, K.; Follen, M.; Aaron, J.; Pavlova, I.; Malpica, A.; Lotan, R.; RichardsKortum, R. Cancer Res. 2003, 63, 1999. (h) Fritzsche, W. Rev. Mol. Biotechnol. 2001, 82, 37. (i) Grubisha, D. S.; Lipert, R. J.; Park, H.-Y.; Driskell, J.; Porter, M. D. Anal. Chem. 2003, 75, 5936. (3) (a) Ulman, A. Chem. Rev. 1996, 96, 1533. (b) Rodriguez, J. A.; Dvorak, J.; Jirsak, T.; Liu, G.; Hrbek, J.; Aray, Y.; Gonza´lez, C. J. Am. Chem. Soc. 2003, 125, 276. (4) (a) Brust, M.; Walker, M.; Bethell, D.; Schiffrin, D. J.; Whyman, R. Chem. Commun. 1994, 801. (b) Brust, M.; Fink, J.; Bethell, D.; Schiffrin, D. J.; Kiely, C. Chem. Commun. 1995, 1655. (c) Chen, S.; Murray, R. W. Langmuir 1999, 15, 682. (d) Aslan, K.; Pe´rez-Luna, V. H. Langmuir 2002, 18, 6059. (5) Schroedter, A.; Weller, H. Angew. Chem., Int. Ed. 2002, 41, 3218. (6) (a) Ingram, R. S.; Hostetler, M. J.; Murray, R. W. J. Am. Chem. Soc. 1997, 119, 9175. (b) Hostetler, M. J.; Templeton, A. C.; Murray, R. W. Langmuir 1999, 15, 3782.
10.1021/la048082i CCC: $30.25 © 2005 American Chemical Society Published on Web 02/05/2005
Dihydrolipoic Acid Capped Gold Nanoparticles
the water solubility is an essential prerequisite for using them in biological media. ω-Thiol carboxylic acid functionalized gold particles are easily dispersible in water8 due to the conversion of the carboxylic acid groups into the negatively charged carboxylate. The reduction of HAuCl4‚3H2O by NaBH4 in the presence of carboxylic acid ended thiolated molecules provides gold nanoparticles fully coated by only the stabilizing species.8a,d However gold particles can be partially functionalized by mercaptoalkanoic acid derivatives through place-exchange reactions.8a-c This protocol allows the formation of multifunctional nanoparticles which are covered by a protecting organic monolayer composed of a mixture of several thiolated molecules. In addition to the stabilizing effect of the gold nanoparticles, the COOH groups can act as anchorage sites for immobilizing organic or biological molecules either by electrostatic interactions or by covalent linkages. Cationic macromolecules can be effectively immobilized on thioctic acid (TA) capped gold nanoparticles9 and amines can be condensed with carboxylic acid groups of the organic layer adsorbed on gold nanoparticle surfaces.8b,10 To prevent thiol desorption, two strategies were reported. The first one consists of forming a reticulated network around the particle.5,11 This efficient method allows further functionalization by organic molecules, but it requires several accurately controlled steps to achieve the protection of the nanoparticles. Easier, the second strategy is based on the use of molecules possessing at least two thiol functions. It has been verified that capping by multithiolated cyclodextrine12 or dithiane epiandrosterone (an endocyclic disulfide)13 or trithiol species7 renders the functionalized particles effectively more stable toward ligand exchange. As it was more pronounced with the increasing number of sulfur atoms per molecule, the enhancement of stability was attributed to the simultaneous anchorage onto gold nanoparticles surface of all the sulfur ends of the molecules. Up to now, this assumption was not obviously confirmed by any spectroscopic characterization owing to the limited choice of efficient techniques for determining the chemical forms of sulfur atoms. Among them, X-ray photoelectron spectroscopy (XPS) is widely used to establish the atomic composition and to study the chemisorptive properties of some organosulfur species adsorbed on gold surfaces. It has been applied to the study of the adsorption of dithiols and disulfides: it showed that the grafting of these molecules onto the gold surface plates by both sulfur atoms depended on their structure. When thiol groups are carried by each extremity of a linear alkyl chain (R,ω-dithiol), S core level XP spectra revealed that these molecules are bound to the gold surface (7) Li, Z.; Jin, R.; Mirkin, C. A.; Letsinger, R. L. Nucleic Acid Res. 2002, 30, 1558. (8) (a) Templeton, A. C.; Chen, C.; Gross, S. M.; Murray, R. W. Langmuir 1999, 15, 66. (b) Templeton, A. C.; Cliffel, D. E.; Murray, R. W. J. Am. Chem. Soc. 1999, 121, 7081. (c) Simard, J.; Briggs, C.; Boal, K. A.; Rotello, V. M. Chem. Commun. 2000, 1943. (d) Chen, S.; Kimura, K. Langmuir 1999, 15, 1075. (9) Maya, L.; Muralidharan, G.; Thundat, T. G.; Kenik, E. A. Langmuir 2000, 16, 9151. (10) Templeton, A. C.; Hostetler, M. J.; Warmoth, E. K.; Chen, S.; Hartshorn, C. M.; Krishnamurthy, V. M.; Forbes, M. D. E.; Murray, R. W. J. Am. Chem. Soc. 1998, 120, 4845. (11) (a) Liz-Marza´n, L. M.; Giersig, M.; Mulvaney, P. Langmuir 1996, 12, 4329. (b) Mine, E.; Yamada, A.; Kobayashi, Y.; Konno, M.; LizMarza´n, L. M. J. Colloid Interface Sci. 2003, 264, 385. (c) Hall, S. R.; Davis, S. A.; Mann, S. Langmuir 2000, 16, 1454. (d) Graf, C.; Vossen, D. L. J.; Imhof, A.; van Blaaderen, A. Langmuir 2003, 19, 6693. (12) Liu, J.; Ong, W.; Roma´n, E.; Lynn, M. J.; Kaifer, A. E. Langmuir 2000, 16, 3000. (13) Letsinger, R. L.; Elghanian, R.; Viswanadham, G.; Mirkin, C. A. Bioconjugate Chem. 2000, 11, 289.
Langmuir, Vol. 21, No. 6, 2005 2527 Chart 1. Thioctic Acid (Disulfide) and Dihydrolipoic Acid (Dithiol)
by only one sulfur atom.14 As the molecular adsorption state is possible for dialkyl disulfide but thermodynamically unstable,15 the adsorption of these molecules involves the cleavage of a S-S bond providing two monothiolated fragments which do not adsorb adjacent to each other.16 However the spiroalkanedithiols17 and the endocyclic disulfides,18 thanks to their peculiar structure, generate chelating monolayers which are more robust than those derived from alkanethiols of similar chain length because both sulfur ends probably attach to gold. Among these molecules that can be grafted to gold by several sulfur ends and can ensure also the immobilization of biomolecules, thioctic acid and its reduced form, the dihydrolipoic acid (DHLA), appear as specially interesting candidates. TA is characterized by an endocyclic disulfide bond while the structure of DHLA is close to that of spiroalkanedithiols (Chart 1). TA is commonly used for electrochemical studies of biomolecules or for the elaboration of biological sensors19 because the carboxylic acid group allows the immobilization of a biomolecule on an electrode surface.20 On the other hand, strongly adherent self-assembling monolayers (SAMs) on gold plates are obtained by adsorption of these endocyclic disulfide molecules. As recently shown by Chen et al.21 TA is able to displace efficiently the chloride and the citrate physisorbed on gold nanoparticles obtained by applying the Frens method.22 The resulting monolayer ensures an electrostatic stabilization of the gold nanoparticles at high pH because TA is negatively charged. Moreover dodecanethiol adsorbed on gold nanoparticles can be substituted by functionalized thioctic acid which is, prior to the exchange reaction, coupled to a porphyrin,23 to the TEMPO (2,2,6,6-tetramethylpiperidin-1-yloxyl) group,16b or to a free radical initiator.24 However TA was rarely employed to synthesize stable and water-soluble gold nanoparticles by applying the Brust method while Porter et al. suggested that the preparation of gold (14) (a) Tour, J. M.; Jones, L.; Pearson, D. L.; Lamba, J. J. S.; Burgin, T. P.; Whitesides, G. M.; Allara, D. L.; Parikh, A. N.; Atre, S. V. J. Am. Chem. Soc. 1995, 117, 9529. (b) Rieley, H.; Kendall, G. K.; Zemicael, F. W.; Smith, T. L.; Yang, S. Langmuir 1998, 14, 5147. (15) Vargas, M. C.; Giannozzi, P.; Selloni, A.; Scoles, G. J. Phys. Chem. B 2001, 105, 9509. (16) (a) Ionita, P.; Caragheorgheopol, A.; Gilbert, B. C.; Chechick, V. J. Am. Chem. Soc. 2002, 124, 9048. (b) Chechik, V.; Wellsted, H. J.; Korte, A.; Gilbert, B. C.; Caldararu, H.; Ionita, P.; Caragheorgheopol, A. Faraday Discuss. 2004, 125, 279. (17) Shon, Y.-S.; Lee, T. R. Langmuir 1999, 15, 1136. (18) Liu, H.; Liu, S.; Echegoyen, L. Chem. Commun. 1999, 1493. (19) (a) Dong, Y.; Abaci, S.; Shannon, C.; Bozack, M. J. Langmuir 2003, 19, 8922. (b) Dong, Y.; Shannon, C. Anal. Chem 2000, 72, 2371. (c) Gadzekpo, V. P. Y.; Xiao, K. P.; Aoki, H.; Buehlmann, P.; Umezawa, Y. Anal. Chem 1999, 71, 5109. (20) Blonder, R.; Willner, I.; Bu¨ckmann, A. F. J. Am. Chem. Soc. 1998, 120, 9335. (21) Lin, S.-Y.; Tsai, Y.-T.; Chen, C.-C.; Lin, C.-M.; Chen, C.-H. J. Phys. Chem. B 2004, 108, 2134. (22) Frens, G. Nat. Phys. Sci. 1973, 241, 20. (23) Beer, P. D.; Cormode, D. P.; Davis, J. J. Chem. Commun. 2004, 414. (24) Mangeney, C.; Ferrage, F.; Aujard, I.; Marchi-Artzner, V.; Jullien, L.; Ouari, O.; El Djouhar, R.; Laschewsky, A.; Vikholm, I.; Sadowski, J. W. J. Am. Chem. Soc. 2002, 124, 5811.
2528
Langmuir, Vol. 21, No. 6, 2005
Roux et al.
Scheme 1. What Is the Grafting Mode of DHLA or TA on Gold Nanoparticles?
nanoparticles could be successfully carried out in the presence of disulfide species.25 The cleavage of the S-S bond was expected since, in addition to the instability of the dialkyl disulfide adsorption state on gold (vide supra), this bond is sensitive to the presence of reducing agents such as NaBH4 (used for gold nanoparticles synthesis).26 TA was already used as a stabilizer for the synthesis of gold nanoparticles according to the Brust protocol.9 As it was not a focus of this early study, there is unfortunately almost no information concerning the surface composition, the interaction between surface gold atoms and the organic layer, and the stability of the resulting colloid in aqueous media. To avoid a competition between the reduction of gold salt and TA, we suggest a replacement of TA by its reduced form DHLA (a dithiolated molecule). This molecule was used to displace trioctylphosphine oxide (TOPO) grafted on CdS/CdSe nanoparticles (QD) which became soluble in water and behaved as an electrostatic anchorage site for proteins, paving also the way to numerous biological applications.27 With its obvious advantages, DHLA was (to our best knowledge) never used to stabilize and functionalize gold nanoparticles. In this paper, we report on the synthesis, the characterization, and the postfunctionalization of DHLA capped gold (Au@DHLA) nanoparticles. They were characterized as solid powder by X-ray diffraction experiments (XRD), thermogravimetric analysis (TGA), high-resolution transmission electron microscopy (HR-TEM), and XPS. The stability of aqueous gold sol with respect to the pH, the aging, and the presence of dithiothreitol (DTT) were investigated by UV-vis spectrophotometry, transmission electron microscopy (TEM), and photon correlation spectroscopy (PCS) experiments. Moreover Au@DHLA nanoparticles constitute a model material to determine the chemical form of sulfur atoms by X-ray absorption near edge structure (XANES) spectroscopy at the sulfur K-edge and deduce the conformation adopted by DHLA (Scheme 1) which is compared to the case of gold nanoparticles synthesized in the presence of TA with different amounts (25) Porter, L. A., Jr.; Ji, D.; Westcott, S. L.; Graupe, M.; Czernuszewicz, R. S.; Halas, N. J.; Lee, T. R. Langmuir 1998, 14, 7378. (26) (a) Yonezawa, T.; Yasui, K.; Kimizuka, N. Langmuir 2001, 17, 271. (b) Gunsalus, I. C.; Borton, L. S.; Gruber, W. J. Am. Chem. Soc. 1956, 78, 1763. (27) Jaiswal, J. K.; Mattoussi, H.; Mauro, J. M.; Simon, S. M. Nat. Biotechnol. 2003, 21, 47.
of NaBH4. Indeed the application field of XPS is restricted to solid samples, excluding the colloidal solutions which is the form required for numerous applications (particularly for the nanostructured coatings and for biological uses). We propose herein sulfur K-edge XANES spectroscopy as an attractive approach for characterizing multithiolated or disulfide species adsorbed on gold surface nanoparticles, not only in the solid state but also in very diluted colloid. Finally the grafting of luminol, an electroluminescent molecule, was carried out through the condensation of its amino group and carboxylic acid of DHLA. The efficiency of the amide-forming coupling reaction was estimated by XPS, solid state 13C nuclear magnetic resonance (NMR) spectroscopy, and spectroelectrochemistry. The preparation of gold nanoparticles functionalized by luminol was especially interesting because their use as biological probes can be envisaged. Experimental Section Chemicals. Tetrachloroauric acid trihydrate (HAuCl4,3H2O), sodium borohydride (NaBH4), thioctic acid (TA), mercaptoundecanoic acid, acetic acid (CH3COOH), sodium hydroxide (NaOH), methanol, luminol (3-aminophthalhydrazide), 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (EDC), N-hydroxysuccinimide (NHS), pentafluorophenol (PFP), dithiothreitol (threo-1,4-dimercapto-2,3-butanediol, DTT) and other organic solvents (reagent grade) were purchased from Aldrich and used as received. For the preparation of an aqueous solution and for the rinsing of gold nanoparticles, only milli-Q water (F > 18 MΩ) was used. The filtration of gold nanoparticles was performed on polymer membrane purchased from Osmonics, Inc. Their pore diameter is around 0.22 µm. DHLA Synthesis. DHLA was synthesized by reduction of TA according to a published procedure.26b NaBH4 (1.59 × 10-3 mol) was added by small portions to an alkaline aqueous solution of TA (1.45 × 10-3 mol in 10 mL of 0.25 N sodium hydroxide solution). The vigorously stirred mixture was cooled and kept at 5 °C. After 2 h, the colorless solution was acidified to pH 1. DHLA was extracted three times from the crude product by 20 mL of diethyl ether. The assembled organic phase was washed by milli-Q water and dried over anhydrous magnesium sulfate. Solvent (diethyl ether) was removed under reduced pressure at room temperature. The purification by chromatography provided pure DHLA (light yellow liquid). Yield: 85%. Functionalized Gold Nanoparticle Preparation. The synthesis, described by Brust et al.,4b consists of reducing HAuCl4‚ 3H2O by NaBH4 in the presence of thiols (stabilizers) which, by adsorption on growing particles, ensures the control of the size and the stability of the colloid.
Dihydrolipoic Acid Capped Gold Nanoparticles For a typical preparation of 5 nm gold particles, 4 × 10-5 mol of HAuCl4‚3H2O dissolved in 10 mL of methanol was placed in a 100 mL round-bottom flask. An 8 × 10-5 mol portion of DHLA in 3.2 mL of methanol and 0.16 mL of acetic acid was added to the gold salt solution under stirring. The mixture turned from yellow to orange. After 5 min, 4 × 10-5 mol of NaBH4 dissolved in 1.1 mL water was added to the orange mixture under vigorous stirring at room temperature. At the beginning of the NaBH4 addition, the solution became first dark brown and then a flocculent black precipitate appeared. The vigorous stirring was maintained for 1 h before adding 5 mL of 1 M aqueous hydrochloric acid solution. After the partial removal of the solvent under reduced pressure and at maximum 40 °C, the precipitate was filtered on a polymer membrane and washed thoroughly and successively with 0.01 N HCl, water, and ether. The resulting black powder (Au@DHLA) was dried and stocked in the solid state or dispersed in 10 mL of 0.01 M NaOH solution (for 8 mg of dry powder). Functionalized gold nanoparticles were also prepared by using TA instead of DHLA as stabilizer but with the same Au/S ratio in each case. Au@TA1 nanoparticles were synthesized under the same conditions as Au@DHLA, whereas Au@TA10 required an amount of NaBH4 (the reducing species) which was 10-fold as large as the one used for the preparation of Au@DHLA. Mercaptoundecanoic acid capped gold nanoparticles (Au@MUA) were prepared by a similar route with the same Au/S ratio as the one used for synthesizing Au@DHLA. Immobilization of Peroxidase on Gold Nanoparticles. Au@DHLA and Au@DHLA-PFP powders were brought into contact with µperoxidase or peroxidase (HRP) solution at a concentration of 1 µg/mL in water. The particles were allowed to react for 1 h at 37 °C and then washed four times with water. The ready to be measured particle precipitates were immersed in a chemiluminescent measurement solution composed of luminol 220 µM, hydrogen peroxide 500 µM, and p-iodophenol 200 µM in a Veronal buffer 30 mM pH 8.5 added to 0.2 M NaCl. At that time, each precipitate was introduced in the CCD light measurement system (Intelligent Dark Box II, Fuji Film) and the emitted light integrated during 10 min. Coupling Luminol to Au@DHLA. The immobilization of luminol on gold nanoparticles may be performed by the formation of amide linkage through the condensation between the amino group of the luminol and the carboxylic acid of DHLA adsorbed on gold. Amide coupling reactions may be accomplished in aqueous solutions by several well-known protocols developed for modifying proteins.28 A combination of the water-soluble carbodiimide EDC and NHS also is an efficient way of creating amide linkages. A 152 µL portion of 0.1 M EDC and 0.1 M NHS aqueous solution was added to Au@DHLA sol prepared by dissolution of 8 mg of powder in 10 mL of alkaline aqueous solution whose pH was adjusted to 8. After 2 h, 1.52 mL of 0.01 M luminol solution (17.7 mg dissolved in 10 mL of 0.01 M NaOH) was added to the mixture. After 2 h, gold nanoparticles were precipitated by lowering the pH down to 3. The precipitate was filtered on polymer membrane and thoroughly washed in order to remove byproducts. After being dried at room temperature, particles were dispersed in Veronal buffer 30 mM for spectroelectrochemical studies. To characterize each step of the functionalization by XPS, NHS was replaced by PFP in the aforementioned protocol. Contrary to the activation by NHS, the reaction was carried out overnight in DMF on the solid powder owing to the low solubility of PFP in aqueous solution. The reaction between COOH groups of Au@DHLA and PFP, assisted by EDC, is expected to yield pentafluorophenyl esters at the surface of the gold nanoparticles. A 4.8 mL portion of anhydrous dimethylformamide (DMF) containing 0.1 M EDC and 0.2 M PFP was added to Au@DHLA nanoparticles immersed in 3 mL of anhydrous DMF. The mixture was stirred for 24 h at room temperature. Afterward the powder was filtered on polymeric membrane, thoroughly washed with anhydrous DMF and with anhydrous ether, and dried by a dry nitrogen stream at room temperature. The resulting powder (28) (a) Bioconjugate techniques; Hermanson, G. T., Ed.; Academic Press: New York, 1996. (b) Staros, J. V.; Wright, R. W.; Swingle, D. M. Anal. Biochem. 1986, 156, 220.
Langmuir, Vol. 21, No. 6, 2005 2529 (Au@DHLA-PFP) was immersed in anhydrous DMF (3 mL) solution in which 5 mL of 0.03 M luminol in DMF was added. The reaction was carried out at room temperarure for 24 h. The powder was then filtered on a polymer and washed successively with 0.01 M HCl aqueous solution, water, DMF, and ether. The black powder (Au@DHLA-luminol) was dried at room temperature. XPS analysis was only performed with freshly prepared Au@DHLA-PFP and Au@DHLA-luminol. UV-vis Spectroscopy Studies. UV-vis absorption spectra were recorded at room temperature with a Shimadzu UV-2401 PC spectrometer for measuring the absorption band of functionalized gold nanoparticles in the 400-800 nm range. The spectral measurements were performed on diluted colloid dispersion of approximately 0.2 g‚L-1 introduced in a standard quartz cuvette. In the case of the study of the displacement of DHLA and MUA by DTT, DTT and NaCl were added to diluted colloid with final concentrations of 0.1 and 0.3 M, respectively. The temperature of the solution was maintained at 40 °C. High-Resolution Transmission Electron Microscopy. HR-TEM was used to obtain detailed structural and morphological information about the samples and was carried out using a JEOL 2010 microscope operating at 200 kV. The samples for HR-TEM were prepared by depositing a drop of a diluted Au@DHLA colloidal solution on a carbon grid and allowing the liquid to dry in air at room temperature. X-ray Diffraction. XRD diagrams were recorded using an D5000 Siemens diffractometer with the Cu KR1 and Cu KR2 X-rays (λ ) 0.15406 nm and λ ) 0.15444 nm). The diffraction pattern was scanned over the 2θ range 10-45° in steps of 0.02° and a counting time of 8 s/step. Thermogravimetric Analysis. TGA was performed with a Setaram G11 device, on ca. 120-130 mg of dry, purified samples, under a nitrogen flow (40 mL‚min-1) at a heating rate of 5 °C‚min-1 in the temperature range 25-700 °C. Size Characterization and ζ-Potential Measurements. Direct determination of the size distribution (in the range 1-100 nm) and of the ζ-potential of the Au@DHLA nanoparticles was performed via a Zetasizer 3000 HS (laser He-Ne (633 nm)) from Malvern Instrument. Prior to the experiment, the sol was diluted to obtain a concentration of around 0.08 g‚L-1 in an aqueous solution containing 0.01 M NaCl and adjusted to the desired pH. Solid-State 13C Nuclear Magnetic Resonance (13C NMR). The solid state 13C NMR studies were done on a Bruker DSX 400 NMR spectrometer (101.6 MHz 13C) at the “Institut de Recherche sur la Catalyse” (IRC, CNRS UPR 5401, Villeurbanne, France). Spectra were obtained with each sample packed in a 4 mm outer diameter rotor, which was spun at 10 kHz. Each spectrum was obtained with a 3 µs 90° 13C pulse. In the experiments, we used a total of 10240 scans. X-ray Photoelectron Spectroscopy (XPS). XPS analyses were carried out at the IRC with a VG Scientific ESCA LAB 200 R using monochromatic Al KR X-ray sources (hν ) 1486.6 eV). The binding energy scales for the monolayers on gold were referenced by setting the Au4f7/2 binding energy to 84.0 eV. Spectra were recorded at a takeoff angle of 35° (angle between the plane of the sample surface and the entrance lens of the analyzer) and with a pass energy of 50 eV. The theoretical analyzer resolution expected with that setting is 1.5 eV. X-ray Absorption Near Edge Spectrocopy (XANES) Setup. This work was carried out at the X-ray Microscopy Beamline (ID21) of the European Synchrotron Radiation Facility (ESRF) in Grenoble (France), which gives access to the 2-7 keV energy range. The scanning X-ray microscope (SXM) of ID21 offers an attractive tool for microspectroscopy at the sulfur K-edge. For this experiment, the SXM was operated at energies around 2472 eV, which corresponds to the sulfur K-edge. The harmonics rejection was ensured by a set of two parallel silicon mirrors deflecting in the horizontal plane with an incident angle of 8 mrad. The energy scan was performed by a fixed-exit double crystal Si(111) monochromator providing a spectral resolution of 0.5 eV. This high-energy resolution is essential to resolve the required sulfur K-edge XANES features. Spectra of sulfur-containing reference compounds were recorded first. The latter are used both for energy calibration of the monochromator and as fingerprints for assigning the resonances
2530
Langmuir, Vol. 21, No. 6, 2005
Roux et al.
Table 1. Characterization of the Au@DHLA Nanoparticles Colloid redispersed after precipitation of a freshly prepared Au@DHLA colloid
freshly prepared Au@DHLA colloid pH 10 (nm)a
dc Dh (nm)b λmax (nm)c ζ (mV)d
pH 8
pH 6
pH 5
5.0 [5.1-8.0] [6.1-9.7] [21-42] [56-92] 514 514 532 -35.1 -33 -25.2
pH 10 6.4 [7.3-11.6] 512
a d : average core diameter determined by TEM. b At least 75% c of Au@DHLA nanoparticles have a hydrodynamic diameter (Dh) comprised between both values in brackets. c λmax: wavelength for the maximum of the surface plasmon band. d ζ potential.
observed in the real samples. The standards used were cystine, cysteine, glutathione, methionine, and chondroitin sulfate. The energy was scanned between 2450 and 2520 eV in 0.175 eV energy steps, with a dwell time varying between 1 and 10 s for the less concentrated samples. The sulfur K-edge study was carried out either for solid samples (pure TA, Au@DHLA nanoparticle powder) or for liquid samples (pure DHLA, Au@DHLA nanoparticles dispersed in an aqueous 0.01 M NaOH solution (0.8 g‚L-1)). Spectroelectrochemical Studies.The spectroelectrochemical studies were performed, in a light-tight box, with a specially designed analytical setup. Briefly, a carbon electrode was used as working electrode (vs Ag/AgCl reference) for triggering the electrochemiluminescent reaction. The very end of the carbon electrode, where the electrochemiluminescent reaction takes place, was facing a photomultiplier tube (Hamamatsu, Japan) recording the light intensity produced. A time correlation, driven by the potentiostat (Radiometer Analytical), between the applied potential scanning and the light intensity acquisition enabled the achievement of spectroelectrochemiluminescence curves. All spectroelectrochemiluminescent measurements were performed in Veronal buffer 30 mM, pH 8.5, added of 30 mM KCl and hydrogen peroxide 500 µM.
Results and Discussion Structure, Morphology, and Composition of Au@DHLA Nanoparticles. The structure, the morphology, the composition, and the colloidal stability of Au@DHLA nanoparticles were characterized by HR-TEM, XRD, TGA, UV-vis spectrophotometry, ζ potential, and PCS. The results are extensively discussed in the Supporting Information and summarized in Table 1. Briefly, the Au@DHLA nanoparticles can be isolated as a black powder which is easily redispersed in an alkaline aqueous solution (e.g., 8 mg in 10 mL of 0.01 M NaOH, then the pH is adjusted to the desired value by adding HCl or in a veronal buffer (pH 8.2) containing 0.01 M NaCl). The resulting colloid is characterized by a deep red Bordeaux wine color and can be stored at pH 8-10 several weeks without particle growth or agglomeration. The stability of the Au@DHLA nanoparticle colloid is strongly dependent on the pH. For pH e 3 precipitation occurs whereas the colloid is stable for pH > 6. Indeed the ζ potential which reflects the surface charge is more and more negative for high pH, inducing a stabilization of the colloid by strong electrostatic repulsion between the particles. However the precipitate, obtained for pH e 3, can be easily redispersed again in an alkaline aqueous solution. This feature is very attractive for further functionalization of Au@DHLA nanoparticles since it allows the extraction from the reaction medium and a careful rinsing of modified Au@DHLA nanoparticles to remove byproducts. The particle size determined by HR-TEM is evaluated to around 5 nm with a narrow size distribution. The nanoparticle diameter measured by PCS is between 5.1
and 8 nm for pH 10 and increases when pH decreases (Table 1). These particles display expected face-centered cubic structure according to the crystallographic data for gold (04-0784, 2002 JCPDS) as revealed by HR-TEM and XRD. Moreover most of the nanoparticles are single crystals since the crystallite size (5.5 nm) determined by XRD pattern is very close to the size estimated from HRTEM. The formation of a concentrated and stable colloid of Au@DHLA nanoparticles with a narrow size distribution constitutes indirect evidence of the presence of a DHLA layer adsorbed on the gold particles. TGA, XPS, and XANES experiments allow confirmation of their immobilization on the gold surface. From the weight loss of adsorbed organic species (9.6% of the total weight of the sample) determined by TGA, the gold atom surface Ausurfto-DHLA ratio is around 4. This is consistent with the results obtained by Murray et al.29 For a gold nanoparticle with diameter of 5.2 nm covered by dodecanethiol (Au@C12S), whose molecular weight is similar to that of DHLA, they determined by TGA that the loss of adsorbed organic species represents 9.7% of the total weight of the sample (versus 9.6% for the Au@DHLA) and the Ausurfto-S ratios are in the same range (2.4 for Au@C12S and 2.1 for Au@DHLA). XPS Study of Au@DHLA. Au@DHLA powder was analyzed by XPS (Figure 1a). The experimental atomic ratios (C/S ) 4.525 and O/S ) 1.075) are consistent with the expected stoichiometric values (C/S ) 4.0 and O/S ) 1.0). The XPS C 1s spectrum of Au@DHLA displays two main peaks at about 284.8 and 289.2 eV (Figure 1a). The line at 284.8 eV originates from the carbon atoms of the aliphatic chain.30 The structure situated at 289.2 eV can be attributed to the carbon atoms of carboxylic acid groups (COOH)30 and confirms the presence of these groups at the surface of the particles. The S 2p core-level XP spectrum, which exhibits a single dissymmetrical peak, could be fit using the S 2p3/2 and S 2p1/2 doublet with a 0.60 peak ratio and a 1.20 eV splitting (Figure 1b), in agreement with previous works reported by several research groups.31,32 The S 2p3/2 and S 2p1/2 peaks are respectively situated at 161.9 and 163.1 eV, and the narrow width at middle height of each peak (1.8 eV) indicates the presence of only one doublet structure. According to the experimental data collected from the adsorption of thiolated molecules on gold plates, these binding energy values are attributed to sulfur atoms bound to the gold surface as thiolate species.31-33 Effectively unbound thiolated species are characterized by higher binding energy values (∼163.5 eV for S 2p3/2 and ∼165 eV for S 2p1/2). As no peak (up to 167 eV) assigned to the oxidized form of sulfur (associated to the presence of sulfinate and sulfonate) and no additional doublet structure attributed to S 2p3/2 and S 2p1/2 peaks of unbound thiols are detected at higher energy, the Au@DHLA S 2p core level spectrum is consistent with the exclusive (29) Hostetler, M. J.; Wingate, J. E.; Zhong, C.-J.; Harris, J. E.; Vachet, R. W.; Clark, M. R.; Londono, J. D.; Green, S. J.; Stokes, J. J.; Wignall, G. D.; Glish, G. L.; Porter, M. D.; Evans, N. D.; Murray, R. W. Langmuir 1998, 14, 17. (30) Beamson, G.; Briggs, D. High-Resolution XPS of Organic Polymers: The Scienta ESCA300 Database; John Wiley & Sons Ltd.: Chichester, 1992. (31) Vance, A. L.; Willey, T. M.; Nelson, A. J.; van Buuren, T.; Bostedt, C.; Terminello, L. J.; Fox, G. A.; Engelhard, M.; Baer, D. Langmuir 2002, 18, 8123. (32) Castner, D. G.; Hinds, K.; Grainger, D. W. Langmuir 1996, 12, 5083. (33) (a) Nuzzo, R. G.; Zegarsky, B. R.; Dubois, L. H. J. Am. Chem. Soc. 1987, 109, 733. (b) Laibinis, P. E.; Whitesides, G. M.; Allara, D. L.; Tao, Y.-T.; Parikh, A. N.; Nuzzo, R. G. J. Am. Chem. Soc. 1991, 113, 7152.
Dihydrolipoic Acid Capped Gold Nanoparticles
Langmuir, Vol. 21, No. 6, 2005 2531
Figure 1. (a) C 1s and (b) S 2p core level XP spectra of Au@DHLA. C 1s core level XP spectra of (c) Au@DHLA-PFP and (d) Au@DHLA-luminol.
presence of grafted thiolate species. Consequently it indicates that DHLA is immobilized onto gold nanoparticles by both S-ends. Unfortunately this statement could not be easily confirmed by the comparison with the case of pure DHLA owing to the liquid state of the latter preventing a direct XPS study. Moreover the low signalto-noise ratio is detrimental for a reliable interpretation of the Au@DHLA S 2p core level spectrum. Sulfur K-Edge XANES Study. Thanks to the large penetration depth of X-rays, XANES experiments allow the characterization of samples in liquid state even at low concentrations in fluorescence mode, which offers higher sensitivity. Consequently the opportunity is offered to compare pure DHLA and DHLA bound to the gold nanoparticles (Au@DHLA) and to determine the chemical form of sulfur atoms in Au@DHLA samples in solid and in liquid states. In particular, thiol (R-S-H) and disulfide (R-S-S-R) spectra exhibit characteristic feature differences.34 For these reasons XANES spectroscopy at the sulfur K-edge has been extensively used for characterizing sulfur oxidation states in biological samples.35 Although it possesses many advantages to establish the chemical form of sulfur atoms, no investigation focused on the characterization of organosulfur compound capped gold (34) Rompel, A.; Cinco, R. M.; Latimer, M. J.; McDermott, A. E.; Guiles, R. D.; Quintanilha, A.; Krauss, R. M.; Sauer, K.; Yachandra, V. K.; Klein, M. P. PNAS 1998, 95, 6122.
nanoparticles by sulfur K-edge XANES spectroscopy was, to our best knowledge, reported. Up to now, only one paper described the use of Au L3,2-edge XANES for studying the influence of the capping molecules on the electronic behavior of gold nanoparticles.36 The sulfur K-edge XANES spectra of Au@DHLA, in solid state or dispersed in aqueous sodium hydroxide (0.01 M) solution, are compared to the pure DHLA (i.e., reduced thioctic acid in absence of gold nanoparticles) spectra (Figure 2a-c). Their study is based on a direct fingerprint approach: the information about the environment of sulfur atoms is retrieved from the shape and energy position of the white line (the most intense structure in the XANES spectrum). The white line of XANES spectra of DHLA spectrum displays a single peak (at 2472.8 eV) (Figure 2a) which can be attributed to the 1s transitions to the molecular orbitals of the S-C and S-H bond (S 1s f S-C and S 1s f S-H transitions, respectively). (35) (a) Pickering, I. J.; Prince, R. C.; Divers, T.; George, G. N. FEBS Lett. 1998, 441, 11. (b) Pickering, I. J.; George, G. N.; Yu, E. Y.; Brune, D. C.; Tuschak, C.; Overmann, J.; Beatty, J. T.; Prince, R. C. Biochemistry 2001, 40, 8138. (c) Sneeden, E. Y.; Harris, H. H.; Pickering, I. J.; Prince, R. C.; Johnson, S.; Li, X.; Block, C. K.; George, G. N. J. Am. Chem. Soc. 2004, 126, 458. (d) Prange, A.; Arzberger, I.; Engemann, C.; Modrow, H.; Schuman, O.; Tru¨per, H. G.; Steudel, R.; Dahl, C.; Hormes, J. Biochim. Biophys. Acta 1999, 1428, 446. (e) Bellacchio, E.; McFarlane, K. L.; Rompel, A.; Robblee, J. H.; Cinco, R. M.; Yachandra, V. K. J. Synchrotron Radiat. 2001, 8, 1056. (36) Zhang, P.; Sham, T. K. Appl. Phys. Lett. 2002, 81, 736.
2532
Langmuir, Vol. 21, No. 6, 2005
Roux et al.
Figure 3. TEM images of (a) Au@DHLA, (b) Au@TA1, and (c) Au@TA10.
Figure 2. Sulfur K-edge XANES spectra of (a) dihydrolipoic acid, (b) Au@DHLA (solid state), (c) Au@DHLA (liquid state), (d) thioctic acid, (e) Au@TA1 (solid state), and (f) Au@TA10 (solid state).
The XANES spectrum recorded on the Au@DHLA nanoparticle powder (Figure 2b) exhibits also only one peak at 2473.6 eV. However pure DHLA and Au@DHLA powder spectra do not superimpose: the position of the white line (2473.6 eV) seen for Au@DHLA (Figure 2b) is 0.8 eV higher in energy compared to the white line position of the pure DHLA spectrum (Figure 2a). This shift toward higher energy is attributed to the interaction between sulfur atoms of adsorbed DHLA and surface gold atoms since the energy position of the peak associated to S 1s f S-C transitions is greatly influenced by sulfur atoms environment. Moreover the absence of an additional peak at 2472.8 eV (characteristic of free DHLA) in the spectrum of Au@DHLA (Figure 2b) clearly shows that most sulfur atoms are consequently placed in the same surroundings: almost each DHLA molecule in Au@DHLA samples is anchored on gold nanoparticles by both sulfur atoms. As in solid state (Figure 2b), the white line of the XANES spectrum of the Au@DHLA colloidal solution (Figure 2c) is mainly characterized by one peak at 2473.6 eV. However, a shoulder is observed at 2472.3 eV and is suspected to be related to unbound sulfur atoms of free DHLA molecules which are partially or totally degrafted from gold nanoparticles. Indeed, the redispersion of Au@DHLA nanoparticle powder in an alkaline solution might have been accompanied by a little desorption of some DHLA molecules from the gold surface. An additional experiment (not shown) revealed that these molecules were partially converted into TA. Although Au@TA1 (Figure 2e) and Au@TA10 (Figure 2f) were prepared with TA as stabilizer upon the gold salt reduction, their XANES spectra are obviously distinguishable from that of TA (Figure 2d). The white line of the pure TA XANES spectrum (Figure 2d) exhibits two peaks at 2472.0 and 2473.6 eV, characteristic for disulfide bonds, which are assigned to S 1s f S-S and S 1s f S-C(S-H) transitions respectively whereas there is only one peak (at 2473.6 eV) in the case of Au@TA1 and Au@TA10. Moreover these spectra (Figure 2e,f) are very similar to that of Au@DHLA (Figure 2b) indicating that Au@TA1 and Au@TA10 nanoparticles are not covered with TA but with DHLA. The conversion of TA into DHLA,
occurring upon the gold nanoparticle synthesis, is arisen from cleavage of the S-S bond, which is probably induced by reducing conditions26 and by the adsorption on gold.15 With regard to the surface composition of gold nanoparticles determined by sulfur K-edge XANES spectroscopy, capping by DHLA is identical to capping by TA. However TEM images show that the nanoparticle dimensions and their size distribution are closely dependent on the oxidation state of the stabilizer and on the amount of the reducing species (Figure 3). Although Au@DHLA and Au@TA1 were obtained by the reduction of gold salt with the same amount of NaBH4 and Au/S ratio, Au@DHLA nanoparticles are smaller (5.0 ( 0.8 nm) and are characterized by a narrower size distribution (Figure 3a,b). It indicates that the presence of a large amount of DHLA is required at the beginning of the gold salt reduction: DHLA, which seems, as expected, to be a better stabilizer than TA, is immediately available for controlling the particle growth. Such a controlled growth can be however achieved with TA only if a large excess of NaBH4 is added as in the case of Au@TA10. To interpret the differences concerning Au@TA1 and Au@TA10, the competition between the particle growth and the conversion of TA into DHLA should be taken into account. Indeed NaBH4 ensures the reduction both of the gold salt and of TA. Moreover, if the amount of NaBH4 is too weak, the reduction rate is lower and the particle growth is relatively more favorable than nucleation, leading to larger particles. From XANES and TEM experiments, we can deduce that the use of DHLA, obtained by reducing TA before the gold nanoparticles synthesis, provides better results than the use of TA unless a large excess of NaBH4 is added which can unfortunately be the root of purification difficulties. Behavior of Gold Nanoparticles in the Presence of DTT. As the lack of stability of functionalized gold colloids (flocculation, precipitation), arising from the departure of the capping molecules, constitutes a crucial issue for their use as biological probes, Au@DHLA nanoparticles appear very interesting owing to the low DHLA desorption revealed by the XANES study. Because DHLA molecules are grafted to the gold surface through both sulfur ends (as shown by the XANES study), their displacement from the nanoparticles by dithiothreitol (DTT) is expected to be less important than in the case of monothiolated molecules such as MUA. The comparative study was performed on gold nanoparticles characterized by the same amount of sulfur atoms for an identical diameter. However the lower propensity of MUA to strongly anchor to the gold surface owing to the presence of a single thiol group (compared to DHLA) should be compensated by the ability of thiols with a long alkyl chain containing 10 or more methylene groups to form ordered and densely packed monolayers because of the collective chain-chain interactions.37 Moreover, for a same amount of sulfur atoms, the number of alkyl chains in (37) Camillone, N., III.; Chidsey, C. E. D.; Liu, G. Y.; Putvinsky, T. M.; Scoles, G. J. Chem. Phys. 1991, 94, 8493.
Dihydrolipoic Acid Capped Gold Nanoparticles
Langmuir, Vol. 21, No. 6, 2005 2533
Figure 4. Time evolution of SPB intensity of Au@DHLA (2) and Au@MUA (1) nanoparticle colloids in an aqueous solution containing 0.1 M DTT and 0.3 M NaCl at 40 °C.
MUA monolayers is twice as large as in DHLA monolayers. MUA monolayers are expected to be more densely packed and therefore to provide an efficient protective barrier. The stability of Au@DHLA and Au@MUA colloids was therefore compared on the basis of UV-vis spectra regularly recorded after the addition of a DTT solution at 40 °C to each colloid. As the position, the width, and the intensity of the plasmon band are sensitive to the size and the concentration of gold nanoparticles, the degradation by DTT of Au@DHLA and Au@MUA at 40 °C is easily followed by UV-vis spectroscopy (Figure 4). Once DTT was added, a progressive discoloration of Au@DHLA and Au@MUA colloids was observed. However the phenomenon is more marked in the case of Au@MUA reflecting therefore a weaker stability of such a colloid. The anchorage on gold nanoparticles of almost each DHLA by both sulfur atoms confers to the colloid a better stability, although the adsorbed layer composed of DHLA molecules is less densely packed. From this comparative study, we can deduce that protection of the gold core by a loose but strongly adherent DHLA monolayer is preferred to the one ensured by a densely packed MUA monolayer. Actually, as the surface of gold nanoparticles is faceted, the chain-chain interactions cannot be well established between adjacent alkyl chains located onto differently oriented planes. The protective barrier is thus not continuous38 and makes the gold surface easily accessible for DTT which induces a fast displacement of MUA since this molecule is bound by only one sulfur atom. Covalent Grafting of Luminol. DHLA is effectively an interesting molecule for the preparation of gold nanoparticles because it is strongly immobilized by both sulfur ends and it allows the binding of reactive organic species or a biomolecule by a condensation reaction between the COOH group carried by DHLA and an amino function.20 The grafting of luminol (through its amino group) on Au@DHLA is a promising strategy to provide original luminescent biological probes. As this molecule emits light under electrochemical stimuli, it allows removal of the residual fluorescence (autofluorescence) of the medium observed when fluorophores were excited by electromagnetic radiation. To improve the yield of the condensation, COOH groups of Au@DHLA were activated by EDC and PFP (the gold particles were therefore further called Au@DHLA-PFP). The resulting fluorinated ester was easily revealed by the appearance of the fluorine peak (Table 2).
Compared to the spectrum of Au@DHLA (Figure 1a), the C 1s core level spectrum of Au@DHLA-PFP is characterized by both an area increase and a shift toward lower binding energies (from 289.2 to 288.3) of the less intense peak (Figure 1c). These changes can be attributed to the contribution of the C-F bonds since the binding energy of these bonds is around 288 eV.30,39 Assuming that adsorbed DHLA molecules form a monolayer, the calculated F/S ratio is equal to 2.5 if the condensation is complete. The experimental ratio is estimated to 1.218, indicating that half of COOH groups are activated. The activation of COOH groups is essential for a further functionalization of Au@DHLA as shown Figure 5. Attempts to immobilize enzymes (µperoxidase (Figure 5a,b) and peroxidase HRP (Figure 5c,d)) on gold nanoparticles were undertaken. An aliquot of aqueous enzyme solution was added to the nanoparticles powder, and the mixture was allowed to react for 1 h. After thorough washing, the presence of the enzyme on the particles was revealed by the light emission produced by a chemiluminescent process (oxidation of luminol in the presence of H2O2) whose µperoxidase and HRP are catalytic species. Among the four samples (µperoxidase or HRP with Au@DHLA-PFP (PFP+), Au@DHLA (PFP-)), light is emitted by the samples activated by PFP (Au@DHLAPFP) whereas the chemiluminescent reaction does not occur with Au@DHLA samples which are not activated. This indicates that the grafting of an enzyme (µperoxidase or HRP) is possible only on Au@DHLA-PFP, i.e., only if COOH groups were activated. As compared with the functionalization of Au@DHLA-PFP by HRP, the light is more intense when µperoxidase was immobilized on Au@DHLA-PFP for a same amount of gold nanoparticles. As µperoxidase corresponds to the active center of the HRP, the steric hindrance is weaker for µperoxidase allowing the tethering of a more important amount of this enzyme on gold particles. Therefore the light emitted by µperoxidase-functionalized gold nanoparticles is more intense. Compared with µperoxidase, luminol is a small molecule. The functionalization of Au@DHLA-PFP should not
(38) Paulini, R.; Frankamp, B. K.; Rotello, V. M. Langmuir 2002, 18, 2368.
(39) Roux, S.; Duwez, A.-S.; Demoustier-Champagne, S. Langmuir 2003, 19, 306.
Figure 5. Photographs after incubation with µperoxidase of (a) Au@DHLA-PFP and (b) Au@DHLA and after incubation with HRP of (c) Au@DHLA-PFP and (d) Au@DHLA.
2534
Langmuir, Vol. 21, No. 6, 2005
Roux et al.
Table 2. Summary of Core-Level Binding Energies binding energy (eV) sample
Au 4f7/2
S 2p
C 1s
O 1s
F 1s
N 1s
Au@DHLA Au@DHLA-PFP Au@DHLA-luminol
83.9 83.9 84.0
163.1 163.0 163.0
284.8, 289.2 285.0, 288.3 284.8, 289.1
533.1 532.8 532.9
X 687.9 688.2
X X 400.5
be handicapped by the steric hindrance. However luminol, in contrast with µperoxidase, possesses only one amino group which is in addition less nucleophilic since the amino group is in conjugation with the aromatic π-electron system. To probe the efficiency of the grafting of luminol on Au@DHLA-PFP, 13C NMR in solid state and XPS experiments were performed. For this study, samples were prepared in such a way that most of COOH groups present on the particles were converted into amide functions resulting from the condensation with NH2 of luminol (a large excess of luminol was used: the experimental amount to the stoichiometric ratio is 50). This strategy was applied for avoiding the superimposition of the spectra of the reacted and unreacted surface groups rendering their interpretation more difficult. The 13C NMR spectrum of gold nanoparticles after reaction with luminol was compared to those of Au@DHLA and pure luminol (Figure 6). Au@DHLA spectrum (Figure 6a) displays peaks between 20 and 50 ppm which are assigned to C sp3 atoms of DHLA and a single peak at 182.3 ppm which is characteristic of C sp2 in carbonyl (CdO) groups of COOH function. In the case of pure luminol (Figure 6b), the peaks appear at higher values between 110 and 165 ppm and are characteristic of C sp2 in an aromatic ring. The peak at 164.3 ppm is attributed to C sp2 of the carbonyl groups present in the luminol. Figure 6c displays the peaks attributed to DHLA and luminol moieties,40 as compared with spectra a and b of Figure 6. However the carbonyl 13 C chemical shift changed from 182.3 to 169.4. This indicates the formation of the amide bond. Solid-state 13C NMR experiments demonstrated that luminol can be immobilized on gold nanoparticles. However the experimental conditions imposed for the preparation of the Au@DHLA-luminol sample did not allow the redispersion of these nanoparticles since according to the solid state 13C NMR data there is no ionizable COOH groups left. Their presence is essential for ensuring the colloidal stability (see Structure, Morphology, and Composition of Au@DHLA Nanoparticles and the Supporting Information file). As the liquid state is required for the spectroelectrochemical study, gold nanoparticles partially functionalized by luminol (i.e., particles with sufficient COOH/COO- groups) were characterized by XPS. They distinguished from the nanoparticles used for the spectroelectrochemical studies by the activation step which is carried out with EDC and PFP instead of EDC and NHS. PFP was chosen for the XPS study because fluorine element is easily recognizable and allows the monitoring of each step of the functionalization. After reaction between Au@DHLA-PFP and luminol, the simultaneous appearance of a small nitrogen peak (N/S ) 0.17) and the marked decrease of the fluorine peak (F/S ) 0.196 while F/S ) 1.218 for Au@DHLA-PFP) reflect the partial replacement of the pentafluorophenoxy groups by the luminol. The grafting of the luminol on a gold surface through the condensation of activated COOH groups of (40) The presence of PFP moieties cannot be totally excluded because of the broadening of the peak at 137.6 ppm. However 13C chemical shift of PFP is confined in the 130-140 ppm region (the 13C NMR spectrum of PFP is available on http://www.sigmaaldrich.com/spectra/fnmr/ FNMR000096.PDF). The presence of peaks outside this region is exclusively from the presence of luminol.
adsorbed DHLA with NH2 function of the luminol was not complete since the experimental N/S ratio (0.17) is 10fold as small as the calculated one (1.5). We can estimate that around 10% of the DHLA present on gold nanoparticles was modified by luminol (such particles were further called Au@DHLA-luminol). The shift of the less intense peak toward higher energy (from 288.3 to 289.1), compared to the C 1s core level spectrum of Au@DHLA-PFP, reflects the decrease of the C-F bond contribution owing to the partial replacement of the pentafluorophenoxy groups (Figure 1d). After redispersion of Au@DHLA-luminol (obtained by coupling Au@DHLA-NHS and luminol) in an aqueous alkaline solution, the electrochemiluminescent intensity and the position of the electrooxidation peaks were determined upon electrochemical cycling between 0 and 1500 mV. The results were compared to the case of Au@DHLA colloid, a luminol solution, and a freshly prepared mixture of luminol and Au@DHLA (Au@DHLA + luminol) in which the coupling reactions were unlikely to occur, as shown by Figure 5, because the COOH groups of Au@DHLA were not activated. Whereas the emission occurs at the same potential (826 mV) for the luminol solution and the mixture, light was emitted at a lower potential (733 mV) for Au@DHLAluminol (Figure 7). As no peak is observed in the case of Au@DHLA (Figure 7), the peak at 733 mV (in the case of Au@DHLA-luminol) is therefore attributed to the presence of luminol grafted to the gold nanoparticles. The 100 mV shift toward lower potentials results probably from the covalent immobilization of luminol onto gold particles since gold nanoparticles do not affect the behavior of nongrafted luminol (the case of the mixture [Au@DHLA + luminol], green curve). Gold seems to favor the oxidation of the grafted luminol. To observe the same intensity as Au@DHLA-luminol (Figure 7, blue curve), the solution of pure luminol (Figure 7, red curve) is nine times more concentrated. We can deduce that a particle is brighter than a single luminol molecule. This feature is very attractive for the biological luminescent markers field because a biomolecule labeled by Au@DHLA-luminol will be more easily detected than a biomolecule conjugated to only one luminol. This luminescence increase, which reaches almost 1 order of magnitude, results from the large amount of luminol immobilized on gold through the coupling with DHLA which is at least larger than 9 fluorophores per particle. Unfortunately the comparison with the XPS analysis, which reveals that 25-30 luminol41 molecules were grafted on a single particle, is not reliable because the experimental conditions of preparation are too different (see Experimental Section). However the number of luminol grafted to Au@DHLA is expected to be larger than 9 because it was demonstrated that the luminescence of dyes grafted to gold was reduced to a large extent. In previous work we observed effectively a decrease of a factor (41) Thanks to TGA, we can estimate that between 250 and 300 DHLA molecules were adsorbed on each gold particle. N/S ratio of Au@DHLA-luminol, determined by XPS, indicates that 10% of DHLA were modified. Therefore we can evaluate that the number of luminol per particle is around 25-30.
Dihydrolipoic Acid Capped Gold Nanoparticles
Langmuir, Vol. 21, No. 6, 2005 2535
Figure 7. Spectroelectrochemical curves of luminol (red), Au@DHLA (orange), mixture of Au@DHLA and luminol (green), and Au@DHLA-luminol (blue). Cyclic voltammetry experiments were carried out in Veronal buffer 30 mM, pH 8.5, added to 30 mM KCl and 500 µM hydrogen peroxide, between 0 and 1.5 V with a potential sweep rate of 100 mV‚s-1.
because of the quenching effect of gold. To limit the light quenching, 4-mercaptophenol plays a crucial role for the adsorption of fluorophore on gold with a low coverage (characteristic of the Frens method) because it prevents bringing the fluorophore and the gold particle in close proximity. When the dye is directly in contact with the particle surface, light emission is no longer observed since the gold nanoparticle is an efficient quencher.2c,2d Although luminol is coupled to flexible DHLA which can adopt a conformation inducing a close proximity between dyes and gold, luminol remains far away from the gold surface because of the high coverage ratio of the particles by DHLA which impedes such a conformation and therefore limits the quenching. Conclusion
Figure 6. Solid-state 13C NMR spectra of (a) Au@DHLA, (b) pure luminol, and (c) Au@DHLA-luminol (δ represents the 13C chemical shift).
of 3 in luminescence of lissamine rhodamine B immobilized via a rigid spacer (4-mercaptophenol) on gold nanoparticles (obtained by the citrate route).42 In the case of Au@DHLA, it implies that the number of grafted luminol is actually around 27. This is consistent with the number of luminol determined by XPS. But only 9 of them are observed
The reduction of gold salt by NaBH4 in the presence of DHLA provides water-soluble Au@DHLA nanoparticles. HR-TEM and XRD experiments confirm the formation of crystalline gold nanoparticles. They were coated, as shown by TGA, by a strongly adherent organic layer whose atomic composition, determined by XPS is very close to the one of pure DHLA. The adsorption of DHLA allows the dispersion of the gold nanoparticles in an aqueous alkaline solution. The resulting colloid can be conserved for several weeks without noticeable alteration when pH g 8, i.e., when most DHLA are negatively charged. The electrostatic stabilization was effectively shown by ζ-potential measurements. Moreover we demonstrated that a DHLA monolayer offers better protection against DTT than a MUA monolayer does. Although the DHLA monolayer is assumed, for the same sulfur atoms number, to be looser, their displacement from the gold surface is more difficult probably because DHLA is bound by both sulfur ends. This feature, often stated to explain the great adherence of multithiolated molecules on gold, was effectively suggested by XPS but demonstrated by sulfur K-edge XANES study. Furthermore a very restricted desorption (42) Chabane Sari, S. M.; Debouttie`re, P. J.; Lamartine, R.; Vocanson, F.; Dujardin, C.; Ledoux, G.; Roux, S.; Tillement, O.; Perriat, P. J. Mater. Chem. 2004, 14, 402.
2536
Langmuir, Vol. 21, No. 6, 2005
was observed thanks to this attractive technique which is perfectly appropriate for the study of organosulfur species capped gold nanoparticles in liquid state. The sulfur K-edge XANES study also shows that the gold nanoparticles are coated by DHLA only, even if TA (the oxidized form of DHLA) was used as stabilizer during the synthesis but their dimension and their size distribution depend on the oxidation state of the stabilizer. Luminol was successfully grafted to gold nanoparticles via an amide coupling reaction, promoted by EDC and NHS (or PFP), between COOH and NH2 groups of DHLA and luminol, respectively. Au@DHLA-luminol emitted light upon electrochemical oxidation at a lower potential than free luminol. As Au@DHLA nanoparticles are water soluble,
Roux et al.
stable when pH > 6, and can be functionalized by luminol, their use as a biological marker can be envisaged since one Au@DHLA-luminol particle is nine times as bright as one luminol. The labeling of biomolecules by Au@DHLAluminol could be performed through electrostatic interactions, amide coupling reactions, or place-exchange reactions. Supporting Information Available: Additional experimental details, UV-vis spectra of nanoparticles, and TEM micrographs of nanoparticles. This material is available free of charge via the Internet at http://pubs.acs.org. LA048082I