alt-mediated elf Assembly of Thioctic Acid on Gold Nanoparticles Introduction The stability of solution-phase nanoparticles can be improved with surface modification via self assembled monolayers (AMs) thereby increasing their usefulness in biology, catalysis, and nanotechnology. 1-10 AMs contain with thiol and/or disulfide groups are widely used because of the strong bond that forms between gold and sulfur. 11 For instance, thioctic acid is ideal for AM formation on gold nanoparticles because of the disulfide and carboxylic acid end group which will be negatively charged at ph 9 (Figure 1A). Utilization of thioctic acid functionalized gold (Au@TA) nanoparticles in any application requires reproducible assembly of thioctic acid on the nanoparticle surface. Reproducible assembly is limited by defect sites in the AM layer which are caused by (1) incomplete adsorption of thioctic acid during the self assembly process, (2) gauche orientations of the thioctic acid carbon chain, (3) or a non-uniform gold nanoparticle surface. 12-14 Herein, we observe that the addition of salt (NaCl) during the self assembly of thioctic acid on the surface of gold nanoparticles increases the packing density of thioctic acid AMs thereby increasing the stability of solution-phase gold nanoparticles. 15 Figure 1 shows a proposed schematic of the self assembly of thioctic acid on gold nanoparticles. We expect that better control in thioctic acid packing density as a result of a saltmediated AM assembly on gold nanoparticles will be fundamental in achieving reproducible covalently functionalized gold nanoparticles. Results and Discussion Optical Characterization of Au@TA. During functionalization, NaCl is added incrementally in 8 hour incubation steps to promote thioctic acid self assembly. Initially, citrate stabilized gold 1
(Au@citrate) nanoparticles are functionalized with thioctic acid and allowed to equilibrate for 16 hours. During the slow addition of salt, a NaCl solution is added drop-wise until the salt concentration reaches 4 mm. This solution is allowed to stir for 8 hours. The NaCl concentration is then increased to 8 and 16 mm, similarly in subsequent steps. Au@TA nanoparticles remain stable up to 16 mm salt concentrations. After each of these incubation periods, excess thioctic acid and NaCl is removed by rinsing the nanoparticles with ph adjusted Nanopure water (ph = 11). amples in which no salt was added but allowed to incubate in thioctic acid solutions were used as controls. Functionalization was evaluated using a variety of techniques including extinction spectroscopy. Noble metal nanoparticles (i.e. copper, gold, silver, etc.) exhibit strong extinction properties in the visible region of the electromagnetic spectrum 16, 17 which are sensitive to changes in nanoparticle shape, size, stability, surrounding medium, and/or surface modification. 18, 19 Extinction spectra arise when the frequency of incident light is in resonance with the oscillation of the conduction band electrons of noble metal nanoparticles. This phenomenon is known as the localized surface plasmon resonance (LPR). 20 Figure 1C shows the LPR spectra of Au@citrate (0 hours) and Au@TA nanoparticles after being functionalized in thioctic acid at varying salt concentrations. The gold nanoparticles exhibit an extinction maximum (λ max ) at ~518 nm prior to functionalization. After functionalization with thioctic acid, the λ max shifts to ~521 nm. This value does not change significantly with increased incubation time and is indicative of stable, electromagnetically isolated nanostructures. Because thioctic acid chemisorbs to the surface of gold nanoparticles, the observed optical properties are consistent with changes in the local solution environment upon thioctic acid conjugation. 2
X-Ray Photoelectron pectroscopy of Au@TA Nanoparticles. Quantitative information regarding the packing efficiency of thiols bound to the nanoparticle surfaces are probed using X- ray photoelectron spectroscopy (XP). The sulfur ( 2p) peak of thioctic acid is peak fitted and the areas are normalized using the gold (Au 4f) area to account for differences in nanoparticle concentration between each sample. To quantitate the XP data, the 2p and Au 4f peak areas were converted to a :Au atomic ratio using the empirical atomic sensitivity factor (F) for each element (F = 0.54 and 4.95 for 2p and Au 4f, respectively). 21 Because the Au atomic signal includes both the signal from the surface and a few atomic layers into the gold nanoparticle, the Au signal must be corrected. The depth of the sample that is being signaled in XP can be calculated using the escape depth as follows: Escape Depth = cos( ) (1) where λ is the inelastic mean free path (IMFP) and is the angle between the surface normal and the direction of the emitted electron. 21, 22 For these experiments, is 0 and λ is 1.78 ± 0.00 2 nm where the IMFP is determined using the NIT Electron Inelastic Mean Free Path Database and the average kinetic energy of the Au 4f peaks. 23-25 To apply this to a nanoparticle, the shell method must also be implemented to calculate the total number of atomic gold layers in a nanoparticle. 5, 26 The shell method models a nanoparticle as a central atom which is surrounded by n shells (i.e. layers) of gold atoms where the number of gold atoms in the nth shell can be calculated using the equation 10n 2 +2. 5, 26 Next, the total number of shells per nanoparticle is calculated by dividing the nanoparticle radius (d Au = r NP /2) by the gold atom diameter (r Au = 2.882Å) (Figure 2). Dividing the escape depth by the diameter of a gold atom will yield the number of atomic layers signaled (N layer ) as follows: N layer 6. 2 rau layers (2) 3
where the number of layers is rounded to the closest whole number of shells in subsequent calculations. Using these equations, a gold nanoparticle (d = 11.6 nm) contains 20 shells total but ~6 are sampled in these XP conditions. The shell method can be applied to correct the to Au atomic ratio (/Au surface ) using the following equation: Au 20 n15 2 (10n 2) 2 10n surface 2 Au (3) where the numerator in the first set of brackets is the number of total gold atoms signaled (n = 15 to 20) and the denominator is the number of surface gold atoms (n = 20). In the second set of brackets, /Au represents the (sensitivity factor) corrected XP signal. In order to distinguish if increased thioctic acid AM packing density arises from the systematic addition of NaCl or from increased incubation time with thioctic acid, the /Au surface atomic ratio for gold nanoparticles incubated with thioctic acid in the absence and presence of NaCl are compared (Figure 3). In both ligand exchange environments, the /Au surface atomic ratio increases systematically with increasing thioctic acid incubation times. Furthermore, these data clearly display that the /Au surface atomic ratio saturates after an incubation period of 72 hours. Additionally, longer incubation times increase the number of thioctic acid molecules on the Au nanoparticle surface, and salt mediates this process. 5, 21-26 By applying an exponential fit to these data, a saturated /Au surface atomic ratio of 0.32 9 and 0.38 8 is calculated for Au@TA nanoparticles incubated in absence and presence of NaCl, respectively. In the absence of NaCl, these values imply that at least three gold atoms interact with one sulfur atom. In comparison, this value decreases to ~2.5 gold atoms interacting with each sulfur atom for AMs prepared in the presence 4
of NaCl. While the difference between the /Au surface atomic ratio is small, significant differences in the number of molecules on nanoparticle surfaces are indicated. Expanding on these data, the packing density of thioctic acid AMs on gold nanoparticle surfaces can be estimated. It should be noted that the surface of ~12 nm gold nanoparticles contain predominately (100) surface planes. 27, 28 As a result, the packing density of atoms on the surface on the nanoparticle (σ hkl ) can be calculated as follows: 4 hkl (4) 1 2 2 2 2 2 Qa ( h k l ) where Q is 2 for (100) and a is the bulk lattice parameter. Next, the packing density of thioctic acid on Au@TA nanoparticle surfaces prepared in the absence and presence of salt can be approximated from XP data as follows: Au TA 100 Packing density = surface (5) where the corrected XP signal (Equation 3), the thioctic acid to sulfur ratio (2 sulfur atoms per thioctic molecule), and the gold atom packing density for a (100) surface plane are found in the first, second, and third brackets, respectively. For Au@TA nanoparticles prepared in the absence of salt, surface coverage is ~1.72 x 10 14 molecules/cm 2 after 16 hours of equilibration time and increases to 1.97 x 10 14 molecules/cm 2 after equilibrating for 72 hours. With systematic NaCl additions, Au@TA nanoparticles equilibrated for 16 hours exhibit thioctic acid packing densities of 1.73 x 10 14 molecules/cm 2 and increases to 2.29 x 10 14 molecules/cm 2 after equilibrating for 72 hours. The packing density calculations demonstrate that the slow addition of salt increases thioctic acid AM packing density by 17% relative to the absence of salt after a 72 hour incubation period. These values agree well with previously reported thioctic acid packing densities on flat gold surfaces which range from 1.8x10 14 2.1x10 14 molecules/cm 2, 29-37 indicating XP is an 5
excellent technique to calculate AM packing density on gold nanoparticles. Debye Length Implications. The LPR and XP measurements suggest that the slow addition of salt increases the chemisorption of thioctic acid on gold nanoparticle surfaces thereby resulting in overall higher AM surface coverages than when no salt is included. We hypothesize that the mechanism of this effect is attributed to changes in the Debye length surrounding the carboxylic acid headgroups and can be calculated using the Debye-Hückel limiting law. Relative permittivities ( ) required for the Debye length calculation are computed using the following expression derived by Fawett et. al.: 38 3 78.45- C b C 2 (6) where is the permittivity decrement (16 L mol -1 when NaCl is the electrolyte), C is the electrolyte concentration, and b is a constant with a value of 5 L 2/3 mol -3/2. The Debye length ranges from 5.6 1.8 nm in NaCl concentrations of 0 and 25 mm, respectively. As expected the NaCl concentration is inversely proportional to Debye length; thus as NaCl concentration increases, defects in the AM layer are more easily accessed and filled by additional thioctic acid molecules. Evaluation of Au@TA Nanoparticle tability. To further investigate how AM packing density impacts nanostructure stability, the flocculation parameter 39-41 for Au@TA nanoparticles incubated for 72 hours in the absence and presence of NaCl was evaluated. Flocculation (formation of aggregates and agglomerates) parameter studies can be used to gain semi-quantitative information about the nanostructure stability by monitoring changes in extinction as a function of solution ph and/or time. 39-41 To monitor flocculation, gold nanoparticles are incubated for 72 hours in the absence and presence of salt, centrifuged, and redispersed to 3.0 m cm -1 sodium acetate and phosphate buffer (ph 5.5 and ph 12, respectively). The solution is stirred, and extinction spectra 6
are collected every 2 seconds. As nanoparticle flocculation increases, the extinction intensity at ~521 nm decreases while a new lower energy band at 650 nm intensifies (Figure 4). Next, the extinction spectra are integrated from 575 800 nm to quantify the degree of nanoparticle flocculation as a function of time (Figure 5). In ph 5.5 buffer, the integrated area increases as a function of time and at different rates for the two nanoparticle samples. To compare the stability of the Au@TA nanoparticles, the integrated data is used to determine when flocculation reached its maximum. Larger flocculation values are indicative of more stable nanostructures. 41 The Au@TA nanoparticles incubated without NaCl flocculated within 26 seconds whereas the Au@TA nanoparticles incubated in the presence of NaCl flocculated in twice the time. ignificantly, these flocculation studies reveal that salt-mediated thioctic acid self-assembly increases the stability of the Au@TA nanoparticles by ~20% vs. controls. Conclusions In summary, gold nanoparticles functionalized with thioctic acid were prepared with the slow addition of NaCl. Extinction spectroscopy, XP, and flocculation studies determined that the selfassembly of thioctic acid on gold nanoparticles increases with increasing NaCl concentration. First, an increase in NaCl decreases the Debye length surrounding the deprotonated carboxylate groups on the assembled thioctic acid molecules thereby facilitating increased AM packing densities. Furthermore, the slow addition of NaCl to gold nanoparticles during thioctic acid self assembly increased subsequent functionalized nanoparticle stability vs. controls as determined from flocculation studies. We expect these results to improve strategies for reproducible AM formation on solution-phase nanostructures. Future studies could be expanded to investigate how nanoparticle shape, size, and radius of curvature impact this self assembly process for ultimate improvements in the reproducible synthesis and use of nanomaterials in a variety of applications. 7
Figure 2. chematic representation of a gold nanoparticle cross section which reveals concentric shells of gold atoms surrounding a central atom. Figure 1. low addition of NaCl to Au@TA nanoparticles. (A) tructure of thioctic acid. (B) chematic of the proposed mechanism for thioctic acid packing with the slow addition of NaCl. (C) Extinction spectra of Au@TA nanoparticles equilibrated for 0-72 hours. The inset shows an enlarged view of the extinction maxima (λ max = 518 nm and ~521 nm for 0 and 16 72 hours, respectively). Figure 3. XP characterization of Au@TA nanoparticles. Comparison of the :Au surface atomic ratio (right-hand y-axis) and packing density (left-hand y axis) vs. incubation time for Au@TA nanoparticles prepared in the presence and absence of NaCl. The solid lines represent exponential fits for the :Au surface atomic ratio vs. incubation time: No NaCl y = -0.19 8 e -x/10.4 + 0.32 9 and With NaCl y = -0.24 0 e -x/18.4 + 0.38 8. 8
Figure 4. Normalized extinction spectra for Au@TA nanoparticles incubated for 72 hours in the (A) absence and (B) presence of 16 mm NaCl. Nanoparticles were centrifuged, dispersed in buffer (ph 5.5), and monitored while stirring as a function of time. Figure 5. Normalized integrated area for Au@TA nanoparticles incubated for 72 hours in the presence and absence of salt. Extinction spectra of Au@TA nanoparticles in buffer (ph=5.5) were integrated from λ = 575-800 nm to semi-quantify flocculation. 9
References (1) Wang, G. L.; Murray, R. W. Nano Lett. 2004, 4, 95-101. (2) Warner, M. G.; Hutchison, J. E. Nat. Mat. 2003, 2, 272-277. (3) DeVries, G. A., et al. cience 2007, 315, 358-361. (4) ardar, R.; humaker-parry, J.. Nano Lett. 2008, 8, 731-736. (5) Daniel, M. C.; Astruc, D. Chem. Rev. 2004, 104, 293-346. (6) Thomas, K. G.; Kamat, P. V. Acc. of Chem. Res. 2003, 36, 888-898. (7) Katz, E.; Willner, I. Ange. Chemie 2004, 43, 6042-6108. (8) Bonnemann, H.; Richards, R. M. Euro. J. Inorg. Chem. 2001, 2455-2480. (9) Ghosh, P.., et al. AC Nano 2008, 2, 1335-1344. (10) Bastus, N. G., et al. AC Nano 2009, 3, 1335-1344. (11) Ulman, A. Chem. Rev. 1996, 96, 1533-1554. (12) Finklea, H. O., et al. J. Phys. Chem. 1996, 100, 18852-18858. (13) Preiner, M. J.; Melosh, N. A. Langmuir 2009, 25, 2585-2587. (14) Lusk, A. T.; Jennings, G. K. Langmuir 2001, 17, 7830-7836. (15) Volkert, A. A., et al. Manuscript ubmitted. (16) Kreibig, U. In Handbook of Optical Properties; Hummel, R. E., Wissmann, P., Eds.; CRC Press: Boca Raton, 1997; Vol. II, pp 145-190. (17) Haes, A. J., et al. MR Bulletinz 2005, 30, 368-375. (18) Haynes, C. L., et al. In Topics in Fluorescence; Lakowicz, J. R., Ed.; Plenum Press: New York, 2003; Vol. 8, pp 1-54. (19) Haynes, C. L., et al. Anal. Chem. 2005, 77, 338A-346A. (20) Mie, G. Ann. Phys. 1908, 25, 377-445. (21) Lumsden, J. B. In Materials Characterization; Whan, R. E., Ed.; American ociety for Metals: Metals Park, OH, 1986; Vol. 10. (22) Baer, D. R.; Engelhard, M. H. J. Electron pectrosc. Relat. Phenom. 2010, 178, 415-432. (23) Giray, C. B., et al. canning 2007, 29, 206-210. (24) Jablonski, A., 1.1 ed.; National Institutes of tandards and Technology: Gaithersburg, MD, 2000. (25) Beilschmidt, H., et al. urf. and Interface Anal. 1994, 22, 120-123. (26) chmid, G. Adv. Eng. Mater. 2001, 3, 737-743. (27) Cao, L., et al. J. Coll. Interface ci. 2006, 293, 69-76. (28) Cademartiri, L.; Ozin, G. Concepts of Nanochemistry; Wiley-VCH: Weinheim, 2009. (29) Wang, Z. L., et al. J. Phys. Chem. B 1998, 102, 6145-6151. (30) Dong, Y. Z., et al. Langmuir 2003, 19, 8922-8926. (31) Madoz, J., et al. J. Am. Chem. oc. 1997, 119, 1043-1051. (32) Kuwabata,., et al. Langmuir 1999, 15, 6807-6812. (33) Mazur, M.; Krysinski, P. Langmuir 2001, 17, 7093-7101. (34) Willey, T. M., et al. Langmuir 2004, 20, 4939-4944. (35) Porter, M. D., et al. J. Am. Chem. oc. 1987, 109, 3559-3568. (36) Rooth, M.; haw, A. M. J. Phys. Chem. C 2007, 111, 15363-15369. (37) chlenoff, J. B., et al. J. Am. Chem. oc. 1995, 117, 12528-12536. (38) Fawcett, W. R.; Tikanen, A. C. J. Phys. Chem. 1996, 100, 4251-4255. (39) Weisbecker, C.., et al. Langmuir 1996, 12, 3763-3772. (40) Mayya, K.., et al. Langmuir 1997, 13, 3944-3947. (41) Volkert, A. A., et al. Chem. Commun. 2011, 47, 478-480. 10