Colloid Stability of Thymine-Functionalized Gold Nanoparticles

Oct 24, 2007 - Anuttam Patra , John Ralston , Rossen Sedev and Jingfang Zhou. Langmuir ... Jingfang Zhou, Rossen Sedev, David Beattie, and John Ralsto...
16 downloads 0 Views 634KB Size
Download PDF
12096

Langmuir 2007, 23, 12096-12103

Colloid Stability of Thymine-Functionalized Gold Nanoparticles Jingfang Zhou, David A. Beattie, John Ralston,* and Rossen Sedev Ian Wark Research Institute, UniVersity of South Australia, Mawson Lakes Campus, Mawson Lakes, Adelaide, SA 5095, Australia ReceiVed July 3, 2007. In Final Form: September 6, 2007 Gold nanoparticles surface-coated with thyminethiol derivatives containing long hydrocarbon chains have been prepared. The diameter of the particles is 2.2 and 7.0 nm, respectively, with a relatively narrow size distribution. Thyminethiol derivatives are attached to the gold particle surfaces with thymine moieties as the end groups. The colloid stability of the gold nanoparticles as a function of the type and concentration of monovalent salt, pH, and particle size was investigated in alkaline, aqueous solutions. The gold particles are stable in concentrated NaCl and KCl solutions, but are unstable in concentrated LiCl and CsCl solutions. The larger gold particles are more sensitive to salt concentration and aggregate at lower salt concentrations. The reversible aggregation and dispersion of the gold particles can be controlled by changing the solution pH. The larger gold particles can be dispersed at higher pH and aggregate faster than the smaller particles, due to stronger van der Waals forces between the larger particles. Hydration forces play an important role in stabilizing the particles under conditions where electrostatic forces are negligible. The coagulation of the gold nanoparticles is attributed to van der Waals attraction and reduced hydration repulsion in the presence of LiCl and CsCl.

I. Introduction The colloid stability of aqueous dispersions can often be described by the Derjaguin-Landau-Verwey-Overbeek (DLVO) theory, in which electrostatic and van der Waals interactions are accounted for.1 Deviations from DLVO behavior are well-known, however, and include steric, hydrophobic, and hydration interactions, among others.2 For example, it has long been realized that silica particles are much more stable than DLVO theory predicts at high ionic strength.3 Healy et al.4 found that amphoteric latex particles are stable even in 3.0 M KCl. Ortega-Vinuesa et al.5 observed that polystyrene particles covered by IgG proteins coagulate with increasing ionic strength, but are restabilized at sufficiently large electrolyte concentrations. Repulsive hydration forces have been widely accepted to explain the above behavior. Hydration forces have been intensively investigated in the past two decades on hydrophilic or charged surfaces, for example, biopolymers,6 amphoteric latex particles,7 particles covered by hydrophilic macromolecules,8,9 mica,10 and metal oxides, such as SiO2.11,12 The surface force apparatus (SFA)10 and atomic force microscope (AFM)13 are commonly used to investigate * Corresponding author. E-mail: [email protected]. (1) Hiemenz, P. C.; Rajagopalan, R. Principles of Colloid and Surface Chemistry, 3rd ed.; Marcel Dekker Inc.: New York, 1997. (2) Lyklema, J. H. Fundamentals of Interface and Colloid Science; Vol IV: Particulate Colloids; Elsevier Academic Press: New York, 2005. (3) Yotsumoto, H.; Yoon, R. H.; Wakamatsu, T.; Ito, S.; Sakamoto, H. Shigen Sozai 1993, 109, 909-15 (in Japanese). (4) Healy, T. W.; Homola, A.; James, R. O.; Hunter, R. J. Polym. Colloids 2 1978, Proc. Symp. Phys. Chem. Colloidal Part, 527-36. (5) Lopez-Leon, T.; Gea-Jodar, P. M.; Bastos-Gonzalez, D.; Ortega-Vinuesa, J. L. Langmuir 2005, 21, 87-93. (6) Rowe, A. J. Biophys. Chem. 2001, 93, 93-101. (7) Valle-Delgado, J. J.; Molina-Bolivar, J. A.; Galisteo-Gonzalez, F.; GalvezRuiz, M. J. Prog. Colloid Polym. Sci. 2004, 123, 255-9. (8) Molina-Bolivar, J. A.; Galisteo-Gonzalez, F.; Hidalgo-Alvarez, R. Colloids Surf., B 1999, 14, 3-17. (9) Valle-Delgado, J. J.; Molina-Bolivar, J. A.; Galisteo-Gonzalez, F.; GalvezRuiz, M. J.; Feiler, A.; Rutland, M. W. Langmuir 2005, 21, 9544-54. (10) Pashley, R. M. J. Colloid Interface Sci. 1981, 80, 153-62. (11) Valle-Delgado, J. J.; Molina-Bolivar, J. A.; Galisteo-Gonzalez, F.; GalvezRuiz, M. J.; Feiler, A.; Rutland, M. W. J. Chem. Phys. 2005, 123, 034708/1034708/12. (12) Vigil, G.; Xu, Z.; Steinberg, S.; Israelachvili, J. J. Colloid Interface Sci. 1994, 165, 367-85.

these forces. Several theoretical models14 have been proposed to elucidate the origin of hydration forces at the molecular level, based on experimental data and computer simulations. Despite the debate over these models, the repulsive hydration force is generally attributed to the organization and orientation of water in the vicinity of a surface, functioning at a short distance with a decay length in the 0.2-1.1 nm range. However, the preexponential factor in the force law may vary by more than an order of magnitude, depending on the nature of the surface in question.11,15 Force measurements also reveal that the hydration force is influenced not only by the hydrophilicity of the surface in question but also by the nature and concentration of the hydrated counterions that surround the surface.16 Israelachvilli and coworkers15 found that the hydration force oscillated between two mica surfaces immersed in aqueous KNO3 solutions, but smoothed out for soft surfaces as well as for rough surfaces, such as those found on many particles. Although the molecular origin of specific ion effects on colloid stability is still not completely understood, the structure-modifying approach15 is generally applied to explain these phenomena qualitatively. Ions are assumed to adsorb onto the particle surface and hence modify water structure in the vicinity of the surface. The hydration force is large for strongly hydrated or “structure making” ions with strongly ordered local water structure; the latter can be reduced remarkably by adding weakly hydrated or “structure breaking” ions, which can cause dispersions to coagulate. Ruckenstein and Manciu17-21 recently proposed a surface dipole theory in which double-layer and hydration (13) Veeramasuneni, S.; Yalamanchili, M. R.; Miller, J. D. Colloids Surf., A 1998, 131, 77-87. (14) Leikin, S.; Parsegian, V. A.; Rau, D. C.; Rand, R. P. Annu. ReV. Phys. Chem. 1993, 44, 369-95. (15) Israelachvili, J. Intermolecular & surface forces, 2nd ed.; Academic Press Limited: London, 1991. (16) Pashley, R. M. Chem. Scr. 1985, 25, 22-7. (17) Ruckenstein, E.; Manciu, M. AdV. Colloid Interface Sci. 2003, 105, 177200. (18) Manciu, M.; Ruckenstein, E. Langmuir 2001, 17, 7061-70. (19) Manciu, M.; Ruckenstein, E. AdV. Colloid Interface Sci. 2004, 112, 10928. (20) Manciu, M.; Ruckenstein, E. Langmuir 2001, 17, 7582-92. (21) Ruckenstein, E.; Manciu, M. Langmuir 2002, 18, 7584-93.

10.1021/la7019878 CCC: $37.00 © 2007 American Chemical Society Published on Web 10/24/2007

Thymine-Functionalized Gold Nanoparticles

interactions couple together. They modified the PoissonBoltzmann formalism by including additional ion-hydration and ion-dispersion interactions which account for specific ion effects on double-layer interactions.17 They discovered that coupling between the double layer and hydration can increase the decay length of the repulsive force, especially at high ionic strength. Surface dipoles can increase or decrease the total repulsion depending on the specific ions involved. Although Ortega-Vinuesa and co-workers5 questioned the validity of this approach, when applied to the influence of some anions on the stability of IgGlatex particles, colloid stability at high ionic strength,18 restabilization mechanisms,22 and specific ions effects23 can generally be well-explained by this theory, especially for cations. The aggregation and dispersion properties of nanosized colloid particles are essential for their optical, electronic, and catalytic applications.24,25 In most cases, however, nanoparticles are surrounded by an organic layer. For colloids whose surfaces are coated by surfactants, polymers, or biomolecules, the nature and chemistry of the organic layer play an important role in stabilizing the particles.2 Non-DLVO forces operate and have a strong influence on stability, especially for small particle sizes.2,15 For coated nanoparticles with a diameter less than 10 nm, the particle size is comparable in dimension to that of the adsorbed organic layer. There are very few studies of the colloid stability of dispersions of this type, prompting the present investigation. Thymine molecules are ionizable and pH-sensitive. Previous studies have shown that the surface pKa of thyminethiol derivatives with long hydrocarbon chains is 11.2.26 If the solution pH is above 11.2, deprotonation of the hydrogen attached to thymine at the N3 position will occur. On the other hand, strong interactions exist between thymine molecules, for example, H-bonding, dipole-dipole interaction, and π-stacking of the thymine rings.27 Therefore, if thymine-containing molecules are attached to particle surfaces, they may potentially cause strong adhesive attractions2,28 when the particles approach and make contact. The purpose of this work is to study the aggregation and dispersion behavior of gold nanoparticles coated by thyminethiol derivatives containing long hydrocarbon chains. We focus on the influence of particle size, solution pH, monovalent salt type, and concentration and on the role of the thymine end group in controlling colloid stability. II. Experimental Methodology 1. Reagents. AR grade reagents were purchased from Aldrich and used without further purification. The thyminethiol derivative, 1-(10-mercaptodecyl)-5-methylpyrimidine-2,4-dione (TSH), was synthesized in our laboratory according to the method we have reported previously.29 AR grade solvents used were obtained from BDH or Chem-Supply. Water was purified by a Millipore Ultrapure water system and had a resistivity of 18.2 MΩ‚cm at 25 °C. 2. Techniques. The samples for transmission electron microscopy (TEM) were prepared by placing one drop of particle dispersion in pH 12.5 water onto standard carbon-coated Formvar films mounted on copper grids (200 mesh). A piece of tissue was used to adsorb water and permit the samples to dry faster. TEM images were obtained with a Philips CM100 electron microscope operating at 100 kV. The (22) Davalos-Pantoja, L.; Ortega-Vinuesa, J. L.; Bastos-Gonzalez, D.; HidalgoAlvarez, R. Colloids Surf., B 2001, 20, 165-75. (23) Peschel, G.; Van, Brevern, O. Prog. Colloid Polym. Sci. 1991, 84, 405-8. (24) Hu, Y.; Dai, J. Miner. Eng. 2003, 16, 1167-72. (25) Snoswell, D. R. E.; Duan, J.; Fornasiero, D.; Ralston, J. J. Colloid Interface Sci. 2005, 286, 526-35. (26) Jang, Y. H.; Sowers, L. C.; Cagin, T.; Goddard, W. A. J. Phys. Chem. A 2001, 105, 274-80. (27) Inaki, Y.; Mochizuki, E.; Yasui, N.; Miyata, M.; Kai, Y. J. Photopolym. Sci. Technol. 2000, 13, 177-82. (28) Liu, J.; Zhang, L.; Xu, Z.; Masliyah, J. Langmuir 2006, 22, 1485-92. (29) Lake, N.; Ralston, J.; Reynolds, G. Lamgmuir 2005, 21, 11922-31.

Langmuir, Vol. 23, No. 24, 2007 12097 size distributions of the gold cores were measured using Universal TEM Imaging Platform software and were based on counting at least 150 individual particle images. Transmittance infrared absorbance spectra in KBr pellet were acquired using a Nicolet Magna-IR 750 spectrometer at a resolution of 2 cm-1. The gold particles were dispersed in pH 12.5 water at a concentration of 0.1 mg/mL. At this concentration, the gold nanoparticles are Rayleigh scatters and the “absorbance” is thus linear with respect to concentration. In addition, the thymine absorbance is less than 1, and is of course proportional to the concentration of thymine molecules in the dispersion. When salt was added, the solution was vigorously shaken and then left for a sufficient time for the gold particles to settle, after which UVvisible absorption measurements were carried out with a Varian Cary 5 UV-vis-NIR spectrophotometer at room temperature. In this manner, the behavior of gold particles dispersed in the aqueous solution could be examined. The extinction coefficient of gold particles in the UV-vis region is caused by the light scattering of gold particles, as well as a surface plasmon band absorption around 525 nm. The thymine molecules have a characteristic absorption peak near 270 nm. UV-vis absorption studies of a series of gold dispersions with different concentrations were performed and it was found that the absorbance was linear with concentration, as expected for Rayleigh scatters.1 Therefore, UV-vis absorption could be applied to monitor the colloid stability of the gold particles. In the present study, the maximum absorption of thymine was chosen as a reference point when accounting for the stability of the gold particles. The data were determined in duplicate or triplicate with a reproducibility of (5% or better. pH measurements were conducted in a Class 100 clean room at a constant temperature of 22 °C. The pH probe (Model IJ44) from Ionode is a combination Li glass bulb and Ag/AgCl electrode. The pH-sensitive glass has a very low sodium ion error and can be corrected using the standard curve provided by the company at high Na+ concentration. The measuring pH range is 0-14 with an accuracy of (0.02 pH. Before each experiment or set of experiments, the pH probe was calibrated with fresh pH 7.0 and 10.0 standard Merck buffer solutions. The zeta potentials of the gold particles in aqueous dispersions were determined using the phase analysis light scattering (PALS) technique. The PALS system was designed and constructed by the Laser Light Scattering and Materials Science Group at the University of South Australia.31 PALS uses a cross-beam technique. Acoustooptic module and Bragg cells are used to offset one of the laser beams relative to the other. The system is capable of reliably measuring electrophoretic mobility down to at least 10-11 m2 s-1 V-1. The measurements were performed three times for each condition and the average value was used in the final data sets. 3. Preparation of Thymine Surface-Coated Au Nanoparticles (AuTSH). Thymine-coated gold nanoparticles were prepared using a modified two-phase transfer method.30 In situ and ex situ (ligand replacement) methods were chosen to obtain different sized gold particles. HAuCl4 (0.12 mmol) dissolved in 20 mL of water was added to 0.48 mmol of tetraoctylammonium bromide (TOAB) dissolved in 20 mL of toluene. The mixture was stirred for 30 min and the water layer was discarded. The organic layer was used as a stock solution. For the in situ method, 0.24 mmol of thyminethiol molecules dissolved in 20 mL of toluene were then added to the stock solution. After the solution was stirred for 10 min, 1.2 mmol of NaBH4, freshly dissolved in 2 mL of water, was quickly added. The solution turned brown immediately and a black powder appeared with time. The solution was stirred overnight under N2 and the black powder was collected. For the ex situ method, 1.2 mmol of NaBH4 dissolved in 5 mL of water was slowly added to the stock solution. A purplish-red solution was obtained and stirred for 3 h under N2. Then, 0.24 mmol of thyminethiol molecules dissolved in 20 mL of toluene were added drop by drop and stirred overnight under N2. (30) Brust, M.; Walker, M.; Bethell, D.; Schiffrin, D. J.; Whyman, R. J. Chem. Soc., Chem. Commun. 1994, 801-2. (31) http://www.unisa.edu.au/laser/Research/PALS.asp.

12098 Langmuir, Vol. 23, No. 24, 2007

Zhou et al.

Figure 1. TEM images and size distributions of Au-TSH nanoparticles: (a) in situ gold; (b) ex situ gold. The black powder was precipitated during reaction. The final products were purified by sonicating in CHCl3 or MeOH and then collected by centrifugation. The process of sonicating and centrifuging was repeated several times to remove any unbound thyminethiol molecules and TOAB in the final product. The gold nanoparticles may be dispersed in highly polar organic solvents such as DMF and DMSO, where they form a transparent dispersion, as well as in high pH (>12) aqueous solutions. In the latter case, deprotonation of the hydrogen attached to thymine occurs and confers stability.29 From the solubility behavior and complementary spectroscopic evidence, thyminethiol molecules are attached to the gold particle surface through the thiol group, leaving the thymine group pendant, although it may exhibit some affinity for gold. 4. Morphology and Structure of Thymine-Coated Gold Nanoparticles. The thyminethiol-coated gold nanoparticles prepared by the in situ and ex situ methods were examined by TEM. The

images and particle size distributions are shown in Figure 1. How the gold particles appear depends on the magnification of the TEM images. They are rather spherical in shape at low magnification, but irregular and highly faceted at high magnification. The average core diameters are 2.2 ( 0.3 and 7.0 ( 1.0 nm for the in situ and ex situ gold particles respectively. During the drying process on the copper grid from high pH water, the gold particles formed nanostructures in some areas, where similar-sized particles tended to form more ordered structures with a hexagonal packing. The average edgeto-edge separation between the metal cores can be calculated from the well-ordered region, yielding a value of about 2.0 nm for both the smaller and larger gold particles. For the thyminethiol derivatives used in this study, the fully extended chain length is about 1.1 nm. The edge-to-edge distance is nearly twice the thickness of the thyminethiol derivatives, which means that interdigitation of the adsorbed organic layer did not occur. This behavior is different from that discovered for alkanethiol-coated gold particles, where inter-

Thymine-Functionalized Gold Nanoparticles

Langmuir, Vol. 23, No. 24, 2007 12099

Figure 2. Transmittance FT-IR spectra of free thyminethiol derivatives (1), 2.2 nm gold particles (2), and 7.0 nm gold particles (3).

Figure 3. UV-vis absorption spectra of free thyminethiol molecules (1), 7.0 nm gold particles (2), and 2.2 nm gold particles (3).

digitation prevails during the drying process and the edge-to-edge distance is generally equal to a single-chain length of the alkanethiols.32 The transmittance FT-IR spectra of the thyminethiol derivatives as well as the composite gold particles in KBr pellet are illustrated in Figure 2. The peak positions, along with the assigned functional groups,33,34 are summarized in Table 1. All of the characteristic peaks of free thyminethiol derivatives can be found in the spectra of the composite gold particles, which means that the thyminethiol derivatives are the essential components in the final product. However, when compared with free thyminethiol molecules, all the absorption peaks for thyminethiol molecules tethered on the gold particle surface become broader. The aromatic C-H and N-H vibrations for these tethered thymines on the particle surface shift to a higher wavenumber. A new peak appeared at 3480 and 3446 cm-1 for the smaller and larger particles, respectively. It is attributed to the formation of hydrogen bonding between thymine units.33,35 The FT-IR spectra confirmed that thyminethiol molecules interact strongly with each other on the surface of the gold particle. The gold nanoparticles were dispersed in pH 12.5 aqueous solution. The in situ gold particles formed a brown, transparent dispersion due to the scattering of the smaller gold particles, whereas the ex situ particles formed a clear, red dispersion resulting from the surface plasmon band absorption of the larger gold particles. The corresponding UV-vis absorption spectra of the thyminethiol derivatives and thymine-coated gold particles are shown in Figure 3. A peak appeared at around 535 nm for the larger particles, which was caused by the gold surface plasmon band absorption and is clear evidence of gold particle formation.36 However, this peak broadened significantly for the smaller particles. A maximum absorption peak around 270 nm occurs for all the samples and is the characteristic absorption peak of thymine. This peak shifted, by about 8 nm, to lower wavelength for the large particles compared with the small particles. This blue shift is attributed to the π-stacking of thymine rings tethered on the larger particles.37,38

PALS technique was applied to measure the electrophoretic mobility, ue, of the gold particles at different particle concentrations. The results obtained are shown in Table 2. The mobility values are very small, as is the corresponding zeta potential, ζ, irrespective of which method of calculation was used.39,40 It is readily seen that the electrostatic interactions are very weak. The monovalent salt influence on the stability of the gold nanoparticles was investigated. The UV-vis absorption spectra of 2.2 nm gold aqueous dispersion (pH ) 12.5) as a function of salt concentration for different monovalent salts are shown in Figure 4. The corresponding normalized maximum absorbance of thymine for each condition is shown in Figure 5. The results showed that NaCl and KCl exerted little influence on colloid stability. These gold particle dispersions were stable, even in saturated NaCl or KCl solutions. On the other hand, the gold particles started to aggregate when LiCl or CsCl was added to the dispersion. Eventually, the gold particles coagulated fully and sedimented in 1.25 M LiCl or 1.5 M CsCl solutions. The same trends were observed for the 7.0 nm gold particles, as shown in Figure 6. These specific ion effects on the coagulation of gold nanoparticles follow the Hofmeister series.41 The results also follow the so-called “lyotropic series”1as coagulants, Cs+ > K+ > Na+, except for Li+ ion. Similar behavior was observed by Healy et al.4 who studied the coagulation behavior of amphoteric polystyrene latex suspensions in the presence of different monovalent salts. When NaCl or KCl was added to dispersions containing different-sized gold particles, both the small and large gold particles were remarkably stable in concentrated NaCl or KCl solutions. The behavior changes markedly for the 2.2 and 7.0 nm gold particles as a function of LiCl and CsCl concentration, as is shown in Figure 7 (combination of Figures 5 and 6). The larger gold particles were more sensitive to LiCl or CsCl concentration and coagulated fully at 0.75 M LiCl or CsCl concentration, a much lower concentration than for the small particles. This behavior reflects the stronger attractive van der Waals forces between the large gold particles, for the van der Waals forces are proportional to particle size. Of course, for the larger composite particles, the gold core will exert a stronger influence on the van der Waals attraction in comparison with their smaller cousins,2 as we will see below. The morphologies of the coagulated small

III. Results and Discussion 1. Stability of Colloidal Gold Nanoparticles. The gold nanoparticles were dispersed in pH 12.5 aqueous solution. The (32) 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. (33) Silverstein, R. M.; Webster, F. X. Spectrometric identification of organic compounds, 6th Ed; John Wiley & Sons, Inc.: New York, 1998. (34) Zhang, S. L.; Michaelian, K. H.; Loppnow, G. R. J. Phys. Chem. A 1998, 102, 461-70. (35) Hamlin, R. M.; Lord, R. C.; Rich, A. Science 1965, 148, 1734-7. (36) Kamat, P. V. Nanoparticles and nanostructured films; Fendler, J., Ed.; Wiley-VCH: Weinheim, 1998. (37) Whittten, D. G.; Chen, L.; Geiger, H. C.; Perlsten, J.; Song, X. J.Phys. Chem. B 1998, 102, 10098-10111. (38) Zhang, J.; Whitesell, J. K.; Fox, M. A. Chem. Mater. 2001, 13, 2323-31.

(39) O’Brien, R. W.; White, L. R. J. Chem. Soc., Faraday Trans. 1978, 274, 1607-18. (40) Delgado, A. V.; Gonzalez-Caballero, F.; Hunter, R. J.; Koopal, L. K.; Lyklema, J. Pure Appl. Chem. 2005, 77, 1753-805. (41) Lopez-Leon, T.; Jodar-Reyes, A. B.; Bastos-Gonzalez, D.; Ortega-Vinuesa, J. L. J. Phys. Chem. B 2003, 107, 5696-708.

12100 Langmuir, Vol. 23, No. 24, 2007

Zhou et al.

Table 1. FT-IR Vibrations of Free Thyminethiol Molecules and Thyminethiol Surface-Coated Gold Nanoparticles band position in cm-1 assignment

free T-C10-SH

2.2 nm Au

7.0 nm Au

CH2 antisymmetric and symmetric stretching CH2 and CH3 bendin secondary amide CdO stretching CdC skeletal in-plane stretching N-H stretching and bending aromatic C-H stretching C-N stretching

2918; 2847 1470; 1429; 1385 1699; 1680 1652 3153; 1639 3010 1356

2920; 2850 1466; 1430; 1385 broad centered at 1678 cannot distinguish 3167; cannot distinguish 3036 1358

2918; 2848 1466; 1431; 1385 broad centered at 1683 cannot distinguish 3165; cannot distinguish 3034 1358

Table 2. Electrophoretic Mobility and Zeta Potential of the Gold Nanoparticles at pH 12.5 Aqueous Solution concentration (mg/mL)

particle size (nm)

electrophoretic mobility (m2/V‚s × 10-9)

zeta potential (mV)

0.5

2.2 7.0

-1.7 ( 2.0 -6.6 ( 1.4

-1.3 ( 1.4 -5.1 ( 1.1

1.0

2.2 7.0

-2.7 ( 1.2 -5.6 ( 2.1

-1.8 ( 0.9 -4.5 ( 1.6

and large gold particles in CsCl solutions are shown in Figure 8. The gold particles formed densely packed aggregates, but retained their individual character, without agglomerating into larger units. The influence of pH on the colloid stability of the gold particles is shown in Figure 9. With increasing pH, the gold particles were dispersed. The further increased the pH, the more dispersed the particles. When the pH value was above 12.3, all the particles were dispersed and transparent dispersions were formed. When the pH was reduced, the particles coagulated and sedimented. This process was reversible and was repeated many times for

both small and large gold particles. However, more 2.2 nm gold particles were dispersed at the same pH than were 7.0 nm diameter particles due to the smaller van der Waals attractions between small gold particles. The gold particles can be dispersed completely at pH 12.3, approximately one pH unit higher than the pKa of the thyminethiol derivatives. This means that almost all of the thymine molecules tethered to the gold particle surface were deprotonated when these particles were fully dispersed. 2. Forces between Gold Nanoparticles. van der Waals forces between two identical spheres immersed in a homogeneous medium are always attractive42 and, in most cases, this attraction is responsible for the aggregation of colloidal particles. According to traditional DLVO theory, the double-layer repulsion forces will keep the particles dispersed. The double-layer repulsion is caused by the charge on the surface, which can be regulated by adjusting solution pH for thymine-coated gold nanoparticles. The surface properties of thymine-coated gold nanoparticles at different pH values are illustrated in Scheme 1. The particle surface is neutral at a dispersion pH less than the pKa of the tethered thymine molecules, where there are no double-layer

Figure 4. Influence of the concentration and type of monovalent salts on the colloidal stability of the 2.2 nm diameter gold particles in pH 12.5 water.

Thymine-Functionalized Gold Nanoparticles

Langmuir, Vol. 23, No. 24, 2007 12101

Figure 5. Influence of monovalent salt concentration on the colloidal stability of the 2.2 nm diameter gold particles in pH 12.5 water.

Figure 6. Influence of monovalent salt concentration on the colloidal stability of the 7.0 nm diameter gold particles in pH 12.5 water.

Figure 7. Comparison of colloid stability between the 2.2 and 7.0 nm diameter gold particles in pH 12.5 water in the presence of LiCl and CsCl.

repulsion forces existing in the system, so that the particles aggregate. When the solution pH increases above the pKa, deprotonation of thymine molecules will occur. The further the pH is from the pKa, the greater the surface potential and the stronger are the double-layer repulsion forces,2 all of which, along with the increased hydration forces and reduced adhesive forces between gold particles,28 lead to enhanced colloid stability. The pKa of thymine is 11.2, so the gold particles are dispersed at pH 12.5. Under these conditions, the thickness of the electrical double layer (Debye-Huckel reciprocal length) at pH 12.5 and (42) Hamaker, H. C. Physica 1937, 4, 1058-1062.

Figure 8. TEM images of the aggregates of 2.2 (a) and 7.0 (b) nm diameter gold nanoparticles in CsCl solutions (pH ) 12.5).

Figure 9. Influence of pH on the colloid stability of the 2.2 and 7.0 nm diameter gold nanoparticles without adding salt.

an ionic strength of 0.03 M is 1.7 nm. The zeta potentials for this system are very small (Table 2), which indicates that electrostatic repulsion plays a minor role in controlling colloid stability. Although it is known43,44 that adsorbed organic layers can reduce the van der Waals attractions or cause steric stabilization, in this study we observed that the stability of gold particles can be controlled by adding salt or adjusting pH. The question arises as to what forces might stabilize the gold nanoparticles at high pH. (43) Bittnera, A. M. Surf. Sci. Rep. 2006, 61, 383-428. (44) Vold, M. J. J. Colloid Sci. 1961, 16, 1-11.

12102 Langmuir, Vol. 23, No. 24, 2007 Scheme 1. Surface Properties of Thymine-Coated Gold Particles at Different p Valuesa

a

Note: The diagram is not drawn to scale.

Scheme 2. Two Identical Spheres P with a Single Identical Adsorbed Layer S in Medium M

Zhou et al.

nanoparticles with diameters of 1.67 and 6.01 nm, respectively, using molecular-dynamics simulations and found that Hamaker’s equation can be used to describe van der Waals forces between spherical nanoparticles quite accurately. The adsorbed organic layer will modify the van der Waals forces between bare gold particles. Vold44 and Ninham and Parsegian,52 among others, developed models to calculate van der Waals forces between coated colloidal particles. A system of two identical spheres with a single identical adsorbed layer in medium M is illustrated in Scheme 2. Based on the generic approach of Usui and Barouch,53 the total van der Waals energy of interaction involves contributions from the organic layer-organic layer interaction, organic layer-particle interaction, and particle-particle interaction, which in this case is given by

Gtotal_vdw ) Gs-s_vdw + Gp-s_vdw + Gp-p_vdw ) 1 R + S A1 2R(R + S) A2 R A3 + + 6 2 D1 2R + S D2 2 D3

[

]

where D is the separation distance between two particles, R is the radius of the particle, S is the thickness of adsorbed layer on the particle surface, and A is the Hamaker constant, where The particle surfaces are negatively charged at a dispersion pH of 12.5. Repulsive hydration forces can of course operate at hydrophilic or negatively charged surfaces,10 and may be modified by certain ions,4,15,19,45,46,47 influencing colloid stability, as observed here. Manciu and Ruckenstein18 considered the role of the hydration force in influencing colloid stability at high ionic strength. If the hydration repulsion is large, the colloids can remain stable at any electrolyte concentration, which reasonably explains the stability of gold particles in this study in saturated NaCl or KCl solutions. Therefore, the forces underlying the colloid stability of gold nanoparticles in high pH water appears to be a weak double-layer repulsion plus a strong hydration repulsion in the absence of added salt. With an increase in ionic strength upon adding salt to the dispersion, the electrostatic forces are screened and are eliminated at about 1.5 M electrolyte concentration.2 Under these conditions, the gold particles are stabilized solely by hydration repulsion forces. We observed that LiCl acted as a more effective coagulant than CsCl for these gold nanoparticle dispersions. The large and small gold nanoparticle dispersions coagulated completely at about 0.75 and 1.25 M LiCl with corresponding pH values of 11.6 and 11.4, respectively. The pH value of the suspension decreased in the presence of LiCl. When the dispersion pH was decreased, protonation of thymine molecules increased, leading to a decrease in surface charge. Israelachvili et al.15 noted that repulsive hydration forces are reduced or eliminated for protonbound surfaces. We also observed that the dispersion became stable when the solution pH was readjusted to above 12.3. CsCl does not reduce the dispersion pH, but certainly does destabilize the dispersion. To help explain the observations, the van der Waals forces between gold nanoparticles were calculated. Hamaker’s equation42 for the calculation of van der Waals attraction is based on continuum theory. Nanoparticles fall in the region between atoms and macroparticles. Fichthorn and Qin48-50 as well as Wang and and co-workers51 calculated the van der Waals attractions between (45) Allen, L. H.; Matijevic, E. J. Colloid Interface Sci. 1969, 31, 287-96. (46) Colic, M.; Fisher, M. L.; Franks, G. V. Langmuir 1998, 14, 6107-12. (47) Colic, M.; Franks, G.; Fisher, M.; Lange, F. Langmuir 1997, 13, 3129-

1/2 2 A1 ) (A1/2 S - AM ) 1/2 1/2 1/2 A2 ) (A1/2 S - AM )(AP - AS ) 1/2 2 A3 ) (A1/2 P - AS )

and

D1 ) D D2 ) D + S D3 ) D + 2S A1, A2, and A3 are the Hamaker constants for the organic layerorganic layer interaction, organic layer-particle interaction, and particle-particle interaction for particles P with adsorbed layer S in medium M. AM, AS, and AP represent the Hamaker constants with respect to vacuum for the medium, adsorbed organic layer, and particle. The corresponding values for water, thyminethiol adsorbed layer with long hydrocarbon chain, and gold core are 4.0 × 10-20, 5.0 × 10-20, and 40 × 10-20 J, respectively.15 The thickness of the adsorbed thymine layer was estimated to be 1.1 nm. The corresponding van der Waals interaction as a function of separation distance for 2.2 and 7.0 nm diameter gold nanoparticles is plotted in Figure 10. The van der Waals interactions contributed from the thyminethiol layer-thyminethiol layer interactions and gold core-thyminethiol layer interactions are only a small component of the total van der Waals interactions. The gold core-gold core interactions predominate in the total interactions for both 2.2 and 7.0 nm gold particles. At a given distance, the Gvdw is greater for the larger particles, but it is attractive for both. Therefore, the forces governing the colloid stability of thymine-functionalized gold nanoparticles in high pH water at different conditions may be described as follows: the Au sols are stable in concentrated NaCl and KCl solutions, as illustrated in Figures 5 and 6. Gvdw is attractive and electrostatic forces are very weak, so the observed colloid stability is due to

35.

(51) Wang, J. C.; Neogi, P.; Forciniti, D. J. Chem. Phys. 2006, 125, 194717(1(48) Qin, Y.; Fichthorn, K. A. J. Chem. Phys. 2003, 119, 9745-54. (49) Qin, Y.; Fichthorn, K. A. Phys. ReV. E 2006, 73, 020401(1-4). (50) Fichthorn, K. A.; Qin, Y. Ind. Eng. Chem. Res. 2006, 45, 5477-81.

6). (52) Ninham, B. W.; Parsegian, V. A. J. Chem. Phys. 1970, 52, 4578-4587. (53) Usui, S.; Barouch, E. J. Colloid Interface Sci. 1990, 137, 281-8.

Thymine-Functionalized Gold Nanoparticles

Langmuir, Vol. 23, No. 24, 2007 12103

by Bostro¨m et al.54 and Ise et al.55 may exist at high ionic strength. AFM studies28,56 revealed that adhesive forces operated between pH-sensitive surfaces in contact at low pH. Thymine end groups can interact strongly on the gold particle surface, as confirmed by FT-IR and UV-vis spectra, which might induce strong adhesive forces when two particles come into contact, contributing potentially to the aggregation of the gold particles at low pH in addition to van der Waals attraction.

IV. Conclusions

Figure 10. Gtotal_vdw interactions as a function of separation distance for 2.2 and 7.0 nm thyminethiol-coated gold nanoparticles.

hydration repulsion. In the presence of LiCl, the dispersions are destabilized, although Li+ ion is generally a structure maker. However, LiCl also reduces the dispersion pH. Thus, the hydration force is weakened as the particle surface is protonated, leading to coagulation of the gold particles. For CsCl, the dispersion pH is not altered, but Cs+ ions are structure breakers. The hydration force is therefore disrupted and weakened, permitting the attractive van der Waals forces to coagulate the gold particles. The larger the gold particles, the greater the attractive Gvdw (Figure 10) and the more sensitive are the gold particles to salt concentration and pH. However, the traditional double-layer repulsion is calculated according to the Poisson-Boltzmann equation, which is inaccurate at concentrations higher than approximately 0.05 M.2,15 Ruckenstein and Manciu17 found that an attractive force can be generated between two surfaces when ion-dispersion interactions are considered in the modified Poisson-Boltzmann approach at high ionic strength. Therefore, double-layer attraction, as observed

Thymine-coated gold nanoparticle dispersions are stable at pH 12.3 and above due to a combination of weak electrostatic repulsion and strongly repulsive hydration forces. The combination of these outweighs the van der Waals attraction, which is dominated by the gold nanocores with the adsorbed organic layer playing a minor role. In the presence of concentrated NaCl and KCl solutions, the dispersions remain stable due to the strong hydration repulsion. Colloid instability occurs in the presence of LiCl and CsCl due to weakened hydration repulsion. The larger gold particles are more sensitive to salt concentration due to greater van der Waals forces. The controlled aggregation and dispersion of nanoparticles are of great interest for applications in nanoscience and nanotechnology with this work providing some new insights. Acknowledgment. Financial support through The Australian Research Council Special Research Centre Scheme is gratefully acknowledged. Discussions with Assoc. Prof. Daniel Fornasiero are warmly acknowledged. LA7019878 (54) Bostrom, M.; Williams, D. M. R.; Ninham, B. W. Phys. ReV. Lett. 2001, 87, 168103(1-4). (55) Ise, N.; Okubo, T.; Sugimura, M.; Ito, K.; Nolte, H. J. J. Chem. Phys. 1983, 78, 536-40. (56) Kane, V.; Mulvaney, P. Langmuir 1998, 14, 3303-3311.