pubs.acs.org/Langmuir © 2010 American Chemical Society

How PEO-PPO-PEO Triblock Polymer Micelles Control the Synthesis of Gold Nanoparticles: Temperature and Hydrophobic Effects )

Poonam Khullar,† Aabroo Mahal,† Vijender Singh,† Tarlok Singh Banipal,‡ Gurinder Kaur,§ and Mandeep Singh Bakshi*,

)

† Department of Chemistry, BBK DAV College for Women, Amritsar 143005, Punjab, India, ‡Department of Applied Chemistry, Guru Nanak Dev University, Amritsar 143005, Punjab, India, §Nanotechnology Research Laboratory, College of North Atlantic, Labrador City, NL A2 V 2K7 Canada, and Department of Chemistry, Mount Saint Vincent University, Halifax, Nova Scotia, B3M 2J6 Canada

Received February 18, 2010. Revised Manuscript Received March 25, 2010 Aqueous micellar solutions of F68 (PEO78-PPO30-PEO78) and P103 (PEO17-PPO60-PEO17) triblock polymers were used to synthesize gold (Au) nanoparticles (NPs) at different temperatures. All reactions were monitored with respect to reaction time and temperature by using UV-visible studies to understand the growth kinetics of NPs and the influence of different micellar states on the synthesis of NPs. The shape, size, and locations of NPs in the micellar assemblies were determined with the help of TEM, SEM, and EDS analyses. The results explained that all reactions were carried out with the PEO-PPO-PEO micellar surface cavities present at the micelle-solution interface and were precisely controlled by the micellar assemblies. Marked differences were detected when predominantly hydrophilic F68 and hydrophobic P103 micelles were employed to conduct the reactions. The UV-visible results demonstrated that the reduction of gold ions into nucleating centers was channeled through the ligand-metal charge -transfer complex (LMCT) and carried out by the surface cavities. Excessive hydration of the surface cavities in the case of F68 micelles produced a few small NPs, but their yield and size increased as the micelles were dehydrated under the effect of increasing temperature. The results concluded that the presence of well-defined predominantly hydrophobic micelles with a compact micelle-solution interfacial arrangement of surface cavities ultimately controlled the reaction.

Introduction Triblock polymers (TBP) belong to a special category of nonionic surfactants with potential industrial applications.1 They possess a unique architecture of three blocks comprising repeating units of polyethylene (PEO) and polypropylene (PPO) oxides. A TBP becomes predominantly hydrophilic when the number of PEO units exceeds the number of PPO units but acquires a predominantly hydrophobic nature when the reverse happens.2 In the aqueous phase, they exist in the form of polydisperse micelles with size ranging from a few to several hundred nanometers depending on the molar mass and concentration of the TBP.3 They also possess the unique property of self-aggregation with respect to temperature.2,4 Greater dehydration of PPO units in comparison to that of PEO units at a particular temperature allows them to aggregate in the form of micelles. The temperature at which this happens is known as the critical micelle temperature (cmt).2,4 Further heating beyond the cmt drastically reduces the hydration of micelles, resulting in a predominantly hydrophobic phase separating out. The temperature where it happens is known as the cloud point (cp).5 Both the cmt *Corresponding author. E-mail: [email protected]. (1) (a) Bahadur, P.; Riess, G. Tenside, Surfactants, Deterg. 1991, 28, 173. (b) Lin, S.-Y.; Kawashima, Y. Pharm. Acta Helv. 1985, 60, 339. (c) Yokoyama, M. Crit. Rev. Ther. Drug Carrier Syst. 1992, 9, 213. (2) Alexandridis, P.; Holzwarth, J. K.; Hatton, T. A. Macromolecules 1994, 27, 2414. (3) (a) Ghosh, S.; Dey, S.; Mandal, U.; Adhikari, A.; Mondal, S. K.; Bhattacharyya, K. J. Phys. Chem. B2007, 111, 13504. (b) Dey, S.; Adhikari, A.; Mandal, U.; Ghosh, S.; Bhattacharyya, K. J. Phys. Chem. B 2008, 112, 5020. (c) Liu, Y.; Chen, S. H.; Huang, J. S. Macromolecules 1998, 31, 6226. (d) Bahadur, P.; Pandya, K. Langmuir 1992, 8, 2666. (e) Jain, N. J.; Aswal, V. K.; Goyal, P. S.; Bahadur, P. J. Phys. Chem. B 1998, 102, 8452. (4) (a) Bakshi, M. S.; Sachar, S. J. Colloid Interface Sci. 2006, 296, 309. (b) Bakshi, M. S.; Kaur, N.; Mahajan, R. K. J. Photochem. Photobiol., A 2007, 186, 349. (c) Bakshi, M. S.; Bhandari, P. J. Photochem. Photobiol., A 2007, 186, 166. (5) Bakshi, M.; Kaur, N.; Mahajan, R.; Singh, J.; Singh, N. Colloid Polym. Sci. 2006, 284, 879.

Langmuir 2010, 26(13), 11363–11371

and cp are specific to a particular TBP concentration in the aqueous phase and are not very prominent in conventional nonionic surfactants because of their monomeric rather than polymeric nature, which provides greater aqueous-phase solubility. PEO and PPO units of TBPs contain ether oxygens, which act as mild reducing agents. Recently, this property has been very well examined by Alexendries et al.6 in the synthesis of gold nanoparticles (NPs). Weak reducing conditions are generally very favorable to shape-controlled synthesis.7 We have tried to correlate this aspect with TBP micellization properties and provide some answers as to how TBP micelles and their physical states are closely related to this property. Below the cmt, TBP exists in the form of a predominantly monomeric or preaggregated state and hence well-defined micelles exist only beyond the cmt.2 A typical TBP spherical micelle8 consists of a PPO core and a PEO corona (6) (a) Sakai, T.; Alexandridis, P. J. Phys. Chem. B 2005, 109, 7766. (b) Sakai, T.; Alexandridis, P. Langmuir 2005, 21, 8019. (c) Sakai, T.; Alexandridis, P. Langmuir 2004, 20, 8426. (d) Sakai, T.; Alexandridis, P. Chem. Mater. 2006, 18, 2577. (7) (a) Gou, L.; Murphy, C. J. Chem. Mater. 2005, 17, 3668. (b) Gole, A.; Murphy, C. J. Chem. Mater. 2004, 16, 3633. (c) Jana, N. R.; Gearheart, L.; Murphy, C. J. J. Phys. Chem. B 2001, 105, 4065. (d) Bakshi, M. S. Langmuir 2009, 25, 12697. (e) Bakshi, M. S.; Possmayer, F.; Petersen, N. O. J. Phys. Chem. C 2008, 112, 8259. (f) Bakshi, M. S.; Possmayer, F.; Petersen, N. O. J. Phys. Chem. C 2007, 111, 14113. (g) Bakshi, M. S.; Possmayer, F.; Petersen, N. O. Chem. Mater. 2007, 19, 1257. (h) Bakshi, M. S.; Sachar, S.; Kaur, G.; Bhandari, P.; Kaur, G.; Biesinger, M. C.; Possmayer, F.; Petersen, N. O. Cryst. Growth Des. 2008, 8, 1713. (i) Bakshi, M. S.; Kaur, G.; Thakur, P.; Banipal, T. S.; Possmayer, F.; Petersen, N. O. J. Phys. Chem. C 2007, 111, 5932. (8) (a) Mortensen, K.; Batsberg, W.; Hvidt, S. Macromolecules 2008, 41, 1720. (b) Kositza, M. J.; Bohne, C.; Alexandridis, P.; Hatton, T. A.; Holzwarth, J. F. Macromolecules 1999, 32, 5539. (c) de Bruijn, V. G.; van den Broeke, L. J. P.; Leermakers, F. A. M.; Keurentjes, J. T. F. Langmuir 2002, 18, 10467. (d) Liang, X.; Guo, C.; Ma, J.; Wang, J.; Chen, S.; Liu, H. J. Phys. Chem. B 2007, 111, 13217. (e) L€of, D.; Schill€en, K.; Torres, M. F.; M€uller, A. J. Langmuir 2007, 23, 11000. (f) Bakshi, M. S.; Kaura, A.; Bhandari, P.; Kaur, G.; Torigoe, K.; Esumi, K. J. Nanosci. Nanotechnol. 2006, 6, 1405. (g) Bakshi, M. S.; Kaura, A.; Kaur, G. J. Colloid Interface Sci. 2006, 296, 370. (h) Bakshi, M. S.; Singh, J.; Kaur, J. J. Colloid Interface Sci. 2005, 287, 704. (i) Bakshi, M. S.; Sachar, S.; Singh, K.; Shaheen, A. J. Colloid Interface Sci. 2005, 286, 369. (j) Bakshi, M. S.; Singh, J.; Kaur, G. J. Colloid Interface Sci. 2005, 285, 403.

Published on Web 04/06/2010

DOI: 10.1021/la100734p

11363

Article

Khullar et al.

because of the greater solubility of the later in the aqueous phase. Upon dehydration, the micellar phase collapses into a kind of reversible vesicular phase with little aqueous-phase solubility. Thus, a difference in the reducing ability of a TBP is expected when the temperature varies over a wide range covering pre- to postmicellar states and then extends to vesicular states. In addition, different aggregated states also act as soft templates in a different manner and are expected to produce different forms of nanomaterials. This will allow us to understand how different aggregated states of TBP macromolecules can be used in the synthesis of different forms of nanomaterials.9 To understand the reducing mechanism of micellar TBP in producing nanomaterials, we have undertaken a systematic study by monitoring the synthesis of Au NPs in a typical ternary mixture of TBP þ gold salt þ water without the use of an additional conventional reducing agent. The reactions were carried out under precise temperature control because the TBP micellar properties are very sensitive to temperature changes. For this purpose, we have selected two TPBs, viz., F68 (PEO78-PPO30-PEO78) and P103 (PEO17-PPO60-PEO17). F68 is predominantly hydrophilic with a greater number of PEO units than PPO whereas P103 is predominantly hydrophobic with a greater number of PPO units. The cp values of both TPBs are listed in the Supporting Information in Table S1. They are greater than 100 °C for F68 but less than 100 °C for P103. A wide range of temperature from 25 to 80 °C has been selected to induce significant dehydration and subsequent structure transitions in the TBP micelles to study their influence on the Au NPs synthesis. The synthesis of NPs has been explained on the basis of both the hydrophilic and hydrophobic nature of TBPs and their effects on the reducing behavior of gold ions with respect to nucleating centers.

Experimental Section Materials. Chloroauric acid (HAuCl4) from Aldrich and triblock polymers F68 (PEO78-PPO30-PEO78) and P103 (PEO17PPO60-PEO17) were purchased from BASF. Doubly distilled water was used for all preparations. Synthesis of Au NPs. Aqueous mixtures (total 10 mL) of TBP (5/10 mM) and HAuCl4 (0.5/1.0/2.0 mM) were placed in screwcapped glass bottles. After the components were mixed at room temperature, the reaction mixtures were kept in a water-thermostatted bath (Julabo F25) at a precise temperature (below/above cp ( 0.1 °C) for 6 h under static conditions. The solution changed from colorless to pink-purple or purple within half an hour and remained the same thereafter in most cases. After 6 h, the samples were cooled to room temperature and kept overnight. They were purified from pure water at least two times to remove unreacted TBP. Purification was done by collecting the Au NPs at 10 00012 000 rpm for 5 min after washing each time with distilled water. Methods. UV-visible measurements were simultaneously carried out with respect to reaction time as well as temperature with the help of a Shimadzu model 2450 (double beam). This instrument is equipped with a TCC 240A thermoelectrically temperature controlled cell holder that allows one to measure the spectrum at a constant temperature within (1 °C. Transmission electron microscopy (TEM) analysis was carried out on a JEOL 2010F at an operating voltage of 200 kV. The samples were prepared by mounting a drop of a solution onto a carbon-coated Cu grid and allowing them to dry in air. It should be mentioned that no staining agent such as methylamine tungstate or Pd/Pt alloy, which is usually employed to observe clearly the micellar or vesicular assemblies in order to build an appropriate contrast against the bright background, has been used in (9) (a) Aizawa, M.; Buriak, J. M. J. Am. Chem. Soc. 2006, 128, 5877. (b) Aizawa, M.; Buriak, J. M. J. Am. Chem. Soc. 2005, 127, 8932. (c) Chai, J.; Buriak, J. M. ACS Nano 2008, 2, 489. (d) Aizawa, M.; Buriak, J. M. Chem. Mater. 2007, 19, 5090.

11364 DOI: 10.1021/la100734p

the present study. Scanning electron microscopy (SEM) analysis was carried out on a Zeiss NVision 40 dual-beam FIB/SEM instrument. Photomicrographs were obtained in bright-field scanning/imaging mode using a spot size of ∼1 nm and a 12 cm camera length. Energy dispersive X-ray spectroscopy (EDS) microanalysis was carried out using an Oxford-INCA atmospheric ultrathin window (UTW), and the data were processed using the Oxford INCA microanalysis suite, version 4.04. The cp measurements on pure aqueous TBPs were made visually for the onset of turbidity by controlled heating (ca. 1 °C/ min) of the well-stirred samples and then simultaneous cooling of the solutions until they were completely clear. The temperatures were measured to a precision of 0.1 °C. Each value determined is the mean of three separate determinations with reproducibility better than (0.5 °C. These values were further confirmed from the variation in the absorbance of 25 μM methyl orange in aqueous TBP solution with respect to temperature.

Results F68-Au NPs. Because the cp values of 5 and 10 mM F68 were higher than 100 °C, two temperatures (40 and 80 °C) were selected to determine the effect of dehydration on the reducing ability of F68 micelles to produce Au NPs. Figure 1a shows typical UVvisible scans at different time intervals of a reaction at 40 °C in the presence of 5 mM F68. Three prominent peaks appear at 220 (see inset), 320, and 545 nm resulting from AuCl4- ions, a ligandmetal charge-transfer complex (LMCT),10 and the surface plasmon resonance (SPR)11 of Au NPs, respectively. The intensity of the 220 nm peak decreases whereas that of the 320 and 545 nm peaks increases with time. For a ternary mixture of F68 þ HAuCl4 þ water, a decrease in the intensity of AuCl4- ions is directly related to their reduction into Au(0) by F68, which progresses with the passage of time.6 Similar behavior is observed when the same reaction is carried out at 80 °C (Supporting Information, Figure S1). Figure 1b illustrates the course of reaction following the variation in the absorbance of 220 and 545 nm peaks at 40 and 80 °C. At both temperatures, the absorbance of AuCl4- ions continuously decreases and that of NPs increases before it equilibrates where both curves approach each other. This happens within 60 min of the reaction with much higher intensity at 80 °C but takes about 90 min at 40 °C. Another contrasting difference is that the presence of sharp peaks of Au NPs at 80 °C (located at 530 nm, Figure S1) in comparison to relatively broad peaks at 40 °C (located at 545 nm) with a difference of 15 nm in peak positions is clearly related to the mode of aggregation of NPs.7 TEM studies help us to understand the characteristic features of Au NPs and micellar assemblies. Figure 1c shows the TEM image of NPs synthesized at 40 °C. Groups of well-defined spherical compound micelles or vesicular aggregates12 of 27 ( 0.7 nm are evident. Each aggregate contains a few dark NPs13 of 2 to 3 nm (indicated by white arrows). Apart from this, the dark (10) Vogler, A.; Kunkely, H. Coord. Chem. Rev. 2001, 221, 489. (11) (a) El-Sayed, M. A.; Eustis, S. Chem. Soc. Rev. 2006, 35, 209. (b) Liz-Marzan, L. M. Langmuir 2005, 22, 32. (c) Rodríguez-Fernandez, J.; Perez-Juste, J.; Javier García de Abajo, F.; Liz-Marzan, L. M. Langmuir 2006, 22, 7007. (12) (a) Yu, Y.; Eisenberg, A. J. Am. Chem. Soc. 1997, 119, 8383. (b) Zhang, L.; Eisenberg, A. Macromolecules 1996, 29, 8805. (c) Yu, K.; Eisenberg, A. Macromolecules 1998, 31, 3509. (d) Luo, L.; Eisenberg, A. Langmuir 2002, 18, 1952. (e) Zupancich, J. A.; Bates, F. S.; Hillmyer, M. A. Macromolecules 2006, 39, 4286. (f) Bai, Z.; Lodge, T. P. J. Phys. Chem. B 2009, 113, 14151. (g) He, Y.; Lodge, T. P. Macromolecules 2007, 41, 167. (h) Liu, C.; Hillmyer, M. A.; Lodge, T. P. Langmuir 2008, 24, 12001. (13) (a) Azzam, T.; Eisenberg, A. Langmuir 2006, 23, 2126. (b) Azzam, T.; Bronstein, L.; Eisenberg, A. Langmuir 2008, 24, 6521. (c) Sidorov, S. N.; Bronstein, L. M.; Kabachii, Y. A.; Valetsky, P. M.; Soo, P. L.; Maysinger, D.; Eisenberg, A. Langmuir 2004, 20, 3543. (d) Liu, G. J. Curr. Opin. Colloid Interface Sci. 1998, 3, 200. (e) Qi, L.; Colfen, H.; Antonietti, M. Nano Lett. 2000, 1, 61. (f) Klingelhofer, S.; Heitz, W.; Greiner, A.; Oestreich, S.; Forster, S.; Antonietti, M. J. Am. Chem. Soc. 1997, 119, 10116.

Langmuir 2010, 26(13), 11363–11371

Khullar et al.

Article

Figure 1. (a) Absorbance vs wavelength scans of the HAuCl4 (0.5 mM) þ F68 (5 mM) þ water ternary reaction carried out at 40 °C with

respect to time from 2 to 345 min. “Blank” refers to the F68 (5 mM) þ water binary mixture without HAuCl4. Absorbances at 220 (inset), 320, and 545 nm are due to AuCl4- ions, the ligand-metal charge-transfer complex, and Au NPs. (b) Plots of normalized intensity at 220 nm (circles) and 545 nm (diamonds) vs reaction time. Filled symbols are for the reaction at 40 °C, and empty symbols are for the reaction at 80 °C. The arrow indicates a reaction time where both absorbances for a reaction at 80 °C merge with each other and refers to the completion of the reaction. A plot with “plus” symbols refers to the intensity variation of the 320 nm peak for the 40 °C reaction. (c) TEM micrograph of F68 micelles loaded with small NPs synthesized in a HAuCl4 (0.5 mM) þ F68 (5 mM) þ water ternary reaction with precise temperature control of 40 ( 0.1 °C. White arrows indicate the presence of distinct NPs of different sizes. (d) Single compound micelle with NPs at different locations. (e) White arrows indicate the presence of inverted micelles along with NPs in a compound micelle. (f) TEM micrograph of F68 micelles loaded with tiny NPs synthesized in a HAuCl4 (0.5 mM) þ F68 (5 mM) þ water ternary reaction with precise temperature control of 80 ( 0.1 °C. (g, h) Close-up images of a single compound micelle loaded with the dust of tiny NPs. (i) TEM image of compound micelles loaded with large NPs synthesized with 10 mM F68 at 40 ( 0.1 °C; micelles loaded with well-defined NPs shown in j-l are synthesized at 80 ( 0.1 °C. (m) Absorbance vs wavelength scans of the HAuCl4 (0.5 mM) þ F68 (10 mM) þ water ternary reaction with respect to reaction temperature from 25 to 70 °C. (n) Normalized intensity (at 220, 320, and 545 nm) versus temperature plots. A plot with symbol  refers to the variation in the wavelength of the 545 nm peak. See details in the text.

areas of each aggregate are loaded with several tiny NPs. This is clearly visible in Figure 1d,e where just one aggregate has been selected to show the presence of NPs of different sizes, as indicated by white arrows. Individual inverted micelles12 (produced upon dehydration) of 8 ( 3 nm can also be seen as white circles in Figure 1e. It seems that the water pool enclosed by an inverted micelle in fact acts as a reaction vessel where an NP grows and ultimately acquires the size of a water pool. When the same reaction is carried out at 80 °C, no clear micelles are observed (Figure 1f). Instead, long strands of micelles that are probably fused bearing clouds of tiny NPs with a few larger NPs are present. The magnified view of a single aggregate indicates that the surface is fully covered by dark tiny dots (Figure 1g,h). Thus, the tiny dots (NPs) produced at 80 °C (Figure 1g,h) are responsible for the sharp absorbance of this sample (Figure S1) in comparison to the broad absorbance of relatively larger NPs synthesized at 40 °C (Figure 1a). When the same reactions are carried out with 10 mM F68 at 40 °C (Figure 1i) and 80 °C (Figure 1j-l), clear, large polyhedral NPs of 14 ( 7 nm with welldefined facets are produced. Both micellar assemblies and NPs are more prominent at 80 °C rather than at 40 °C, and the NPs are completely entrapped by the compound micelles. It is known that the TBP micelles are never monodisperse like conventional surfactant micelles, and their polydispersity increases with concentration.12 A slight increase in the temperature significantly dehydrates the micelles and induces structural transitions12f-h that ultimately lead to compound micelles, inverted micelles, or Langmuir 2010, 26(13), 11363–11371

vesicle formation.14 Thus, greater dehydration at 80 °C produces more prominent aggregates rather than at 40 °C and is quite true if we compare Figure 1i with Figure 1j-l. TEM studies clearly point to a significant temperature effect on the synthesis of Au NPs and can be best understood if the synthesis is monitored with respect to temperature. Figure 1m shows a temperature effect from 25 to 70 °C on the synthesis of Au NPs in aqueous 10 mM F68. Again, three peaks are present at 220, 320, and 545 nm because of AuCl4- ions, LMCT,10 and Au NPs, respectively. The peak at 220 nm is not shown in Figure 1m to impart better clarity to the other two peaks. The variation in the intensity of all peaks is presented in Figure 1n. Peaks at 220 and 320 nm are interrelated because they belong to Au(III) and Au(I) species, respectively. (A detailed mechanism will be discussed later in the Discussion section.) As the temperature increases to above 25 °C, both peaks show a slight decrease up to 32 °C whereas the 545 nm peak shows a simultaneous increase with a red shift of 10 nm. The temperature range from 25 to 32 °C belongs to the premicellar phase2 where no well-defined micelles are present in the bulk; therefore, the reduction is mainly carried out with F68 monomers or preaggregates. As soon as the micelles are formed at around 32 °C (the cmt of 10 mM F68 is ∼34 °C),2 a significant increase in the intensity of the 320 nm peak happens because of the formation of the LMCT10 between the AuCl4- ions and the PEO (14) (a) Ma, J.-h.; Guo, C.; Tang, Y.-l.; Liu, H.-z. Langmuir 2007, 23, 9596. (b) Chen, S.; Yang, B.; Guo, C.; Ma, J.-h.; Yang, L. R.; Liang, X.; Hua, C.; Liu, H.-z. J. Phys. Chem. B 2008, 112, 15659.

DOI: 10.1021/la100734p

11365

Article

surface cavities.6 A simultaneous reduction of Au(I) to Au(0) causes a rapid increase in the 545 nm peak. It continues until a complete reduction is achieved at around 50 °C, and thereafter both peaks at 220 and 320 nm show a regular fall due to the slow depletion of both Au(III) and Au(I) species to produce Au(0). This is indicated by the predominantly constant intensity of the 545 nm peak with almost constant wavelength. Thus, Figure 1n suggests a collective, active role of Au(III) (220 nm peak) and Au(I) (320 nm peak) species in the synthesis of Au NPs, which will again be discussed in the Discussion section. P103-Au NPs. Unlike F68, P103 is predominantly hydrophobic with PPO = 60 and PEO = 34 units and a cp of 52 °C for 5 mM P103 (Figure S2). Thus, two reaction temperatures (40 and 55 °C) were selected for the synthesis of Au NPs below and above the cp. Figure 2a shows a UV-vis scan of a reaction carried out at 55 °C (i.e., above the cp). Aqueous P103 does not show any absorbance in the visible region. However, as soon as 0.5 mM of HAuCl4 is added, two prominent peaks appear at 220 nm (not shown in Figure 2a) and 320 nm15 within 2 min of the reaction whereas two more peaks appear at 545 and 730 nm in another 10 min. The latter peaks are due to the Au NPs in their different locations in the aqueous bulk phase. As the reaction progresses further up to 345 min, there is about a 10 nm red shift in the 545 nm peak but a much more pronounced red shift of 100 nm in the 730 nm peak. Figure 2b illustrates a systematic variation in the normalized intensities of all four peaks (i.e., 220, 320, 545, and 730 nm). The intensity of the 220 and 320 nm peaks shows a sudden increase within 2 min of the reaction and then remains constant for approximately 40 min. The second one then shows a drastic fall before attaining a constant value, and the first one shows a slight dip. The intensity of the 545 and 730 nm peaks also shows an instantaneous increase within 2 min of the reaction but then follows a less significant increase for up to 40 min before attaining a constant value. The variations in the intensity of all peaks are interrelated. cp is a temperature where reversible phase separation occurs in the aqueous phase because of the formation of large vesicles when dehydrated micelles undergo fusion among themselves. Because such assemblies are already present in the bulk at 55 °C, the addition of HAuCl4 will instantaneously direct the AuCl4- ions to the reducing sites (surface cavities formed by PEO-PPO-PEO units6) to form the ligand-metal charge-transfer complex. This causes a sudden increase in the intensity of the 220 and 320 nm peaks and the simultaneous formation of Au nucleating centers.15 The nucleating centers in turn produce the absorbance due to SPR at 545 nm. The absorbance at 730 nm originates from the self-aggregation of independent large colloidal NPs produced by the growing nucleating centers that reaches their limiting value after 40 min. An instant fall in the intensity of the 320 nm peak after 40 min indicates the completion of the reaction. Figure 2c simultaneously monitors the variation in the wavelength of the 545 and 730 nm peaks. Peak 730 nm shows a huge red shift of 100 nm in comparison to just 10 nm for the 545 nm peak and can be attributed to the significant self-aggregation of the NPs in the bulk phase.7 It is very pronounced within 40 min of the reaction when a dramatic increase in the number density of tiny nucleating centers takes place as demonstrated in Figure 2b. Below the cp (at 40 °C), a similar situation exists (Figure 2d) and clearly shows how two Au NP absorbances emerge from a single absorbance. In the beginning of the reaction, only a single absorbance appears at 540 nm, which becomes increasingly prominent and red shifts to 600 nm within 15 min. It then (15) Chen, S.; Guo, C.; Hu, G. H.; Wang, J.; Ma, J.-h.; Liang, X. F.; Zheng, L.; Liu, H.-z. Langmuir 2006, 22, 9704.

11366 DOI: 10.1021/la100734p

Khullar et al.

bifurcates into two broad absorbances. The first prominent one shows a blue shift, and the second, much broader one red shifts to even beyond 700 nm. Their variation is illustrated in Figure 2e, where one can find a fine correlation between the increase in the intensity and the shift in their respective wavelengths. In both cases, a marked increase in the intensity and a simultaneous shift in the wavelength occur within 40 min of the reaction as observed at 55 °C. This means that the reaction at 40 °C (Figure 2d) simply speeds up at 55 °C (Figure 2a) but is precisely controlled by the micellar assemblies. TEM studies help us to understand how micellar assemblies below and above the cp control the reactions in both places. Figure 3a,b shows the TEM images of the reaction products carried out at 55 °C. Figure 3a shows several spherical vesicular assemblies14 of 41 ( 0.9 nm with some scattered NPs, and Figure 3b shows several groups of large fused NPs of 15.2 ( 7.5 nm (size distribution histogram, Figure S3) mostly in conjunction with these assemblies. Because no staining agent has been used to detect low-contrast vesicular assemblies in Figure 3a, thus their clear image is considered to be due to only the electron diffraction from the adsorbed tiny nucleating centers/NPs in the PEO cavities. EDS analysis confirms the presence of elemental gold (Figure S4). These nucleating centers are in fact responsible for the formation of the large NPs shown in Figure 3b and hence clearly authenticate the presence of two prominent absorbances in Figure 2a. However, the sample prepared at 40 °C shows only large, clear, dark micelles of 103 ( 23 nm (Figure 3c), where micelles rather than vesicles are expected because the cmt of 5 mM P103 is far less than 40 °C2 and 40 °C is less than cp = 52 °C. A close-up image of one of the micelles (Figure 3d) shows that it is loaded with tiny NPs. The NPs have completely covered the entire micellar surface and have provided a dark contrast. Even budding of the fresh, large NPs (indicated by a white arrow) from the surface of the micelle due to the nucleation among the tiny nucleating centers is also evident and ultimately produces several interconnected monodisperse NPs of 10.0 ( 3.5 nm (size distribution histogram, Figure S5) as shown in Figure 3e. These findings are very much in line with the UV-visible results presented in Figure 2d and explain the bifurcation of the initial absorbance very well. A blue shift of 40 nm from 600 to 560 nm in the absorbance of the first main peak suggests that the synthesis of well-defined independent NPs (Figure 3e) starts when micelles are completely covered with nucleating centers within 15 min of the reaction (or when the bifurcation takes place in the main absorbance, Figure 2d). The independent NPs, when separate from the micelles, generate their own SPR, which causes a blue shift in their absorbance. This mechanism can be better understood by increasing the amount of HAuCl4 from 0.5 to 2 mM with 10 mM P103 in a similar reaction. Figure 3f shows an SEM image of large micelles of 611 ( 155 nm loaded with much larger NPs of 61.4 ( 17.1 nm (size distribution histogram, Figure S6). They are predominantly polyhedral geometries and are randomly arranged on the surface of each micelle (Figure 3g) with EDS emission due to only elemental gold (Figure 3h). The proposed mechanism is further authenticated in the aqueous bulk phase by precisely varying the reaction temperature from room temperature to well above the cp (i.e., 25-70 °C). UV-visible scans with respect to temperature are shown in Figure 2f. The trend is very similar to the combined trends of Figure 2a,d in the sense that below the cp (52 °C) Figure 2f follows the trend of Figure 2d and above the cp it follows that of Figure 2a. A plot of intensity or wavelength versus temperature (Figure 2g) allows us to determine the influence of cp on the synthesis of Au NPs. Figure 2g demonstrates that the intensity of Langmuir 2010, 26(13), 11363–11371

Khullar et al.

Article

Figure 2. (a) Absorbance vs wavelength scans of the HAuCl4 (0.5 mM) þ P103 (5 mM) þ water ternary reaction carried out at 55 °C with respect to time from 2 to 345 min. Peaks at 545 and 730 nm show red shifts of 10 and 100 nm, respectively. (b) Plots of normalized intensity (at 220, 320, and 545 nm, and 730 nm) vs reaction time. (c) Plots of the normalized wavelength of 545 and 730 nm peaks vs reaction time. (d) Absorbance versus wavelength scans of the HAuCl4 (0.5 mM) þ P103 (5 mM) þ water ternary reaction carried out at 40 °C with respect to time from 2 to 345 min. Note the bifurcation in the initial broad absorbance into two broad absorbances after 15 min of reaction. The left peak blue shifts and the right peak red shifts with the passage of time. (e) Plots of intensity (at final 560 and 730 nm peaks in d, circles) and wavelengths of both peaks (diamonds) vs reaction time. Empty symbols represent the 560 nm peak (left peak), and filled symbols represent the 730 nm (right peak). (f) Absorbance vs wavelength scans of the HAuCl4 (0.5 mM) þ P103 (5 mM) þ water ternary reaction with respect to reaction temperature from 25 to 70 °C. Again note a bifurcation in the initial broad absorbance of the reaction into two broad absorbances (left and right of the main absorbance) as temperature exceeds 30 °C. (g) Plots of intensity (at final 560 and 730 nm peaks in f, circles) and wavelengths of both peaks (diamonds) vs reaction temperature. Empty symbols represent the 560 nm peak (left peak), and filled symbols represent the 730 nm peak (right peak). The three parts of this Figure are indicated as I, II, and III. See the details in the text.

the main peak increases instantaneously as the temperature increases from 25 °C and then splits into two around 30 °C. The intensity of bifurcated peaks remains essentially constant within 30-52 °C (cp) and then increases thereafter. Likewise, the wavelength of the respective peaks also remains predominantly constant within 30-52 °C and then shows a red shift in both cases Langmuir 2010, 26(13), 11363–11371

before achieving a constant value. On the basis of the variation in both properties, Figure 2g can be explained as a phase diagram with three different regions. Region I represents a temperature range of 25-30 °C where micelles are already present in the solution because the cmt of 5 mM P103 is well below 25 °C.2 Region II extends from 30-52 °C and represents the clouding DOI: 10.1021/la100734p

11367

Article

Khullar et al.

Figure 3. (a) TEM micrograph of P103 vesicles loaded with tiny Au NPs synthesized in a HAuCl4 (0.5 mM) þ P103 (5 mM) þ water ternary reaction with precise temperature control at 55 ( 0.1 °C. Distinct large NPs are also visible. (b) Groups of fused large polyhedral NPs in conjunction with the vesicles of P103 also prepared in the same reaction mentioned in part a. (c) TEM micrograph of P103 micelles loaded with tiny Au NPs synthesized in a HAuCl4 (0.5 mM) þ P103 (5 mM) þ water ternary reaction with precise temperature control at 40 ( 0.1 °C. (d) Single micelle completely covered by tiny NPs. The white arrow indicates a few budding NPs from the surface of the micelle. (e) Groups of several fused NPs also produced in the same reaction mentioned in part c. (f) SEM micrograph of much larger P103 micelles loaded with well-defined, distinct Au NPs synthesized in a HAuCl4 (2 mM) þ P103 (10 mM) þ water ternary reaction with precise temperature control at 40 ( 0.1 °C. (g) High-resolution SEM image showing a single micelle loaded with several polyhedral NPs and (h) its EDS spectrum. 11368 DOI: 10.1021/la100734p

Langmuir 2010, 26(13), 11363–11371

Khullar et al.

Article

Figure 4. (a) Plots of intensity at 560 nm versus temperature for the HAuCl4 þ P103 (5 mM) þ water ternary reaction with different amounts of HAuCl4 and (b) for the HAuCl4 (0.5 mM) þ P103 þ water ternary reaction with different amounts of P103. (c) Plots of intensity at 545 nm vs temperature for the HAuCl4 þ F68 (5 mM) þ water ternary reaction with different amounts of HAuCl4 and (d) for the HAuCl4 (0.5 mM) þ F68 þ water ternary reaction with different amounts of F68.

phenomenon. Region III represents the presence of vesicles and their aggregates. Regions I and III show an increase in the intensity of both peaks as well as a red shift in their absorbance, but both properties remain constant in region II. As the temperature increases, micelles and vesicles present at around 25 and 52 °C in regions I and III, respectively, tend to merge to produce larger respective assemblies with greater surface area. That in turns enhances the reduction process and thereby increases the intensity. Subsequent occupation of the surface cavities or confinement of the tiny NPs causes a red shift in the absorbance. But this is not so in region II, where micelles start getting dehydrated at around 30 °C and achieve a predominantly hydrophobic environment. This pushes the hydrophile-lipophile balance (HLB) to a lower value, thus inducing phase separation. This process continues until the lowest HLB value is attained at the cp while keeping the overall concentration of the micellar phase constant in aqueous bulk phase.16 Thus, the overall surface area needed to carry out the surface reduction of gold ions into nucleating centers also remains constant with effectively no formation of fresh nucleating centers in region II, and hence the intensity and the wavelength of the respective peaks remain constant. Relative Comparison on the Basis of Micellar Properties. The effect of cp on the synthesis of Au NPs has been further evaluated on the basis of the systematic variation in the HAuCl4 and TBP concentrations. A constant 5 mM P103 and varying amounts of HAuCl4 keep the cp region constant between 30 and 52 °C with cp = 52 °C (Figure 4a). However, a constant 0.5 mM HAuCl4 and decreasing amounts of P103 shift the cloud-point region to a higher temperature range (Figure 4b). The cp’s of 5, 2, (16) (a) Materna, K.; Milosz, I.; Miesiac, I.; Cote, G.; Szymanowski, J. Environ. Sci. Technol. 2001, 35, 2341. (b) Pennell, K. D.; Adinolfi, A. M.; Abriola, L. M.; Diallo, M. S. Environ. Sci. Technol. 1997, 31, 1382.

Langmuir 2010, 26(13), 11363–11371

and 1 mM P103 are 52, 59, and 78 °C, respectively (Supporting Information, Table S1), and are accurately reproduced in Figure 4b. In other words, as long as the concentration of P103 is constant, the varying amount of HAuCl4 is fully dependent on the cp region (Figure 4a) and hence the Au NPs synthesis is completely controlled by the micellar assemblies of P103 in all three regions of a phase diagram similar to that of Figure 2g. Likewise, a systematic decrease in the amount of P103 regularly shifts the cp to a higher temperature range (Figure 4b) irrespective of the amount of HAuCl4 used. However, a similar dependence of the synthesis of Au NPs on the micellar assemblies of F68 can also be observed even if there is no cp region within the comparable temperature range. A constant F68 concentration of 5 mM keeps the maximum at around 40 °C with varying amounts of HAuCl4 (Figure 4c) because at 40 °C the total number of micelles is completely saturated with the nucleating centers and the surface reduction of gold ions into atoms is complete. However, as the number of micelles is increased with 10 mM F68 (Figure 4d), the maximum shifts to a higher temperature range irrespective of the amount of HAuCl4 present. This further confirms the previous conclusion that it is the micelles or micellar assemblies that control the reduction of gold ions into nucleating centers.

Discussion The results clearly show a significant difference from the mechanistic point of view in the synthesis of Au NPs when predominantly hydrophilic F68 and hydrophobic P103 are used. It is the surface cavities formed by the PEO-PPO-PEO groups that are mainly responsible for the reduction of gold ions.6 However, the surface cavities17 are formed only when TBP monomers (17) Mu~noz-Bonilla, A.; Ibarboure, E.; Papon, E.; Rodriguez-Hernandez, J. Langmuir 2009, 25, 6493.

DOI: 10.1021/la100734p

11369

Article

Khullar et al.

Scheme 1. Demonstration of the Overall Redox Process Taking Place in the Surface Cavities at the Micelle-Solution Interface of TBP Micellesa

a

See the details in the text.

aggregate in the form of either preaggregates or well-defined micelles or vesicles. Here, the amphiphilic property of a particular TBP is considered to be related to proper surface cavity formation. The micelle of a predominantly hydrophilic F68 will produce a concentric corona with a greater radius lined with a greater number of surface cavities (due to a greater number of PEO units) rather than a predominantly hydrophobic P103.18 However, a micelle from a larger corona such as that from F68 is more hydrated than that from P103 and will experience excessive hydration of its surface cavities. It will lead to a poor reduction process, which is why micelles of F68 (Figure 1c) are loaded with relatively fewer NPs than are micelles of P103 (Figure 3d) at 40 °C. A more hydrophobic environment caused by a greater number of PPO units will result in compact micelles19 with a relatively less hydrated corona and a compact arrangement of surface cavities. However, the latter arrangement can also be achieved at a relatively higher concentration (Figure 1i) and temperature (Figure 1j-l) even for the hydrated micelles of a predominantly hydrophilic F68. Both factors will produce less hydrated and more compact micelles with a compact micellesolution interfacial arrangement19 of surface cavities. The following mechanism will help us to understand these aspects in terms of a redox process. A redox reaction that involves the oxidation of TBP and the reduction of AuCl4- ions initiates the synthesis of Au nucleating centers in the PEO-PPO-PEO surface cavities (Scheme 1). The reduction of Au(III) (Scheme 1a) to Au(0) (Scheme 1d) proceeds through the formation of either Au(II) (Scheme 1b) or Au(I) (Scheme 1c) species.10 Au(II) (Scheme 1a) is usually unstable; therefore, it is converted to Au(I)20 (Scheme 1c), and the Au(I) species thus produced are involved in the LMCT to produce (18) Yang, L.; Alexandridis, P. Langmuir 2000, 16, 4819. (19) Liu, Y.; Chen, S. H.; Huang, J. S. Macromolecules 1998, 31, 2236. (20) (a) Herring, F. G.; Hwang, G.; Lee, K. C.; Mistry, F.; Phillips, P. S.; Willner, H.; Aubke, F. J. Am. Chem. Soc. 1992, 114, 1271. (b) Elder, S. H.; Lucier, G. M.; Hollander, F. J.; Bartlett, N. J. Am. Chem. Soc. 1997, 119, 1020.

11370 DOI: 10.1021/la100734p

Au(0) (Scheme 1d) when the reducing power of the reducing agent is sufficiently maintained. Thus, in the multistep proposed redox process, the formation of the LMCT seems to be the ratedetermining step. In those reactions, where systematic LMCT formation is observed, they result in the formation of fine Au NPs.21 This is all related to the amount of reducing agent (TBP) and its easily accessibility to the oxidizing agent (AuCl4- ions) because this reaction is a site-specific redox reaction and takes place at the micelle-solution interface. A strongly hydrated corona of F68 is expected to decrease the rate of the reduction because AuCl4- ions cannot effectively reach the surface cavities. Thus, an energetically lower transition state (i.e., LMCT) cannot be achieved (Figure 1b, 320 nm peak with a constant value of intensity after 40 min, Au(I) is not fully converted into Au(0)) and hence results in a poor yield of Au NPs (Figure 1c). However, because the temperature induces dehydration, it not only leads to LMCT formation (Figure 1n, with the intensity of the peak at 320 nm running through a maximum) but also produces finefaceted Au NPs (Figure 1j-l). This happens only when micellar surfaces19 are lined with the proper surface cavities in well-defined micelles. An exceptionally high temperature of 80 °C dehydrates the micelles to such an extent that they merge with each other to produce compound or inverted micelles with entrapped pools of the aqueous phase (Scheme 1g) as reaction vessels. Thus, surface cavities lining these pools will then undergo the redox reaction to produce Au NPs. Once this is done, the reverse solubilization of a compound or inverted micelles into the aqueous bulk phase with decreasing temperature is not achieved because now the HLB of the vesicular assembly loaded with NPs further falls below its equilibrium value. However, P103 produces well-defined micelles with relatively less hydrated surface cavities than those of F68; therefore, there is clear LMCT formation (Figure 2a). The LMCT is completely (21) Vogler, A.; Quett, C.; Kunkely, H. Ber. Bunsen-Ges. Phys. Chem. 1988, 92, 1486.

Langmuir 2010, 26(13), 11363–11371

Khullar et al.

converted to Au(0) as evident from the dramatic decrease in the peak intensity of the 320 nm peak to a minimum value after 40 min (Figure 2b), resulting in complete coverage of the micelles with tiny Au NPs (Figure 3d). Further nucleation among the growing nucleating centers produces well-defined NPs (Figure 3e) and is even accelerated by the increase in the micellar surface area (with 10 mM P103) and a higher concentration of HAuCl4, 2 mM, thus resulting in the formation of much larger micelles completely covered with well-defined large NPs (Figure 3f,g).

Concluding Remarks A collective review of all of the above results clearly suggests that the reduction is carried out by the surface cavities produced by the compact arrangement of TBP monomers in the corona layer of TBP micelles. Because the redox reaction is a site-specific reaction and takes place only at the micelle-solution interface, the extent of hydration of the surface cavities is the rate-determining step. Greater hydration screens the approach of gold ions and hence reduces the nucleation process, but the same reaction is

Langmuir 2010, 26(13), 11363–11371

Article

accelerated when an appropriate dehydration is achieved. Dehydration not only allows the greater compactness among the surface cavities but also brings the growing nucleating centers close enough to produce large NPs. The large NPs obviously cannot be supported by the soft micelles and hence find their way into the bulk phase. If the micelles are large enough, especially when a higher concentration of a TBP is used, then they tend to carry them. Therefore, the overall shape and structure of a TBP micelle is the central issue in the successful redox process and are strongly related to the temperature variation. Acknowledgment. These studies were partially supported by financial assistance from UGC (ref no. 34-323/2008(SR)) and CSIR (ref no. 01(2102)/07/EMR-II), New Delhi, India. Supporting Information Available: Size distribution histograms, UV-visible spectra, and EDS analysis. This material is available free of charge via the Internet at http://pubs. acs.org.

DOI: 10.1021/la100734p

11371