Chapter Interaction of osmium clusters with metallic nanoparticles and surfaces One aspect of nanoparticles (NPs) that is different from bulk materials is the percentage of atoms that are interfacial For nanoparticles smaller than nm, the majority of atoms are located at the interface between the particle and the surrounding environment, that is, at the surface For this reason, there is a strong resemblance between nanoparticles and self-assembled monolayers (SAMs) Since SAMs form by self-assembly, it does not matter what the size of the particles are, i.e., chemistry controls the process The structure of SAMs, however, differs greatly depending on the curvature of a given surface SAMs on nanoparticles simultaneously stabilize the reactive surface of the particle and present organic functional groups at the particle-solvent interface In particular, monolayer-protected clusters (MPCs) of gold and silver show good stability, tunable solubility and relative ease of characterization; capping reagents for such MPCs have included alkanethiolates, unsaturated carboxylates, amines and isocyanates The results in the previous chapter show that both the support and the functional groups on the triosmium clusters will affect how they interact Furthermore, such interactions are difficult to study We have thus been interested in the preparation of nanoparticles with the appropriate substrates to serve as molecular models Ideally, such a model molecular system should be monodispersed and have suitable spectroscopic handles Some related examples include that of metallic nanoparticles functionalized with borane clusters, transition metal carbonyls such as Cr(CO)4(L) 72 and Re(CO)3(L)Br (L = 2,3-bis(2'-pyridyl)pyrazine), as well as the recent synthesis of silver nanoparticles stabilized by the water soluble organometallic surfactant [Os3 (CO)10(μ-H){μ-S(CH2)10COO}]Na In this chapter, a set of osmium clusters were deposited onto gold or silver substrates, or used as stabilizing agents for gold or silver NPs We hope that a comparison can be made to examine if the NPs protected by osmium clusters can be used as models for the supported osmium clusters on gold or silver surfaces 3.1 Anchoring of osmium clusters onto gold or silver The preparation of gold or silver compounds containing triosmium clusters with heterometallic bonds have been reported by many groups The gold starting materials used are gold phosphine cationic units and they tend to attach to the osmium core via a μ3 or μ2 bonding mode When several gold phosphine cationic units are coordinated to the osmium framework, Au-Au bonds may be formed, depending on the distance between the gold phosphine cationic units It is obvious that the gold atom shows good affinity to the osmium core in all these osmium-gold clusters Osmium-silver clusters also show a similar bonding mode In this section, the deposition of the osmium clusters Os3(CO)10(μ-H)(μ-OH), 1, Os3(CO)10(μ-H)2, 22 and Os3(CO)10(μ-H)(μ-SCOPh), 33 onto a gold or silver substrate are compared with their deposition onto Au or Ag NPs 73 3.1.1 Interaction of osmium clusters with gold or silver nanoparticles Silver nanoparticles In order to avoid interference by surfactants, Ag NPs were prepared by the photolysis of AgNO3 in a toluene solution of osmium clusters without any other surfactants The modified Ag NPs can be redispersed in ethanol, water and dichloromethane The surface plasmon (SP) band energy of NPs is sensitive to the electronic and optical properties of the particles’ surfaces and of the protecting monolayers Uncapped silver colloids in water exhibit an absorption maximum at 390 nm, and Ag NPs protected by different surfactants show red shifts which depend on the functional group of the surfactant used This red shift increases with the particle size, and the geometrical shape plays a major role in determining the plasmon resonance The UV- vis absorption spectra of 1, 22, 33 and their corresponding modified Ag NPs in toluene are shown in Figure 3.1 The absorption band of in toluene shows a peak at 308 nm, and another broad peak at 356 - 445 nm, respectively In contrast, the UV absorption band of the 1-modified Ag NPs shows a very broad hump at around 333 nm to 700 nm, with a maximum at 463 nm This red shift of the absorption band of 1-modified Ag NPs can be attributed to the adsorption of osmium clusters It is possible that the broad UV absorbance peak of the Ag NPs was due to a large size or shape distribution of the Ag NPs formed However, the peak corresponding to cannot be detected, possibly due to the small amount of osmium clusters on the Ag NPs A control experiment in which finely ground AgNO3 in toluene was irradiated 74 for 24 h with a tungsten lamp did not show any colour change nor an absorption peak in the UV-vis spectrum, indicating that no silver particles were formed in the absence of the osmium clusters Thus, induced the formation of Ag NPs The formation of the silver colloids may be attributed to the visible light induced photooxidation of by excited Ag+, resulting the formation of Ag atoms (Ag0) Subsequent agglomeration Normalized absorbance of Ag0 produced silver colloids b a c e d f 300 400 500 600 700 Wavelength (nm) Figure 3.1 Absorption spectra of (a) 1-modified Ag NPs, (b) 22-modified Ag NPs, (c) 33-modified Ag NPs and (d-f) the corresponding precursor clusters Similar results were observed for clusters 22 and 33 except that for the former, the UV-vis spectrum exhibited a smooth curve with a maximum at 466 nm, while the latter showed a broad peak centered at 523 nm This indicated that the size of the former Ag NPs were smaller and more uniform than the 1-modified Ag NPs, while the latter are the largest Thus, it is clear that different osmium clusters lead to different 75 sizes of Ag NPs, although the factors are unclear Figure 3.2 shows TEM images of the modified Ag NPs Small spherical particles (1.1 ± 0.3 nm) together with some big particles (12.1 nm ± 2.6 nm) and some irregularly shaped NPs were observed in the TEM image of 1-modified Ag NPs (Figure 3.2a) Closer inspection of some of the TEM images of the larger particles provide visual evidence for the adsorption of cluster The inset of Figure 3.2a shows elongated Ag NPs with a layer of about nm thickness surrounding them Since surfactants were not used in the preparation of the NPs, the presence of these shells is proposed to be a multilayer of clusters adsorbed on the silver NPs, as the triosmium cluster has an approximate width of less than nm EDX results from both small and big particles show the presence of Ag (characteristic peaks at keV) and Os (characteristic peaks at 2, 8, 9, 10 and 12 keV), thus confirming the association of with the Ag particles (Figure 3.3) a b c 20 nm 20 nm 50 nm Figure 3.2 TEM images of (a) 1-modified Ag NPs (Inset: Enlarged image of elongated Ag NPs.), (b) 22-modified Ag NPs (Inset: Enlarged image of an irregularly shaped Ag NPs.), (c) 33-modified Ag NPs 76 Figure 3.3 EDX for Ag NPs protected by Similarly, the TEM image of 22-modified Ag NPs (Figure 3.2b) shows small, uniformly sized, spherical particles (1.4 ± 0.4 nm) together with some irregularly shaped NPs The inset shows an irregularly shaped Ag NP with a layer of about nm thickness surrounding it, suggesting again the presence of a multilayer of clusters adsorbed on the Ag NPs The TEM image of 33-modified Ag NPs (Figure 3.2c) shows that spherical particles with average size about 10.0 nm (± 2.8 nm), together with some bigger particles (~20-50 nm), were obtained EDX analyses of both modified Ag NPs also show the presence of osmium and silver, and for 33, sulfur as well Thus the 33-modified Ag NPs are bigger than those modified with or 22, and the cluster used influences the size of Ag NPs formed It is possible that the association of Ag NPs with 22 is faster than with or 33, thus accounting for the smaller size observed with 22-modified Ag NPs Presumably, the rate association of Ag NPs with clusters is generally very fast, hence accounting for the small size of most of the Ag NPs obtained As indicated in the earlier chapter, ToF-SIMS appears to be a good characterization 77 technique for clusters on surfaces When Os3(CO)12, 3(Os) or Ru3(CO)12, 3(Ru) was coated onto a gold or silver substrate, the molecular ion and its attendant fragmentation pattern was observed in the ToF-SIMS spectrum However, if the substrate was washed with dichloromethane, no signal attributable to the parent clusters or their fragments was detected, thus allowing us to affirm that any species detected in the ToF-SIMS spectrum of cluster-modified Ag NPs were not surface anchored species The negative ion ToF-SIMS spectrum of Ag NPs protected by is shown in Figure 3.4a The most intense cluster of peaks could be attributed to the series [Os3(CO)x(μ-H)]- (x ≦ 9) and [Os3(CO)x(μ-H)(μ-O)]- (x ≦ 8) The molecular ion, or fragments, containing silver atoms were not observed The intensities were quite low due to a low concentration of osmium clusters on the NPs However, the presence of peaks that could be attributed to a triosmium moiety indicated that was anchored onto the Ag NPs successfully For the 22-modified Ag NPs, other than the two series of ions [Os3(CO)x(μ-H)]- (x ≦10) and [Os3(CO)x(μ-H)(μ-O)]- (x ≦ 9), the fragment [Os3(CO)10Ag]- was also detected (Figure 3.4b) This showed that the osmium clusters were possibly bonded to the Ag NPs 78 400 -CO -CO -CO -CO -CO a -CO -CO -CO -CO -CO Intensity 757 713 741 685 673 701 200 729 [Os (CO) (μ-O)]- 771 [Os3(CO)9(μ-H)]785 813 796 825 618 700 800 900 Mass (m/z) 3500 3000 -CO b 785 - Intensity 2500 [Os3(CO)8(μ-OH)] 813 2000 960 1500 829 1000 847 [Os3(CO)10Ag]979 1020 920 111 500 800 1000 1170 1441 1200 Mass (m/z) Figure 3.4 ToF-SIMS spectra (negative ion mode) of Ag NPs modified with (a) and 22 (b) The appearance of the ToF-SIMS spectrum of 33-modified Ag NPs is very different (Figure 3.5) Peaks corresponding to triosmium clusters were not observed Instead, intense peaks at m/z = 192, 137, 105, 77 and 28, assignable to Os, SCOPh, COPh, Ph and CO, respectively, were observed An intense peak assignable to Ag was also 79 observed Careful examination of the spectrum showed that the highest mass fragment could be found at m/z = 616, which may be assignable to a diosmium moiety Os2(CO)6S2H2, on the basis of the isotopic pattern This may be due to a low level of clusters on the modified Ag NPs; other evidence (see later) suggest that 33 was Intensity deposited as a trinuclear species thus ruling out decomposition on the surface Intensity Ag- Os SCOPh 137 188 192 Mass (m/z) COPh - 100 110 120 130 Mass (m/z) Figure 3.5 ToF-SIMS spectrum (negative ion mode) of Ag NPs modified with 33 Inset shows the cluster of ion at m/z 192 Figure 3.6 shows the IR spectra in the carbonyl region for the osmium clusters and those of the modified Ag NPs All the modified Ag NPs are characterized by vibrations at about 2125(w), 2032(s) and 1951(s) cm-1 The essentially identical patterns and band maxima indicate that the osmium clusters are likely to be bound to the surfaces in the same fashion The pattern is similar to that of the osmium species 80 Os3(CO)10(μ-H)(μ-Zn) on the ZnO surface discussed in the previous chapter (Figure 2.11), suggesting significant interaction between the osmium cluster core and the Ag NPs a b Transmittance c d e f 2200 2150 2100 2050 2000 1950 1900 1850 Wavenumber (cm-1) Figure 3.6 Solid-state IR spectra of (a) 1, (b) 1-modified Ag NPs, (c) 22, (d) 22-modified Ag NPs, (e) 33, and (f) 33-modified Ag NPs Gold nanoparticles Au NPs were prepared by laser ablation in ethanol; 10 no other surfactants were added The concentration of the Au NPs was estimated to be x 10-5 M (based on the number of gold atoms), by comparing the intensity of the plasmon band to the intensity of the band from NPs produced using chemical reduction methods 11 The choice of ethanol over water as a solvent was to enable dissolution of the osmium clusters, hence facilitating their direct adsorption onto the Au NPs 81 ether The resultant nanoparticles remained spherical, with sizes of about nm for the Ag+ complex, and about nm for Na+ and K+ complexes Figure 3.32 TEM and HR-TEM micrographs of dried Ag NPs capped with 36 The UV-vis spectrum (toluene solution) of the Ag NPs capped with 36 displayed a broad and unsymmetrical surface plasmon band at 450 nm (Figure 3.33); this is red shifted compared to unprotected silver particles (390 nm) In contrast, the surface plasmon band for that complexed with Ag+ was blue shifted to 411 nm (Figure 3.32d) No emission band was observed when a toluene solution of the uncomplexed NPs was excited at 450 nm (Figure 3.34a) In contrast, visible luminescence at 488 nm was observed for that complexed with Ag+ (Figure 3.34b); similar emission was also observed for those complexed with Na+ and K+ The observations here demonstrate clearly that the emissions are associated with aggregation of Ag NPs, but these aggregates are presumably fairly dynamic in nature as they not show up on the TEM images; this is in contrast to that reported for the gold analogues.37 113 Normalized Absorbance d a c b 400 500 600 700 Wavelength (nm) Figure 3.33 UV–vis spectra of silver nanoparticles capped with 36: (a) in toluene, (b) after reflux with aqueous sodium nitrate, (c) after reflux with aqueous potassium PL nitrate, and (d) after reflux with aqueous silver nitrate b a 480 510 540 Wavelength (nm) Figure 3.34 Emission spectra of Ag NPs capped with 36: (a) in toluene, and (b) after reflux with aqueous silver nitrate The excitation wavelength was 450 nm Interestingly, photolysis of 37 with AgNO3 in toluene afforded small NPs (~2 nm) 114 and some aggregates (Figure 3.35), which showed a surface plasmon band at 573 nm but no fluorescence (Figure 3.36) The IR spectrum of 37-modified Ag NPs suggested that 37 is bonded to the Ag NPs via a μ-Ag mode (the vibrational bands of vCO at 2105w, 2026s, 1943s) It would therefore appear that the Ag NPs are stabilized through some sort of direct interaction with the triosmium cluster Figure 3.35 TEM micrographs of Ag NPs capped with 37 b c 400 600 a 800 Wavelength (nm) Figure 3.36 Absorption spectra of (a) 37, (b) 37-modified Ag NPs and (c) emission spectrum of 37-modified Ag NPs 115 3.3 Conclusion Our study has shown that the structure of the osmium cluster species deposited onto silver or gold, be in nanoparticles or substrates, are mostly the same However, their mode of interaction with the surface depends on the nature of linking groups present For clusters without intervening linkers, the interaction is via a μ-Ag/Au mode, i.e direct interaction with the osmium core For clusters linked to a functionalized group, such as –SH, -OH and –COOH, the various interactions depicted below (Scheme 3.8) may be present, depending on the nature of the functional group and/or the linkers Os Os Os Os Os Os E Os E linker E' E linker E' H H Os Os linker S (CH2)2 Os Os Os E' O OH Ag E = S, O E' = S, O E=S E' = COO Ag E = O, S E' = S, O, COO Scheme 3.8 3.4 Experimental 3.4.1 Synthesis of osmium clusters 3.4.1.1 General procedure The general procedure for the synthesis and characterization of osmium clusters are the same as that described in Chapter Os3(CO)10(µ-H)(µ-SCOPh), 33 was synthesized according to a literature method 42 116 3.4.1.2 Synthesis of Os3Au2(CO)9(µ3-S)(PPh3)2, 34 The cluster Os3(CO)10(µ-H)(µ-S ͡ SH) (0.035 mmol) was stirred with NaOMe (0.035 mmol, prepared by adding 0.71 g Na into 50 ml methanol) and Ph3PAuCl (17.3 mg, 0.035 mmol) in dichloromethane (30 ml) at room temperature overnight The colour turned from yellow to red The solvent was removed under reduced pressure and the product was purified by TLC to yield 34 as a red solid H NMR (CDCl3): δ 7.56-7.46 (m, 30H, Ph) 31 P NMR (CDCl3): δ 55 FAB-MS: m/z 1774 (M+), calcd: 1773 νCO (hexane): 2092m, 2037s, 2009s, 1985w cm-1 Elemental analysis: Calcd for C45H30Au2O9Os3P2S: C, 30.45; H, 1.69; S, 1.80 Found: C, 30.31; H, 2.38; S, 1.86 Table 3.2 Quantity of osmium clusters used and the yield of 34 from different starting materials Starting material 14a 15a 16a Amount used (mg) 15.5 20.2 20.6 Ph3PAuCl (mg) 9.9 10.0 9.8 Yield (%) 34 (21 mg) 35 (21 mg) 29 (18 mg) 117 3.4.1.3 Synthesis of Os3(μ-H)(CO)10[(6-mercaptohexyloxy)methyl-15-crown-5], 37 Synthesis of (6-bromohexyloxy)methyl-15-crown-5 (35) A suspension of 2-(hydroxymethyl)-15-crown-5 (0.41 g, 1.64 mmol) and NaOH (77 mg, 1.93 mmol) in THF was stirred at room temperature After 30 min, 1, 6-dibromohexane (1.27 ml, 8.50 mmol) was added, and the mixture was stirred overnight The reaction was then quenched with methanol, the solvent removed by rotary evaporation, and the residue was taken up in CH2Cl2 (100 ml) and washed several times with water After drying over Na2SO4, the solvent was evaporated and the residue purified by column chromatography on silica, using a solvent gradient from hexane/EtOAc (1:1, v/v) to EtOAC, to yield a pale yellow oil Yield: 0.38 g (56 %) H NMR (CDCl3): δ 1.30-1.65 (m, 6H), 178-1.93 (m, 2H), 3.30-3.45 (m, 7H), 3.52-3.78 (m, 18 H) Synthesis of (6-mercaptohexyloxy)methyl-15-crown-5 (36) To a stirred solution of 35 (400 mg, 0.97 mmol) in ethanol (40 ml) was added thiourea (405 mg, 5.32 mmol) The reaction mixture was heated under reflux for h, and the solvent was then removed under reduced pressure The resulting solid residue was mixed with KOH (497 mg, 8.72 mmol) and deionized water (40 ml), and then refluxed for h This was then acidified with M aq HCl and then extracted with dichloromethane The organic layer was separated, dried over MgSO4, and the solvent was then distilled off under reduced pressure The residue was purified by TLC on 118 silica gel plates (hexane/EtOAc, 3:1, v/v, as eluant) to yield 36 as a pale yellow oil Yield: 200 mg (56 %) H NMR (CDCl3): δ 1.27-1.45 (m, 4H), 1.50-1.70 (m, 4H), 2.50 (q, 2H), 3.32-3.45 (m, 4H), 3.50-3.60 (m, 19H) FAB-MS: m/z 366.51 (M+), calcd: 366 Elemental analysis: Calcd for C17H34O6S: C, 55.71; H, 9.35; S, 8.75 Found: C, 55.45; H, 9.39; S, 9.21 % Synthesis of Os3(μ-H)(CO)10[(6-mercaptohexyloxy)methyl-15-crown-5] (37) The cluster Os3(CO)10(NCCH3)2 (200 mg, 0.21 mmol) was stirred with 35 (48 mg, 0.13 mmol) in dichloromethane (30 ml) at room temperature for 12 h The solvent was removed under reduced pressure and the product was purified by TLC to yield 37 as a yellow oil Yield: 140 mg (57 %) νCO (DCM): 2108 w, 2066s, 2057w, 2020s, 1996m cm-1 νCH2: 2922m, 2875m cm-1 H NMR (CDCl3): δ 1.25-1.47 (m, 4H), 1.50-1.68 (m, 4H), 2.32 (t, 2H, CH2S), 3.35-3.44 (m, 4H), 3.50-3.65 (m, 19H), -17.41 (s, 1H, OsHOs) FAB-MS: m/z 1218.2 (M+), calcd: 1216 Elemental analysis: Calcd for C27H34SO16Os3: C, 26.64; H, 2.82; S, 2.63 Found: C, 27.12; H, 2.90; S, 2.45 % 119 3.4.2 Characterization UV-vis spectra were recorded using a Shimadzu UV-1601 PC spectrometer TEM images were recorded on a JEOL JEM 3010 TEM at an accelerating voltage of 300 kV TEM samples were prepared by placing a drop of the NPs onto a carbon-coated Cu grid Energy dispersive X-ray (EDX) studies were made with an EDX analyzer attached to the TEM Voltage and spot size were adjusted to make the Counts per Second (CPS) in the range of about 1000 Solid and solution IR spectra were obtained using an IR Prestige-21 Fourier Transform Infrared Spectrometer (Shimadzu) Surface IR spectra were recorded using an IR Prestige-21 Fourier Transform Infrared Spectrometer (Shimadzu) equipped with AIM-8800 automatic infrared microscope Photoluminescence spectra were recorded with PerkinElmer Instruments (LS55) The ToF-SIMS spectra were recorded on an ION-TOF SIMS instrument, using bunched 69 Ga+ ion pulses with impact energy of 25 KeV 3.4.3 Synthesis of nanoparticles protected with osmium clusters 3.4.3.1 Silver nanoparticles A solution of AgNO3 (0.12 mmol) and osmium clusters (4.1 μmol) in toluene was photolysed for about h The brown precipitate obtained was collected by centrifuge and washed with toluene and water in turn to get rid of excess osmium clusters and AgNO3 120 3.4.3.2 Gold nanoparticles A Nd:YAG pulsed laser (1064 nm, 10 ns, 200 mJ/pulse) was focused onto a piece of gold foil, and placed in ethanol for while the solvent was continuously stirred The experiment was performed in air A 532 nm laser beam (10 ns, 50 mJ/pulse) from the same laser was then focused into the solution itself to induce melting and boiling of the nanoparticles to give a narrow range of nanoparticulate size Depending on the number of NPs produced, about 20 mL of a 1.7 x 10-5 M osmium clusters solution in ethanol was then added to 24 mL of the dispersed Au NPs (about x 10-5 M) in ethanol After stirring, the yellow solution was centrifuged (Thermo Eelectron Corporation) for in order to separate the NPs from the excess osmium clusters After centrifugation, a yellow solution with precipitated black solids were obtained, which was later shown to be a combination of gold nanoparticles adsorbed with osmium cluster The solid was dried and kept in the dark, or it could be dispersed in ethanol to be used for characterization studies 3.4.4 Deposition of osmium clusters onto substrates The Au surfaces on glass were cleaned in ethanol and immersed in a 3:1 mixture of H2SO4/H2O2 at 398 K for minutes to remove any organic contaminants, followed by rinsed with deionized water, then dried in air 43 Ag substrate were cleaned in ethanol and followed by reduced in H2 gas at 100oC A substrate was soaked in a dichloromethane solution (5 ml) of clusters (3 mg) for d, washed with dichloromethane, and then dried in vacuo for h The surface then was analysed by ToF-SIMS and IR 121 3.5 Crystal data for 34 Empirical formula C45 H30 Au2 O9 Os3 P2 S Formula weight 1773.22 Temperature 223(2) K Wavelength 0.71073 Å Crystal system Monoclinic Space group P21/n Unit cell dimensions a = 12.6385(6) Å α= 90° b = 27.3964(12) Å β= 108.9930(10)° c = 13.9385(6) Å γ= 90° Volume 4563.5(4) Å3 Z Density (calculated) 2.581 Mg/m3 Absorption coefficient 14.899 mm-1 F(000) 3216 Crystal size 0.15 x 0.10 x 0.06 mm3 Theta range for data collection 2.03 to 26.37° Index ranges -15