Supporting Information for: I. Materials and Methods
S1
Supporting Information for: Direct Imaging of the Ligand Monolayer on an Anion-Protected Metal Nanoparticle through Cryogenic Trapping of its Solution-State Structure Alevtina Neyman, † Louisa Meshi,‡ Leila Zeiri, † and Ira A. Weinstock*,† Department of Chemistry, Ben Gurion University of the Negev, Beer Sheva, 84105, Israel, Ilse Katz Institute for Nanoscale Science and Technology, Ben Gurion University of the Negev, Beer Sheva, 84105, Israel
Contents: I. Materials and methods II. Comments on high-magnification Cryo-TEM. III. Characterization of α-AlW11O399–-protected Ag(0) NPs, including spectroscopic data that confirm the structural integrity of α-AlW11O399-, and additional TEM and Cryo-TEM images. IV. Additional comments on Cryo-TEM images of POM-stabilized metal(0) NPs. V. Additional characterization and TEM and Cryo-TEM images of citrate-stabilized Ag(0) NPs. VI. TEM and Cryo-TEM images obtained before and after reaction of α-PW11O397– with citrateprotected Ag(0) NPs. VII. Surface-enhanced Raman (SER) spectroscopy. VIII. References
I. Materials and Methods Materials For Ag(0) nanoparticle synthesis and characterization, sodium borohydride (Fisher; general grade), trisodium citrate (Sigma; USP), sodium chloride (Frutarom; analytical grade), silver nitrate (J.T. Baker; A.C.S. reagent), deuterium oxide (99.9% D) (Bio Lab, Ltd.), and potassium bromide (EDE, spectral grade for spectroscopy) were used as received. Sodium tungstate dihydrate was purchased from Merck (extra pure). Additional reagent grade salts, acids and diethylether for polyoxometalate synthesis were obtained from commercial sources and used as received. α-K7PW11O391 and α-K9AlW11O392-4 were prepared using published methods. All water used was of high purity (18.2 MΩ resistivity) from a Millipore Direct-Q waterpurification system. Methods. pH Measurements. pH values were measured using a EuTech pH 510 Bench-Top pH meter, or a TES-1380 pH meter equipped with a Ag/AgCl electrode (Thermo Electron, UK), and temperature compensation. Prior to use the pH meter was calibrated using standard reference solutions (pH 4.00 and 7.00). Spectroscopy Analyses. UV-visible. UV-vis spectra were acquired using a Hewlett-Packard 8452A spectrophotometer equipped with a diode-array detector. For all measurements the solutions were diluted as needed to avoid over-saturation by the strongly absorbing ligand-tometal charge transfer bands of the polyoxometalates in the 190-300 nm region. Nuclear Magnetic Resonance (NMR). Al-27 NMR2,3 and 31P NMR5 spectra were acquired on a Bruker 500 MHz spectrometer (5 mm BBI and 5 mm QNP probes respectively). Chemicalshift values were externally referenced to 0.10 M AlCl3 ([Al(H2O)6]3+, or to 1.0 M
S2 H3PO4, both set to δ = 0 ppm. Internal lock signals were tuned using D2O. Spectral data were processed using the NMR software package, Mnova version 5.1. Infra-Red. Fourier-transform infra-red (FTIR) spectra were acquired using a Nicolet Impact 410 spectrophotometer (KBr pellets). Raman. The Raman instrument comprised a Jobin-Yvon LabRam HR 800 micro-Raman system, equipped with a liquid-N2-cooled detector. The excitation source was a He-Ne laser (633 nm) with a power of 5 mW on the sample. (Spectra of the POM-protected Ag(0) nanoparticle (NP) solutions obtained at 514 nm were not improved relative to 633 nm.) The laser was focused with an x50 long-focal-length objective on a spectroscopic cell containing the sample, to a spot of about 2 µm. The measurements were taken with the 600 g mm-1 grating and a confocal microscope with a 100-µm hole. The exposure time for the surface-enhanced Raman spectrum (SERS) show in Figure 4 (text) was 30 minutes. Fluorescence. Fluorescence spectra were acquired using a Jobin Yvon Fluorolog-3 spectrophotometer. No fluorescence arising from the POM-stabilized Ag(0) solutions was observed. Dynamic Light Scattering (DLS). DLS data was collected at 25 °C on an ALV-CGS-8F instrument (ALV-GmbH, Germany) with appropriately diluted samples. The CONTIN method was used to obtain hydrodynamic radii (Rh). The dependence of Rh on measurement angle was investigated from 30 to 120 degrees (in increments of 30 degrees) and appeared to be negligible—consistent with spherically shaped particles. Experimental uncertainties in total scattered intensities and in Rh were about 1 to 3%. The DLS data were obtained, primarily, to ensure that colloidal solutions of POM-protected Ag(0) NPs contained significant concentrations of particles in the size range of interest (ca. 10 to 20 nm). Transmission Electron Microscopy (TEM). Samples for Dry TEM and for high-resolution TEM (HRTEM) were prepared by pipetting 5-10 µL of the aqueous sample solution onto Cu grids covered with thin carbon-support films and dried in air. TEM data was obtained using a FEI Tecnai 12 G2 electron microscope (120kV) equipped with a Gatan slow-scan camera. HRTEM data were obtained using a JEOL JEM-2010 instrument (200kV) equipped with a Noran energy dispersive spectroscopy (EDS) system and a Gatan slow-scan camera. Cryogenic TEM. The cryogenically frozen samples were prepared using a fully automated vitrification device (“Vitrobot”). First, 5 µL of the sample solution was placed by pipette onto a glow-discharged Cu grid covered with a lacey-carbon film held inside a 100% humidity chamber. The grid was then mechanically “blotted” and immediately plunged into liquid ethane (b.p. 185K) cooled by liquid nitrogen (b.p. 77K). Data were collected on the FEI Tecnai 12 G2 instrument (120kV) and the Gatan slow-scan camera. Syntheses. α-K9AlW11O39-Protected Ag(0) Nanoparticles. α-K9AlW11O39 (the K+ salt of 1),2-4 was specifically chosen for its stability at borate-buffer pH values of between 8 and 9 (this salt is prepared at pH values of 85). Therefore, unlike α-SiW11O398- or α-PW11O397-, 1 can be used instead of citrate during BH4- reduction of Ag+ (Creighton method7). The ratio of Ag+ to α-K9AlW11O39 (1) was 1:1, and NaBH4 was added at 1.6 times molar excess relative to Ag+ (i.e., the ratio of BH4- to Ag+ was 1.6). Specifically, 0.039 g (1.25 x 10-5 mol) 1 was dissolved in 4.5 mL of pure water. Then, 0.5 mL of freshly prepared aqueous AgNO3 solution (0.025 M) was added to the solution of 1, and the clear, colorless solution was stirred vigorously at room temperature for 1 hr. After cooling a freshly prepared 0.1 M aqueous solution of NaBH4 in an ice bath, 200 µL (2.0 x 10-5 mol, 1.6 equivalents) were added in 3 equal (66.7
S3 µL) portions at 2-3 seconds intervals, with vigorous stirring. The resultant dark-brown solution was stirred for another 30 min at room temperature. The final pH values of these solutions were typically between 8 and 8.5. The solution was filtered 3 times through a 0.45 µm PVDF nonsterile Millipure syringe filters, then another 3 times through a 0.22 µm filter. The solution was characterized by UV-vis, FTIR, 27Al NMR, surface enhanced Raman spectroscopy, dynamic light scattering, transmission and high-resolution transmission electron microscopy (TEM and HRTEM), electron dispersive spectroscopy (EDS) and X-ray diffractions (both in conjunction with HRTEM) and by cryogenic-TEM (cryo-TEM). These data and images are presented in Figures 2-5 of the text, and below. The spectroscopic characterization was carried out with due attention to the integrity of 1 during the above procedure (see data and discussion of this in section III., below). Citrate-Protected Ag0 Nanoparticles (NPs). Citrate-stabilized Ag(0) NPs were prepared according to a variation of methods reported by Creighton 6 and Tong.7 Briefly, 0.0425 g (2.5 x 10-4 mol) of AgNO3 and 0.0735 g (2.5 x 10-4 mol) of tri-sodium citrate, Na3C6H5O7, were dissolved in 100 ml of pure (18.2 MΩ) water and stirred for 1 h at room temperature (between 22 and 25 °C). Meanwhile, 0.051 g (4.0 x 10-4 mol) of NaBH4 was dissolved in 4 mL of pure water, cooled in ice-bath, and added quickly with vigorous stirring. The color of the solution immediately turned dark green-brown. The mixture was then stirred for 30 min, filtered 3 times through a 0.45 µm polyvinylidene difluoride (PVDF) non-sterile Millipure syringe filter, then another 3 times through a 0.22 µm filter. Finally, 10 portions of 1.5 mL each (in 1.5 mL Eppendorf vials) were centrifuged for 6 min at 5000 rpm. The clear yellow supernatant solution was decanted and stored in the dark at room temperature for analysis by TEM and cryo-TEM. Reaction of Citrate-Protected Ag0 NPs with α-K7PW11O39. Using a variation of Tong’s method,7, the pH of 10 mL of the above citrate-protected Ag(0) NP solution was first adjusted to 5.5 with dilute H2SO4. Then, 0.075 g (2.5 x 10-5 mol) of solid α-K7PW11O39 was added. The solution was stirred at room temperature for 1 hr, then filtered and stored in the dark at room temperature for analysis by TEM and cryo-TEM.
II. Comments on High-Magnification Cryo-TEM. In cryo-TEM, the normal methods used to achieve image contrast in traditional (dry-sample) TEM studies are not applicable. For cryo-TEM, “phase contrast” is used. To enable this, the “focus” must be slightly physically below the object being imaged. A more serious limitation of the technique, relevant to the data provided here, is that high-magnification images are difficult to obtain due to the tendency of the vitreous water to change when warmed by the electron beam. The magnification typically used for cryo-TEM imaging of meso-scale and relatively large nanoscale objects is x30,000. At this relatively modest magnification, the energy of the electron beam is dispersed over a relatively large visual field. The smaller energy flux (e.g., per nm3) within the sample limits the extent to which the beam induces changes to the vitreous water, and allows more time for focusing and obtaining high-quality images of a specific object. Hence, at lower magnification, a number of images of the same object can be obtained and the “under focus” method optimized for clarity and contrast. The images shown below were obtained at x150,000, near the limit of the instrument and operator, whose experience and skill play an important role. The limitation is that optimized images must be obtained with very few attempts allowed before heating of the vitreous water in the visual field obviates further data acquisition at that location on the grid. The small collection
S4 of images provided here shows the variation in contrast and focus typical of this technique when used at high (x150,000) magnification. The polyoxometalate anions used in this work are ca. 1.2 nm in diameter. In future work, larger POMs will be used to investigate POM interactions with smaller metal(0) nanoparticles. This should provide nano-assemblies that, while based on 1 or 2 nm diameter metal(0) particles, are relatively large due to the dimensions of the associated POM clusters (1 to 4 nm for each POM cluster if one includes the larger oxomolybdate clusters planned for inclusion in future work).
III. Characterization of α-AlW11O399–-Protected Ag(0) NPs, including Spectroscopic Data that Confirm the Structural Integrity of α-AlW11O399-, and Additional TEM and Cryo-TEM images. Structural Integrity of α-K9AlW11O39 (1) as a Protecting Ligand for Ag(0). The UV-vis spectrum of the dark brown solution featured a broad absorbance at 425 nm, due to the surface plasmon resonance of the Ag(0) nanoparticles (Figure S1, A). The FTIR spectrum obtained after concentrating the solution to dryness was identical to that of pure 1 (KBr pellet; Figure S1, B). The 27Al NMR spectrum of the solution (not shown),2-4 was also the same as that of pure 1.
B
A
nm
cm-1
Figure S1. Characterization and effects of drying 1-protected Ag(0) nanoparticles. A: UV-vis spectra of aq. solutions of K91 (black line) and of 1-protected Ag(0) (red line). B: FTIR spectra of pure K91 (black line) and of the dark solid obtained by drying the 1-protected Ag(0) solution (red line).
S5
aa-86-1 (k9AlW11O39+Ag0NPs) 1.20E+00
34.683 1.00E+00
8.00E-01
6.00E-01 5.091 4.00E-01
2.00E-01
0.00E+00 0.01
0.10
1.00
10.00
100.00
1000.00
10000.00
-2.00E-01 hydrodinamic radius (nm)
Figure S2. Hydrodynamic radii (Rh) calculated from dynamic light scattering (DLS) data using the CONTIN method. The DLS data were obtained to ensure that colloidal solutions of POM-protected Ag(0) NPs contained significant concentrations of particles in the size range of interest (ca. 10 to 20 nm diameter).
20 nm
Figure S3. Enlarged Transmission electron microscopy (TEM) images of α-K9AlW11 O39 (1) stabilized Ag0 nanoparticles after drying on a carbon-film coated Cu grid. Nanoparticles of Ag0 lie in a grey, variable-density "sea"8 of 1. The image in the left panel is Figure 2A in the text.
S6 Comment to Figure S3: In cases where excess POM anions are removed, dry TEM images nonetheless fail to represent the true solution-state structures. See discussion immediately following Figure S6, below.
a
Figure S4. (a) Enlarged dry HRTEM image of 1-protected Ag(0) (Figure 2B from the text), and energy dispersive spectroscopy results for (b) the location at the end of the red arrow in the HRTEM image, and (c) the location at the end of the yellow arrow in the image.
Table 1. Quantitative EDS Analysisa of the Objects Indicated by Arrows in Figure S4 Elementsb EDS spectrum taken from 1 EDS spectrum taken from (region indicated by yellow the Ag particle (indicated arrow on Figure S4) by red arrow on Figure S4) K 22.32 ± 2.65 13.22 ± 1.94 Ag 0 63.84 ± 5.15 W 77.68 ± 4.11 18.53 ± 2.20 a b The results are shown in atom-%. Cu and C, whose intensities arose from the grid rather than from the sample itself, are not included.
S7
Figure S5. Diffraction pattern taken from the Ag(0) nanoparticles shown by the red arrow in Figure S4 (above) and in Figure 2B in the text. The line spacings and rings are characteristic of polycrystalline Ag(0).
S8
a
b
100 nm
d
c
e
10 nm Figure S6. α-K9 AlW11O39-protected Ag(0) nanoparticles as imaged by cryo-TEM. (a) Nanoparticles of Ag0 lying separately from one another (compare with Figure S3). (b)-(d) Ag(0) particles with their ligand shells. The insets to (b)-(d) are expanded views of the particles in each panel. The ligands shells are shown in views that differ in the location of the focus, and hence in apparent size, resolution and contrast. The smaller nanostructure in panel d is near of the limits of the method for obtaining focused images. Due to its smaller size, however, the intensity from the smaller Ag(0) core is less, and with adjustment to brightness and contrast (panel e), possible evidence for POMs in the plane perpendicular to the electron beam (i.e., on the upward and downward looking faces of the particle) might be discerned.
S9 IV. Additional Comments on Cryo-TEM Images of POM-Stabilized Metal(0) NPs. In all cases, the Ag(0) particles in Figure S6 (b-d) are surrounded by regularly placed objects with diameters of from ca. 1.4 to 1.6 nm (the crystallographic diameter of a Keggin anion is 1.2 nm). Small particles, attributed to individual molecules of “free” α-K9AlW11O39 are seen randomly distributed about the visual field. The same randomly distributed particles are seen by cryo-TEM after adding α-K7PW11O39 to citrate-stabilized Ag(0), as shown in Figure 5B in the text, and in section V. below. These “solution-state” structures provide no evidence for outer-sphere associations of POM anions. By contrast, for one of the most thoroughly characterized POM-stabilized Metal(0) NPs, multiple lines of evidence9 suggest that 2 nm diameter Ir(0) nanoparticles are each stabilized by ca. 33 P2W15Nb3O629- and/or P4W30Nb6O12316- anions. Four lines of evidence indicate that some of the anions are adsorbed on the Ir(0) surface, while the majority (probably 2/3, or ca. 20 POM anions) are present as outer-sphere ligands. In dry TEM images.9 however, no POM anions are observed in an organized fashion on or near the Ir(0) particles, nor are POMs seen in greater abundance near the Ir(0) particles. It is not surprising that the dried samples differ from the solution structures carefully determined by combinations of complementary (non-micrographic) analytical methods. However, it highlights the need for the “solution-state” imaging of POMprotected metal(0) NPs demonstrated in the present work.
V. Additional Characterization and TEM and Cryo-TEM images of CitrateStabilized Ag(0) NPs. Ag-citrate NPs yellow (6 filtrations+centrifuge 5000rpm 6min) concentrated 1.20
13.22 1.00
0.80
0.60
0.40
90.04
0.20 1.94 0.00 0.01
0.10
1.00
10.00
-0.20 radius (nm)
100.00
1000.00
10000.00
S10 Figure S7. Hydrodynamic radii calculated from dynamic light scattering (DLS) data using the CONTIN method. As for the 1-protected Ag(0) particles, DLS was here used to ensure that the solutions contained reasonable concentrations of particles of an appropriate size range for cryo-TEM.
a
b
Figure S8. Transmission electron microscopy (TEM) of Ag0-citrate coated nanoparticles after drying on a carbon-film coated Cu grid. (a) General view with low magnification. (b) High magnification— individual, round smooth particles lie separately on the grid.
a
b
c
Figure S9. Cryo-TEM images of the citrate-protected Ag0 nanoparticles, ca. 10-20 nm in diameter. All three images show round smooth particles on the Cu grid. The high-magnification (x150,000) cryo-TEM image in Figure 5A in the text was obtained from this sample.
S11
VI. TEM and Cryo-TEM Images Obtained Before and After Reaction of αPW11O397– with Citrate-Protected Ag(0) NPs.
100 nm
20 nm
Figure S10. Dry-TEM images of citrate-protected Ag 0 nanoparticles after addition of α-K7PW11 O39 (2). When dried, the Ag(0) particles (dark) are embedded within a “sea” of POM salt (grey), on the white background of the grid.
50 nm 20 nm
Figure S11. Cryo-TEM images of the citrate-protected Ag0 nanoparticles after addition of α-K7PW11 O39 (2). As for 1-protected Ag(0) NPs, cryogenic trapping of the solution-state structures avoids aggregation of the POM, and individual POM anions are observed on the surfaces of the Ag(0) particles. In both cases, substantial displacement of citrate by 2, or at least co-association of 2, is observed. Figure 5B in the text is a separate image of the particle in the above-left panel.
S12
VII. Surface-Enhance Raman (SER) Spectroscopy. In the SER spectrum (Figure 4 in the text), most of the bands observed in the FTIR spectrum of solid K91 (Figure S1, right panel) and in the Raman spectrum of solid K91 (Figure 4 in the text) are observed. Two bands due to the AlO4 moiety at the center of the Cs-symmetry anion (ca. 755 and 704 in IR spectra of K91)3 are considerably weaker in the SER spectrum than they are in the FTIR spectrum. However, the terminal W=O bonds give rise to two bands in the SER spectrum, at 945 and 930 cm-1, indicative of the Cs symmetry mono-defect Keggin anion. The same two bands are observed in the FTIR and Raman spectra of solid K91. These observations provide additional evidence that the anions actually on the surface of the Ag(0) Nps are intact 1. The bands observed in the SER spectrum are red-shifted to lower wavenumbers, an effect that has been attributed10 to direct interaction between oxometalates and the Ag(0) nanoparticle surface. Also, the bands between 750 and 900 cm-1, assigned to vibrational modes involving bridging oxygen atoms, are dramatically enhanced relative to the spectrum of solid K91, and relative to the intensities of the vibrations assigned to terminal W=O. On roughened Ag electrodes, Gewirth observed this pattern of selective enhancement for α-SiW12O404-,11 but not for α-SiW11O398-,12 despite the fact that STM images clearly showed that both anions form ordered arrays on Ag(111). This ambiguity is not uncommon in SER studies, but rather, points out the need for caution when SERS data to assign orientations of molecules to metal surfaces. This is particularly true for 1, where because of its relatively low symmetry, all its vibrational modes are both IR and Raman active. Therefore, we believe more SERS spectra, involving additional POM anions and both Au and Ag, is needed before conclusions regarding the orientation of 1 to the Ag(0) surface can be definitively drawn. This work is currently in progress.
VIII. References (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
Geletii, Y. V.; Hill, C. L.; Bailey, A. J.; Hardcastle, K. I.; Atalla, R. H.; Weinstock, I. A. Inorg. Chem. 2005, 44, 8955-8966. Weinstock, I. A.; Cowan, J. J.; Barbuzzi, E. M. G.; Zeng, H.; Hill, C. L. J. Am. Chem. Soc. 1999, 121, 4608-4617. Cowan, J. J.; Bailey, A. J.; Heintz, R. A.; Do, B. T.; Hardcastle, K. I.; Hill, C. L.; Weinstock, I. A. Inorg. Chem. 2001, 40, 6666-6675. Cowan, J. J.; Hill, C. L.; Reiner, R. S.; Weinstock, I. A. In Inorganic Syntheses; Coucouvanis, D., Ed.; John Wiley& Sons, Inc.: New York, 2002; Vol. 33, p 18-26. Kozik, M.; Baker, L. C. W. J. Am. Chem. Soc. 1990, 112, 7604-7611. Creighton, J. A.; Blatchford, C. G.; Albrecht, M. J. Farad. Trans. 2 1979, 75, 790-798. Lica, G. C.; Browne, K. P.; Tong, Y. J. Cluster Sci. 2006, 17, 349-359. Maayan, G.; Neumann, R. Catal. Lett. 2008, 123, 41-45. Lin, Y.; Finke, R. G. J. Am. Chem. Soc. 1994, 116, 8335-8353. Siiman, O.; Feilchenfeld, H. J. Phys. Chem. 1988, 92, 453-64. Teague, C. M.; Li, X.; Biggin, M. E.; Lee, L.; Kim, J.; Gewirth, A. A. J. Phys. Chem. B 2004, 108, 1974-1985. Kim, J.; Gewirth, A. A. Langmuir 2003, 19, 8934-8947.
Supporting Information for: Direct Imaging of the Ligand Monolayer on an Anion-Protected Metal Nanoparticle through Cryogenic Trapping of its Solution-State Structure Alevtina Neyman, † Louisa Meshi,‡ Leila Zeiri, † and Ira A. Weinstock*,† Department of Chemistry, Ben Gurion University of the Negev, Beer Sheva, 84105, Israel, Ilse Katz Institute for Nanoscale Science and Technology, Ben Gurion University of the Negev, Beer Sheva, 84105, Israel
Contents: I. Materials and methods II. Comments on high-magnification Cryo-TEM. III. Characterization of α-AlW11O399–-protected Ag(0) NPs, including spectroscopic data that confirm the structural integrity of α-AlW11O399-, and additional TEM and Cryo-TEM images. IV. Additional comments on Cryo-TEM images of POM-stabilized metal(0) NPs. V. Additional characterization and TEM and Cryo-TEM images of citrate-stabilized Ag(0) NPs. VI. TEM and Cryo-TEM images obtained before and after reaction of α-PW11O397– with citrateprotected Ag(0) NPs. VII. Surface-enhanced Raman (SER) spectroscopy. VIII. References
I. Materials and Methods Materials For Ag(0) nanoparticle synthesis and characterization, sodium borohydride (Fisher; general grade), trisodium citrate (Sigma; USP), sodium chloride (Frutarom; analytical grade), silver nitrate (J.T. Baker; A.C.S. reagent), deuterium oxide (99.9% D) (Bio Lab, Ltd.), and potassium bromide (EDE, spectral grade for spectroscopy) were used as received. Sodium tungstate dihydrate was purchased from Merck (extra pure). Additional reagent grade salts, acids and diethylether for polyoxometalate synthesis were obtained from commercial sources and used as received. α-K7PW11O391 and α-K9AlW11O392-4 were prepared using published methods. All water used was of high purity (18.2 MΩ resistivity) from a Millipore Direct-Q waterpurification system. Methods. pH Measurements. pH values were measured using a EuTech pH 510 Bench-Top pH meter, or a TES-1380 pH meter equipped with a Ag/AgCl electrode (Thermo Electron, UK), and temperature compensation. Prior to use the pH meter was calibrated using standard reference solutions (pH 4.00 and 7.00). Spectroscopy Analyses. UV-visible. UV-vis spectra were acquired using a Hewlett-Packard 8452A spectrophotometer equipped with a diode-array detector. For all measurements the solutions were diluted as needed to avoid over-saturation by the strongly absorbing ligand-tometal charge transfer bands of the polyoxometalates in the 190-300 nm region. Nuclear Magnetic Resonance (NMR). Al-27 NMR2,3 and 31P NMR5 spectra were acquired on a Bruker 500 MHz spectrometer (5 mm BBI and 5 mm QNP probes respectively). Chemicalshift values were externally referenced to 0.10 M AlCl3 ([Al(H2O)6]3+, or to 1.0 M
S2 H3PO4, both set to δ = 0 ppm. Internal lock signals were tuned using D2O. Spectral data were processed using the NMR software package, Mnova version 5.1. Infra-Red. Fourier-transform infra-red (FTIR) spectra were acquired using a Nicolet Impact 410 spectrophotometer (KBr pellets). Raman. The Raman instrument comprised a Jobin-Yvon LabRam HR 800 micro-Raman system, equipped with a liquid-N2-cooled detector. The excitation source was a He-Ne laser (633 nm) with a power of 5 mW on the sample. (Spectra of the POM-protected Ag(0) nanoparticle (NP) solutions obtained at 514 nm were not improved relative to 633 nm.) The laser was focused with an x50 long-focal-length objective on a spectroscopic cell containing the sample, to a spot of about 2 µm. The measurements were taken with the 600 g mm-1 grating and a confocal microscope with a 100-µm hole. The exposure time for the surface-enhanced Raman spectrum (SERS) show in Figure 4 (text) was 30 minutes. Fluorescence. Fluorescence spectra were acquired using a Jobin Yvon Fluorolog-3 spectrophotometer. No fluorescence arising from the POM-stabilized Ag(0) solutions was observed. Dynamic Light Scattering (DLS). DLS data was collected at 25 °C on an ALV-CGS-8F instrument (ALV-GmbH, Germany) with appropriately diluted samples. The CONTIN method was used to obtain hydrodynamic radii (Rh). The dependence of Rh on measurement angle was investigated from 30 to 120 degrees (in increments of 30 degrees) and appeared to be negligible—consistent with spherically shaped particles. Experimental uncertainties in total scattered intensities and in Rh were about 1 to 3%. The DLS data were obtained, primarily, to ensure that colloidal solutions of POM-protected Ag(0) NPs contained significant concentrations of particles in the size range of interest (ca. 10 to 20 nm). Transmission Electron Microscopy (TEM). Samples for Dry TEM and for high-resolution TEM (HRTEM) were prepared by pipetting 5-10 µL of the aqueous sample solution onto Cu grids covered with thin carbon-support films and dried in air. TEM data was obtained using a FEI Tecnai 12 G2 electron microscope (120kV) equipped with a Gatan slow-scan camera. HRTEM data were obtained using a JEOL JEM-2010 instrument (200kV) equipped with a Noran energy dispersive spectroscopy (EDS) system and a Gatan slow-scan camera. Cryogenic TEM. The cryogenically frozen samples were prepared using a fully automated vitrification device (“Vitrobot”). First, 5 µL of the sample solution was placed by pipette onto a glow-discharged Cu grid covered with a lacey-carbon film held inside a 100% humidity chamber. The grid was then mechanically “blotted” and immediately plunged into liquid ethane (b.p. 185K) cooled by liquid nitrogen (b.p. 77K). Data were collected on the FEI Tecnai 12 G2 instrument (120kV) and the Gatan slow-scan camera. Syntheses. α-K9AlW11O39-Protected Ag(0) Nanoparticles. α-K9AlW11O39 (the K+ salt of 1),2-4 was specifically chosen for its stability at borate-buffer pH values of between 8 and 9 (this salt is prepared at pH values of 85). Therefore, unlike α-SiW11O398- or α-PW11O397-, 1 can be used instead of citrate during BH4- reduction of Ag+ (Creighton method7). The ratio of Ag+ to α-K9AlW11O39 (1) was 1:1, and NaBH4 was added at 1.6 times molar excess relative to Ag+ (i.e., the ratio of BH4- to Ag+ was 1.6). Specifically, 0.039 g (1.25 x 10-5 mol) 1 was dissolved in 4.5 mL of pure water. Then, 0.5 mL of freshly prepared aqueous AgNO3 solution (0.025 M) was added to the solution of 1, and the clear, colorless solution was stirred vigorously at room temperature for 1 hr. After cooling a freshly prepared 0.1 M aqueous solution of NaBH4 in an ice bath, 200 µL (2.0 x 10-5 mol, 1.6 equivalents) were added in 3 equal (66.7
S3 µL) portions at 2-3 seconds intervals, with vigorous stirring. The resultant dark-brown solution was stirred for another 30 min at room temperature. The final pH values of these solutions were typically between 8 and 8.5. The solution was filtered 3 times through a 0.45 µm PVDF nonsterile Millipure syringe filters, then another 3 times through a 0.22 µm filter. The solution was characterized by UV-vis, FTIR, 27Al NMR, surface enhanced Raman spectroscopy, dynamic light scattering, transmission and high-resolution transmission electron microscopy (TEM and HRTEM), electron dispersive spectroscopy (EDS) and X-ray diffractions (both in conjunction with HRTEM) and by cryogenic-TEM (cryo-TEM). These data and images are presented in Figures 2-5 of the text, and below. The spectroscopic characterization was carried out with due attention to the integrity of 1 during the above procedure (see data and discussion of this in section III., below). Citrate-Protected Ag0 Nanoparticles (NPs). Citrate-stabilized Ag(0) NPs were prepared according to a variation of methods reported by Creighton 6 and Tong.7 Briefly, 0.0425 g (2.5 x 10-4 mol) of AgNO3 and 0.0735 g (2.5 x 10-4 mol) of tri-sodium citrate, Na3C6H5O7, were dissolved in 100 ml of pure (18.2 MΩ) water and stirred for 1 h at room temperature (between 22 and 25 °C). Meanwhile, 0.051 g (4.0 x 10-4 mol) of NaBH4 was dissolved in 4 mL of pure water, cooled in ice-bath, and added quickly with vigorous stirring. The color of the solution immediately turned dark green-brown. The mixture was then stirred for 30 min, filtered 3 times through a 0.45 µm polyvinylidene difluoride (PVDF) non-sterile Millipure syringe filter, then another 3 times through a 0.22 µm filter. Finally, 10 portions of 1.5 mL each (in 1.5 mL Eppendorf vials) were centrifuged for 6 min at 5000 rpm. The clear yellow supernatant solution was decanted and stored in the dark at room temperature for analysis by TEM and cryo-TEM. Reaction of Citrate-Protected Ag0 NPs with α-K7PW11O39. Using a variation of Tong’s method,7, the pH of 10 mL of the above citrate-protected Ag(0) NP solution was first adjusted to 5.5 with dilute H2SO4. Then, 0.075 g (2.5 x 10-5 mol) of solid α-K7PW11O39 was added. The solution was stirred at room temperature for 1 hr, then filtered and stored in the dark at room temperature for analysis by TEM and cryo-TEM.
II. Comments on High-Magnification Cryo-TEM. In cryo-TEM, the normal methods used to achieve image contrast in traditional (dry-sample) TEM studies are not applicable. For cryo-TEM, “phase contrast” is used. To enable this, the “focus” must be slightly physically below the object being imaged. A more serious limitation of the technique, relevant to the data provided here, is that high-magnification images are difficult to obtain due to the tendency of the vitreous water to change when warmed by the electron beam. The magnification typically used for cryo-TEM imaging of meso-scale and relatively large nanoscale objects is x30,000. At this relatively modest magnification, the energy of the electron beam is dispersed over a relatively large visual field. The smaller energy flux (e.g., per nm3) within the sample limits the extent to which the beam induces changes to the vitreous water, and allows more time for focusing and obtaining high-quality images of a specific object. Hence, at lower magnification, a number of images of the same object can be obtained and the “under focus” method optimized for clarity and contrast. The images shown below were obtained at x150,000, near the limit of the instrument and operator, whose experience and skill play an important role. The limitation is that optimized images must be obtained with very few attempts allowed before heating of the vitreous water in the visual field obviates further data acquisition at that location on the grid. The small collection
S4 of images provided here shows the variation in contrast and focus typical of this technique when used at high (x150,000) magnification. The polyoxometalate anions used in this work are ca. 1.2 nm in diameter. In future work, larger POMs will be used to investigate POM interactions with smaller metal(0) nanoparticles. This should provide nano-assemblies that, while based on 1 or 2 nm diameter metal(0) particles, are relatively large due to the dimensions of the associated POM clusters (1 to 4 nm for each POM cluster if one includes the larger oxomolybdate clusters planned for inclusion in future work).
III. Characterization of α-AlW11O399–-Protected Ag(0) NPs, including Spectroscopic Data that Confirm the Structural Integrity of α-AlW11O399-, and Additional TEM and Cryo-TEM images. Structural Integrity of α-K9AlW11O39 (1) as a Protecting Ligand for Ag(0). The UV-vis spectrum of the dark brown solution featured a broad absorbance at 425 nm, due to the surface plasmon resonance of the Ag(0) nanoparticles (Figure S1, A). The FTIR spectrum obtained after concentrating the solution to dryness was identical to that of pure 1 (KBr pellet; Figure S1, B). The 27Al NMR spectrum of the solution (not shown),2-4 was also the same as that of pure 1.
B
A
nm
cm-1
Figure S1. Characterization and effects of drying 1-protected Ag(0) nanoparticles. A: UV-vis spectra of aq. solutions of K91 (black line) and of 1-protected Ag(0) (red line). B: FTIR spectra of pure K91 (black line) and of the dark solid obtained by drying the 1-protected Ag(0) solution (red line).
S5
aa-86-1 (k9AlW11O39+Ag0NPs) 1.20E+00
34.683 1.00E+00
8.00E-01
6.00E-01 5.091 4.00E-01
2.00E-01
0.00E+00 0.01
0.10
1.00
10.00
100.00
1000.00
10000.00
-2.00E-01 hydrodinamic radius (nm)
Figure S2. Hydrodynamic radii (Rh) calculated from dynamic light scattering (DLS) data using the CONTIN method. The DLS data were obtained to ensure that colloidal solutions of POM-protected Ag(0) NPs contained significant concentrations of particles in the size range of interest (ca. 10 to 20 nm diameter).
20 nm
Figure S3. Enlarged Transmission electron microscopy (TEM) images of α-K9AlW11 O39 (1) stabilized Ag0 nanoparticles after drying on a carbon-film coated Cu grid. Nanoparticles of Ag0 lie in a grey, variable-density "sea"8 of 1. The image in the left panel is Figure 2A in the text.
S6 Comment to Figure S3: In cases where excess POM anions are removed, dry TEM images nonetheless fail to represent the true solution-state structures. See discussion immediately following Figure S6, below.
a
Figure S4. (a) Enlarged dry HRTEM image of 1-protected Ag(0) (Figure 2B from the text), and energy dispersive spectroscopy results for (b) the location at the end of the red arrow in the HRTEM image, and (c) the location at the end of the yellow arrow in the image.
Table 1. Quantitative EDS Analysisa of the Objects Indicated by Arrows in Figure S4 Elementsb EDS spectrum taken from 1 EDS spectrum taken from (region indicated by yellow the Ag particle (indicated arrow on Figure S4) by red arrow on Figure S4) K 22.32 ± 2.65 13.22 ± 1.94 Ag 0 63.84 ± 5.15 W 77.68 ± 4.11 18.53 ± 2.20 a b The results are shown in atom-%. Cu and C, whose intensities arose from the grid rather than from the sample itself, are not included.
S7
Figure S5. Diffraction pattern taken from the Ag(0) nanoparticles shown by the red arrow in Figure S4 (above) and in Figure 2B in the text. The line spacings and rings are characteristic of polycrystalline Ag(0).
S8
a
b
100 nm
d
c
e
10 nm Figure S6. α-K9 AlW11O39-protected Ag(0) nanoparticles as imaged by cryo-TEM. (a) Nanoparticles of Ag0 lying separately from one another (compare with Figure S3). (b)-(d) Ag(0) particles with their ligand shells. The insets to (b)-(d) are expanded views of the particles in each panel. The ligands shells are shown in views that differ in the location of the focus, and hence in apparent size, resolution and contrast. The smaller nanostructure in panel d is near of the limits of the method for obtaining focused images. Due to its smaller size, however, the intensity from the smaller Ag(0) core is less, and with adjustment to brightness and contrast (panel e), possible evidence for POMs in the plane perpendicular to the electron beam (i.e., on the upward and downward looking faces of the particle) might be discerned.
S9 IV. Additional Comments on Cryo-TEM Images of POM-Stabilized Metal(0) NPs. In all cases, the Ag(0) particles in Figure S6 (b-d) are surrounded by regularly placed objects with diameters of from ca. 1.4 to 1.6 nm (the crystallographic diameter of a Keggin anion is 1.2 nm). Small particles, attributed to individual molecules of “free” α-K9AlW11O39 are seen randomly distributed about the visual field. The same randomly distributed particles are seen by cryo-TEM after adding α-K7PW11O39 to citrate-stabilized Ag(0), as shown in Figure 5B in the text, and in section V. below. These “solution-state” structures provide no evidence for outer-sphere associations of POM anions. By contrast, for one of the most thoroughly characterized POM-stabilized Metal(0) NPs, multiple lines of evidence9 suggest that 2 nm diameter Ir(0) nanoparticles are each stabilized by ca. 33 P2W15Nb3O629- and/or P4W30Nb6O12316- anions. Four lines of evidence indicate that some of the anions are adsorbed on the Ir(0) surface, while the majority (probably 2/3, or ca. 20 POM anions) are present as outer-sphere ligands. In dry TEM images.9 however, no POM anions are observed in an organized fashion on or near the Ir(0) particles, nor are POMs seen in greater abundance near the Ir(0) particles. It is not surprising that the dried samples differ from the solution structures carefully determined by combinations of complementary (non-micrographic) analytical methods. However, it highlights the need for the “solution-state” imaging of POMprotected metal(0) NPs demonstrated in the present work.
V. Additional Characterization and TEM and Cryo-TEM images of CitrateStabilized Ag(0) NPs. Ag-citrate NPs yellow (6 filtrations+centrifuge 5000rpm 6min) concentrated 1.20
13.22 1.00
0.80
0.60
0.40
90.04
0.20 1.94 0.00 0.01
0.10
1.00
10.00
-0.20 radius (nm)
100.00
1000.00
10000.00
S10 Figure S7. Hydrodynamic radii calculated from dynamic light scattering (DLS) data using the CONTIN method. As for the 1-protected Ag(0) particles, DLS was here used to ensure that the solutions contained reasonable concentrations of particles of an appropriate size range for cryo-TEM.
a
b
Figure S8. Transmission electron microscopy (TEM) of Ag0-citrate coated nanoparticles after drying on a carbon-film coated Cu grid. (a) General view with low magnification. (b) High magnification— individual, round smooth particles lie separately on the grid.
a
b
c
Figure S9. Cryo-TEM images of the citrate-protected Ag0 nanoparticles, ca. 10-20 nm in diameter. All three images show round smooth particles on the Cu grid. The high-magnification (x150,000) cryo-TEM image in Figure 5A in the text was obtained from this sample.
S11
VI. TEM and Cryo-TEM Images Obtained Before and After Reaction of αPW11O397– with Citrate-Protected Ag(0) NPs.
100 nm
20 nm
Figure S10. Dry-TEM images of citrate-protected Ag 0 nanoparticles after addition of α-K7PW11 O39 (2). When dried, the Ag(0) particles (dark) are embedded within a “sea” of POM salt (grey), on the white background of the grid.
50 nm 20 nm
Figure S11. Cryo-TEM images of the citrate-protected Ag0 nanoparticles after addition of α-K7PW11 O39 (2). As for 1-protected Ag(0) NPs, cryogenic trapping of the solution-state structures avoids aggregation of the POM, and individual POM anions are observed on the surfaces of the Ag(0) particles. In both cases, substantial displacement of citrate by 2, or at least co-association of 2, is observed. Figure 5B in the text is a separate image of the particle in the above-left panel.
S12
VII. Surface-Enhance Raman (SER) Spectroscopy. In the SER spectrum (Figure 4 in the text), most of the bands observed in the FTIR spectrum of solid K91 (Figure S1, right panel) and in the Raman spectrum of solid K91 (Figure 4 in the text) are observed. Two bands due to the AlO4 moiety at the center of the Cs-symmetry anion (ca. 755 and 704 in IR spectra of K91)3 are considerably weaker in the SER spectrum than they are in the FTIR spectrum. However, the terminal W=O bonds give rise to two bands in the SER spectrum, at 945 and 930 cm-1, indicative of the Cs symmetry mono-defect Keggin anion. The same two bands are observed in the FTIR and Raman spectra of solid K91. These observations provide additional evidence that the anions actually on the surface of the Ag(0) Nps are intact 1. The bands observed in the SER spectrum are red-shifted to lower wavenumbers, an effect that has been attributed10 to direct interaction between oxometalates and the Ag(0) nanoparticle surface. Also, the bands between 750 and 900 cm-1, assigned to vibrational modes involving bridging oxygen atoms, are dramatically enhanced relative to the spectrum of solid K91, and relative to the intensities of the vibrations assigned to terminal W=O. On roughened Ag electrodes, Gewirth observed this pattern of selective enhancement for α-SiW12O404-,11 but not for α-SiW11O398-,12 despite the fact that STM images clearly showed that both anions form ordered arrays on Ag(111). This ambiguity is not uncommon in SER studies, but rather, points out the need for caution when SERS data to assign orientations of molecules to metal surfaces. This is particularly true for 1, where because of its relatively low symmetry, all its vibrational modes are both IR and Raman active. Therefore, we believe more SERS spectra, involving additional POM anions and both Au and Ag, is needed before conclusions regarding the orientation of 1 to the Ag(0) surface can be definitively drawn. This work is currently in progress.
VIII. References (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
Geletii, Y. V.; Hill, C. L.; Bailey, A. J.; Hardcastle, K. I.; Atalla, R. H.; Weinstock, I. A. Inorg. Chem. 2005, 44, 8955-8966. Weinstock, I. A.; Cowan, J. J.; Barbuzzi, E. M. G.; Zeng, H.; Hill, C. L. J. Am. Chem. Soc. 1999, 121, 4608-4617. Cowan, J. J.; Bailey, A. J.; Heintz, R. A.; Do, B. T.; Hardcastle, K. I.; Hill, C. L.; Weinstock, I. A. Inorg. Chem. 2001, 40, 6666-6675. Cowan, J. J.; Hill, C. L.; Reiner, R. S.; Weinstock, I. A. In Inorganic Syntheses; Coucouvanis, D., Ed.; John Wiley& Sons, Inc.: New York, 2002; Vol. 33, p 18-26. Kozik, M.; Baker, L. C. W. J. Am. Chem. Soc. 1990, 112, 7604-7611. Creighton, J. A.; Blatchford, C. G.; Albrecht, M. J. Farad. Trans. 2 1979, 75, 790-798. Lica, G. C.; Browne, K. P.; Tong, Y. J. Cluster Sci. 2006, 17, 349-359. Maayan, G.; Neumann, R. Catal. Lett. 2008, 123, 41-45. Lin, Y.; Finke, R. G. J. Am. Chem. Soc. 1994, 116, 8335-8353. Siiman, O.; Feilchenfeld, H. J. Phys. Chem. 1988, 92, 453-64. Teague, C. M.; Li, X.; Biggin, M. E.; Lee, L.; Kim, J.; Gewirth, A. A. J. Phys. Chem. B 2004, 108, 1974-1985. Kim, J.; Gewirth, A. A. Langmuir 2003, 19, 8934-8947.
