Controlling the Distribution of Supported Nanoparticles by Aqueous

Feb 11, 2013 - Lei , Y.; Mehmood , F.; Lee , S.; Greeley , J.; Lee , B.; Seifert , S.; Winans , R. E.; Elam , J. W.; Meyer , R. J.; Redfern , P. C.; T...
0 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/cm

Controlling the Distribution of Supported Nanoparticles by Aqueous Synthesis Tamara M. Eggenhuisen,† Heiner Friedrich,†,‡ Fabio Nudelman,‡ Jovana Zečević,† Nico A. J. M. Sommerdijk,*,‡ Petra E. de Jongh,† and Krijn P. de Jong*,† †

Inorganic Chemistry and Catalysis, Debye Institute for Nanomaterials Science, Utrecht University, Utrecht, The Netherlands Laboratory of Materials and Interface Chemistry and Soft Matter CryoTEM Research Unit, Department of Chemical Engineering and Chemistry, Eindhoven University of Technology, Eindhoven, The Netherlands

‡

S Supporting Information *

ABSTRACT: Synthesis of supported nanoparticles with controlled size and uniform distribution is a major challenge in nanoscience, in particular for applications in catalysis. Cryo-electron tomography revealed with nanometer resolution the 3D distribution of phases present during nanoparticle synthesis via impregnation, drying, and thermal treatment with transition metal salt precursors. By conventional methods a nonuniform salt distribution led to clustered metal oxide nanoparticles (NiO, Co3O4). In contrast, freezedrying restricted solution mobility during drying and a more uniform nanoparticle distribution was obtained. By this fundamental insight into catalyst preparation, controlled synthesis of supported catalysts was achieved in a way that is also applicable for other nanostructured materials.

KEYWORDS: nanoparticle synthesis, heterogeneous catalyst, cryogenic electron tomography, freeze-drying

■

INTRODUCTION Nanoparticles play a key role in many present day applications such as catalysis, biomedical imaging, fuel cells, and solar cells. The synthesis of functional nanostructures by simple, costeffective, and environmentally friendly methods is crucial for their large scale application.1,2 To exploit the exceptional properties of nanoparticles in multiphase systems, they are often stabilized within carriers that warrant their function and accessibility.3,4 Supported nanoparticles are for example widely studied and applied for the energy efficient production of fuels and chemicals.5,6 For applications relying on the surface properties of these nanoparticles, their controlled, equally spaced deposition within a porous carrier is of high importance.7−9 For the preparation of deactivation resistant heterogeneous catalysts, the deposition of equally sized and maximally spaced sub-10 nm particles within a porous support material is considered a holey grail.10−13 However, the controlled deposition of a high number density of 1−10 nm sized particles has remained largely elusive. State-of-the-art preparation routes that apply lithography14 or printing15 or use metal−organic frameworks16 and singlecrystal surfaces17 can be used for small scale model systems. Within the field of heterogeneous catalysis, the impregnation and drying method is a seemingly simple bottom-up approach to prepare supported transition metal or metal oxide nanoparticles and is commonly applied in both academia and industry.18 This procedure generally uses aqueous transition metal nitrate precursor solutions and produces little waste, but © 2013 American Chemical Society

control over particle size, shape, and distribution during the consecutive synthesis steps is problematic.19 Clustering of nanoparticles and broad particle size distributions occur in porous supports as silica and alumina20,21 as well as with different transition metals at varying concentrations.22,23 Catalyst preparation and the derived catalyst structure play a determining role for the activity, selectivity, and stability in a catalytic process. Therefore, it is not surprising that many groups report on methods improving control over nanoparticle size and distribution over the support, for example, utilizing precursors other than transition metal nitrates, alternative decomposition pathways, or self-assembly methods.24−28 In addition, by applying varying drying methods, smaller nanoparticles with improved distributions were obtained, especially for macroscopic support bodies.29,30 In an earlier paper we have reported that freeze-drying leads to monodisperse particle sizes and homogeneous distributions in an industrially relevant cobalt catalyst.31 Therefore, freeze-drying seems to improve the catalyst structure, while still using the simple impregnation method with cheap precursors. However, a link between the distribution of the system in the solute state, the dried state, and after activation has not been established. In this paper we address the link between the preparation stages to obtain fundamental insight into the genesis of Received: November 23, 2012 Revised: February 6, 2013 Published: February 11, 2013 890

dx.doi.org/10.1021/cm3037845 | Chem. Mater. 2013, 25, 890−896

Chemistry of Materials

Article

broadening was not corrected for strain effects. TEM microscopy was performed using a Tecnai 12 operated at 120 kV. Particle size analysis by electron microscopy was performed on 2.7 nm thick slices cutting through the tomographic reconstruction parallel to the pore direction. Over 90 particles were counted on at least three slices cutting through different pores. The volume averaged particle diameter is reported. Analysis of an 8.1 nm thick slice, i.e., cumulating slices up to ∼one pore diameter thick, did not affect the obtained average. Cryo-Electron Tomography: In Situ and Lab Scale Drying Experiments. The effect of drying on the cobalt nitrate salt distribution was visualized in two sets of experiments. For the in situ drying experiment, tilt series were acquired of the same support particle as impregnated and after subsequent drying. SBA-15 was impregnated to incipient wetness with a saturated aqueous Co(NO3)2 solution and applied on a TEM grid, which was subsequently immersed into liquid nitrogen. For the conventional drying experiment finder grids (Cu, 200 mesh, type F1 Electron Miscroscopy Sciences) were used. After acquisition of the tilt series of an impregnated particle, the grid was removed from the microscope sample holder and heated to 60 °C for 16 h on a hot plate. The grid was then reimmersed in liquid nitrogen and reinserted into the microscope sample holder. The same particle was located on the finder grid to acquire the second tilt series. In situ freeze-drying was performed on a Quantifoil R2/2 carbon film supported on square mesh grids (Cu, 200 mesh). A tilt series was acquired of an impregnated SBA-15 particle, and subsequently the temperature of the grid holder inside the ultrahigh vacuum microscope column was allowed to rise to room temperature overnight. Prior to acquiring the tilt series of the freeze-dried particle, the system was cooled again to ∼−170 °C. Image analysis was performed on 14.8 nm thick slices through the reconstructions perpendicular through the body of the particle. Empty or partially empty pores were detected by a gray value intensity upper limit integrated over 10 nm circular areas fixed on an hexagonal array.36 For clarity of the three-dimensional reconstructions, empty pore detections at the rim of the particle were cut off. Samples prepared on lab scale were impregnated and conventionally dried in an oven as described in the Catalyst Preparation section. Freeze-drying of samples on lab scale for cryo-ET was performed using a Sublimator 400 Freeze-dryer (Zirbus). Samples were frozen by cooling to −45 °C (1 °C/min, 3 h) and were dried at 0.02−0.03 mbar for 4 days on a shelf with a temperature of −30 °C. The temperature was increased to −10 °C for 1 h and to 20 °C for 1 day at 0.01−0.02 mbar. A small amount of powder was deposited on a TEM grid, which was then immersed in liquid nitrogen and transferred to the cryoholder of the CryoTitan microscope (FEI, www.cryoTEM.nl) to perform cryo-electron tomography. The freeze-dried sample was handled under a dry N2 atmosphere at all times. TEM grids were glow discharged prior to sample preparation using a Cressington 208 carbon coater. Tomography labels were added by applying 10 nm sized colloidal gold particles (Aurion, PAG conjugated) from aqueous suspension. The calcined catalysts were applied onto a 2/2 Quantifoil grid from a suspension in ethanol. The other samples were prepared by depositing a small amount of powder on the grid, after which the excess was shaken off. FD-Co(NO3)2/ SBA-15 was applied under a dry N2 atmosphere in a N2-drybox. Cryotomography was performed on the TU/e CryoTitan (FEI) equipped with a field emission gun and operated at 300 kV acceleration voltage. Cryo-TEM images were acquired on a Gatan a 2k × 2k Gatan CCD camera using zero-loss energy filtering with a slit width of 20 eV. Tilt series were recorded from ±65° at 1° tilt increments at a nominal magnification of 24k resulting in a pixel size of 0.37 nm. Electron tomography of calcined catalysts was performed on a Tecnai 20 microscope operated at 200 kV. Tilt series were recorded from ±75° at 2° tilt increments at a nominal magnification of 29k with a 2k × 2k Gatan CCD camera resulting in a pixel size of 0.27 and 0.54 nm after binning by a factor of 2. Alignment of the TEM tilt series and 3D reconstructions were performed in IMOD software.37

nanoparticle distribution using (cryo-) electron tomography visualizing the three-dimensional distribution of phases at the nanometer level.32−34 We chose freeze-drying versus drying in a muffle oven to study the effect of drying on the precursor salt and the nanoparticle distribution in heterogeneous catalysts. Both methods represent the extremes of expected salt redistribution during drying: immobilization and agglomeration, respectively. Our results showed that the distribution of the transition metal nitrate precursor in a mesoporous silica support (SBA-15) formed the fingerprint for the final nanoparticle distribution, when combined with the NO-assisted decomposition method.19 The derived approach was successfully applied, yielding model catalysts with significantly improved structural characteristics by rational design. Our strategy was demonstrated for the synthesis of catalysts with more uniform Co3O4 and NiO nanoparticle distributions in a mesoporous silica support but is applicable to a much broader range of catalytic nanoparticles and carriers.

■

EXPERIMENTAL METHODS

Catalyst Preparation. SBA-15 with rodlike morphology was prepared according to the procedure described by Sayari et al.35 A total of 4.0 g of P123 was dissolved in 130 mL of demineralized H2O and 20 mL of HCl (37 wt %, Aldrich) at 39 °C. After the addition of 7.5 g of tetraethylorthosilicate (98%, Aldrich), the solution was stirred for 6 min and subsequently kept static at 39 °C for 24 h. After a hydrothermal treatment at 100 °C for 48 h, the white product was filtered, washed, and dried at 60 °C for 3 h and 120 °C for 6 h. Calcination was performed at 550 °C for 6 h (heating rate 1 °C/min). The structural properties of the different batches of SBA-15 were determined using N2 physisorption at −196 °C (Tristar 3000, Micromeritics): dp = 11 nm, Vp = 1.15 cm3/g, and SBET = 845 m2/g (see SI4, Supporting Information, for porosity characterization of individual batches). Catalyst preparation was performed by incipient wetness impregnation with a saturated aqueous Co(NO3)2 solution (4.2 M, >99% Co(NO3)2·6H2O, Sigma-Aldrich) leading to a cobalt metal loading of 18 wt %, or with a saturated aqueous solution of Ni(NO3)2 (4.3 M, Ni(NO3)2·6H2O 99%, Acros), leading to a nickel metal loading of 19 wt %. Platinum was added as a reduction promoter for the cobalt based catalyst via coimpregnation (Pt(NH3)4(NO3)2, Sigma-Aldrich) to obtain 0.05 wt % Pt. Conventional drying was performed in a preheated muffle oven for 16 h at 60 °C for the cobalt catalysts and 120 °C for the nickel catalyst. Freeze-drying for NiO and Co3O4−Pt catalysts was performed using an Epsilon 2-4 Lyophilizer (Martin Christ). Samples were frozen with liquid nitrogen, after which the sample vials were placed on a precooled shelf at −55 °C. Primary drying was performed at a pressure of 0.02 mbar and a shelf temperature of −50 °C for 4.5 days. The temperature was increased stepwise to 20 °C over 8 h, during which the pressure did not increase. Freeze-dried samples were transferred and stored under a dry N2 atmosphere to prevent rehydration. The dried catalysts were thermally treated to decompose the nitrate salt in a plug-flow reactor at 350 °C (1 °C/min, 1 h) in a flow of 1% NO/N2 or N2 (100 mL/50 mg catalyst). Catalyst Characterization. Thermogravimetric analysis (TGA, Q50 TA Instruments) measured the weight loss during heating to 500 °C at 10 °C/min for 30 min. The weight loss was corrected for the loss of water from a pure silica reference sample between 150 and 500 °C, while water desorbed below 150 °C was considered to be included in the precursor phase. The residual water was expressed in moles H2O/mole metal. XRD patterns were recorded for all catalysts between 10 and 60° 2θ with a Bruker-AXS D2 Phaser X-ray diffractometer using Co Kα12 radiation (λ = 1.790 Å). The volume averaged Co3O4 and NiO crystallite sizes were determined using line broadening analysis by a fitting procedure in Eva2 software (Bruker AXS) and the Scherrer equation with shape factor k = 0.9. Line 891

dx.doi.org/10.1021/cm3037845 | Chem. Mater. 2013, 25, 890−896

Chemistry of Materials

Article

Figure 1. 3D reconstruction showing empty pore sections (in white) present in a single mesoporous silica particle after impregnation and after in situ drying on electron microscopy grids. a, In situ conventional drying at 60 °C in oven. b, In situ freeze-drying. c, d, Cartoon showing local pore filling after impregnation and drying treatments. All scale bars are 200 nm. Please refer to the movie Figure 1_3D-insitu.mov for the 3D reconstruction models (see SI1, Supporting Information; for details on image analysis see SI2, Supporting Information).

Figure 2. Co(NO3)2 salt distribution in mesoporous silica particles after impregnation and different drying treatments visualized by cryo-electron tomography. a, Prepared on mg scale by conventional drying at 60 °C in oven; open triangle and rectangle in slices B and C indicate pores with alternating filling. b, Prepared on mg scale by freeze-drying. a, b, All scale bars are 50 nm. 2.2 nm thick slices from cryo-ET reconstructions parallel to the pores. Lines A, B, and C indicate positions of the 14.8 nm thick cross sections perpendicular to the pore direction. The 3D distribution of salt within the pores is best appreciated by the illustrating movies Figure 2a_Conv.mov and Figure 2b_Freeze.mov (see SI1, Supporting Information).

■

RESULTS AND DISCUSSION

pores of the SBA-15 particle in Figure 1a were homogeneously filled after impregnation with cobalt nitrate solution. However, after conventional drying, many open or partially open pore sections were present throughout the particle. The SBA-15 particle in Figure 1b contained some empty pores already after impregnation, which were ascribed to local deficiency of impregnation solution, in line with our previous report.38 Nevertheless, in this case, after freeze-drying the salt distribution perfectly matched the solution distribution after impregnation. This demonstrates the successful immobilization of precursor phase during the sublimation of the water in contrast to the redistribution occurring during conventional drying. Impregnation and drying on lab scale of SBA-15 with a cobalt nitrate solution yielded similar results (Figure 2). After

Particle filling before and after drying was studied with cryoelectron tomography directly using in situ drying on a transmission electron microscope grid. The distribution of the impregnation solution and of the remaining precursor salt after drying was visualized within the same SBA-15 carrier particle. Rodlike SBA-15 particles35 with a narrow pore size distribution and a uniform morphology (∼1 μm × 0.3 μm) were used as model supports to exclude transport effects induced by variations in pore length or diameter. 3D reconstructions of the tomograms revealed the pore-filling directly after impregnation and after either conventional drying (Figure 1a) or freeze-drying (Figure 1b) within the same carrier particle and is schematically represented in Figure 1c,d. The 892

dx.doi.org/10.1021/cm3037845 | Chem. Mater. 2013, 25, 890−896

Chemistry of Materials

Article

Figure 3. Effect of drying on Co3O4 nanoparticle distribution in a Co−Pt on mesoporous silica catalyst after calcination in a flow of 1% NO/N2. a, Prepared via conventional drying at 60 °C. b, Prepared via freeze-drying. a,b, Scale bar: 50 nm. Close-up of boxed aream, scale bar: 10 nm. 2.7 nm thick slices through reconstructions obtained by electron tomography.

Figure 4. Effect of drying on NiO nanoparticle distribution in mesoporous silica after calcination in a flow of 1% NO/N2. a, Prepared via conventional drying at 120 °C in oven. b, Prepared via freeze-drying. c, Close-up of boxed area in a; NiO nanoparticles are accentuated in pink for clarity. d, Close-up of boxed area in b; NiO nanoparticles are accentuated in blue for clarity. a−d, 2.7 nm thick slices through reconstructions from electron tomography. Scale bars are 50 nm (a,b) and 10 nm (c,d). Please refer to movies Figure 4a_NiOconv.mov and Figure 4b_NiOfreeze.mov to observe the 3D distribution of nanoparticles (see SI1, Supporting Information).

conventional drying the pores were partially filled with plugs of cobalt nitrate salt with empty and filled pore sections occurring randomly throughout the particle (Figure 2a). Furthermore, by comparing the pores in the rectangular and triangular areas marked in slices B and C of Figure 2a it becomes clear that both empty and filled sections exist within a single pore. In contrast, freeze-drying (Figure 2b) led to a uniform salt distribution. The tomogram shows that the pore system was homogeneously filled, with exception of the pores at the outer rim of the particle. The metal loading of the two samples was equal; however, the freeze-dried particle (Figure 2b) has a higher degree of pore filling. This may be due to variations in the

metal loading of individual particles as discussed above for the in situ experiment. An alternative explanation is that the salt residue after either conventional drying or freeze-drying is present in the form of nanoparticles and networks thereof with differences in packing below the resolution limit of Cryo-ET. This would appear as if the freeze dried precursor salt would have a lower density, thus occupying a larger volume, than the conventionally dried salt. This explanation is qualitatively supported by the lower contrast of the salt phase with respect to the silica wall seen in Figure 2b vs 2a. Cryo-ET analysis of several SBA-15 particles prepared by conventional lab-scale ex situ drying showed a significant variance in the level of salt 893

dx.doi.org/10.1021/cm3037845 | Chem. Mater. 2013, 25, 890−896

Chemistry of Materials

Article

Figure 5. Particle size distribution of Co3O4−Pt/SBA-15 and NiO/SBA-15 catalysts prepared by conventional drying or freeze-drying and calcination in a flow of NO/N2 as determined from the electron tomograms.

redistribution, which was attributed to variations in the local drying rate (see SI3, Figure S4, Supporting Information). Indeed, in some cases plug-wise filling was observed (cf. Figure 2a), while other cases showed deposition of small aggregates throughout the pore system (cf. Figure 1a). In contrast, labscale prepared freeze-dried SBA-15 particles all showed the same uniform salt distribution. Rational synthesis of catalysts with controlled nanoparticle distribution was performed by combining freeze-drying for a uniform salt distribution with an NO-assisted nitrate decomposition method.19 As opposed to more standard procedures which use N2 or air, calcination in an NO/N2 gas flow restricts mobility of the precursor phase during decomposition, allowing control over the final nanoparticle size.39 Figure 3 illustrates the 3D nanoparticle distribution in a platinum promoted cobalt catalyst, which is an industrially relevant catalyst system for the Fischer−Tropsch synthesis.21 After conventional drying and cobalt nitrate decomposition in NO/N2, the Co3O4 nanoparticles were concentrated within several pores (Figure 3a), leaving the remaining pores empty as is typical for hydrated transition metal nitrates calcined in a NO-containing atmosphere.32 After freeze-drying, the distribution of cobalt oxide nanoparticles was more uniform (Figure 3b), even though a few pores at the outer rim of the particle were empty. This clearly demonstrates the potential of freeze-drying for the preparation of more sintering resistant ex-nitrate prepared cobalt Fischer−Tropsch catalysts. The versatility of the method was demonstrated by extending the scope from the Co−Pt system to ex-nitrate preparation of NiO nanoparticles via the same synthesis route. For NiO, the use of freeze-drying yielded a uniform distribution of the nanoparticles and showed a significant improvement compared to conventional drying (Figure 4). Figure 4d suggests gaps in the silica pore walls. Although the pore walls of SBA-15 are porous, the gaps observed were relatively large. Nevertheless, this was found not to be representative of the freeze-drying procedure but rather linked to the SBA-15 parent pore structure. The structural integrity of the support was confirmed by N2-physisorption (see SI4, Supporting Information). Significantly, ET showed that an even smaller average particle size was obtained for NiO (2.7−3.5 nm) than for Co3O4−Pt (5.1−5.5 nm) (Figure 3−5, Table 1). For both Co3O4−Pt and NiO, freeze-drying resulted in slightly smaller nanoparticles, as can be observed in the histograms displayed in Figure 5. The

Table 1. Average Crystallite and Average Particle Sizes As Determined with XRD and Electron Tomography (ET), respectively, for Nickel Oxide and Platinum Promoted Cobalt Oxide Catalysts after Calcination in a Flow of 1% NO/N2 or in a Flow of N2 via Impregnation and Conventional Drying (CD) or Freeze Drying (FD) calcination in 1% NO/N2

a

catalyst

dXRD (nm)

CD-NiO/SBA-15 FD-NiO/SBA-15 CD-Co3O4-Pt/SBA-15 FD-Co3O4-Pt/SBA-15

3.4 2.8 5.2 4.6

calcination in N2

dET (nm)

dXRD (nm)

± ± ± ±

8.4 6.5 9.7 7.4

3.5 2.7 5.5 5.1

0.6 0.5 1.3a 1.2a

ET particle size determined after reduction at 450 °C and passivation.

particle sizes were determined from slices cutting through the tomogram, as the high metal loading and overlapping pores hindered particle size analysis by TEM. The particle sizes corresponded very well to the average crystallite size determined by XRD, while the formation of some nickel or cobalt (hydro)silicates during calcination cannot be excluded.40,41 Reducibility of metal nanoparticles on oxidic supports is potentially hampered by such amorphous oxidic species, as they are difficult to reduce. However, previous XANES analysis of Co3O4−Pt/SiO2 catalysts with similar particle sizes showed that degrees of reduction of >90% were achieved using conventional reduction conditions.21 Furthermore, the diffraction patterns of the catalysts prepared by conventional drying and freeze-drying showed similar intensity, indicating similar crystallinity of the metal oxides. Considering the low salt-support interaction in nitrate−silica systems which generally causes aggregation,23 the small nanoparticle sizes obtained by calcination in NO/N2 are remarkable. Freeze-drying yielded generally a uniform distribution of the NiO nanoparticles significantly different from conventional drying (Figure 4). The NO-assisted calcination led to very small NiO nanoparticles (2.8 nm, Figure 4d) while by more standard calcination larger crystallites were formed (6.5 nm, Table 1). The effect of N2 on the aggregation of the active phase during the decomposition of hydrated Co(NO3)2 and Ni(NO3)2 is well-known.19,39,42 Indeed, the calcination of nickel nitrate under N2 after conventional drying caused the nickel oxide nanoparticles (average diameter 8.4 nm, Table 1) to form 10−30 nm clusters in between which the pores appear 894

dx.doi.org/10.1021/cm3037845 | Chem. Mater. 2013, 25, 890−896

Chemistry of Materials

Article

Figure 6. TEM micrographs of NiO nanoparticles on SBA-15 after calcination in an N2 flow. a, Prepared by conventional drying at 120 °C. Scale bar: 100 nm. b, Prepared by freeze-drying. Scale bar: 100 nm. c, Close-up of boxed area in a. Scale bar: 25 nm. d, Close-up of boxed area in b. Scale bar: 25 nm.

lab scale experiments. On the basis of these findings a rational synthesis approach for catalysts with more uniform nanoparticle distributions was composed. By combining freeze-drying with the NO-assisted decomposition method, 3 nm NiO particles were obtained with a uniform distribution over the support. In contrast, conventional procedures utilize only a fraction of the pores of the carrier. The versatility of the method was illustrated by the preparation of platinum promoted cobalt catalysts for the Fischer−Tropsch synthesis and a nickel based hydrogenation catalyst. Our results not only have implications for the field of catalysis, they also open an avenue to exploit the exceptional properties of supported nanoparticles for applications in other fields of materials science.

empty (Figure 6a,c). Still, both the nickel (Figure 6b,d) as well as the Pt promoted cobalt catalyst (SI5, Figure S6, Supporting Information) prepared by freeze-drying showed a lesser extent of clustering as compared to conventional drying. Clearly, freeze-drying improves the nanoparticle distribution in the carrier, even when standard calcination conditions induce mobility of the active phase. This can be explained by the hydration state of the nitrate precursor which was found to be directly dependent on the applied drying method. For cobalt nitrate 3−7 H2O molecules per mole of Co remained after conventional drying, depending on the details of handling. However, samples freeze-dried contained only 0.3 residual water molecules per mole of cobalt. For Ni(NO3)2, which is known to partially decompose to Ni3(NO3)2(OH)4 at 120 °C, the observed residual weight loss corresponded to ∼75% Ni3(NO3)2(OH)4 and ∼25% Ni(NO3)2·6H2O, in line with previously reported results.39 However, after freeze-drying only 0.7 residual molecules of water remained. So, after conventional drying, cobalt nitrate retained crystal water, while freeze-dried samples were almost fully dehydrated.

■

■

ASSOCIATED CONTENT

S Supporting Information *

The following media files are available: movie Figure 1_3Dinsitu.mov showing the 3D reconstruction models of empty pore sections, movies Figure 2a_Conv.mov and Figure 2b_Freeze.mov showing the 3D distribution of salt within the pores after drying and the ET reconstructions of NiO/SBA-15 catalysts in movies Figure 4a_NiOconv.mov and Figure 4b_NiOfreeze.mov. PDF containing descriptions of the movies, cryo-ET and image analysis, cryo-ET reconstructions, structural characterization, and TEM. This material is available free of charge via the Internet at http://pubs.acs.org.

CONCLUSIONS

Detailed insight into the link between the elementary steps of the aqueous synthesis of supported nanoparticles was obtained with (cryo-) electron tomography. It allowed for the first time a direct comparison of the impact of drying on the local salt distribution within a single support particle. While conventional drying led to a nonuniform deposition of salt, freeze-drying immobilized the precursor phase leading to a uniform distribution. This concept derived from in situ drying experiments on individual support particles was confirmed by

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected] (K.P.dJ.), [email protected] (N.A.J.M.S.). 895

dx.doi.org/10.1021/cm3037845 | Chem. Mater. 2013, 25, 890−896

Chemistry of Materials

Article

Notes

(25) Sietsma, J. R. A.; Meeldijk, J. D.; Versluijs-Helder, M.; Broersma, A.; van Dillen, A. J.; de Jongh, P. E.; de Jong, K. P. Chem. Mater. 2008, 20, 2921. (26) Marceau, E.; Che, M.; Cejka, J.; Zukal, A. ChemCatChem 2010, 2, 413. (27) Eggenhuisen, T. M.; den Breejen, J. P.; Verdoes, D.; de Jongh, P. E.; de Jong, K. P. J. Am. Chem. Soc. 2010, 132, 18318. (28) Kong, L.; Wei, W.; Zhao, Q.; Wang, J.-Q.; Wan, Y. ACS Catal. 2012, 2, 2577. (29) Lekhal, A.; Glasser, B. J.; Khinast, J. G. Chem. Eng. Sci. 2001, 56, 4473. (30) van de Loosdrecht, J.; Barradas, S.; Caricato, E. A.; Ngwenya, N. G.; Nkwanyana, P. S.; Rawat, M. A. S.; Sigwebela, B. H.; van Berge, P. J.; Visagie, J. L. Top. Catal. 2003, 26, 121. (31) Eggenhuisen, T. M.; Munnik, P.; Talsma, H.; de Jongh, P. E.; De Jong, K. P. J. Catal. 2012, 297, 306. (32) Friedrich, H.; Sietsma, J. R. A.; de Jongh, P. E.; Verkleij, A. J.; de Jong, K. P. J. Am. Chem. Soc. 2007, 129, 10249. (33) Nudelman, F.; Pieterse, K.; George, A.; Bomans, P. H. H.; Friedrich, H.; Brylka, L. J.; Hilbers, P. A. J.; de With, G.; Sommerdijk, N. A. J. M. Nat. Mater. 2010, 9, 1004. (34) Nudelman, F.; de With, G.; Sommerdijk, N. A. J. M. Soft Matter 2011, 7, 17. (35) Sayari, A.; Han, B. H.; Yang, Y. J. Am. Chem. Soc. 2004, 126, 14348. (36) Crocker, J. C.; Grier, D. G. J. Colloid Interface Sci. 1996, 179, 298. (37) Kremer, J. R.; Mastronarde, D. N.; McIntosh, J. R. J. Struct. Biol. 1996, 116, 71. (38) Eggenhuisen, T. M.; van Steenbergen, M. J.; Talsma, H.; de Jongh, P. E.; de Jong, K. P. J. Phys. Chem. C 2009, 113, 16785. (39) Wolters, M.; Munnik, P.; Bitter, J. H.; de Jongh, P. E.; de Jong, K. P. J. Phys. Chem. C 2011, 115, 3332. (40) Louis, C.; Cheng, Z. X.; Che, M. J. Phys. Chem. 1993, 97, 5703. (41) Trujillano, R.; Lambert, J. F.; Louis, C. J. Phys. Chem. C 2008, 112, 18551. (42) Sietsma, J. R. A.; Friedrich, H.; Broersma, A.; Versluijs-Helder, M.; van Dillen, A. J.; de Jongh, P. E.; de Jong, K. P. J. Catal. 2008, 260, 227.

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS M. van Steenbergen and H. Talsma (Utrecht University) and H. de Waard (University of Groningen) are acknowledged for their technical support and useful discussions on freeze-drying. The Electron Microscopy Unit at Utrecht University is acknowledged for their excellent facilities. J. Hilhorst (Utrecht University) is acknowledged for the image analysis performed for Figure 1. T.M.E. and K.P.dJ. thank The Netherlands Organization for Scientific Research (NWO) for funding (TOP 700.57.341), P.E.dJ. thanks NWO-Vidi, K.P.dJ, J.Z., and H.F. thank NRSCC for funding, and N.S. and H.F. thank NWO-Vici.

■

REFERENCES

(1) Steele, B. C. H.; Heinzel, A. Nature 2001, 414, 345. (2) Raimondi, F.; Scherer, G. G.; Kotz, R.; Wokaun, A. Angew. Chem., Int. Ed. 2005, 44, 2190. (3) Astruc, D.; Lu, F.; Aranzaes, J. R. Angew. Chem., Int. Ed. 2005, 44, 7852. (4) Lu, A. H.; Salabas, E. L.; Schuth, F. Angew. Chem., Int. Ed. 2007, 46, 1222. (5) Lei, Y.; Mehmood, F.; Lee, S.; Greeley, J.; Lee, B.; Seifert, S.; Winans, R. E.; Elam, J. W.; Meyer, R. J.; Redfern, P. C.; Teschner, D.; Schlogl, R.; Pellin, M. J.; Curtiss, L. A.; Vajda, S. Science 2010, 328, 224. (6) Turner, M.; Golovko, V. B.; Vaughan, O. P. H.; Abdulkin, P.; Berenguer-Murcia, A.; Tikhov, M. S.; Johnson, B. F. G.; Lambert, R. M. Nature 2008, 454, 981. (7) Bartholomew, C. H. Appl. Catal., A 2001, 212, 17. (8) Saib, A. M.; Moodley, D. J.; Ciobica, I. M.; Hauman, M. M.; Sigwebela, B. H.; Weststrate, C. J.; Niemantsverdriet, J. W.; van de Loosdrecht, J. Catal. Today 2010, 154, 271. (9) Xin, H. L.; Mundy, J. A.; Liu, Z.; Cabezas, R.; Hovden, R.; Kourkoutis, L. F.; Zhang, J.; Subramanian, N. P.; Makharia, R.; Wagner, F. T.; Muller, D. A. Nano Lett. 2011, 12, 490. (10) Bell, A. T. Science 2003, 299, 1688. (11) Somorjai, G. A.; Park, J. Y. Top. Catal. 2008, 49, 126. (12) Torres Galvis, H. M.; Bitter, J. H.; Khare, C. B.; Ruitenbeek, M.; Dugulan, A. I.; de Jong, K. P. Science 2012, 335, 835. (13) Prieto, G.; Zecevic, J.; Friedrich, H.; de Jong, K. P.; de Jongh, P. E. Nat. Mater. 2012, 12, 34. (14) Chai, J. A.; Huo, F. W.; Zheng, Z. J.; Giam, L. R.; Shim, W.; Mirkin, C. A. Proc. Natl. Acad. Sci. U.S.A. 2010, 107, 20202. (15) Kraus, T.; Malaquin, L.; Schmid, H.; Riess, W.; Spencer, N. D.; Wolf, H. Nat. Nanotechnol. 2007, 2, 570. (16) Houk, R. J. T.; Jacobs, B. W.; El Gabaly, F.; Chang, N. N.; Talin, A. A.; Graham, D. D.; House, S. D.; Robertson, I. M.; Allendorf, M. D. Nano Lett. 2009, 9, 3413. (17) Subbaraman, R.; Tripkovic, D.; Strmcnik, D.; Chang, K.-C.; Uchimura, M.; Paulikas, A. P.; Stamenkovic, V.; Markovic, N. M. Science 2011, 334, 1256. (18) Neimark, A. V.; Kheifez, L. I.; Fenelonov, V. B. Ind. Eng. Chem. Prod. Res. Dev. 1981, 20, 439. (19) Sietsma, J. R. A.; Meeldijk, J. D.; den Breejen, J. P.; VersluijsHelder, M.; van Dillen, A. J.; de Jongh, P. E.; de Jong, K. P. Angew. Chem., Int. Ed. 2007, 46, 4547. (20) Arslan, I.; Walmsley, J. C.; Rytter, E.; Bergene, E.; Midgley, P. A. J. Am. Chem. Soc. 2008, 130, 5716. (21) den Breejen, J. P.; Sietsma, J. R. A.; Friedrich, H.; Bitter, J. H.; de Jong, K. P. J. Catal. 2010, 270, 146. (22) van der Meer, J.; Bardez, I.; Bart, F.; Albouy, P. A.; Wallez, G.; Davidson, A. Microporous Mesoporous Mater. 2009, 118, 183. (23) Jiao, L.; Regalbuto, J. R. J. Catal. 2008, 260, 329. (24) van Dillen, A. J.; Terorde, R. J. A. M.; Lensveld, D. J.; Geus, J. W.; de Jong, K. P. J. Catal. 2003, 216, 257. 896

dx.doi.org/10.1021/cm3037845 | Chem. Mater. 2013, 25, 890−896