Iridium Nanocrystal Synthesis and Surface Coating-Dependent

shape distribution. The Ir nanocrystals were tested for their ability to catalyze the hydrogenation of 1-decene as a model reaction. ... Stefanos ...
2 downloads 0 Views 824KB Size
Download PDF
NANO LETTERS

Iridium Nanocrystal Synthesis and Surface Coating-Dependent Catalytic Activity

2005 Vol. 5, No. 7 1203-1207

Cynthia A. Stowell and Brian A. Korgel* Department of Chemical Engineering, Texas Materials Institute and Center for Nano and Molecular Science and Technology, The UniVersity of Texas, Austin, Texas 78712-1062 Received April 7, 2005; Revised Manuscript Received May 14, 2005

ABSTRACT Iridium (Ir) nanocrystals were synthesized by reducing (methylcyclopentadienyl)(1,5-cyclooctadiene)Ir with hexadecanediol in the presence of four different capping ligand combinations: oleic acid and oleylamine, trioctylphosphine (TOP), tetraoctylammonium bromide (TOAB), and tetraoctylphosphonium bromide (TOPB). The oleic acid/oleylamine-capped nanocrystals were of the highest quality, with the narrowest size and shape distribution. The Ir nanocrystals were tested for their ability to catalyze the hydrogenation of 1-decene as a model reaction. The oleic acid/oleylamine and TOP-capped nanocrystals were both catalytically dead. TOAB and TOPB-coated nanocrystals both catalyzed 1-decene hydrogenation, with the TOPB-coated nanocrystals exhibiting the highest turnover frequencies. Recycling through several catalytic reactions increased the catalytic activity, presumably as a result of ligand desorption and increased exposure of the metal surface, with ligand desorption eventually leading to precipitation and significantly decreased activity.

The catalytic properties of solid-supported metal nanocrystals have been studied for decades because of their high surface area-to-volume ratios and ability to catalyze many industrially important reactions;1-5 however, some recent reports have found that nanoscale materials can in some cases give rise to new unique catalytic activity.6 These findings, enabled to a large extent by significant recent advances in metal nanocrystal synthetic control,7-10 have renewed interest in developing a better fundamental understanding of metal nanocrystal catalytic properties by studying metal nanocrystals as “model” materials.10 Iridium (Ir) is a particularly interesting transition-metal catalyst; it is well known for promoting alkene, aldehyde, and ketone hydrogenation,11-18 with more reactions possible when used in a bimetallic system.19,20 Furthermore, there have been reports of unusually high catalytic activity from molecular Ir clusters.21 A few reports of Ir nanocrystal synthesis exist in the literature;12,15-17,22 however, existing synthetic methods for Ir nanocrystals are not well developed, and better-quality particles are needed for catalytic studies. Here we demonstrate the synthesis of high-quality Ir nanocrystals by reducing (methylcyclopentadienyl)(1,5-cyclooctadiene)Ir or ((MeCp)Ir(COD)), an organometallic compound developed as a chemical vapor deposition (CVD) precursor for Ir films, in the presence of organic capping ligands to * Corresponding author. E-mail: [email protected]. Phone (512) 471-5633. Fax: (512) 471-7060. 10.1021/nl050648f CCC: $30.25 Published on Web 06/02/2005

© 2005 American Chemical Society

control the particle size.23 The nanocrystal synthesis was carried out with four different capping ligands: oleic acid/ oleylamine, trioctylphosphine (TOP), tetraoctylammonium bromide (TOAB), and tetraoctylphosphonium bromide (TOPB). The ligand chemistry greatly influences the nanocrystal size and size distribution and the catalytic actiVity of the nanocrystals. Oleic acid/oleylamine-stabilized Ir nanocrystals were synthesized under nitrogen by hexadecanediol reduction of the Ir precursor in dioctyl ether in the presence of oleic acid and oleylamine at 290 °C following the general procedures developed by Sun et al. for FePt (see Supporting Information for synthetic details).24 After the synthesis, the Ir nanocrystals were precipitated with hexanol and isolated by brief centrifugation. The supernatant was discarded, and the nanocrystals were redispersed in chloroform and reprecipitated with hexanol and centrifuged. This “washing” step was repeated again to ensure removal of organic and molecular byproducts from the nanocrystals. The nanocrystals redisperse in various nonpolar organics solvents including chloroform, cyclohexane, and toluene. The oleic acid/oleylamine-coated Ir nanocrystals obtained after purification ranged from 1.5 to 5 nm in diameter. Sizeselective precipitation using chloroform/ethanol as the solvent/ antisolvent pair could be used to narrow the size distribution if desired. Figure 1 shows transmission electron microscopy (TEM), X-ray diffraction (XRD), and small-angle X-ray

Figure 2. Conversion of 1-decene to decane at 75 °C and 3 psig of H2 in the presence of dispersed Ir nanocrystals: (]) 2 nm oleic acid/oleylamine-coated; (×) 4 nm oleic acid/oleylamine-coated; (+) TOP-coated; (O) 1.5 nm TOAB-coated, TOF ) 5 s-1; (0) 5 nm TOPB-coated, TOF ) 270 s-1.

Figure 1c shows SAXS data for oleic acid/oleylaminecoated Ir nanocrystals dispersed in hexane. Unlike TEM, which can be used to probe the size of only a very small number of nanocrystals in the sample, SAXS provides an accurate measure of the average nanocrystal size and size distribution by probing >1010 nanocrystals.25-28 In SAXS, the scattered X-ray intensity, I(q), is measured as a function of the wave vector, q ) (sin(θ)4π)/λ, which depends on the scattering angle 2θ and the X-ray wavelength (λ ) 0.154 nm). In a nanocrystal dispersion, the particles are generally sufficiently dilute and nonaggregating such that scattering results from “isolated” nanocrystals of radius R and I(q) is a function of only the size distribution, N(R), and the shape factor for a sphere, P(qR) ) [3(sin(qR) - (qR)cos(qR))/ (qR)3]2:25-28 I(q) ∝

Figure 1. Oleic acid/oleylamine-coated Ir nanocrystals. (A) TEM image of ∼4 nm diameter size-selected Ir nanocrystals. (B) XRD Ir nanocrystals, with peaks corresponding to fcc Ir (JCPDS file 461044). (C) SAXS of two different sizes of Ir nanocrystals isolated by size-selective precipitation from the same reaction: (0) Rav ) 1.09 nm, σ ) (0.283 nm; (O) Rav ) 1.92 nm, σ ) (0.635 nm; the curves correspond to the best fits of eq 1 to the data that were used to obtain Rav and σ. (D) HRTEM of a 5 nm oleic acid/ oleylamine-coated Ir nanocrystal.

scattering (SAXS) data of the Ir nanocrystals. TEM and XRD confirm that the nanocrystals are crystalline face-centered cubic (fcc) Ir. The lattice spacings observed by TEM match fcc Ir. Analysis of the XRD peak broadening using the Scherrer equation gives an estimate of the particle diameter of 5.3 nm, which corresponds well with the average diameter of 5 nm determined from TEM images. 1204

∫N(R) P(qR)R6 dR

(1)

Assuming that the nanocrystal size distribution is Gaussian, N(R) ) 1/(σx2π) exp[-(R - Rav)2/2σ2], eq 1 can be fit to the data in Figure 1c to obtain the average radius, Rav, and the standard deviation, σ. The Ir nanocrystals measured in Figure 1c by SAXS have diameters of Rav ) 1.09 nm (σ ) (0.283 nm) and 1.92 nm (σ ) (0.635 nm). Batch hydrogenation reactions were carried out in a 100 mL round-bottom flask with a 1000:1 decene/Ir nanocrystal mass ratio, with the liquid volume totaling 15 mL under 3 psig (1.22 × 105 Pa) of H2 at 75 °C. As the reaction proceeded, aliquots were removed from the flask every minute for the first 20 min, every 5 min for the next 40 min, and every 10 min after that. The aliquots were placed in screw cap gas chromatography vials, and the nanocrystals were later recovered by vacuum evaporation of the liquid. The decene/decane conversion was measured using either a Hewlett-Packard 5890 series II or a Finnigan MAT GCQ gas chromatography mass spectrometer (GC/MS). Figure 2 shows the 1-decene/decane conversion in the presence of 2 and 4 nm oleic acid/oleylamine-coated Ir nanocrystals. It is clear that the oleic acid/oleylamine-coated Ir nanocrystals did not catalyze 1-decene hydrogenation. Oleic acid/oleylamine particles of 1.5 and 5 nm diameter were also tested and did not convert a measurable amount of decene after 5 h. Nano Lett., Vol. 5, No. 7, 2005

Figure 3. TEM images of Ir nanocrystals synthesized with different capping ligands: (A) TOAB (1.5-3 nm diameter); (B) TOPB (2-5 nm diameter); and (C) TOP (10-100 nm diameter).

reaction conditions due to poor capping, leading to diminished reactivity over time due to a decreased surface areato-volume ratio.30 There are several capping ligands, such as trioctylphosphine (TOP), tetraoctylammonium bromide (TOAB), and tetraphosphonium bromide (TOPB), that are known to be “weak” binders to metal and semiconductor nanocrystal surfaces.31,32 Because there is little predictive knowledge about the influence of capping ligand chemistry on metal nanocrystal catalysis, Ir nanocrystals were synthesized using all three of these ligands and then studied for their catalytic activity. TOP-coated Ir nanocrystals were synthesized in neat TOP under nitrogen at 290 °C, with the addition of hexadecanediol to reduce the (MeCp)Ir(COD). TOAB and TOPB-coated Ir nanocrystals were synthesized in dioctyl ether under nitrogen at 270 °C, again using hexadecanediol as a reducing agent for (MeCp)Ir(COD). All three capping ligands stabilized Ir nanocrystals; however, the quality of the nanocrystals varied widely. Figure 3 shows TEM images of TOP, TOAB, and TOPB-coated Ir nanocrystals. The TOP-capped particles were the poorest qualitysextremely polydisperse, with diameters ranging from 10 to 100 nm and irregular shapes. TOABcapped Ir nanocrystals were crystalline and relatively sizemonodisperse, with diameters ranging from 1.5 to 3 nm. TOPB-capped Ir nanocrystals were also crystalline but were larger and slightly more polydisperse than the TOAB-capped nanocrystals, averaging 4 nm in diameter. The capping ligand chemistry significantly affects the Ir nanocrystal catalytic activity. Figure 2 shows the conversion of 1-decene to decane as a function of reaction time in the presence of TOP-, TOAB-, and TOPB-capped Ir nanocrystals. TOP-capped Ir did not promote 1-decene hydrogenation, but both TOAB- and TOPB-coated Ir nanocrystals were catalytically active. TOAB-capped nanocrystals of ∼1.5 nm diameter converted 42% of the 1-decene after 230 min and 5 nm diameter TOPB-coated Ir nanocrystals converted 100% of the decene after 60 min. Turnover frequencies (TOF), defined as the number of molecules n reacted at each available catalytic site per time t, TOF )

The oleic acid/oleylamine-coated Ir nanocrystals were of relatively high quality in terms of crystallinity, narrow size and shape distribution, and good dispersibility; however, they exhibited no ability to catalyze the hydrogenation of 1-decene to decane. Because Ir is a well-known hydrogrenation catalyst and recent reports have demonstrated the catalytic activity of Ir nanocrystals in ionic liquids,16-18 the absence of catalytic activity of oleic acid/oleylamine-coated Ir nanocrystals prompted the exploration of weaker-binding capping ligands to determine if the capping ligands were in fact preventing catalysis on the nanocrystal surface. Recently, Li and ElSayed noted that very good capping ligand stabilization appears to diminish catalytic nanocrystal activity in some cases,29 and Narayanan and El-Sayed showed that nanocrystals can also ripen and precipitate under the catalytic Nano Lett., Vol. 5, No. 7, 2005

( )( ) 1 dn Nact dt

(2)

were calculated assuming that all surface atoms are active with Nact ) AP/(AUCn), where AP is the surface area of the average particle size, AUC is the surface area of an Ir unit cell face, and n is the number of Ir atoms in a unit cell face. Although only a small portion of surface Ir atoms can actually serve as catalytic active sitessmany will be bonded to capping ligands and unavailable for catalysissit is common practice to take the total number of surface atoms as the number of active catalytic sites when the value is not known.33 Using eq 2 and the average nanocrystal diameter measured using TEM, TOF ) 4 s-1 for ∼1.5 nm TOABcoated nanocrystals, and TOF ) 270 s-1 for ∼5 nm TOPBcoated nanocrystals. 1205

Figure 5. 1-Decene hydrogenation using recycled TOPB-coated Ir nanocrystals: (O) first reaction, TOF ) 0 s-1; (0) second reaction, TOF ) 24 s-1; (]) third reaction, TOF ) 36 s-1; (4) fifth reaction, TOF ) 119 s-1; (+) sixth reaction, TOF ) 75 s-1.

Figure 4. TOAB-coated Ir nanocrystals after 1-decene hydrogenation reactions. (A) Hydrogenation of 1-decene to decane as a function of catalyst recycling: (O) first reaction, TOF ) 5 s-1; (0) second reaction, TOF ) 14 s-1; (]) third reaction, TOF ) 50 s-1; (×) fourth reaction, TOF ) 124 s-1; (4) fifth reaction, TOF ) 38 s-1. (B) TEM images of TOAB-coated Ir nanocrystals before catalysis and (C) after four hydrogenation reactions.

Recycling the nanocrystals through several hydrogenation reactions changed their catalytic activity. Figure 4 shows 1-decene conversion to decane in the presence of 1.5 nm diameter TOAB-coated Ir nanocrystals used in sequential reactions at 75 °C for 230 min. The nanocrystal catalytic activity increased with subsequent reactions until the fifth reaction cycle, when the catalytic activity dropped significantly. During the first reaction cycle, the nanocrystals achieve only a 42% conversion of 1-decene and a TOF of 4 s-1. A second hydrogenation yielded 100% conversion in the same 230 min period with TOF ) 13 s-1. The third and 1206

fourth cycles led to 100% conversion in shorter times and significantly higher TOFs: 50 and 124 s-1, respectively. Additionally, the lag time in the initial stages of the reaction shortened with each subsequent reaction. The increase in TOF indicates that the nanocrystal surfaces become increasingly catalytically active with each reaction, most likely as a result of ligand desorption and reduced surface capping. It is well known that capping ligands on nanocrystal surfaces are in dynamic equilibrium with free solvated ligands, and ligand desorption is always a concern when considering nanocrystal dispersion stability and particle aggregation. The ligand desorption kinetics depend on the binding strength between the ligand functional group and the inorganic nanocrystal surface. Because of their relatively weak binding, TOAB and TOPB can be stripped from the nanocrystal surfaces by excessive precipitation with antisolvent (5 to 10 precipitations), preventing redispersibility, whereas oleic acid/oleylamine are bound relatively strongly to the Ir nanocrystal surfaces and could be reprecipitated many times and still be redispersed. In the fifth reaction, the catalytic activity decreased significantly, reaching 100% conversion after 150 min with a TOF ) 38 s-1. A comparison of TEM images of nanocrystals before and after catalysis (Figures 4B and 4C) reveals that the nanocrystals exhibit an increasing amount of particle aggregation. TEM also showed that the Ir particles increase slightly in size (because of Ostwald ripening) during each reaction, which was accounted for in the TOF calculations in each reaction. The reduced TOF after the fifth cycle (the average Ir particle size is 3 nm) is consistent with the decreased available surface area per particle for catalysis due to aggregation. For comparison, the TOF calculated by considering the available surface area-to-volume ratio of particles as “aggregates” with an average “diameter” of 10 nm gives a value of TOF ) 127 s-1, which is the same as that measured in the fourth reaction cycle. The nanocrystals appear to become better catalysts as the capping ligand surface coverage decreases, until the dispersion becomes unstable and the particles aggregate. The same increase in catalytic activity was observed for TOPB-coated Ir nanocrystals used in subsequent reactions, as shown in Figure 5. The TOF increases with each subsequent reaction. In comparison to the nanocrystals used Nano Lett., Vol. 5, No. 7, 2005

in Figure 2, these TOPB-coated Ir nanocrystals were not initially catalytically active. However, recycling led to catalytic activity that improved from 0 to 119 s-1 after the fifth reaction. Also similar to the TOAB-coated nanocrystals, the TOF decreased after five reactions, slowing to 75 s-1 on the sixth reaction. The catalytic activity appears to be a strong function of ligand coverage and also ligand chemistry. The oleic acidcapped Ir nanocrystals were not found to be catalytically active in any of our experiments. TOPB-coated nanocrystals were more reactive than TOAB-coated nanocrystals, which appears to relate to the difference in ligand binding strength: Ir-N has a shorter equilibrium bond length than Ir-P, indicating that Ir-N has a larger binding energy and stronger attachment to the surface.34,35 The expected difference in ligand binding strength is also consistent with the observed smaller diameter of TOAB-capped nanocrystals relative to that of TOPB-capped nanocrystals. Essentially, more Ir surface atoms are exposed for catalysis with TOPB coating. Ir nanocrystal catalysis of 1-decene hydrogenation to decane is extremely sensitive to capping ligand chemistry and ligand coverage. “Good” capping ligands (i.e., those that stabilize robust nanocrystals with very narrow size distributions) appear to be poor choices for catalytic applications. However, the nanocrystals must be of reasonably high quality with good dispersibility in multiple reaction cycles, so the ligands must be strong enough binders to stabilize nanocrystals but “weak” enough to provide reactant access to the metal surface. These studies reveal that capping ligands play a central role in the catalytic activity of transition-metal nanocrystals and cannot be ignored. Acknowledgment. We thank Medhi Moini for assistance with the GC/MS measurements. We also thank the NSF, the Welch Foundation, and the Texas Higher Education Coordinating Board for financial support. Supporting Information Available: Detailed descriptions of synthetic materials characterization methods. This material is available free of charge via the Internet at http:// pubs.acs.org. References (1) Kralik, M.; Biffis, A. J. Mol. Catal. A: Chem. 2001, 177, 113-138. (2) Lewis, L. N. Chem. ReV. 1993, 93, 2693-2730.

Nano Lett., Vol. 5, No. 7, 2005

(3) Aiken, J. D., III; Finke, R. G. J. Mol. Catal. A: Chem. 1999, 145, 1-44. (4) Bonnemann, H.; Brijoux, W.; Brinkmann, R.; Dinjus, E.; Joussen, T.; Korall, B. Angew. Chem., Int. Ed. Engl. 1991, 30, 1312-1314. (5) Roucoux, A.; Schulz, J.; Patin, H. Chem. ReV. 2002, 102, 37573778. (6) Valden, M.; Lai, X.; Goodman, D. W. Science 1998, 281, 16471650. (7) Narayanan, R.; El-Sayed, M. A. Nano Lett. 2004, 4, 1343-1348. (8) Narayanan, R.; El-Sayed, M. A. J. Phys. Chem. B 2004, 108, 57265733. (9) Kim, S. W.; Park, J.; Jang, Y.; Chung, Y.; Hwang, S.; Hyeon, T.; Kim, Y. W. Nano Lett. 2003, 3, 1289-1291. (10) Son, S. U.; Jang, Y.; Yoon, K. Y.; Kang, E.; Hyeon, T. Nano Lett. 2004, 4, 1147-1151. (11) Lin, Y.; Finke, R. G. Inorg. Chem. 1994, 33, 4891-4910. (12) Yee, C. K.; Jordan, R.; Ulman, A.; White, H.; King, A.; Rafailovich, M.; Sokolov, J. Langmuir 1999, 15, 3486-3491. (13) Hayek, K.; Goller, H.; Penner, S.; Rupprechter, G.; Zimmermann, C. Catal. Lett. 2004, 92, 1-9. (14) Lin, Y.; Nomiya, K.; Finke, R. G. Inorg. Chem. 1993, 32, 60406045. (15) Mevellec, V.; Roucoux, A.; Ramirez, E.; Phillippot, K.; Chaudret, B. AdV. Synth. Catal. 2004, 346, 72-76. (16) Fonseca, G. S.; Umpierre, A. P.; Fichtner, P. F. P.; Teixeira, S. R.; Dupont, J. Chem.sEur. J. 2003, 9, 3263-3269. (17) Fonseca, G. S.; Scholten, J. D.; Dupont, J. Synlett 2004, 1525-1528. (18) Dupont, J.; Fonseca, G. S.; Umpierre, A. P.; Fichtner, P. F. P.; Teixeira, S. R. J. Am. Chem. Soc. 2002, 124, 4228-4229. (19) Okada, M.; Nakamura, M.; Moritani, K.; Kasai, T. Surf. Sci. 2003, 523, 218-230. (20) Shiblisma, N. Int. J. Hydrogen Energy 1993, 18, 141-147. (21) Anderson, J. R.; Howe, R. F. Nature 1977, 268, 129-130. (22) Shah, P. S.; Husain, S.; Johnston, K. P.; Korgel, B. A. J. Phys. Chem. B 2001, 105, 9433-9440. (23) Hoke, J. B.; Stern, E. W.; Murray, H. H. J. Mater. Chem. 1991, 1, 551-554. (24) Sun, S.; Murray, C. B.; Weller, D.; Folks, L.; Moser, A. Science 2000, 287, 1989-1992. (25) Small-Angle X-ray Scattering; Glatter, O.; Kratky, O., Eds.; Academic Press: New York, 1982. (26) Guinier, A.; Fournet, G. Small-Angle Scattering of X-rays; Wiley: New York, 1955. (27) X-ray Characterization of Materials; Lifshing, E., Ed.; WileyVCH: New York, 1999. (28) Korgel, B. A.; Fitzmaurice, D. Phys. ReV. B 1999, 59, 14191-14201. (29) Li, Y.; El-Sayed, M. A. J. Phys. Chem. B 2001, 105, 8938-8943. (30) Narayanan, R.; El-Sayed, M. A. J. Am. Chem. Soc. 2003, 125, 83408347. (31) Saunders: A. E.; Sigman, M. B.; Korgel, B. A. J. Phys. Chem. B 2004, 108, 193-199. (32) Mattoussi, H.; Cumming, A. W.; Murray, C. B.; Bawendi, M. G.; Ober, R. J. Chem. Phys. 1996, 105, 9890-9896. (33) Dumonteil, C.; Lacroix, M.; Geantet, C.; Jobic, H.; Breysse, M. J. Catal. 1999, 187, 464-473. (34) Dahlenburg, L.; Gotz, R. Eur. J. Inorg. Chem. 2004, 888-905. (35) Housecroft, C. E.; O’Neill, M. E.; Wade, K.; Smith, B. C. J. Organomet. Chem. 1981, 213, 35-43.

NL050648F

1207