Organometallic Route to Surface-Modified ZnO Nanoparticles Suitable

Jun 20, 2012 - ABSTRACT: Well-defined ZnO nanoparticles with bound carboxylate surface-functionalization and narrow size distribution were prepared vi...
0 downloads 0 Views 3MB Size
Article pubs.acs.org/cm

Organometallic Route to Surface-Modified ZnO Nanoparticles Suitable for In Situ Nanocomposite Synthesis: Bound Carboxylate Stoichiometry Controls Particle Size or Surface Coverage Katherine L. Orchard,† Milo S. P. Shaffer,*,† and Charlotte K. Williams*,† †

Department of Chemistry, Imperial College London, London SW7 2AZ, U.K. S Supporting Information *

ABSTRACT: Well-defined ZnO nanoparticles with bound carboxylate surface-functionalization and narrow size distribution were prepared via an efficient organometallic hydrolysis route, occurring at ambient temperature and without postsynthesis refinement. Depending on the reaction conditions, the nanoparticles’ degree of surface coverage or diameter was controlled independently. The method was used for the in situ preparation of well-dispersed ZnO/epoxy resin nanocomposites. KEYWORDS: ZnO nanoparticles, polymers, organic−inorganic hybrid composites, hydrolysis organo-zinc



INTRODUCTION ZnO is an important functional material;1 in the bulk, it is a wide bandgap semiconductor, applied in lasing, field effect transistors, gas sensors, and in photovoltaics.2−5 The optoelectronic properties of ZnO nanoparticles depend on the particle size, defect concentration, and surface species; thus control of these factors is of paramount importance.6 Zinc oxide and mixed metal oxides are also important in catalysis, for example in the synthesis of methanol from syn-gas; once again, the surface chemistry critically impacts performance.7−9 Control of surface chemistry is also critical in the preparation of ZnO-polymer nanocomposites.10,11 These materials are applied in electronics, for example as dielectrics,12 diodes,13 and as the active layer in photovoltaics.5 They are also applied to protect polymers from UV radiation,14,15 and as luminescent materials. In order to optimize the nanocomposite bulk properties, it is usual to maximise the dispersion of ZnO nanoparticles in the polymer matrix, thereby increasing the particle−polymer interfacial area.16 The development of fabrication methods that minimize particle aggregation remains a key research goal; in particular, in situ routes whereby the nanoparticles are synthesized directly within the polymer/ prepolymer mixture are attractive as they can minimize hard agglomerates often formed during handling of particles synthesized ex situ.17 In situ nanocomposite syntheses require preparations for nano-ZnO that are compatible with the polymer chemistry, i.e. those which operate under mild conditions, are tolerant of chemical functionality, and which generate few products. One common route to ZnO nanoparticles is via the alkaline hydrolysis of zinc halides,18−21 often accomplished in alcoholic solvents. Such ‘sol−gel’ syntheses have been used to prepare nanocomposites using certain thermoplastic matrices;14 however, the method is not generally applicable due to the presence © 2012 American Chemical Society

of salt byproduct and the lack of compatibility with basesensitive polymer functionalities, common in reactive thermosets. Indeed, we have previously established that sol−gel routes are incompatible with in situ syntheses in epoxy resins, due to the base reacting with the epoxy prepolymer.22 The preparation of zinc oxide nanoparticles by the hydrolysis of organo-zinc precursors, a field pioneered by Chaudret and co-workers, is a highly attractive method as it operates under ambient conditions and, if diethyl zinc is applied, generates a gaseous byproduct.6,23−26 Alkylzinc alkoxide precursors have also been extensively investigated to prepare ZnO nanoparticles7,8,27−31; a recent thorough study32 was carried out into the mechanism of hydrolysis of [MeZn(OtBu)]4. The organometallic hydrolyses are chemically tolerant toward a variety of polymer matrices; our research group recently reported the “one-pot”, in situ preparation of bulk ZnO-epoxy resin nanocomposites with improved thermal conductivity, via the hydrolysis of diethylzinc.22 Others have hydrolyzed organo-zinc precursors in conjugated polymers to produce efficient photovoltaics;5 the synthetic route obviates byproduct contamination and avoids polymer degradation. Effective modification of the nanoparticle surfaces, either in nanocomposites or as materials in their own right, still remains highly challenging.33 The most common method to control ZnO surface chemistry is via the application of surfactant ligands which are usually applied in great excess. Excess ligands or reactive small molecules are particularly undesirable in nanocomposites where weak interfaces and plasticization by free surfactant significantly reduce performance. We reasoned that a modification of the conventional organo-zinc hydrolysis Received: February 2, 2012 Revised: May 10, 2012 Published: June 20, 2012 2443

dx.doi.org/10.1021/cm300058d | Chem. Mater. 2012, 24, 2443−2448

Chemistry of Materials

Article

version 1.40 g (W. Rasband, National Institutes of Health); particles were measured manually. X-ray Diffractometry (XRD) was performed using an X’Pert Pro diffractometer (PANalytical B. V., The Netherlands) and X’Pert Data Collector software, version 2.2b. The instrument was used in the theta/theta reflection mode, fitted with a nickel filter, 0.04 rad Soller slit, 10 mm mask, 1/4° fixed divergence slit, and 1/2° fixed antiscatter slit. Samples were analyzed with a step size of 0.0042°, at a scanning speed of 0.028° s−1. The diffraction patterns were analyzed using Fityk (version 0.9.0; Marcin Wojdyr, 2010): the peaks were fitted to a Pseudo-Voigt function using the Levenberg−Marquardt algorithm, and the particle size was calculated using the fitted full-width half-maximum. Zinc Oxide Nanoparticle Synthesis. In general, the nanoparticle precursor solutions were prepared such that the total concentration of zinc species was 0.15 M; for the rapid hydrolysis method, the total concentration of zinc species was 0.15 M after the addition of the water solution. For both the rapid and slow hydrolysis methods, two precursor stock solutions were prepared: (A) 1.0 M ZnEt2 in toluene and (B) 1.0 M [Zn] in toluene. Stock solution B was prepared by mixing ZnEt2 and Zn(OOCR)2 in toluene in the appropriate ratio to give a carboxylate loading [OOCR]/[Zn] of 0.20 or 0.33 and equilibrating for 2−4 h (OOCR = hexanoate and dodecanoate) or 16 h (OOCR = stearate). The hexanoate and dodecanoate solutions were clear after 2 h; the stearate solutions required gentle heating to achieve full dissolution. After the equilibration period, the solutions were made up to volume using a volumetric flask. Proportions of each stock solution were then mixed and made up to the correct volume with the appropriate solvent in an inert atmosphere glovebox (hexane or toluene; see Table S1, SI). For both types of nanoparticle synthesis, after the required reaction time, the particles were precipitated using excess acetone and centrifuged (10000 rpm; 15 min), and the liquid was decanted. The particles were washed by resuspending in a small amount of fresh toluene and reprecipitating with excess acetone. Centrifugation was repeated, and the wet product paste was dried in vacuo for 16 h. Typical Synthesis, Rapid Hydrolysis Method. A solution of distilled water in HPLC grade acetone (0.86 M, 2.30 mL, 1.98 mmol H2O) was added dropwise to the precursor solution (total addition time 4 min); a gel stage was observed after approximately 75% of the water solution was added, lasting 5−10 s. The solution was stirred for a further 2 h. Slow Hydrolysis Method. In order to maintain a constant, reproducible humidity atmosphere and to minimize solvent loss during the experiment, the samples were placed into a glass tank which had been equilibrated for at least 18 h with a saturated salt solution, and the lid was sealed with vacuum grease. Two relative humidities were chosen: 32% (CaCl2•6H2O) and 11% (LiCl).38 The humidity was verified with a digital hygrometer during equilibration and was found not to change during removal and replacement of the lid. The reactions were carried out at 20 ± 2 °C which corresponds to a change in absolute humidity of ±0.5 g−3 (±0.02 mmoldm−3) for the 32% solution and ±0.2 g−3 (±0.01 mmoldm−3) for the 11% solution, which was considered negligible.38 The precursor solutions were prepared and sealed in vials. The vials were brought out of the glovebox, the lids were removed, and the vials were placed in the controlled humidity chamber. The solutions were allowed to equilibrate for 15 min before stirring for a further 24 h. No gel stage was observed. Nanocomposite Synthesis. (See Table S2, SI, for reagent quantities.) The epoxy prepolymer (DGEBA) was dried under vacuum at 65 °C for 6 h prior to addition of the organometallic zinc precursor solution. For the surface-modified particle composites, the loading of carboxylate [OOCR]/[Zn] was 0.2. The ZnEt2/“EtZn(OOCR)” precursor mixtures were equilibrated for 18 h prior to use, and both the stearate and benzoate derivatives gave clear solutions after gentle heating. The precursor solution was added to the dried DGEBA and stirred to form a homogeneous solution. A solution of distilled water (2 equiv) in acetone (2−4 mL) was added dropwise, and the mixture

method could enable efficient ZnO surface capping without the need for any excess surfactant. The hydrolysis of a mixture of organo-zinc reagent, R2Zn, and an additional, substoichiometric quantity of a zinc precursor with both a hydrolyzable organoligand (R) and a nonhydrolyzable ligand (X), RZnX, should lead to ZnO nanoparticles with full surface coverage. The approach should be useful for simple, modified-nanoparticle production but is particularly applicable for in situ nanocomposite systems. Related concepts have been applied using modified silicon (and other) alkoxide complexes to prepare amorphous, hybrid inorganic−organic silica materials34 but have not been developed for zinc oxide synthesis. Ethylzinc carboxylates were chosen as the heteroleptic precursor species as the carboxylate group was predicted to be unreactive toward water. We have previously studied the (solvent-dependent) solution behavior and solid state structures of a series of ethylzinc carboxylates.19,35 Herein, we describe the hydrolysis of mixtures of diethyl zinc and ethylzinc carboxylates to prepare carboxylate-capped ZnO nanoparticles and demonstrate the successful application of the reaction to the in situ synthesis of well-dispersed nanoparticles within an epoxy resin.



EXPERIMENTAL SECTION

All reactions involving air-sensitive reagents were conducted under a nitrogen atmosphere using either standard Schlenk techniques or in a nitrogen-filled glovebox. Materials. Solvents were distilled from either sodium and stored under nitrogen. Unless otherwise stated, solvents were freshly degassed prior to use by performing at least three freeze−pump− thaw cycles. Diethylzinc was purchased from Aldrich, vacuum distilled, and stored in an ampule, under nitrogen, at −38 °C. Diethyl zinc, in common with many other organometallic compounds, is pyrophoric and must be handled with appropriate precautions. The epoxy resin system used was produced by Huntsman Advanced Materials and consisted of a diglycidyl ether of bisphenol A (DGEBA) and an amine hardener, thermally cured. The DGEBA resin (CY219) was donated by Moldlife Ltd. The hardener (XB3473) was purchased from Robnor Resins. The composites were cured in silicone molds, prepared using Elastosil M4641 (Amber Composites) Parts A and B. Zinc bis(dodecanoate) and zinc bis(hexanoate) were prepared according to the method described by Berkesi et al.,36 from reaction with the respective acid with Zn(OH)2 in boiling octane. Zinc bis(stearate) was prepared by the method described by Ekwunife et al.,37 from reaction of in situ prepared potassium stearate with zinc chloride in an ethanol solution. All zinc bis(carboxylates) were found to be of sufficient purity by elemental analysis. Unless otherwise stated, all other reagents were purchased from commercial suppliers and used as received. Instrumentation. Infrared (IR) spectroscopy was carried out using a Perkin-Elmer Spectrum 100 Fourier Transform IR spectrometer: powder samples were analyzed using the Attenuated Total Reflection (ATR) accessory. Thermogravimetric analysis (TGA) was carried out using a Perkin-Elmer Pyris 1 TGA machine, under a flow of dry air, from 50 to 800 °C, at a heating rate of 10 °C/min. Optical absorption spectra were collected on a Perkin-Elmer Lambda 950 spectrophotometer using toluene as the solvent and a nanoparticle concentration of 10 μg mL−1. Photoluminescence measurements were carried out using a CaryEclipse spectrometer, using chloroform as the solvent and a nanoparticle concentration of 1 μg mL−1. High-resolution Transmission Electron Microscopy (HRTEM) was carried out using a JEOL 2010 microscope. Nanoparticle samples were drop-cast (CHCl3 solution) onto 300-mesh, holey carbon-coated copper films (Agar Scientific) and imaged at an operating voltage of 200 kV. Thin slices (∼90 nm thick) of nanocomposite samples were cut using a diamond knife microtome, placed directly onto 300-mesh copper films (Agar Scientific) and imaged at an operating voltage of 100 kV. Digital images were analyzed for particle sizing using the software ImageJ, 2444

dx.doi.org/10.1021/cm300058d | Chem. Mater. 2012, 24, 2443−2448

Chemistry of Materials

Article

stirred for 2 h. Volatiles were removed in vacuo for 10 min at 25 °C and a further 30 min at 65 °C. Hardener was then added and mixed well, and the mixture was degassed at 65 °C for 15 min and transferred to the molds. The DGEBA/ZnO/hardener mixture (100:23 DGEBA:hardener, by weight) was degassed in the molds by heating to 100 °C under vacuum and then cured by heating under air at 120 °C for 2 h and 140 °C for 2 h.

nm; XRD; see Figure S1 (b), SI), whereas the capped particles showed no change in size, morphology, or dispersibility on heating, demonstrating the stability imparted by the capping ligands. XRD analysis (Figure 1(a)) confirmed the formation of crystalline ZnO (wurtzite) with extra, broad peaks at low angle



RESULTS AND DISCUSSION The general form of the nanoparticle synthesis is shown in Scheme 1. Scheme 1. General Form of the Nanoparticle Synthesis

Briefly, the organometallic zinc precursors were mixed in dry toluene, and, under inert atmosphere, diluted distilled water in acetone (one equivalent per ethyl group; two equivalents per diethylzinc) was added dropwise over 4 min. Growth was continued for 2 h before isolation of the particles by centrifugation. The size and quality of the nanoparticles was equivalent (TEM, XRD, UV−vis) regardless of whether the ethylzinc carboxylate precursor was isolated and dried prior to use or generated in situ from diethylzinc and the zinc bis(carboxylate); for the majority of experiments, the in situ approach was preferred for ease of handling and in order to maintain the “one-pot” nature of the method. In the absence of carboxylate ([OOCR]/[Zn] = 0), an opaque nanoparticle suspension was formed that precipitated instantly when stirring was discontinued. The particles could be resuspended in chloroform by sonication and were found to be largely agglomerated, forming aggregates on the order of 50 nm−2 μm, by TEM (Figure S1 (a), SI); regions between the particles had an appearance similar to that of sintered particles, indicating that the particles had aggregated and become permanently fused during synthesis. The introduction of carboxylate after the synthesis did not generate individualized primary nanoparticles, even with sonication, as the initial agglomeration process is irreversible. The formation of agglomerated particles, in the absence of surfactant, is consistent with the findings of Kahn et al.23−25 The standard surface-functionalized nanoparticle synthesis conditions utilized a precursor ratio of diethylzinc to zinc bis(stearate) (ZnSA2) of 9:1, in toluene, resulting in a carboxylate loading, [OOCR]/[Zn], of 0.2 (R = (CH2)16CH3; n = 0.11 in Scheme 1). The functionalized nanoparticles formed rapidly and remained stably dispersed, at a consistent size (TEM, UV/vis), in the reaction growth mixture, for several hours; significant ripening only occurred at longer reaction times (Figure S2, SI). The resulting nanoparticle suspensions (typical reaction 2 h) could be reversibly dissolved and precipitated on addition of excess toluene and acetone, respectively. Once isolated and dried, the particles formed a free-flowing powder (89% yield) that could be easily redissolved in toluene or chloroform by gentle heating, to form clear solutions; toluene solutions (1 mgmL−1) precipitated a white film after standing for two days, but the film was easily redissolved on heating. Uncapped particles (described above) were found to increase in size on heating the dried powder to 100 °C for 6 h (average crystallite size increased from 4 to 10

Figure 1. Example data for standard product (9:1 diethylzinc:zinc bis(stearate) (ZnSA2), in toluene): a) X-ray diffraction pattern, confirming ZnO formation (reference lines from PDF 036-1451, ICDD PDF4+ database); extra peak due to organic component marked with “*”; b) UV−vis spectrum, toluene solution; c) representative TEM image with (inset) individual particle showing (101) lattice planes; d) representative electron diffraction pattern; e) size distribution histogram, measured from multiple TEM images.

due to the organic component (vide inf ra). TEM showed the particles to be roughly spherical, nonagglomerated, and nearly monodisperse, with a narrow size-distribution (standard deviation, σ = 15%, Figure 1(c)). The average size of the particles was estimated from the XRD pattern using the Scherrer equation39 and from the UV−vis absorption spectrum using the relation derived by Meulenkamp (see the SI for details);18 the measured size was very reproducibly 3−4 nm with excellent agreement between the size measurement techniques (mean of 10 independent syntheses was 3.6 ± 0.4 nm by TEM, 3.1 ± 0.4 nm by XRD, and 3.6 ± 0.2 nm by UV− vis spectroscopy). The photoluminescence spectrum of the particles (excitation wavelength, λ = 360 nm) showed a broad peak centered at 520 nm (Figure S3, SI), which is consistent with the green luminescence previously observed for ZnO nanoparticles, attributed to the presence of oxygen vacancies.23−25 It has been suggested that low temperature, organometallic methods can lead to a higher vacancy concentration that may be helpful for catalytic applications.7,8 Infrared (IR) spectroscopy confirmed the presence of carboxylate groups; the carboxylate antisymmetric and symmetric vibrational modes were broadened and shifted compared to ZnSA2, at 1550 and 1418 cm−1, respectively (1537 and 1398 cm−1, respectively for ZnSA2). The difference, Δ, between the two modes, 132 cm−1, indicates that the stearate groups adopt a bridging arrangement on the nanoparticle surface.40 The presence of a high carboxylate content in the TGA and IR analyses, combined with the solubility of the 2445

dx.doi.org/10.1021/cm300058d | Chem. Mater. 2012, 24, 2443−2448

Chemistry of Materials

Article

ZnSA2 were also present between 5−25° 2θ. Diffuse peaks in the range 18−28°, associated with lateral chain packing,41 appear gradually as surface coverage increases and the stearate chains become more ordered. The IR spectrum for sample a matched that of ZnSA2 (Figure S6, SI), supporting the XRD assignment. The IR carboxylate absorbances broadened and shifted to higher frequency with decreasing carboxylate loading, suggesting that the predominant form of the stearate groups in these samples was a bound, bridging state on the nanoparticle surfaces, rather than as Zn(SA)2. Qualitatively, dispersibility of the dried nanoparticles decreased with decreasing carboxylate content, correlating with the proposed decrease in surface coverage. It can be concluded that, for these rapid hydrolysis experiments, the particle size is primarily determined by the nucleation step (number of nuclei formed); the particles grow to 3−4 nm with the carboxylate groups being either distributed over the surfaces to give a (partial) surface coverage ([OOCR]/ [Zn] ≤ 0.2) or segregated into fully capped nanoparticles and excess Zn(OOCR)2 ([OOCR]/[Zn] > 0.2). The ability to systematically control surface coverage is useful for many applications where surface access is important, such as catalysis. In contrast, when the hydrolysis rate is slowed down, by allowing the water to diffuse in from a vapor of controlled humidity, particle size was dependent on carboxylate loading (Figure 3). For a carboxylate loading of 0.2, soluble, largely

nanoparticles in organic solvents, was taken as good evidence for surface-functionalization. A broad, weak absorption centered at approximately 3400 cm−1 was also present in the IR spectrum, which may be due either to small amounts of residual moisture or to additional surface functionalization with −OH groups. No meaningful information could be gathered from the 1H NMR spectrum of the particles, which showed only very broad peaks in the alkyl region. For a constant carboxylate loading of 0.2, decreasing the carboxylate alkyl chain length suggested a slight increase in average particle size but the effect was not significant (stearate, dodecanoate, hexanoate; average size = 3.6 ± 0.2, 3.9 ± 0.4, and 4.1 ± 0.4 nm, respectively; TEM), and no change in particle morphology was observed. TGA indicated a significant organic content, reducing with chain length, as expected (38, 32, and 20 wt % for stearate, dodecanoate, and hexanoate, respectively; expected content 41, 33, and 22 wt %, respectively; Figure S4, SI). The percentage organic component was used to estimate the surface coverage (surface area capped by carboxylate ligands) based on the measured average particle size (see the SI for calculation); the estimated surface coverage was above 80% for each alkyl chain length investigated. It was hoped that, by varying the loading of carboxylate, the size of the particles could be controlled; however, for the standard synthesis conditions, the particle size was found to be largely independent of carboxylate loading: the overall size change was less than 1 nm (Figure 2; TEM images and

Figure 2. Average particle diameter (d), measured by TEM (squares), XRD (diamonds), and UV−vis spectroscopy (circles), and calculated surface coverage (S; gray triangles) for ZnO nanoparticles prepared via the standard hydrolysis method, with varying carboxylate loadings. The error bars shown represent the standard deviation for multiple repeat reactions; calculated error bars for S (surface coverage) are smaller than the data points as shown (±2%).

Figure 3. Effect of carboxylate loading and relative humidity on the average particle diameter (d) of ZnO nanoparticles prepared by the slow hydrolysis method (measured by XRD); triangles: 32%RH, squares: 11%RH. The error bars represent the standard error in the mean for multiple repeat reactions of each data point. Unlike the rapid hydrolysis experiments, the particle size was independent of solvent and carboxylate alkyl chain length (within error); therefore, the figure collates all data collected.

associated size distributions are available in Figure S7, SI). The implied corollary, that the degree of surface coverage can be adjusted, is nevertheless interesting. The total carboxylate content (TGA) matched the intended loading for each sample which, based on the measured average particle size, corresponds to a decrease in surface coverage with decreasing loading (Figure 2, triangles). The calculated coverage for the 0.33 loading sample was greater than 100% (154%), a result that can be attributed to the presence of excess Zn(SA)2. Indeed, XRD (Figure S5, SI) showed crystalline ZnO (wurtzite) for all carboxylate loadings, but at the highest loading of carboxylate (sample a, [SA]/[Zn] = 0.33) distinctive sharp peaks matching

spherical, 4 nm nanoparticles were obtained. However, at lower carboxylate loadings, particle size increased, consistent with growth limited by carboxylate termination rather than nucleation. The effect was stronger at lower relative humidity (i.e., reaction rate); however, polydispersity of particle size and shape also increased with decreasing carboxylate loading, and the particles were difficult to redisperse in organic solvents once dried. This dependence on humidity may explain a recent, though brief, report26 of a concentration dependent size effect in the case of the hydrolysis of dicyclohexyl zinc in the presence of amine-terminated PEG, by atmospheric moisture. 2446

dx.doi.org/10.1021/cm300058d | Chem. Mater. 2012, 24, 2443−2448

Chemistry of Materials

Article

stoichiometry dependent on solvent conditions.19,35 The arrangement of these complexes during the hydrolysis of the excess diethyl zinc must involve significant carboxylate mobility. The resulting nanoparticles show high crystallinity and the ability to form colloidal solutions reversibly. If the reaction is performed under nucleation-controlled conditions, near monodisperse, functionalized ZnO particles, of size 3−4 nm, are formed with controlled degrees of surface coverage; if the reaction is slowed down to be limited by carboxylate termination, nanoparticle size is controlled. The new method occurs under conditions very well suited to the fabrication of well-dispersed functional ZnO-epoxy resin nanocomposites and should be generally suitable for a wide range of polymeric nanocomposites. Similar approaches for the in situ production of functionalized nanoparticles could be applied to other metals, as well as other chalcogenides (using, for example, H2S or H2Se in place of water42). Combinations of different metal precursors, as widely used in thin film growth, may provide a convenient route to surface functionalized, doped nanoparticles, relevant to (opto)electronic applications.

To demonstrate the compatibility of the new organometallic route with cross-linking resins, in situ ZnO/epoxy resin nanocomposites were prepared. A zinc precursor mixture with carboxylate loading of 0.2 (as the standard synthesis conditions) was added to an epoxy prepolymer of the diglycidyl ether of bisphenol A (DGEBA) and hydrolyzed. The volatile compounds were removed, and the ZnO/epoxy prepolymer was mixed with an amine hardener, before casting and curing. In our previous report, the use of diethylzinc as a single precursor resulted in nanocomposites with a homogeneous distribution of individual particles and small agglomerates (Figure 4 (a) and (d)).22 When stearate-capped ZnO



ASSOCIATED CONTENT

* Supporting Information S

In detail, all experimental protocols, the method for estimation of particle surface coverage from TGA, the method for the estimation of particle size from UV−vis spectroscopy, UV−vis and PL spectra, TGA thermograms, XRD data, and IR spectra, as well as TEM and associated size distributions. This material is available free of charge via the Internet at http://pubs.acs.org.

Figure 4. Representative TEM images of ZnO/epoxy nanocomposites of comparable loading (1 vol%): a) and d) with uncapped particles (ZnEt2 used as single precursor), b) and e) stearate-capped particles, and c) and f) benzoate-capped particles. Scale bars: a), b), and c) 50 nm; d) 20 nm; e) and f) 10 nm.



AUTHOR INFORMATION

Corresponding Author

*C.K.W.: e-mail [email protected]. M.S.P.S.: e-mail m.shaff[email protected].

nanocomposites were prepared, large agglomerated regions of nanoparticles formed (Figure 4 (b) and (e)): although both the stearate-capped nanoparticles and the epoxy prepolymer are soluble in toluene, they phase segregate on curing. As shown in Figure 5(e), although the particles are agglomerated, they are separated from one another (not permanently fused), indicating segregation due to the surface-modifying alkyl chains. When the capping agent was benzoate, well-dispersed and well-distributed ZnO nanoparticles were formed (Figure 4 (c) and (f)). The nanoparticle sizes in both Figure 4 (e) and (f) are measured to be 3−4 nm, in good agreement with the other nanoparticles synthesized using this method. It is proposed that the use of the aromatic benzyl group improves compatibility between the epoxy prepolymer (which has aromatic functionality in the polymer chain) and the surface capping groups on the zinc oxide nanoparticles, thus improving dispersion quality in the nanocomposite.

Notes

The authors declare no competing financial interest.

■ ■

ACKNOWLEDGMENTS The EPSRC are acknowledged for funding the research (Grants: EP/H046380/1, EP/C544838/1). REFERENCES

(1) Ozgur, U.; Alivov, Y. I.; Lui, C.; Teke, A.; Reshchikov, M. A.; Dogan, S.; Avrutin, V. J. Appl. Phys. 2005, 98, 041301. (2) Sun, B.; Sirringhaus, H. Nano Lett. 2005, 5, 2408. (3) Polarz, S.; Roy, A.; Lehmann, M.; Driess, M.; Kruis, F. E.; Hoffmann, A.; Zimmer, P. Adv. Funct. Mater. 2007, 17, 1385. (4) Fallert, J.; Dietz, R. J. B.; Sartor, J.; Schneider, D.; Klingshirn, C.; Kalt, H. Nat. Photonics 2009, 3, 279. (5) Beek, W. J. E.; Slooff, L. H.; Wienk, M. M.; Kroon, J. M.; Janssen, R. A. J. Adv. Funct. Mater. 2005, 15, 1703. (6) Kahn, M.; Glaria, A.; Pages, C.; Monge, M.; Macary, L. S.; Maisonnat, A.; Chaudret, B. J. Mater. Chem. 2009, 19, 4044. (7) Polarz, S.; Neues, F.; van den Berg, M. W. E.; Grunert, W.; Khodeir, L. J. Am. Chem. Soc. 2005, 127, 12028. (8) Polarz, S.; Strunk, J.; Ischenko, V.; van den Berg, M. W. E.; Hinrichsen, O.; Muhler, M.; Driess, M. Angew. Chem., Int. Ed. 2006, 45, 2965. (9) Schimpf, S.; Rittermeier, A.; Zhang, X. N.; Li, Z. A.; Spasova, M.; van den Berg, M. W. E.; Farle, M.; Wang, Y. M.; Fischer, R. A.; Muhler, M. ChemCatChem 2010, 2, 214. (10) Hussain, F.; Hojjati, M.; Okamoto, M.; Gorga, R. E. J. Compos. Mater. 2006, 40, 1511. (11) Winey, K. I.; Vaia, R. A. MRS Bull. 2007, 32, 314.



CONCLUSIONS In conclusion, we report a versatile new method for synthesizing ZnO nanoparticles, that avoids free or excess surfactant and operates at ambient temperature, in organic solvents. The method takes advantage of heteroleptic organometallic zinc complexes featuring hydrolyzable alkyl groups and nonhydrolyzable carboxylate coligands to deliver the carboxylate groups to the particle surfaces. The growth mechanism must involve a number of complex steps worthy of further study; diethyl zinc and zinc bis(carboxylates) are known to react directly to form well-defined compounds with a specific 2447

dx.doi.org/10.1021/cm300058d | Chem. Mater. 2012, 24, 2443−2448

Chemistry of Materials

Article

(12) Singha, S.; Thomas, M. J. IEEE Trans. Dielectr. Electr. Insul. 2008, 15, 12. (13) Verbakel, F.; Meskers, S. C. J.; Janssen, R. A. J. J. Phys. Chem. C 2007, 111, 10150. (14) Li, Y.-Q.; Fu, S.-Y.; Mai, Y.-W. Polymer 2006, 47, 2127. (15) Ge, J.; Zeng, X.; Tao, X.; Li, X.; Shen, Z.; Yun, J.; Chen, J. J. Appl. Polym. Sci. 2010, 118, 1507. (16) Caseri, W. Mater. Sci. Technol. 2006, 22, 807. (17) Rong, M. Z.; Zhang, M. Q.; Ruan, W. H. Mater. Sci. Technol. 2006, 22, 787. (18) Meulenkamp, E. A. J. Phys. Chem. B 1998, 102, 5566. (19) Orchard, K. L.; Harris, J. E.; White, A. J. P.; Shaffer, M. S. P.; Williams, C. K. Organometallics 2011, 30, 2223. (20) Spanhel, L. J. Sol-Gel Sci. Technol. 2006, 39, 7. (21) Niederberger, M. Acc. Chem. Res. 2007, 40, 793. (22) Gonzalez-Campo, A.; Orchard, K. L.; Sato, N.; Shaffer, M. S. P.; Williams, C. K. Chem. Commun. 2009, 27, 4034. (23) Kahn, M.; Monge, M.; Snoeck, E.; Maisonnat, A.; Chaudret, B. Small 2005, 1, 221. (24) Kahn, M. L.; Cardinal, T.; Bosquet, B.; Monge, M.; Jubera, V.; Chaudret, B. ChemPhysChem 2006, 7, 2392. (25) Monge, M.; Kahn, M. L.; Maisonnat, A.; Chaudret, B. Angew. Chem., Int. Ed. 2003, 41, 5321. (26) Rubio-Garcia, J.; Coppel, Y.; Lecante, P.; Mingotaud, C.; Chaudret, B.; Gauffre, F.; Kahn, M. L. Chem. Commun. 2011, 47, 988. (27) Driess, M.; Merz, K.; Rell, S. Eur. J. Inorg. Chem. 2000, 2000, 2517. (28) Hambrock, J.; Rabe, S.; Merz, K.; Birkner, A.; Wohlfart, A.; Fischer, R. A.; Driess, M. J. Mater. Chem. 2003, 13, 1731. (29) Polarz, S.; Regenspurger, R.; Hartmann, J. Angew. Chem., Int. Ed. 2007, 46, 2426. (30) Lizandara-Pueyo, C.; Siroky, S.; Wagner, M. R.; Hoffmann, A.; Reparaz, J. S.; Lehmann, M.; Polarz, S. Adv. Funct. Mater. 2011, 21, 295. (31) Boyle, T. J.; Bunge, S. D.; Andrews, N. L.; Matzan, L. E.; Sieg, K.; Rodriguez, M. A.; Headley, T. J. Chem. Mater. 2004, 16, 3279. (32) Lizandara-Pueyo, C.; van den Berg, M. W. E.; De Toni, A.; Goes, T.; Polarz, S. J. Am. Chem. Soc. 2008, 130, 16601. (33) Li, S.; Toprak, M. S.; Jo, Y. S.; Dobson, J.; Kim, K.; Muhammed, M. Adv. Mater. 2007, 19, 4347. (34) Sanchez, C.; Belleville, P.; Popall, M.; Nicole, L. Chem. Soc. Rev. 2011, 40, 696. (35) Orchard, K. L.; White, A. J. P.; Shaffer, M. S. P.; Williams, C. K. Organometallics 2009, 58, 5828. (36) Berkesi, O.; Dreveni, I.; Andor, J. A. Inorg. Chim. Acta 1992, 195, 169. (37) Ekwunife, M. E.; Nwachukwu, M. U.; Rinehart, F. P.; Sime, S. J. J. Chem. Soc., Faraday Trans. 1975, 71, 1432. (38) Lide, D. R. The CRC Handbook of Chemistry and Physics, 73rd ed.; CRC Press: 1992. (39) Cullity, B. D. Elements of X-ray Diffraction; 2nd ed.; AddisonWesley: 1978. (40) Nakamoto, K. In Infrared and Raman Spectroscopy of Inorganic and Coordination Compounds Part B: Applications on Coordination, Organometallic, and Bioinorganic Chemistry; 1997; p 59. (41) Taylor, R. A.; Ellis, H. A.; Maragh, P. T.; White, N. A. S. J. Mol. Struct. 2006, 787, 113. (42) Gallagher, D.; Heady, W. E.; Racz, J. M.; Bhargava, R. N. J. Mater. Res. 1995, 10, 870.

2448

dx.doi.org/10.1021/cm300058d | Chem. Mater. 2012, 24, 2443−2448