Highly Shape-Selective Synthesis, Silica Coating, Self-Assembly, and

Dec 14, 2009 - ... Department of Physics, Nanjing University,. Nanjing 210093, China and ‡Department of Physics, Boise State University, Boise, Idah...
0 downloads 0 Views 2MB Size
pubs.acs.org/Langmuir © 2009 American Chemical Society

Highly Shape-Selective Synthesis, Silica Coating, Self-Assembly, and Magnetic Hydrogen Sensing of Hematite Nanoparticles Jianhui Zhang,*,†,‡ Aaron Thurber,‡ Charles Hanna,‡ and Alex Punnoose‡ †

National Laboratory of Solid State Microstructures, Department of Physics, Nanjing University, Nanjing 210093, China and ‡Department of Physics, Boise State University, Boise, Idaho 83725 Received September 18, 2009. Revised Manuscript Received November 17, 2009

The open forced hydrolysis method and controllable silica growth based on bound water to polyvinylpyrrolidone molecules have been developed for the highly shape (including rhombohedra, semispheres, and rods) selective synthesis, selfassembly, and uniform silica coating (in the unprecedented range of 5-200 nm) of hematite nanoparticles. The open system realizes the direct short-range self-assembly of hematite semispheres in their growth process. The bound water method has been extended to coat gold nanoparticles with tunable silica shell and directly assemble the cores into one-dimensional, dimer, and trimer nanostructures during the coating process. The silica coating increases the particle stability and monodispersity even as hematite is modified into ferromagnetic Fe3O4. The hematite@silica core-shell spheres are assembled into long-range ordered structures with considerable photonic bandgap for the first time due to their high monodispersity. By exploiting the hematite antiferromagnetism caused by the superexchange interaction via intervening oxygen ions that are sensitive to hydrogen, a novel hydrogen sensing based on magnetization variations is achieved in the hematite assemblies. Weakening the antiferromagnetism by reducing the hematite size and/or covering the hematite surface by silica coating suppresses the sensitivity to hydrogen, showing that the antiferromagnetic spin variations on the hematite surface are responsible for the gas sensing.

Introduction Iron oxide has been widely applied in magnetic recording media, catalysts, pigments, gas sensors, optical devices, rechargeable lithium batteries, and electromagnetic devices.1-9 To enhance the performance in existing applications and to explore novel applications, morphology-selective synthesis of iron oxide has been intensively studied because its physicochemical properties are strongly size- and shape-dependent. For example, a hydrothermal method was modified by using polyisobutylene bissuccinimide (L113B) or span80 as a soft template to prepare hematite (R-Fe2O3) nanotubes and nanorods, where shape-dependent magnetic properties were observed.9 Size-dependent magnetic properties and electrochemical properties were also revealed in hematite nanorods (with gradients in size and porosity) made by a hydrothermal method improved by adopting a high concentration of inorganic salts and high temperature.10 With the assistance of polyvinylpyrrolidone (PVP), Zheng et al. made single-crystalline quasicubic hematite nanoparticles that are superior catalysts to the other forms of nano- and microsized *Corresponding author. E-mail: [email protected]. (1) Bronstein, L. M.; Huang, X.; Retrum, J.; Schmucker, A.; Pink, M.; Stein, B. D.; Dragnea, B. Mater. Chem. 2007, 19, 3624. (2) Morales, M. P.; Gonzalez-Carre~no, T.; Serna, C. J. J. Mater. Res. 1992, 7, 2538. (3) Sadakane, M.; Horiuchi, T.; Kato, N.; Takahashi, C.; Ueda, W. Chem. Mater. 2007, 19, 5779. (4) Kiwi, J.; C€artzel, M. J. Chem. Soc., Faraday Trans. 1 1987, 83, 1101. (5) Syiroky, K.; Jiresova, J.; Hudec, L. O. Thin Solid Films 1994, 245, 211. (6) Neri, G.; Bonavita, A.; Galvagno, S.; Siciliano, P.; Capone, S. Sens. Actuators, B 2002, 82, 40. (7) Matijevic, E.; Scheiner, P. J. Colloid Interface Sci. 1978, 63, 509. (8) Chen, J.; Xu, L.; Li, W.; Gou, X. Adv. Mater. 2005, 17, 582. (9) Liu, L.; Kou, H. Z.; Mo, W. L.; Liu, H. J.; Wang, Y. Q. J. Phys. Chem. B 2006, 110, 15218. (10) Wu, C. Z.; Yin, P.; Zhu, X.; Ouyang, C. Z.; Xie, Y. J. Phys. Chem. B 2006, 110, 17806. (11) Zheng, Y.; Cheng, Y.; Wang, Y.; Bao, F.; Zhou, L.; Wei, X.; Zhang, Y.; Zheng, Q. J. Phys. Chem B 2006, 110, 3093.

Langmuir 2010, 26(7), 5273–5278

hematite catalysts in terms of activation temperature, conversion efficiency, and thermal stability in the catalytic reaction.11 Hollow hematite spheres with higher photocatalytic properties than other hematite crystals were synthesized by a CTAB (hexadecyltrimethylammonium bromide)-assisted hydrothermal method.12 γ-Fe2O3 particles were coated with silica to increase their stability or dispersion in a wide pH range in aqueous solutions and organic media.13 Fe3O4 particles were coated with mesoporous silica for targeted drug delivery and multiphase separation.14 The forced hydrolysis of ferric salt in a sealed system is the most popular method for the shape-selective synthesis of iron oxide with shapes such as spheres, cubes, ellipsoids, and rods.7 To obtain more controllable shapes and sizes, a variety of additives such as phosphate,2 L113B or span80,9 PVP,11 sodium dodecylsulfonate, sodium dodecylebenzenesulfonate, CTAB,12,15 hexadecyipyridinium chloride,16 NaClO3,17 and PEG18 have been used to improve this method, and many distinctive iron oxides such as ellipsoidal,7 rodlike and tubelike,9,15 quasicubic,1 hollow spheric,12 rhombohedral,16 cantaloupe-like,17 and ringlike18 hematite, and β-FeOOH rods15 have been synthesized. However, the residual additives will contaminate the final product.15 Here we developed an open forced hydrolysis method because opening the reaction system facilitates the volatilization of the byproduct HCl, which not only accelerates the hydrolysis and reduces the reaction time but also helps one tune the product shape which is sensitive to the ratio of [Fe3þ]/[Cl-].7 (12) Li, L.; Chu, Y.; Liu, Y.; Dong, L. J. Phys. Chem. C 2007, 111, 2123. (13) Klotz, M.; Ayral, A.; Guizard, C.; Menager, C.; Cabuil, V. J. Colloid Interface Sci. 1999, 220, 357. (14) Zhao, W.; Gu, J.; Zhang, L.; Chen, H.; Shi, J. J. Am. Chem. Soc. 2005, 127, 8916. (15) Wang, X.; Chen, X.; Gao, L.; Zheng, H.; Ji, M.; Tang, C.; Shen, T.; Zhang, Z. J. Mater. Chem. 2004, 14, 905. (16) Jing, Z.; Wu, S. Mater. Lett. 2004, 58, 3637. (17) Zhu, L.-P.; Xiao, H.-M.; Fu, S.-Y. Cryst. Growth Des. 2007, 7, 177. (18) Li, L.; Chu, Y.; Liu, Y. Nanotechnology 2007, 18, 105603.

Published on Web 12/14/2009

DOI: 10.1021/la903544a

5273

Article Scheme 1. Proposed Silica Coating Procedure in WPN System

A variety of routes such as microemulsion,19 molecular templating,20 and modified st€ober methods13,14,21-26 have been developed to coat iron oxide with silica for high stability, drug delivery, multiphase separation, and possible photonic crystal with magnetically controlled bandgap. However, these methods are limited by aggregations, deformed spherical shape, polydispersity, and additional surface modification processes used to suppress the formation of new nuclei. Here, to effectively break through these limitations, thus obtaining monodisperse hematite@silica core-shell particles, a two-phase system (water/ PVP/n-pentanol or WPN) has been successfully developed to confine the silica growth on the hematite surface. The development of the WPN system is spurred by the following two motivations: (i) water can be bound by PVP in aqueous solutions with a high concentration of PVP, wherein water does not act as a solvent in the usual sense. Indeed, it is customary to refer to such water as “bound” water.27 Therefore, we can confine the hydrolysis of the silica precursor along with PVP by binding water with PVP in the WPN system, where the aqueous solution of highly concentrated PVP is formed since PVP tends to entirely exists in the water phase of the water/n-pentanol system.28 (ii) PVP has been extensively used as the stabilizer and/or affinity agent in the nanotechnology field due to its excellent adsorption ability. For example, we managed to coat silica/polystyrene colloids with uniform metal, alloy, and TiO2 while preventing agglomeration29-32 by using PVP. In our WPN reaction system, as shown in Scheme 1, the direct adsorption of PVP molecules with bound water onto the hematite particles not only can successfully protect the hematite particles from agglomeration but also can realize the direct growth of silica on the hematite particles by consuming the water bound to PVP with the hydrolysis of the silica precursor. The self-assembly of colloids or nanoparticles has been intensely investigated for its importance in a wide variety of applications such as biological assays, biochemical sensors, paints, (19) Santra, S.; Tapec, R.; Theodoropoulou, N.; Dobson, J.; Hebard, A.; Tan, W. Langmuir 2001, 17, 2900. (20) Wu, P.; Zhu, J.; Xu, Z. Adv. Funct. Mater. 2004, 14, 345. (21) Lu, Y.; Yin, Y.; Mayers, B. T.; Xia, Y. Nano Lett. 2002, 2, 183. (22) Hao, L.; Zhu, C.; Jiang, W.-Q.; Chen, C.-N.; Hu, Y.; Chen, Z.-Y. J. Mater. Chem. 2004, 14, 2929. (23) Yang, C.; Wang, G.; Lu, Z.; Sun, J.; Zhuang, J.; Yang, W. J. Mater. Chem. 2005, 15, 4252. (24) Han, Y.-S.; Jeong, G.-Y.; Lee, S.-Y.; Kim, H.-K. J. Solid State Chem. 2007, 180, 2978. (25) Phylipse, A. P.; van Bruggen, M. P. B.; Pathmamanoharan, C. Langmuir 1994, 10, 92. (26) Dai, Z.; Meiser, F.; M€ohwald, H. J. Colloid Interface Sci. 2005, 288, 298. (27) de Dood, M. J. A.; Kalkman, J.; Strohh€ofer, C.; Michielsen, J.; van der Elsken, J. J. Phys. Chem. B 2003, 107, 5906. (28) Cui, Y. D.; Yi, G. B.; Liao, L. W. The Synthesis and Applications of Polyvinylpyrrolidone; Science: Beijing, 2001. (29) Zhang, J.; Liu, J.; Wang, S.; Zhan, P.; Wang, Z.; Ming, N. Adv. Funct. Mater. 2004, 14, 1089. (30) Zhang, J.; Wang, S.; Liu, J.; Wang, Z.; Ming, N. J. Mater. Res. 2005, 20, 965. (31) Zhang, J.; Zhan, P.; Liu, H.; Wang, Z.; Ming, N. Mater. Lett. 2006, 60, 280. (32) Zhang, J.; Liu, H.; Wang, Z.; Ming, N. J. Solid State Chem. 2007, 180, 1291. (33) Zhang, J. H.; Zhan, P.; Wang, Z. L.; Zhang, W. Y.; Ming, N. B. J. Mater. Res. 2003, 18, 649. (34) Zhang, J.; Liu, H.; Wang, Z.; Ming, N. B. J. Appl. Phys. 2008, 103, 013517. (35) Yi, G.-R.; Manoharan, V. N.; Michel, E.; Elsesser, M. T.; Yang, S.-M.; Pine, D. J. Adv. Mater. 2004, 16, 1204.

5274 DOI: 10.1021/la903544a

Zhang et al.

ceramics, photolithography, and photonic crystals.33-35 Here we accomplished direct short-range ordered assembly of hematite semispheres in their synthesis process by simply modifying the ferric chloride concentration. We also tried to directly assemble the hematite nanoparticles coated with silica into ordered structures such as one-dimensional structures, dimer, and trimer clusters in their coating process by adjusting the coating conditions. Finally, we managed to assemble the as-prepared monodisperse hematite-silica core-shell spheres into long-range ordered colloidal crystals with considerable photonic bandgap by using the negative pressure assembly method developed by ourselves to assemble colloids.34 It is well-known that hematite has an antiferromagnetic nature arising from the superexchange interactions between Fe3þ spins in adjacent layers, mediated by the intervening layer of oxygen ions.36 The intervening oxygen ions on the hematite surface will be easily reduced or removed by hydrogen, thus weakening the antiferromagnetic coupling and increasing the magnetization. This magnetization enhancement on the surface is insignificant and undetectable in bulk hematite. However, when hematite is prepared in nanparticle form, the increased specific surface area makes the surface magnetization enhancement significant and detectable. Namely, the magnetization enhancement on the hematite surface is detectable when introducing the reducing gas, hydrogen. By modifying the nanoparticle size, here the magnetic hydrogen sensing is successfully observed in assemblies of 220 nm hematite semispheres. Compared to the conventional hydrogen sensing based on the variations of electrical properties, the novel magnetic hydrogen sensing developed here overcomes limits such as low sensitivity and complex electrode contacts8 and provides rapid and reliable response without physical contact.

Experimental Section Materials and Characterization. FeCl3 (g98.5%), hydrochloric acid (HCl, 38-40%), PVP (average molecular weight Mn = 30 000), tetraethyl orthosilicate (TEOS) (SiO2, g27.9%), NH3 solution (25-28%), n-pentanol (g99%), and anhydrous denatured ethanol (g99.7%) were used as received. TEM images were obtained on JEM-200CX (JEOL, Japan) transmission electron microscopes. The sample was placed on a carbon-coated copper grid for TEM observation. SEM images were obtained on a LEO1530 VP (Phillips, The Netherlands) instruments. X-ray powder diffraction (XRD) patterns were recorded using an X-ray diffractometer (Rigaku D/max-RA, Japan) with Cu KR radiation (λ=1.5418 A˚). Transmission spectra were recorded on a U-3410 spectrophotometer (Hitachi, Japan). Sample films (with thickness of ∼10 μm measured by a optical microscope) were deposited on the 12  4 mm quartz sample holder for the magnetization measurements with and without 5% H2 flow in Ar (at 100 mL/min) by using a vibrating sample magnetometer (VSM, Lakeshore model 7404) equipped with a high-temperature oven. The VSM oven was continuously purged through its normal purge system with ultrapure nitrogen at 100 mL/min.

Synthesis and Self-Assembly of Hematite Nanoparticles. The reaction solutions of HCl (0.0005 M) and FeCl3 with concentrations ranging from 0.005 to 0.045 M were aged at 100 °C in Pyrex tubes for 24 h. To avoid large losses of the reaction solution via evaporation, the tube was closed by a glass stopper. A piece of filter paper was inserted between the tube and stopper to keep the reaction solution open to atmosphere during the entire reaction process. The direct short-range self-assembly of the as-formed hematite nanoparticles was achieved when the concentration of FeCl3 was controlled to be around 0.03 M. (36) Robinson, P.; Harrison, R. J.; McEnroe, S. A.; Hargraves, R. B. Am. Mineral. 2004, 89, 725.

Langmuir 2010, 26(7), 5273–5278

Zhang et al.

Figure 1. TEM (a, c-e) and SEM (b) images of hematite nanoparticles made using 0.005 (a), 0.01 (b), 0.02 (c), 0.03 (d), and 0.045 M (e) FeCl3. Inset of (a) shows the corresponding SAED pattern. (f) The corresponding XRD patterns for the samples shown in a (1), c (2), and e (3).

Synthesis and Self-Assembly of Hematite@Silica CoreShell Nanostructures. The as-prepared hematite particles were dispersed in 4 mL of ethanol solution of ammonia-water (20 vol %). The obtained suspension was added into 20 mL of n-pentanol solution of PVP (0.2 g) under stirring. After 30 min, the ethanol solution of TEOS (20 vol %) was added. The dilution of TEOS with ethanol can help one suppress the formation of new nuclei and the aggregation or adhesion of particles.33 The reaction solution was then stirred for 12 h and centrifuged. The obtained product was washed three times with ethanol by centrifugation and ultrasonication. In general, the surface charge of the hematite nanoparticles made from the forced hydrolysis method is positive. If they are dispersed into ammonia-water, the OH- ions in ammoniawater will neutralize the positive surface charge and lead to the aggregation of nanoparticles. Therefore, ammonia-water was greatly diluted by ethanol before the dispersion of the as-prepared hematite particles to prevent aggregation. However, by reasonably controlling the ammonia-water volume ratio in ethanol, the aggregation caused by ammonia-water can be purposely used to facilitate self-assembly of colloidal clusters. Here, we tried to directly assemble the as-formed hematite-silica core-shell particles into one-dimensional, dimer, and trimer cluster structures during the coating processes by simply increasing the ammoniawater volume ratio to 40%. The long-range ordered assembly of the as-prepared core-shell colloids was realized by using the negative pressure assembly method.34 In brief, a glass substrate was held vertically in the ethanol suspension of the as-prepared colloids (1 wt %) in a Bunsen flask by using a small paper clip. Then, the flask was placed in a vacuum drier connected with a water-jet air pump. After the pump was opened, the self-assembly of colloids under negative pressure started with a typical ethanol evaporation rate of 0.002 cm3/min.

Results and Discussion As shown in Figure 1a, the product made using 0.005 M FeCl3 is dominated by monodisperse rhombohedral particles with average edge length of 45 nm. The selected area electron Langmuir 2010, 26(7), 5273–5278

Article

diffraction (SAED, see inset of Figure 1a) pattern obtained from a single particle lying flat on the support film, with the electron beam perpendicular to the rhombic facets, clearly shows the single-crystal nature of the rhombohedral particles. With increasing the concentration of FeCl3 from 0.005 to 0.01 M, the product remains rhombohedral shape and increases in average edgelength from 45 to 70 nm. The SEM image (Figure1b) clearly shows the rhombohedral nature, with the particles comprising of six rhombic facets. The sequential increase of the concentration of FeCl3 suppresses the formation of the rhombohedral shape and induces the formation of the spherical particles. The pure monodisperse semispherical particles with average diameter of 120 nm are obtained by using 0.02 M FeCl3 (see Figure 1c). By further increasing the concentration of FeCl3 from 0.02 to 0.03 M, the particles improve in spherical shape and monodispersity and increase in average diameter to 150 nm. Furthermore, some short-range ordered structures of the semispheres arranging in the hexagonal closed-packed way (see Figure 1d) are formed. Interestingly, there is barely interstice among the semispheres in the self-assemblies. There is some precipitation at the bottom of the tube after this synthesis reaction is completed. So we deduce that these assemblies may arise from the ordered deposit of the semispheres on the tube bottom (due to their large diameter and high monodispersity) while they are still growing. The postgrowth of the semispheres will fill their interstice in the assemblies. The average diameter of the semispheres can be further increased from 150 to 220 nm by prolonging the reaction time to 48 h. When the concentration of FeCl3 is increased above 0.03 M, the growth of the spherical shape is suppressed and the rodlike shape begins to appear. With a concentration of 0.045 M FeCl3, the product is dominated by rodlike particles (see Figure 1e). The XRD patterns have been recorded for the typical samples with rhombohedral (Figure 1a), semispherical (Figure 1c), and rodlike (Figure 1e) morphologies to identify their crystal structure. As shown in Figure 1f, all the samples can be indexed to a pure hexagonal structure of hematite with cell constants of a = 0.5035 nm and c = 1.3740 nm (JCPDS No. 33-0664). No diffraction peaks from impurities are found in the samples. Compared with conventional forced hydrolysis methods in a sealed system2,7 for the synthesis of hematite, the open system demonstrated here accelerates the hydrolysis reaction, reduces the reaction time from 2 to 10 days to 24 h, expands the strict formation condition for hematite from 0.02-0.04 to 0.005-0.045 M FeCl3, improves the product monodispersity, and produces the novel ordered self-assemblies of hematite semispheres and the distinctive rhombohedral structure which usually only forms in the presence of additives.16 The sample shown in Figure 1c has been first chosen as the core for the silica coating due to its good spherical shape. As shown in Figure 2a-d, every hematite particle (with dark contrast) is uniformly coated with silica (with light contrast). By simply modifying the amount of hematite and/or TEOS solution, the silica shell can be continuously tuned from 10 to 200 nm. With increasing silica shell thickness, the core-shell complex particles obviously improve in monodispersity and spherical shape. We also coated the hematite samples with irregular shapes including the short-range ordered (shown in Figure 1d) and rodlike (shown in Figure 1e) structures with uniform silica shell. As shown in Figure 2e,f, all the irregular hematite nanoparticles can be well coated with uniform silica shell regardless of the shape such as dimer, Y-like, and X-like morphologies. All the hematite samples tend to tightly stick to the vessel wall if they are dispersed in water or ethanol (they cannot flake away even under intense sonication) and quickly aggregate into large DOI: 10.1021/la903544a

5275

Article

Zhang et al.

Figure 3. TEM images of the hematite clusters assembled directly by modifying the ammonia-water concentration during silica coating process. (a) One-dimensional clusters (0.02 g of rhombohedral hematite nanoparticles was dispersed in 2 mL of ethanol solution of ammonia-water (40 vol %), 0.1 mL of TEOS solution). (b) Dimer and trimer clusters (0.01 g of hematite semispheres was dispersed in 2 mL of ethanol solution of ammonia-water (40 vol %), 1 mL of TEOS solution).

Figure 2. TEM images of the semispherical hematite particles (Figure 1c) coated with a 10 nm (a, 0.015 g of hematite, 0.1 mL of TEOS solution), 60 nm (b, 0.015 g of hematite, 1 mL of TEOS solution), 80 nm (c, 0.01 g of hematite, 1 mL of TEOS solution), and 200 nm (d, 0.001 g of hematite, 1 mL of TEOS solution) silica shell. (e) and (f) show the TEM images of the irregular hematite particles coated with a 10 nm silica shell. (e) Short-range ordered structures shown in Figure 1d (0.03 g of hematite, 0.1 mL of TEOS solution) and (f) rodlike structures shown in Figure 1e (0.02 g of hematite, 0.1 mL of TEOS solution).

clusters and deposit onto the vessel bottom if they are dispersed in ammonia-water. However, all the hematite@silica core-shell particles made here can be well dispersed in water, ethanol, and ammonia-water and do not stick to the vessel wall or each other. Even if the hematite cores are changed into Fe3O4 ferromagnetic nanoparticles, the uniform silica coating still effectively prevents them from aggregating due to magnetic attraction and allows them to be easily dispersed in water or ethanol (Supporting Information, Figure S1). Compared with the previous silica coating routes,13,14,19-26 the coating method reported here offers many advantages. (1) In our WPN system, the hydrolysis reaction of TEOS is kept slow and confined around the core surface due to the “bound” water, and the formation of new nuclei is suppressed. So, the additional surface modification process usually used to forbid the formation of new nuclei is avoided. (2) The aggregation and/or adhesion of the cores and the final core-shell composite particles are effectively suppressed. (3) The silica coating greatly improves the particles’ spherical shape and monodispersity. (4) The silica shell thickness can be continuously tuned in the unprecedented large range of 10-200 nm by simply controlling the amount of TEOS. (5) The route opens a general method for the uniform silica coating of particles regardless of the core composition, size, and shape. Here, it has been successfully extended to coat gold nanoparticles with uniform and tunable thickness (Supporting Information, Figure S2). During the process of coating gold nanoparticles with silica, we noticed that the gold nanoparticles tend to self-assemble into dimer or trimer cluster nanostructures (see Figure S2), which can be attributed to ammonia-water because gold nanoparticles 5276 DOI: 10.1021/la903544a

Figure 4. SEM images for the colloidal crystals (opals) made from the hematite@silica core-shell particles shown in Figure 2b (a) and Figure 2c (b). Curves 1, 2, and 3 in (c) show the corresponding transmission spectra of the opals shown in (a) and (b) and the thin films made from the hematite cores, respectively.

made here quickly aggregate into clusters once they are dispersed in ammonia-water. Therefore, we tried to directly assemble the hematite nanoparticles into cluster structures during their silica coating process by modifying the ammonia concentration of their ethanol suspensions because ammonia-water may also induce aggregation of hematite. As seen in Figure 3, by reasonably adjusting the ammonia-water concentration and/or the hematite amount, the direct self-assembly of the rhombohedral and semispherical hematite nanoparticles into one-dimensional, dimer, or trimer nanostructures is realized in the silica coating process. We are trying to prepare pure one-dimensional, dimer, trimer, or new clusters by optimizing the coating conditions because these ordered clusters (especially for gold nanoparticles) have wide applications in electronic, optical (such as tuning the surface plasmon resonance), and catalysis fields. Furthermore, the silica shell makes them robust and stable in a variety of circumstances. By using the negative pressure assembly method,34 the monodisperse hematite@silica core-shell spheres have been successfully assembled into three-dimensional long-range ordered colloidal crystals. Figures 4a and 4b show the opals made from the core-shell complex particles shown in Figures 2b and 2c, respectively. As seen in the figures, all the particles are well ordered in the hexagonal closed-packed form in both the opals, which further confirms that the as-prepared core-shell complex particles have good monodispersity and fine spherical shape. The corresponding transmission spectra at normal incidence of the opals shown in Figures 4a and 4b have been investigated to detect the photonic bandgap (PBG). For comparison, the transmission spectra of the thin films made from the hematite cores have also been recorded. As shown in Figure 4c, an asymmetric dip is observed in the transmittance of the opals shown in Figure 4a. The PBG effects arising from Bragg diffraction due to the periodicity in the location of the composite particles in the opal should be the main factors in the formation of the dip. However, Langmuir 2010, 26(7), 5273–5278

Zhang et al.

Article

Figure 6. HRTEM images of 220 nm hematite semispheres before (a) and after (b) a 5 nm silica coating by using 0.03 g of hematite and 0.1 mL of TEOS solution.

Figure 5. Magnetization response of the thin films made of 220 nm hematite semispheres when subjected to a periodic 5% H2 flow in Ar at 300 °C in 50 G field (a) and at 200 °C in 5 kG field (b). (c) and (d) show the corresponding XRD patterns and hysteresis loops before and after H2 sensing at 300 °C, respectively. The XRD peaks marked with (/) are from the quartz sample holder. The hysteresis loops before H2 sensing at 20, 100, and 200 °C are also shown in (d). Inset of (d) shows the hysteresis loops before sensing at 20 °C for the thin films made of 45 nm (shown in Figure 1a) or 70 nm (shown in Figure1b) hematite nanoparticles.

the absorption of hematite cores (see curve 3) makes the PBG dip deformed and shallow in intensity. Compared with the opal shown in Figure 4a, the opal shown in Figure 4b has larger periodicity in its lattice due to its bigger core-shell particles. Correspondingly, its PBG dip appears in the longer wavelength region wherein the absorption of the hematite cores is weak and presents better symmetry and larger depth. As shown in Figure 5a, the magnetization of the thin films made of 220 nm hematite semispheres is highly sensitive to 5% H2 flow in Ar at 300 °C under a small magnetic field of 50 G. When the hydrogen gas is turned on, the sample magnetization increases rapidly. Once the hydrogen flow is stopped, the sample magnetization drops quickly. The sample magnetization is also highly sensitive to the hydrogen gas even when the temperature is reduced to 200 °C (see Figure 5b). Both the XRD patterns of the sample before and after hydrogen sensing (Figure 5c) can be indexed to a pure hexagonal structure of hematite. Namely, no structural or chemical phase change is caused by the hydrogen sensing. This is further confirmed by the corresponding magnetization hysteresis loops before and after H2 sensing. As shown in Figure 5d, the hydrogen sensing has no influence on the magnetization hysteresis loop. All of these results strongly suggest that the hydrogen sensing observed here should arise from the antiferromagnetic nature of hematite. As seen in Figure 5d, the saturation magnetization of the sample greatly increases and then slightly decreases when the temperature is increased from 20 to 300 °C; this variation trend demonstrates the canted antiferromagnetism of hematite.9,37 We also investigated the hydrogen sensing of the thin films made of hematite nanoparticles with smaller sizes (including the samples shown in Figure 1a-d) or 5 nm silica shell. These samples show no detectable hydrogen sensing. This further suggests that the hydrogen sensing arise from the antiferromagnetic spin variations on the hematite surface because the following reasons. (1) As shown in Figure 5d, with reducing the hematite nanoparticle size from 220 to 45 nm, the saturation magnetization reduces and the hysteresis loop (37) Lin, S. T. Phys. Rev. 1959, 116, 1447.

Langmuir 2010, 26(7), 5273–5278

closes, showing the hematite magnetism changes from weakly ferromagnetic-like (canted antiferromagnetic) to superparamagnetic as observed commonly in most antiferromagnetic materials.9,38 In other words, the excessive reduction of the hematite size will greatly weaken or destroy the total antiferromagnetic interactions on the hematite surface and the relevant hydrogen sensing although it will also largely increase the antiferromagnetic spin ratio between the hematite surface and inside. (2) As shown in Figure S1e, the hematite core cannot be protected by the silica shell from hydrogen gas and can be reduced to ferromagnetic Fe3O4 at 450 °C. Namely, hydrogen gas can still contact the hematite even after it is coated with uniform silica in the sensing experiment. So the weakening of hydrogen sensing should arise from the hematite surface-structure variations caused by the porous silica coating. Because of the incomplete crystal structure on the hematite surface (Figure 6a), there must be some defects and/or vacancies in the intervening layer of oxygen ions that are sensitive to hydrogen gas. The abundant oxygen of the amorphous silica shell (Figure 6b) should compensate most of these defects or vacancies, thus weakening the hydrogen sensing.

Conclusions We report the highly shape-selective synthesis, silica coating, self-assembly, and magnetic hydrogen sensing of hematite nanoparticles. Besides the hematite semispheres and rods, the distinctive monodisperse rhombohedral hematite with tunable size and the novel ordered self-assemblies of hematite semispheres have been made by developing an open forced hydrolysis method. A new silica coating concept based on “binding” water to PVP has been demonstrated to confine the hydrolysis reaction of silica precursor around the core surface to realize the uniform silica coating with tunable shell thickness in the unprecedented range of 5-200 nm, regardless of the core sizes, shapes, and composition. This method effectively overcomes the limits (including aggregation, deformed spherical shape, polydispersity, and additional surface modification) in the previous silica coating routes and can directly assemble both hematite and gold nanoparticles into clusters such as one-dimensional, dimer, and trimer structures during the coating process. It may open a general approach for the uniform silica coating and direct assembly into ordered clusters in the coating process of both oxide and metal nanoparticles. The hematite-silica core-shell spheres with high monodispersity and spherical shape are successfully assembled into three-dimensional ordered structures with obvious PBG. Magnetic hydrogen sensing is observed in the thin films made of 220 nm hematite semispheres even at low magnetic field (50 G) or low temperature (200 °C). The mild sensing conditions make the magnetic (38) Raming, T. P.; Winnubst, A. J. A.; Van Kats, C. M.; Philipse, A. P. J. Colloid Interface Sci. 2002, 249, 346.

DOI: 10.1021/la903544a

5277

Article

hydrogen sensor practical. Compared with the conventional electrical sensor, the novel magnetic sensor should have higher stability and sensitivity because it removes the need for electrode contacts. Acknowledgment. This work was supported at Nanjing University by NSFC under Project 10404013 and at Boise State University in part by ARO Grant (W911NF-09-1-0051),

5278 DOI: 10.1021/la903544a

Zhang et al.

NSF-CAREER program (DMR-0449639), DoE-EPSCoR program (DE-FG02-04ER46142), and NSF-RUI grant (DMR0840227). Supporting Information Available: Synthesis and characterizations of Fe3O4@silica and Au@silica core-shell structures. This material is available free of charge via the Internet at http://pubs.acs.org.

Langmuir 2010, 26(7), 5273–5278