J. Phys. Chem. B 2006, 110, 16937-16940
16937
Excited Carrier Dynamics of r-Cr2O3/r-Fe2O3 Core-Shell Nanostructures Gang Xiong, Alan G. Joly,* Gary P. Holtom, Chongmin Wang, David E. McCready, Kenneth M. Beck, and Wayne P. Hess Pacific Northwest National Laboratory, P.O. Box 999, K8-88, Richland, Washington 99352 ReceiVed: April 24, 2006; In Final Form: June 30, 2006
In this work R-Cr2O3/R-Fe2O3 core-shell polycrystalline nanostructures were synthesized by using R-Cr2O3 nanoparticles as seed crystals during aqueous nucleation. The formation of R-Fe2O3 polycrystallites on R-Cr2O3 surfaces was confirmed by X-ray diffraction, transmission electron microscopy, and energy-dispersive X-ray analysis. The excited-state relaxation dynamics of as-grown core-shell structures and “pure” R-Fe2O3 particles of the same size were measured with femtosecond transient absorption spectroscopy. The results show the carrier lifetimes decay within a few picoseconds regardless of sample. This is likely due to fast recombination/ trapping of carriers to defects and iron d-states.
I. Introduction Hematite (R-Fe2O3) is a natural mineral with low toxicity and good chemical stability. It has a band gap of 2.2 eV, in the middle of the solar spectrum, making R-Fe2O3 a potential photocatalyst.1,2 However, the photon-electron conversion efficiency is less than 1.8%.3 The excited-state lifetimes have been measured to be on the order of a few picoseconds in bulk crystals, thin films, and nanoparticles.4,5 The fast excited carrier lifetimes in R-Fe2O3 are believed due to nonradiative recombination and trapping mediated by the d states and defect states in the band gap.4,5 Recently R-Cr2O3/R-Fe2O3 (0001) epitaxial heterostructures have been investigated both experimentally and theoretically.6,7 The results show a significant valence band offset at the R-Cr2O3/R-Fe2O3 interface to form a type-II heterojunction. The R-Cr2O3/R-Fe2O3 type-II heterojunction should allow segregation of excited electrons and holes at the R-Fe2O3 and R-Cr2O3 side of the interface, respectively. Carrier charge separation may result in increased carrier lifetimes leading to improved catalyst performance. R-Cr2O3 and R-Fe2O3 have the same crystal structure (R3hc) and similar lattice constants. The lattice mismatch between R-Cr2O3 (0001) and R-Fe2O3 (0001) is only 2% (in plane) and 0% (out of plane), and epitaxial growth of R-Fe2O3 (0001) on R-Cr2O3 (0001) has been achieved by using molecular beam epitaxy.6 It may be possible to synthesize R-Cr2O3/R-Fe2O3 core-shell nanostructures with epitaxial interfaces, to provide an alternative method to band-offset engineering. In addition, composite nanoparticles are technologically important in catalysis due to their large surface/volume ratio. Although such a core-shell structure has multiple interfaces with different crystal orientations, the average valence band offset value is about 0.5 eV. Therefore the overall core-shell nanostructure can still be regarded as a type-II heterojunction.7,8 Similar type-II coreshell structures such as CdTe/CdSe have been reported.9,10 Here we report the synthesis of type-II heterojunction R-Cr2O3/R-Fe2O3 polycrystalline core-shell nanostructures and investigate the charge carrier lifetimes and dynamics using femtosecond transient absorption spectroscopy. Although dif* Address correspondence to this author. E-mail:
[email protected].
ferences in the transient spectra can be found between the R-Cr2O3/R-Fe2O3 core-shell and the “pure” R-Fe2O3 nanoparticle sample of the same size, no solid evidence supports the premise that the R-Cr2O3/R-Fe2O3 interface significantly alters the excited carrier lifetimes. We believe that the d f d transitions and defect states in these materials result in fast carrier recombination and trapping. Therefore, modification of the excited carrier lifetimes by band-offset engineering is limited because of short free carrier lifetimes due to rapid trapping. II. Experimental Section Commercial R-Cr2O3 nanopowders (99%) were purchased from Sigma-Aldrich. The average particle size was verified to be approximately 30 nm by high-resolution X-ray diffraction (XRD) analysis and transmission electron microscopy (TEM). The R-Cr2O3 nanopowders were used as seeds in the growth of the R-Cr2O3/R-Fe2O3 core-shell structures. Our synthesis method was inspired by the work of Penners et al., who used R-Fe2O3 seeds to grow larger R-Fe2O3 particles in aqueous solution with very narrow size distribution.11 By adjusting the ratio between the number of seeds (R-Fe2O3) and reactant concentration (FeCl3 and HClO4), one can precisely control the size of the final product. For the synthesis of R-Cr2O3/R-Fe2O3 core-shell structures in our work, we used R-Cr2O3 as seeds instead of R-Fe2O3. The seeded-growth procedure for the core-shell structures can be described as follows: First 800 mL of a water solution containing 5 × 10-3 M FeCl3 and 9 × 10-2 M HClO4 was first refluxed at 100 °C for 30 min, then 56 mg of R-Cr2O3 seed crystals were dispersed into the solution, and the mixture was incubated at 100 °C for 30 h. The solution was magnetically stirred during incubation. The precipitate was then separated from the solution by centrifugation. For comparison, “pure” R-Fe2O3 nanoparticles of the same size were grown from R-Fe2O3 seeds under the same conditions. The R-Fe2O3 seeds were prepared by using a slightly different chemical approach also given by Penners et al. in the same publication.11 The synthesized R-Cr2O3/R-Fe2O3 core-shell and “pure” R-Fe2O3 nanoparticles were characterized by XRD, TEM, and energy-dispersive X-ray (EDX) analysis. In the XRD analysis, the samples were measured on a high-resolution, double-crystal
10.1021/jp062507n CCC: $33.50 © 2006 American Chemical Society Published on Web 08/05/2006
16938 J. Phys. Chem. B, Vol. 110, No. 34, 2006
Xiong et al.
Figure 2. Transmission electron microscopy images of (a) R-Cr2O3 nanoparticles used as seeds for the growth of the core-shell structures; (b) R-Fe2O3 nanoparticles of average size of 70 nm grown in the same seeded method but with smaller R-Fe2O3 as the nuclei; (c) R-Cr2O3/ R-Fe2O3 core-shell nanoparticles; and (d) high-resolution TEM image showing the polycrystalline core-shell structures. EDX microanalysis shows different elemental composition at the core and shell region. Figure 1. XRD spectra of R-Cr2O3 and R-Cr2O3/R-Fe2O3 core-shell nanoparticles.
Philips X’Pert MRD system. The experimental diffraction data were analyzed with Jade software (Materials Data, Inc., Livermore, CA) and compared with the Powder Diffraction File, PDF-2 database (International Centre for Diffraction Data, Newtown Square, PA). TEM analysis was carried out on a JEOL JEM 2010F microscope with a specified point-to-point resolution of 0.194 nm. The operating voltage on the microscope was 200 keV. Elemental distribution was determined by using EDX spot analysis. Prior to the optical measurements, the samples were suspended in deionized water and diluted to have an optical density of roughly 0.5 in the visible region. The optical extinction spectra from 200 to 800 nm were measured with a UV-vis spectrometer. The time-resolved absorption measurements were performed with a regeneratively amplified Ti:Sapphire laser system. The output pulse at 814 nm and 1 kHz repetition rate was split into excitation and probe beams with a 10% beam splitter. The excitation beam was frequency doubled in a 1 mm BBO crystal to produce 407 nm (3.05 eV) pulses. The probe beam was sent through a computer-controlled variable delay line and then focused into a cell of water to produce a whitelight continuum. The probe photon energy was frequency selected by using 10 nm band-pass interference filters and then focused with the pump pulse onto the sample. The transmitted probe beam intensity was monitored by using an amplified Si photodiode. The time resolution of the transient absorption experiment was determined to be about 300 fs fwhm. III. Results A. Structural Characterization. The as-grown R-Cr2O3/RFe2O3 core-shell and Fe2O3 nanoparticle samples were characterized by XRD, TEM, and EDX. The XRD spectra were taken from 5° to 80° (2θ). The corresponding XRD spectra are shown in Figure 1. For comparison purposes, only the spectra for the R-Cr2O3 seeds and the core-shell sample in the range from 30 ° to 60 ° are plotted. For the core-shell sample, the corresponding R-Fe2O3 and R-Cr2O3 diffraction peaks are near each other due to their similar lattice constants for the same crystal structure. The R-Cr2O3 peaks are significantly weaker in intensity because of the absorption of the R-Fe2O3 shells. Figure 2 shows TEM images of the R-Cr2O3 seeds (a), the comparison R-Fe2O3 sample (b), and the R-Cr2O3/R-Fe2O3 core-shell structures (c and d). The R-Cr2O3 nanoparticles are found to be approximately 30 nm in diameter, in agreement with the XRD peak profile analysis using Scherrer’s equation. The size of the as-grown R-Fe2O3 particles is about 70 nm. The R-Cr2O3/R-Fe2O3 core-shell nanoparticles are also roughly 70
Figure 3. The EDX results of the R-Cr2O3/R-Fe2O3 core-shell nanoparticles. Left: The average elemental ratio of all the particles. Right: The average elemental ratio measured at the core regions.
nm in diameter with an average shell thickness of approximately 20 nm (Figure 2d). The core-shell structured particle is featured by a singlecrystal R-Cr2O3 core, which is covered by a multidomain R-Fe2O3 shell. Each domain of the R-Fe2O3 shell grows along a facet of the R-Cr2O3 seed. Due to the small difference in lattice constants between R-Fe2O3 and R-Cr2O3, the shell does not uniformly encapsulate the seed particle (Figure 2d). Therefore, it is difficult to resolve the interfaces between the R-Cr2O3 core and the R-Fe2O3 shell in the TEM images. EDX spot analysis for multiple individual core-shell particles as illustrated in Figure 2d revealed an average atomic ratio of 93:7 (Fe:Cr) for the entire particle, while in the center region the ratio is measured to be 84:16. This can be seen in Figure 3. Furthermore, EDX results show that the outer region of the particle consists of only R-Fe2O3. Together this indicates that the final product has a core-shell structure. The 93:7 average atomic ratio measured for the final product agrees reasonably well with the relative volume estimated from the dimensions of the core and shell. B. Optical Extinction Spectra. The optical extinction spectrum of R-Cr2O3 seed crystals is displayed in Figure 4a. According to the literature, the optical band gap of R-Cr2O3 is approximately 3.4 eV,5,12 yet there is some absorption in the visible region, due to internal d-d transitions of the Cr3+ ions.13,14 The average seed particle diameter is 30 nm, thus Mie scattering has little effect on the observed spectra of this sample. The extinction spectra of the R-Cr2O3/R-Fe2O3 core-shell and pure R-Fe2O3 samples are plotted in Figure 4b. For the R-Fe2O3 sample, an optical band gap of approximately 2.2 eV (564 nm) is evident. Some broad weak features at 550 and 720 nm are
R-Cr2O3/R-Fe2O3 Core-Shell Nanostructures
J. Phys. Chem. B, Vol. 110, No. 34, 2006 16939
Figure 4. The optical extinction spectra of R-Cr2O3, R-Fe2O3, and R-Cr2O3/R-Fe2O3 core-shell nanoparticles.
Figure 5. The transient absorption spectra of R-Fe2O3 and R-Cr2O3/R-Fe2O3 core-shell nanoparticles.
also observed, due to ligand field transitions.5,15 These weak bands can also be found in the spectrum of the core-shell sample. The additional absorption from 600 to 800 nm (below the optical band gap of R-Fe2O3) in the core-shell spectrum can be attributed to absorption from R-Cr2O3 (Figure 4a). The R-Fe2O3 and R-Cr2O3/R-Fe2O3 samples are considerably larger than the R-Cr2O3 seeds, therefore Mie scattering may not be completely ignored in the R-Fe2O3 and R-Cr2O3/R-Fe2O3 spectra. Although Mie scattering usually produces a small red shift of the optical transmission edge,16,17 it does not significantly affect our interpretation of the experimental results. C. Transient Absorption Spectra. Our 3.04 eV pump photon mainly excites the band-to-band transitions in R-Fe2O3, while some excitation of d-d metal states in iron and chromium is also possible. Control experiments performed on colloidal R-Cr2O3 seeds found that the colloidal R-Cr2O3 sample produced transient signals of at least 1 order of magnitude smaller than the corresponding transient signals of R-Fe2O3 and R-Cr2O3/ R-Fe2O3 core-shell samples. Typical transient responses for both R-Fe2O3 and R-Cr2O3/R-Fe2O3 samples with use of probe wavelengths of 680 and 600 nm are displayed in Figure 5. Probe wavelengths in the red below the band gap of either material are expected to probe primarily excited electron dynamics in these materials.5 The results shown in Figure 5 reveal that the transient absorption in both samples shows an initial pulse-width limited feature followed by a decay of about 5 ps. The initial ultrafast transient can be attributed to band filling and band gap renormalization following initial population of the conduction band. The longer picosecond decays characterize the relaxation dynamics of the excited carriers. Figure 5 also reveals that the absorption does not recover completely within 10 ps,
indicating a longer duration component that can be attributed to carrier trapping. We have previously observed that this component in R-Fe2O3 films persists for hundreds of picoseconds.5 The differences in the overall transient absorption spectra between R-Fe2O3 and R-Cr2O3/R-Fe2O3 may be related to the R-Cr2O3/R-Fe2O3 heterostructures; however, we cannot rule out possible small contributions from defect and d-state absorption of R-Cr2O3 in the visible region. Although the carrier relaxation time varies slightly between the R-Fe2O3 and the R-Cr2O3/RFe2O3 particles, the differences are relatively small. Therefore the results do not show any direct evidence of charge separation at R-Cr2O3/R-Fe2O3 interfaces as compared with the results obtained from the “pure” R-Fe2O3 particles. These results are consistent with previous measurements on R-Fe2O3 (0001) epitaxial thin films grown on a thin R-Cr2O3 (0001) buffer layer where carrier lifetimes were determined to be of the order of a few picoseconds.5 In these previous studies, a relatively thick (100 nm) layer of R-Fe2O3 may have obscured the effects of the type-II junction whereas in the present materials the R-Fe2O3 layer is relatively thin (20 nm) and interface effects should be more apparent in the transient absorption measurements. The fact that we did not observe any significant change in relaxation dynamics indicates that the excited carriers likely undergo trapping or fast nonradiative recombination, as reported recently in R-Fe2O3 films and nanoparticles.4,5 This result agrees with initial photochemistry studies on R-Cr2O3/R-Fe2O3 (0001) heterostructures, which demonstrated that R-Cr2O3 promotes hole-mediated photodecomposition of adsorbed trimethylacetic acid on its own, and that reactivity enhancement due to the valence band offset is not observed.18
16940 J. Phys. Chem. B, Vol. 110, No. 34, 2006 IV. Summary We have synthesized type-II heterostructure R-Cr2O3/R-Fe2O3 core-shell polycrystalline nanoparticles with controllable shell thickness. The excited-state relaxation dynamics were studied with transient absorption spectroscopy. The results were compared with “pure” Fe2O3 nanoparticles of similar size. Although some differences can be found between their transient absorption spectra, there is no evidence that charge separation at the R-Cr2O3/R-Fe2O3 interface leads to increased carrier lifetimes or increased activities. We attribute this to fast recombination and carrier trapping, which limits the lifetimes of free carriers and thus charge separation effects due to the type-II band offset electronic effects. Acknowledgment. This research is supported by the U.S. Department of Energy, Office of Basic Energy Sciences, Chemical Sciences Division. Experiments were performed in the W. R. Wiley Environmental Molecular Sciences Laboratory, a national scientific user facility sponsored by the Office of Biological and Environmental Research located at Pacific Northwest National Laboratory (PNNL). PNNL is operated for the U.S. Department of Energy by Battelle. We thank John Jaffe for valuable discussions.
Xiong et al. References and Notes (1) Pulgrin, C.; Kiwi, J. Langmuir 1995, 11, 519. (2) Faust, B.; Hoffmann, M.; Bahnemann, D. J. Phys. Chem. 1989, 93, 6371. (3) Cai, S.; Jiang, D.; Tong, R.; Jin, S.; Zhang, J.; Fujishima, A. Electrochim. Acta 1991, 36, 1585. (4) Cherepy, N. J.; Liston, D. B.; Lovejoy, J. A.; Deng, H.; Zhang, J. Z. J. Phys. Chem. B 1998, 102, 770. (5) Joly, A. G.; Williams, J. R.; Chambers, S. A.; Xiong, G.; Hess, W. P.; Laman, D. M. J. Appl. Phys. 2006, 99, 053521. (6) Chambers, S. A.; Liang, Y.; Gao, Y. Phys. ReV. B 2000, 61, 13223. (7) Jaffe, J. E.; Dupuis, M.; Gutowski, M. Phys. ReV. 2004, 69, 205106. (8) Jaffe, J. E. Private communication. (9) Gur, I.; Fromer, N. A.; Geier, M. L.; Alivisatos, A. P. Science 2005, 310, 462. (10) Li, J.; Wang, L.-W. Appl. Phys. Lett. 2004, 84, 3648. (11) Penners, N. H. G.; Koopal, L. K. Colloids Surf. 1986, 19, 337. (12) Misho, R. H.; Murad, W. A.; Fattahallah, G. H. Thin Solid Films 1989, 169, 235. (13) Blazey, K. W. Solid State Commun. 1972, 11, 371. (14) McClure, D. J. Chem. Phys. 1963, 38, 2289. (15) Sherman, D. M.; Waite, T. D. Am. Mineral. 1985, 70, 1262. (16) Hsu, W. P.; Matijevic, E. Appl. Opt. 1985, 24, 1623. (17) Aden, A. L.; Kerker, M. J. Appl. Phys. 1951, 22, 1242. (18) Chambers, S. A.; Williams, J. R.; Henderson, M. A.; Joly, A. G.; Varela, M.; Pennycook, S. J. Surf. Sci. 2005, 587, L197.
