Synthesis and Characterization of Eu-Doped Cadmium Selenide

Substitution of Cd(II) sites by Eu ions in 5.0 nm ± 0.25 nm CdSe, Cd1-xEuxSe (x = 0−0.374), can be achieved by modification of a lyothermal, single...
0 downloads 0 Views 91KB Size
Download PDF
NANO LETTERS

Synthesis and Characterization of Eu-Doped Cadmium Selenide Nanocrystals

2002 Vol. 2, No. 12 1443-1447

Orlando E. Raola and Geoffrey F. Strouse* Department of Chemistry and Biochemistry, UniVersity of California Santa Barbara, Santa Barbara, California 93106-1905 Received September 20, 2002; Revised Manuscript Received October 11, 2002

ABSTRACT Substitution of Cd(II) sites by Eu ions in 5.0 nm ± 0.25 nm CdSe, Cd1-xEuxSe (x ) 0−0.374), can be achieved by modification of a lyothermal, single source precursor method. The Eu guest ion occupies a tetrahedral lattice site as a Eu(III) defect ion based on analysis of the XPS and XAS data. XRD and XAS measurements show a linear contraction of the lattice parameters for increasing Eu(III) concentration consistent with statistical substitution at both core and surface sites in the lattice by a random ion displacement mechanism. On the basis of a Vegard’s law analysis, the Eu(III) ion is substituted in the tetrahedral cationic lattice site without formation of vacancies or phase segregation of the Eu ion in the lattice.

The incorporation of transition metal ion defect centers into the core of semiconductor materials has fascinated researchers because of the potential for phosphor, magnetic, and electronic applications.1-5 It has been suggested that confinement effects arising in nanoscale materials may allow tuning of these defect ions by modulating the nature of coupling to the host lattice.6 The effects of quantum confinement on the exchange interactions in dilute magnetic quantum dot (DMQD) systems have already been shown to influence magnetic exchange properties of Mn2+ (S ) 5/2) and Co2+ (S ) 3/2) in II-VI materials.2,7-13 While several researchers have probed doped nanocrytalline materials, a systematic study of the controlled displacement of core atoms by guest ions, coupled to analysis of the guest ion site, site symmetry, and valence state have yet to be done. Intentionally incorporating a defect ion by random ion displacement of the cation site in the core of a CdSe nanoparticle offers a convenient platform to probe the influence of quantum confinement on the interplay of magnetic and electronic degrees of freedom of the guest ion and host lattice. In this letter, we present evidence for the preparation of a series of Eu(III)-doped CdSe nanocrystal (NC) samples prepared by application of a single source precursor lyothermal strategy. Vegard’s law analysis of pXRD data, coupled to X-ray absorption (XAS) analysis, provides direct evidence of Eu(III) guest ion incorporation into a 5.0 nm wurtzite-type CdSe host lattice.14 The Eu ion appears to form a solid solution with the CdSe sample with random ion displacement of the cadmium cation site by the substitutional * Corresponding author. E-mail: [email protected]. 10.1021/nl025805e CCC: $22.00 Published on Web 11/06/2002

© 2002 American Chemical Society

guest Eu ion up to levels of ∼37%. Similar doping levels have been observed previously for transition metal ion doping of II-VI materials.15 Synchrotron XAS measurements confirm the oxidation state of the Eu ion, and analysis of the EXAFS region confirms the geometry about the Eu(III) ions with four Se nearest neighbors corresponding to incorporation into the cationic site in a hexagonal lattice. It is well known that vacancy formation due to charge imbalance and lattice strain can self-limit inclusion of guest ions into a host lattice. At nanoscale sizes, further complications arise from the propensity for the ion to migrate to less strained surface sites, rather than incorporate as substitutional ions in the crystal lattice. Recent studies in nanocrystalline II-VI semiconductor materials have confirmed the difficulty in obtaining random ion displacement, particularly for lyothermally prepared CdSe, CdS, and ZnSe. In materials doped with Mn and Co, doping occurs primarily at the surface or near-surface sites in these materials due to defect center migration to the surface at the lyothermal growth temperatures.4,12 The apparent difficulty in controlling doping presents a unique challenge for chemists to develop a methodology to incorporate transition metal ion defect centers into the core of semiconductor materials and, more importantly, necessitates the development of analytical tools to probe site occupation between core, surface, and interstitial sites, as well as measure formation of vacancies in the host system. Understanding these issues is important to analyzing the exchange coupling between a guest ion and host lattice, which is influenced by the coordination around the guest ion and its site occupation in the system.

To achieve random ion displacement of Eu ions into the core of CdSe, Eu/CdSe materials were prepared by a modification of a previously reported growth methodology, employing Li4[Cd10Se4(SPh)16] as a single source precursor.16 This strategy has been shown to allow core doping of CdSe by Co(II).13 Briefly, to a solution of degassed hexadecylamine (HDA) at 60 °C under nitrogen, EuCl2, and Li4[Cd10Se4(SPh)16] were combined using standard airless techniques. Nanoparticle growth was initiated by raising the temperature to 220-240 °C (1 °C/min). At the desired size, ∼5.0 nm, the reaction was cooled by 20 °C, left overnight to narrow the size distribution by annealing the nanocrystals, and isolated by precipitation through the addition of 100 mL of anhydrous methanol. Excess HDA is removed by suspension in toluene followed by reprecipitation in methanol and isolation by centrifugation, as previously described.16 Varying the initial stoichiometry of the reaction allows control of the Eu concentration in the final materials. This technique allows isolation of 5.0 ( 0.25 nm Eu/CdSe materials, based on TEM analysis (Figure 1 C), in which the Eu ion displaces the Cd ion between 0 and 37.4 mole percent (CdEuxSe, x ) 0.059, 0.106, 0.248, 0.374 ( 0.002) based on ICP-AE measurements measured at 382.0 nm for the Eu atom, and 214.4 nm for the Cd atom. Nanocrystalline CdSe above 2 nm typically adopts a wurtzite-type structure rather than a zinc blende lattice, although a pressure-induced phase transition to the NaCl lattice has been observed.17 Inspection of the pXRD data in Figure 1 shows a wurtzite lattice is obtained at all Eu doping levels. This can be confirmed by inspection of the TEM data in Figure 1C, which shows a representative image of an individual CdEu0.374Se nanocrystal looking down the 〈100〉 direction that exhibits lattice fringes consistent with the assignment of a highly crystalline wurtzite structure. The observation of wurtzite geometry is surprising, since pure Eu-Se exhibits either a cubic NaCl-type structure for EuSe with Eu in the +2 oxidation state or an orthorhombic Sc2S3type structure, which constitutes a superstructure of the NaCltype lattice, for Eu2Se3 with Eu in the +3 oxidation state. In both structures the Eu ion is coordinated to six Se atoms in the first shell. Tetrahedral site occupation for f-elements is rarely observed, since the octahedral environment around the Eu ion tends to stabilize the f-element crystal field. This would suggest significant strain may arise in the lattice around the Eu ions, and may limit the stability of Eu ions that migrate to the nanomaterial surface. A random ion displacement model can be used to describe the Eu ion substitution into the cationic Cd site in a wurtzite lattice symmetry. This allows the use of a Vegard’s law analysis, an empirical law that relates the statistical substitution of a guest ion into the host lattice with the experimentally observed degree of lattice change with increasing defect ion concentration. Statistical substitution into a lattice site is predicted to lead to a lattice contraction for smaller ions and a lattice expansion for larger ions. Isolation of the defect ion at only surface sites or at interstial sites will result in insignificant lattice shifts arising primarily from strain effects. The shift in the a-lattice parameter in Figure 1B shows a 1444

Figure 1. (A) Selected X-ray powder diffractograms of Eu-doped CdSe, showing the shift in the position of the (110) reflection from wurtzite-type nanocrystals. (a) CdSe; (b) CdEu0.248Se; (c) CdEu0.374Se where the asterisk indicates reflections from the Si internal standard added to the powder for calibration of the XRD lines. (B) Vegard’s law plot of the a lattice parameter vs molar fraction of dopant for Eu/CdSe. (C) TEM image of a 5 nm CdEu0.374Se nanocrystal showing diffraction fringes corresponding to the 〈110〉 atomic planes.

lattice contraction of approximately 6% as the Eu ion concentration in the doped material increases from 0.059 to 0.374 mole percent. This is clearly seen as a movement of the (110) reflection toward higher 2θ in the powder diffractogram shown in Figure 1A. The observation of a linear contraction with increasing Eu concentration is in accord with the predictions for Vegard’s law behavior, suggesting displacement occurs at the lattice sites situated both in the core and on the surface through a purely statistical process, and not isolation of the defect ion only at the surface sites through an ion-migration pathway, as previously observed.18,19 Prediction of strain and site occupation of a guest ion in a host lattice requires comparison of the ionic radii in similar Nano Lett., Vol. 2, No. 12, 2002

Figure 2. (A) XPS Survey spectrum of (a) CdSe, (b) CdEu0.374Se. (B) Expanded XPS region showing the Eu 3d and CdMNN lines for (a) CdSe; (b) CdEu0.374Se; (c) Eu2O3.

coordination environments. This presents a problem for estimation of the lattice change for substitution of the Cd center by the Eu guest ion in a wurtzite lattice, as Td Eu(II) and Eu(III) are not reported. Lattice mismatch and variance in coordination number have been seen for Mn(II) doping of bulk crystals of II-VI materials, where doping levels in excess of 40% have been observed with a smaller lattice contraction than predicted.15 An indirect estimation of the lattice mismatch for the Eu ion can be achieved by comparison of the standard ionic radii for the Eu and Cd ions in a NaCl lattice, in which each element is six coordinate. According to the standard reported data for ionic radii,20 the effective ionic radius for Eu(II) is larger than Cd(II) (117 pm vs 95 pm); however, substitution of Eu(III) for Cd(II) (95 pm vs 95 pm). The observation of a 6% lattice contraction in the XRD data with increasing concentration suggests that Eu (III) is the more likely guest ion rather than Eu(II) due to the observation of lattice contraction. The difference in contraction relative to the ionic radii arises from the coordination difference between the empirically compared NaCl and experimentally observed wurtzite crystalline lattice. To further analyze the shift; the oxidation state of the Eu must be identified. Identification of the oxidation state of the Eu ion in the doped lattice can be achieved by inspection of the X-ray photoelectron spectra (XPS) in Figure 2, which shows contributions from Se, Cd, and Eu.21 The energy of the transitions for the 3d5/2 and 3d3/2 of Eu between 1120 and 1180 eV allows assignment of the Eu valence as Eu(III) in the doped Eu/CdSe samples by comparison to a bulk sample of Eu2O3. Eu(II) lines are found to shift to lower energy in the XPS. The CdMNN lines and the Se 3p lines at 1109 and 162 eV are observed to be invariant with doping (Figure 2B). This suggests that the Cd(II) and Se sites are largely unperturbed by the presence of the Eu guest ion in the lattice. Nano Lett., Vol. 2, No. 12, 2002

Figure 3. XANES spectra of Eu/CdSe and model compounds: (a) CdEu0.059Se; (b) CdEu0.106Se; (c) CdEu0.248Se; (d) CdEu0.374Se; (e) Eu2O3; (f) EuSe.

Similar observations have been made in transition metal ion doping of bulk lattices of CdSe and ZnSe.7 The narrowing of the line width for the CdMNN lines with doping apparent in Figure 2B is indicative of lower crystalline disorder (strain) in the doped sample, relative to the undoped material, potentially due to the pinning of the strain at the defect site. To verify the assignment of Eu(III) by XPS, X-ray absorption spectroscopy (XAS) measurements were performed at the Eu-LIII edge around 6977 eV. In Figure 3, the X-ray absorption near edge absorption spectrum (XANES) at the Eu-LIII edge shows a narrow and intense “white line” that has been attributed to the Eu transition 2p3/2f5d (Figure 1445

Figure 4. Eu-LIII edge EXAFS plotted as FT magnitude for Eu/ CdSe: (a) CdEu0.059Se; (b) CdEu0.106Se; (c) CdEu0.248Se; (d) CdEu0.374; (e) EuSe; (f) simulated EXAFS for a 2 nm-CdSe cluster (wurtzite structure) with the central Cd atom replaced by Eu.

3). The position of the XANES line is a good indicator of the oxidation state of Eu, the divalent ion appearing around 7 eV down from the trivalent ion.22-24 Comparison of the data to Eu2O3 and EuSe, as model compounds, provides unequivocal proof of Eu(III) in these materials.25 This confirms the assignment of the Eu oxidation state suggested by the lattice contraction observed in the pXRD data in Figure 1. Europium(III) is a surprising dopant for the Cd2+ site, due to the potential production of a high level of Cd vacancies in order to charge balance the lattice. A high degree of Cd vacancies would be expected to significantly perturb the lattice crystallinity, which is inconsistent with the observed CdMNN line widths in the XPS data, as well as the observed TEM images for individual Eu/CdSe samples. A more likely possibility for formation of Eu(III) is a post-process oxidation driven by the low redox potential necessary to convert Eu(II) to Eu(III). Post-oxidation has been shown by Vercaemst26 to readily occur for Eu(II) compounds handled in air. Consistent with this prediction XPS intensity analysis for the samples does not support a vacancy induced stoichiometric variation in the ratios of cation to anion in the lattice. Further information about the Eu site occupation can be gained by inspection of the extended X-ray absorption fine structure (EXAFS) region of the Eu-LIII edge XAS data. The Fourier transform of the EXAFS region for the doped nanocrystals and a EuSe standard are shown in Figure 4. Analysis of the FT plots suggests that the Eu center is situated in a wurtzite geometry surrounded by four Se atoms. In the 1446

Eu/CdSe materials, the observed radial distance between Eu and the first shell of Se atoms calculated from the FT transformation off the EXAFS region range from 227 pm for the sample with the lowest Eu concentration (x ) 0.059) to 218 pm for the highest Eu concentration (x ) 0.374). To verify the assignment of Td Eu site occupation, a fit of the shell distance at the limit of one Eu atom displacing the central Cd cation in a 2 nm CdSe cluster was carried out using the FEFF 7 code.27 The model predicts a Se shell at 224 pm, in agreement with the experimental data for the lowest doping level. The model only fits the first backscattering shell around the absorbing atom. The observed experimental shift in the Eu-Se distance is equivalent to a 4% lattice contraction between the highest and lowest doping concentration, which matches exactly the value expected from the ionic radii data and is within experimental error of the XRD measurement.20 A complete analysis of the EXAFS data requires further data collection of the XAS at the Cd and Se edges, which is currently underway. In conclusion, we have shown that using lyothermal techniques we can introduce Eu as a guest ion into the crystal lattice of CdSe nanocrystals. Following isolation of the CdSe nanoparticle, conversion of the Eu center to Eu(III) is observed due to the oxidative instability of Eu(II). The guest Eu(III) ion is statistically distributed between the core and surface of the nanoparticle as a substitutional impurity. Acknowledgment. Support for this work was provided by a NSF CAREER Award (DMR-9875940) and by a NSF Graduate Research Fellowship (O.E.R.). This work made use of MRL Central Facilities supported by the MRSEC Program of the National Science Foundation under award No. DMR00-80034. Portions of this research were carried out at the Stanford Synchrotron Radiation Laboratory, a national user facility operated by Stanford University on behalf of the U.S. Department of Energy, Office of Basic Energy Sciences. References (1) Alivisatos, A. P. Pure Appl. Chem. 2000, 72, 3-9. (2) Wolf, S. A.; Awschalom, D. D.; Buhrman, R. A.; Daughton, J. M.; von Molnar, S.; Roukes, M. L.; Chtchelkanova, A. Y.; Treger, D. M. Science 2001, 294, 1488-1495. (3) Bhargava, R. N. J. Lumin. 1996, 70, 85-94. (4) Mikulec, F. V.; Kuno, M.; Bennati, M.; Hall, D.; Griffin, R.; Bawendi, M. G. J. Am. Chem. Soc. 2000, 122, 2532-2540. (5) Chen, W.; Malm, J. O.; Zwiller, V.; Huang, Y. N.; Liu, S. M.; Wallenberg, R.; Bovin, J. O.; Samuelson, L. Phys. ReV. B 2000, 61, 11021-11024. (6) Pileni, M. P. Catal. Today 2000, 58, 151-166. (7) Furdyna, J. K. J. Appl. Phys. 1988, 64, R29-R64. (8) Hoffman, D. M.; Meyer, B. K.; Ekimov, A. I.; Merkulov, I. A.; Efros, A. L.; Rosen, M.; Couino, G.; Gacoin, T.; Boilot, J. P. Solid State Commun. 2000, 114, 547-550. (9) Mattoussi, H.; Mauro, J. M.; Goldman, E. R.; Anderson, G. P.; Sundar, V. C.; Mikulec, F. V.; Bawendi, M. G. J. Am. Chem. Soc. 2000, 122, 12142-12150. (10) Norris, D. J.; Vlasov, Y. A. AdV. Mater. 2001, 13, 371-376. (11) Ladizhansky, V.; Hodes, G.; Vega, S. J. Phys. Chem. B 2000, 104, 1939-1943. (12) Radovanovic, P. V.; Gamelin, D. R. J. Am. Chem. Soc. 2001, 123, 12207-12214. (13) Hanif, K. M.; Meulenberg, R. W.; Strouse, G. F. J. Am. Chem. Soc. 2002, 124, 11495-11502. Nano Lett., Vol. 2, No. 12, 2002

(14) The TEM micrographs were registered with the samples supported on an amorphous carbon grid. Powder diffractograms were acquired using silicon as an internal standard. (15) Samarth, N.; Furdyna, J. K. Proc. IEEE 1990, 78, 990-1003. (16) Cumberland, S. L.; Hanif, K. M.; Javier, A.; Khitrov, G. A.; Strouse, G. F.; Woessner, S. M.; Yun, C. S. Chem. Mater. 2002, 14, 15761584. (17) Chen, C. C.; Herhold, A. B.; Johnson, C. S.; Alivisatos, A. P. Science 1997, 276, 398-401. (18) Vegard, L. L.; Vegard, Z. P. Z. Phys. 1921, 5, 17. (19) Hines, M. A.; GnyotSionnest, P. J. Phys. Chem. 1996, 100, 468471. (20) Shannon, R. D. Acta Crystallogr. 1976, A32, 751-767. (21) The samples for XPS were prepared by pressing a small amount of the free-flowing powder against adhesive graphite tape. The XPS was collected in a Kratos Axis Ultra with an efficient charge balancing system. (22) Antonio, M. R.; Soderholm, L.; Song, I. J. Appl. Electrochem. 1997, 27, 784-792.

Nano Lett., Vol. 2, No. 12, 2002

(23) Bornick, R. M.; Stacy, A. M.; Taylor, R. D.; Kwei, G. H. J. Solid State Chem. 1999, 144, 252-254. (24) Shimizugawa, Y.; Umesaki, N.; Qiu, J. R.; Hirao, K. J. Synchrotron Radiat. 1999, 6, 624-626. (25) Both XANES and EXAFS regions were recorded in fluorescence mode at beam line 7-3 of Stanford Synchrotron Radiation Laboratory. We used a 30-element Ge and the measurements were conducted at room temperature. The appropriate amount of sample was sandwiched in Kapton tape. The data collected were processed using the standard package EXAFSPAK by G. N. George http:// www-ssrl.slac.stanford.edu/exafspak.html, 2001, and the Athena interactive EXAFS data processing package by B. Ravel http:// feff.phys.washington.edu/∼ravel/software/exafs/aboutathena.html. (26) Vercaemst, R.; Poelman, D.; Van Meirhaeghe, R. L.; Fiermans, L.; Laflere, W. H.; Cardon, F. J. Lumin. 1995, 63, 19-30. (27) Ankudinov, A. L. RelatiVistic Spin-dependent X-ray Absorption Theory; University of Washington, Ph.D. Thesis, 1996.

NL025805E

1447