NANO LETTERS
Highly Luminescent Ultranarrow Mn Doped ZnSe Nanowires
2009 Vol. 9, No. 2 745-750
Patrick T. K. Chin, Jan W. Stouwdam, and Rene´ A. J. Janssen* Molecular Materials and Nanosystems, EindhoVen UniVersity of Technology, P.O. Box 513, 5600 MB EindhoVen, The Netherlands Received November 2, 2008; Revised Manuscript Received January 2, 2009
ABSTRACT Colloidal Mn-doped ZnSe nanowires with diameters of 1-3 nm and lengths up to 200 nm were prepared from Li4[Zn10Se4(SPh)16] clusters and manganese stearate. The nanowires exhibit optical properties that depend on size, shape, and doping level. The manganese photoluminescence is slightly polarized perpendicular to the long axis and reaches a quantum yield of 40% after passivating the crystals with a CdSe shell.
Colloidal semiconductor nanocrystals with dimensions below the bulk exciton radius exhibit size-related optical and electronic properties.1-3 Their stable, tunable, bright, and narrow photoemission makes them favorable emitters for application in biomedical labeling,4 LEDs,5,6 and lasers.7,8 By introducing transition metal dopants, quantum size effects and atomic band transitions are combined, resulting in a novel class of materials with new electronic, optical, and magnetic properties.9 Spherical colloidal ZnS10,11 and ZnSe12-15 nanocrystals have frequently been doped with Mn2+ ions, but anisotropic-doped nanocrystals are rather exceptional. Thus far Mn-doped ZnS and ZnSe nanowires have been produced by templated growth and thermal deposition techniques,16-18 but the diameters (10-50 nm) of these one-dimensional structures were larger than the Bohr exciton radius of the bulk semiconductors. In fact, even undoped ZnSe wires with diameters below 10 nm are rare. Although various one-dimensional ZnSe nanowires19,20 and nanorods21 have been synthesized their dimensions were generally larger than the spatial confinement regime.22,23 Only recently, Panda et al.24 demonstrated the formation of ultranarrow (1.3 nm) nanorods and nanowires in a lyothermal synthesis at moderate temperatures. Here we report the reproducible solution-based synthesis to create the first colloidal Mn-doped wurtzite ZnSe nanowires with diameters in the confinement regime that have optical properties that depend on size, shape, and doping level. The ZnSe-Mn nanowires were prepared by reacting manganese stearate (MnSt2) with preformed Li4[Zn10Se4(SPh)16] single source clusters or by combining Li2[Zn4(SPh)10] clusters with elemental selenium.25,26 The latter method enabled the control over the Zn-Se ratio. The ZnSe-Mn nanowires are highly crystalline and can be * To whom correspondence should be addressed. E-mail: r.a.j.janssen@ tue.nl. 10.1021/nl8033015 CCC: $40.75 Published on Web 01/15/2009
2009 American Chemical Society
prepared with ultranarrow diameters from 1 to 3 nm, that is, well below the bulk exciton Bohr radius (3.8 nm) of ZnSe. By forming a CdSe shell, photoluminescence (PL) quantum yields up to 40% were obtained. The doped nanowires combine quantum size and shape effects with atomic band transitions, making them particular interesting for anisotropic optical studies and for anisotropic spin related studies. ZnSe-Mn nanoparticles were synthesized by slow heating of a mixture of Li4[Zn10Se4(SPh)16] and MnSt2 (5 mol% Mn compared to Zn) in hexadecylamine (HDA) from 353 to 563 K during 1 h (for full experimental details see Supporting Information). The particle growth is clearly reflected in the in the red shift of the absorption spectrum (Figure 1a) with increasing temperature, corresponding to an increase in particle diameter. The efficiency of Mn doping was confirmed by ICP-AES elemental analysis, which showed that the doping level in ZnSe-Mn nanocrystals corresponded to the ratios of Li2[Zn4(SPh)10] and MnSt2 added. The nanocrystals isolated at 543 K show PL emission at 590 nm (Figure 1b) with a quantum yield of about 3%. This emission is attributed to the 4T1 f 6A1 transition of Mn2+.14 At ∼380 nm, a weak ZnSe exciton emission is visible (Figure 1b) that exhibits a small Stokes shift compared to the corresponding exciton absorption (Figure 1a). During the initial stages of the particle growth, the absorption spectra show no sharp exciton absorption peak. Instead, an absorption peak between 310 and 320 nm is present, well below the onset of the absorption at ∼340 nm. This peak is due to very small, magic-sized, ZnSe clusters with a preferred thermodynamic stability27,28 that eventually disappear with increasing reaction time and temperature. The PL intensity of the exciton and Mn2+ emission show a significant increase with particle growth (Figure 1b). The maximum wavelength of the Mn2+ emission varies during the different stages of particle growth (Figure 2a). Initially there is a small blue
Figure 1. Evolution of the absorption (a) and PL (b) spectra during the growth of ZnSe-Mn particles with increasing temperature.
shift from 610 nm (sample 443 K) to 590 nm (sample 523 K). It has been established that Mn2+ ions in a ZnSe host matrix show a typical luminescence maximum at 590 nm,13 which shifts to 610 nm in the presence of thiolate ligands,29 as a result of a change in crystal field splitting. Hence the experiment in Figure 2a may indicate that the Mn2+ ions become effectively shielded from the surface, which is capped by thiophenols,30 upon growth of the ZnSe particle. The emission of particles heated above 543 K exhbits again a small red shift, which can be caused by decomposition of the thiophenol surfactants that decompose at high temperatures,31 introducing sulfide ions in ZnSe host crystal, leading to a change in crystal field. The decomposition of the thiophenol surfactants may also result in a decrease in surface passivation, leading to the observed decrease in PL intensity at 563 K compared to 543 K (Figure 1b). To confirm that the Mn2+ ions are actually embedded in the ZnSe host matrix we studied the nanocrystals with EPR spectroscopy. The X-band EPR spectrum of ZnSe-Mn at 120 K shows a six line spectrum superposed on a broad background (Figure 2b). The six line spectrum results from the hyperfine interaction of the unpaired electrons with the 55 Mn nuclear spin (I ) 5/2). The 55Mn hyperfine splitting is sensitive to the ions surrounding the Mn center and can be used to determine the position of the Mn atoms in the crystal. 746
Figure 2. (a) Normalized PL spectra of the Mn emission taken at different stages during particle growth. (b) EPR spectrum for ZnSe-Mn nanowires.
The hyperfine splitting of 63.0 × 10-4 cm-1 determined from the EPR spectrum is in close agreement with literature data for Mn in bulk ZnSe (61.7 × 10-4 cm-1).32,33 When the Mn2+ ions are located on the particle surface, an increase of the hyperfine splitting to 76.6 × 10-4 cm-1 is expected.15,9 The broad background in Figure 2b most likely results from the relative high Mn concentration ≈1.4%, which results in dipole-dipole interactions between the Mn centers that broaden the spectra.33 The EPR measurement and the blueshifted Mn emission at 590 nm lead to the conclusion that at reaction temperatures between 523 and 543 K most of the Mn atoms are located inside the ZnSe host. The X-ray diffraction pattern of the ZnSe-Mn nanocrystals is shown in Figure 3a. Within the angular resolution of the experiment, wurtzite and zinc blende structured ZnSe have several overlapping diffraction peaks, especially the (002), (110), and (112) peaks of wurtzite are close to the (111), (220), and (311) peaks of zinc blende. However the (100) and (101) peaks for wurtzite in Figure 3a have no equivalent for zinc blende and give credence to assume that the ZnSe-Mn nanocrystals have a wurtzite structure. On the other hand, some weak peaks expected for wurtzite are not visible, viz. the (102) at ∼36° and the (103) at ∼50°. This may be caused by attenuation of the (102) and (103) peaks Nano Lett., Vol. 9, No. 2, 2009
Figure 3. (a) Powder X-ray (λ ) 1.54056 Å) diffraction pattern of ZnSe-Mn (5% Mn-Zn). (b) TEM image of Mn-doped ZnSe nanodots and nanorods (scale bar is 50 nm). (c) TEM image of ZnSe-Mn nanorods made with extra TBP-Se (scale bar is 50 nm), the inset shows a high resolution TEM image (scale bar is 10 nm).
by the combination of a small diameter, resulting also in broad (110) and (112) peaks and some stacking faults producing short zinc blende domains. The high intensity of the (002) peak, together with broad (110) and (112) peaks, suggest an anisotropic growth in the (002) direction with a narrow particle width.
The shape and dimensions of the particles were further investigated with TEM (Figure 3b). This revealed that a mixture of spherical and rod-shaped particles was formed using the preformed Li4[Zn10Se4(SPh)16] clusters, the width of the rods and the diameter of the spherical particles was the same, explaining the clear exciton absorption peak in the absorption and spectrum (Figure 1a). The nanowire diameter observed by TEM increased from ∼1 to ∼3 nm (563 K, 1 h) with increasing reaction time and temperature. In a second experiment, the concentration of Se in the reaction mixture was increased by adding tributylphosphine selenium (TBPSe) to the Li4[Zn10Se4(SPh)16] clusters at the start of the reaction to obtain a 1:1 Zn-Se molar ratio. This resulted in the formation of mainly ZnSe-Mn nanorods with aspect ratios up to 17 (Figure 3c). The high resolution images show that the rods are single crystalline. Because TBPSe is not very reactive in combination with Li4[Zn10Se4(SPh)16], we used Li2[Zn4(SPh)10] clusters containing only Zn and combined these with elemental selenium and MnSt2 (5 mol% Mn compared to Zn) in HDA solution to study the influence of the Se concentration on the formation of ZnSe and ZnSe-Mn doped spherical or rodshaped nanocrystals in more detail. Heating this mixture from room temperature resulted in a rapid incorporation of Se into the clusters. The particle growth and formation took place during the following hour, when the reaction mixture was heated to 563 K (4 K/5 min). Figure 4 shows TEM images of the particles formed with different Zn-Se ratios. At high Zn-Se ratios (1:0.1), predominantly spherical particles were formed, but formation of nanowires rapidly set in when the amount of Se was increased. The highest yield of nanowires was obtained for a Zn-Se ratio of 1:0.3. Together with the increase in wire formation also the length and aspect ratio of the wires increased to 200 nm and 80, respectively. When the Se concentration was further increased, spherical particles started to appear again and at a Zn-Se ratio of 1:1 only spherical particles were formed. At even higher Se concen-
Figure 4. TEM images of ZnSe-Mn nanocrystals synthesized with different Zn-Se ratios. Nano Lett., Vol. 9, No. 2, 2009
747
Figure 6. Polarized PL spectra from macroscopically aligned ZnSe-Mn nanowires in cyclohexane, measured parallel and perpendicular to the flow direction. The spectra were recorded with linearly polarized excitation at 366 nm.
Figure 5. (a) PL and (b) absorption spectra of ZnSe-Mn-CdSe nanowires at different stages during CdSe shell growth. The labels a-f correspond to alternate additions of either Se or Cd to the mixture (cumulative amounts are given).
tration, uncontrolled anisotropic growth occurred. We note that the Mn concentration showed no influence on the formation of the nanocrystals; without Mn similar ratios between spherical and rod shaped crystals were found. The formation of the nanowires as shown in Figure 4 may involve different mechanisms, that is, oriented attachment of small crystals34-36 or specific binding of surfactants to different facets of wurtzite crystals, giving rise to oriented crystal growth.24 The formation of nanowires by oriented attachment can occur when dipoles are created in the ZnSe cluster when opposite crystal facets are terminated with zinc or selenium. This mechanism is very likely operative here because the dot versus wire formation strongly depends on the Zn-Se ratio. As the reactivity of the zinc and selenium source is not the same, an optimum nanowire formation is found at 1:0.3 Zn-Se. A further indication for this mechanism is the fact that at all Zn-Se ratios used there is always a certain fraction of spherical particles formed. This is consistent with the notion that oriented attachment is a statistical effect and not all ZnSe clusters will form wires. Furthermore, several TEM images (e.g., Figure 4, sample 1:0.8) clearly show pearl necklace-like nanowires. This is direct evidence of the oriented attachment mechanism, 748
considering that spherical particles with similar diameter are present.34-38 The other mechanism, anisotropic growth,39,40 has been previously been suggested by Panda et al.24 for wurtzite ZnSe nanowires made from zinc acetate and selenourea in the presence of weak binding octadecylamine. Here, anisotropic growth was explained by the stronger binding of the octadecylamine to the long side walls and the zinc terminated tips of the rods, compared to the selenide terminated tip, forcing one-dimensional growth. This growth mechanism implies that the kinetics of the alkylamine adsorption to the mineral surface is significantly slower than the growth of the ZnSe mineral phase.24 Although in the anisotropic growth the Zn-Se concentration will be an important parameter in both mechanisms, the presence of spherical particles together with the pearl necklace-like wires indicates that oriented attachment is expected to be the main mechanism for wire formation. To increase the PL quantum yield of the ZnSe-Mn nanocrystals, the surface was passivated by the formation of CdSe on the nanocrystal surface. The CdSe shell growth was performed in a mixture of HDA and trioctylphosphine oxide (TOPO) at 493 K, using TBPSe and anhydrous cadmium acetate (CdAc2) dissolved in trioctylphosphine (TOP). The addition of first Se followed by Cd resulted immediately in a strong increase luminescence, but only after the addition of both precursors. Subsequent alternate additions of Se and Cd resulted in a further increase of PL quantum yield up to ∼40% (∼24 µmol Se and ∼25 µmol Cd to ∼145 µmol (21 mg) of ZnSe-Mn nanocrystals). The effect of adding either Se or Cd was studied in more detail by monitoring the absorption and luminescence spectra at time intervals of ∼15 min between each addition at 493 K. The PL spectrum (Figure 5a) reveals that the major increase in luminescence intensity occurs upon the addition of Cd whereas the addition of Se only results in a much smaller increase in PL intensity. The PL increase by adding Cd, however, is somewhat variable as can be seen from the small decrease when the second aliquot of Cd is added. The third aliquot causes a 2-fold increase, resulting in a total 55-fold increase in Mn emission intensity compared to the original Nano Lett., Vol. 9, No. 2, 2009
ZnSe-Mn core emission. The absorption spectrum shows also a clear red shift with increasing CdSe shell growth and, again the major red shift is observed when Cd is added, while addition of Se has no significant influence on the absorption spectrum. (Figure 5b). After passivating the nanowires with a thin layer of CdSe the PL quantum yield increased from ∼3 to 40%. After size selective precipitation (see Supporting Information) to remove the smaller quantum dots, the nanowires remained highly luminescent and showed no remarkable difference in quantum yield. The absorption and luminescence of anisotropic nanocrystals is known to show a clear polarization.41-43 To study anisotropic optical properties of dopant emission, we aligned size-selected ZnSe-Mn-CdSe nanowires using a 1 mm flow cell. Dynamic alignment of the 1D wires along the flow direction in the cell was achieved by applying a continuous flow of ZnSe-Mn-CdSe nanowires dispersed in cyclohexane. The PL spectra obtained for the excitonic and atomic Mn emission using linearly polarized excitation at 366 nm were corrected for both the polarization dependence of light source and detector caused by the gratings.44 The corrected polarized PL spectra (Figure 6) show a weak polarization dependence of the Mn emission. The highest PL intensity is found perpendicular to the flow direction and hence orthogonal to the long axis of the nanowires. A polarization dependence of both photoexcitation and photoemission has previous been shown for CdSe, CdS, InP, and InAs nanorods and nanowires.6,41,45-48 The emission in these materials however is polarized parallel along the long axis of the rod or wire. In contrast, Golan et al.43 established a polarization perpendicular to the long axis of narrow wurtzite ZnSe wires and discussed the origin of is effect this in terms of a possible change in the order of the valence sub-bands. We note that in our experiment the polarized emission corresponds to an atomic 4T1 f 6A1 Mn transition. The origin of the polarization is unclear at present. The small diameter of the wires likely excludes that dielectric effects47 are important and suggests that the explanation be sought in the electronic structure of the wire. In summary we synthesized ultranarrow Mn doped ZnSe nanowires with an aspect ratio above 80. The nanowire formation is controlled by the ratio between Zn and Se in the reaction mixture. The ZnSe-Mn nanowires are highly monodisperse in the width direction and show an almost pure Mn2+ emission at 590 nm. PL and EPR spectroscopy indicate that the manganese dopant is incorporated in the ZnSe matrix, rather than at the surface. The PL quantum yield of the ZnSe-Mn nanowires is 40% after passivation with a thin CdSe shell. Macroscopic alignment of the nanowires in a continuous flow cell revealed a small polarization of the manganese emission perpendicular to the long axis. The nanowires presented in this work reveal the combined effects of size, shape, shell, and transition metal doping on the photoluminescence in a single nanocrystal. Acknowledgment. This research has been supported by NanoNed and by the Interreg program OLED+. Nano Lett., Vol. 9, No. 2, 2009
Supporting Information Available: Experimental details on the cluster synthesis, particle synthesis, and characterization. Additional results on the size selective precipitations to separate the spherical and wire shaped particles. This material is available free of charge via the Internet at http:// pubs.acs.org. References (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27) (28) (29) (30) (31) (32) (33) (34) (35)
Alivisatos, A. P. J. Phys. Chem. 1996, 100, 13226–13239. Alivisatos, A. P. Science 1996, 271, 933–937. Brus, L. E. J. Chem. Phys. 1983, 79, 5566–557. Michalet, X.; Pinaud, F. F.; Bentolila, L. A.; Tsay, J. M.; Doose, S.; Li, J. J.; Sundaresan, G.; Wu, A. M.; Gambhir, S. S.; Weiss, S. Science 2005, 307, 538–544. Colvin, V. L.; Schlamp, M. C.; Alivisatos, A. P. Nature 1994, 370, 354–357. Hikmet, R. A. M.; Chin, P. T. K.; Talapin, D. V.; Weller, H. AdV. Mater. 2005, 17, 1436–1439. Klimov, V. I.; Mikhailovski, A. A.; Xu, S.; Malko, A.; Hollingsworth, J. A.; Leaterdale, C. A; Eisler, H. J.; Bawendi, M. G. Science 2000, 290, 314–317. Kazes, M.; Lewis, D. Y.; Ebenstein, Y.; Mokari, T.; Banin, U. AdV. Mater. 2002, 14, 317–321. Norris, D. J.; Efros, A. L.; Erwin, S. C. Science 2008, 319, 1776– 1779. Becker, W. G.; Bard, A. J. J. Phys. Chem. 1983, 87, 4888–4893. Bhargava, R. N.; Gallagher, D.; Hong, X.; Nurmikko, A. Phys. ReV. Lett. 1994, 72, 416–419. Pradhan, N.; Goorskey, D.; Thessing, J.; Peng, X. J. Am. Chem. Soc. 2005, 127, 17586–17587. Suyver, J. F.; Wuister, S. F.; Kelly, J. J.; Meijerink, A. Phys. Chem. Chem. Phys. 2000, 2, 5445–5448. Norris, D. J.; Yao, N.; Charnock, F. T.; Kennedy, T. A. Nano Lett. 2001, 1, 3–7. Hofmann, A.; Graf, C.; Boeglin, C.; Ru¨hl, E. ChemPhysChem 2007, 8, 2008–2012. Ramanathan, S.; Patibandla, S.; Bandyopadhyay, S.; Anderson, J.; Edwards, J. D. Nanotechnology 2008, 19, 195601. Radovanovic, P. V.; Barrelet, C. J.; Gradecˇak, S.; Qian, F.; Lieber, C. M. Nano Lett. 2005, 5, 1407–1411. Jia, T.; Baba, M.; Huang, M.; Zhao, F.; Qiu, J.; Wu, X.; Ichihara, M.; Suzuki, M.; Li, R.; Xu, Z.; Kuroda, H. Solid State Commun. 2007, 141, 635–638. Duan, X. F.; Lieber, C. M. AdV. Mater. 2000, 12, 298–302. Xiang, B.; Zhang, H. Z.; Li, G. H.; Su, F. H.; Wang, R. M.; Xu, J.; Lu, G. W.; Sun, X. C.; Zhao, Q.; Yu, D. P. Appl. Phys. Lett. 2003, 82, 3330–3332. Cozzoli, P. D.; Manna, L.; Curri, L.; Kudera, S.; Giannini, C.; Striccoli, M.; Agostiano, A. Chem. Mater. 2005, 17, 1296–1306. Lv, R.; Cao, C.; Zhai, H.; Wang, D.; Liu, S.; Zhu, H. Solid State Commun. 2004, 130, 241–245. Zhu, Y. C.; Bando, Y. Chem. Phys. Lett. 2003, 377, 367–370. Panda, A. B.; Acharya, S.; Efrima, S. AdV. Mater. 2005, 17, 2471– 2474. Dance, I. G.; Choy, A.; Scudder, M. L. J. Am. Chem. Soc. 1984, 106, 6285–6295. Cumberland, S. L.; Hanif, K. M.; Javier, A.; Khitrov, G. A.; Strouse, G. F.; Woessner, S. M.; Yun, C. S. Chem. Mater. 2002, 14, 1576– 1584. Rogach, A. L.; Katsikas, L.; Kornowski, A.; Su, D.; Eychmu¨ller, A.; Weller, H. Ber. Bunsen-Ges. Phys. Chem. 1997, 101, 1668–1670. Wuister, S. F.; Driel, F.; Meijerink, A. J. Lumin. 2003, 102, 327– 332. Pradhan, N.; Peng, X. J. Am. Chem. Soc. 2007, 129, 3339–3347. Chin, P. T. K.; Stouwdam, J. W.; van Bavel, S. S.; Janssen, R. A. J. Nanotechnology 2008, 19, 205602. Archer, P. I.; Santangelo, S. A.; Gamelin, D. R. J. Am. Chem. Soc. 2007, 129, 9808–9818. Ludwig, G. W.; Woodbury, H. H. Electron Spin Resonance in Semiconductors In Solid State Physics; Seitz, F., Turnbull, D., Eds.; Academic Press: New York, 1962; Vol. 13, p 298. Kennedy, T. A.; Glaser, E. R.; Klein, P. B.; Bhargava, R. N. Phys. ReV. B 1995, 52, R14356–R14359. Cho, K.-S.; Talapin, D. V.; Gaschler, W.; Murray, C. B. J. Am. Chem. Soc. 2005, 127, 7140–7147. Pradhan, N.; Xu, H.; Peng, X. Nano Lett. 2006, 6, 720–724. 749
(36) Ribeiro, C.; Lee, E. J. H.; Longo, E.; Leite, E. R. ChemPhysChem 2006, 7, 664–670. (37) Tang, Z.; Kotov, N. A.; Giersig, M. Science 2002, 297, 237–240. (38) Pacholski, C.; Kornowski, A.; Weller, H. Angew. Chem., Int. Ed. 2002, 41, 1188–1191. (39) Yu, W. W.; Wang, Y. A.; Peng, X. G. Chem. Mater. 2003, 15, 4300– 4308. (40) Manna, L.; Wang, L. W.; Cingolani, R.; Alivisatos, A. P. J. Phys. Chem. B 2005, 109, 6183–6192. (41) Hu, J. T.; Li, L. S.; Yang, W. D.; Manna, L.; Wang, L. W.; Alivisatos, A. P. Science 2001, 292, 2060–2063. (42) Manna, L.; Scher, E. C.; Alivisatos, A. P. J. Cluster Sci. 2002, 13, 521–523. (43) Somobrata, A.; Panda, A. B.; Efrima, S.; Golan, Y. AdV. Mater. 2007, 19, 1105–1108 .
750
(44) Lakowicz, J. R. Principles of Fluorescence Spectroscopy, 3rd ed.; Springer: New York, 2006. (45) Peng, X.; Manna, L.; Yang, W.; Wickham, J.; Scher, E.; Kadavanich, A.; Alivisatos, A. P. Nature 2000, 404, 59–61. (46) Talapin, D. V.; Koeppe, R.; Go¨tzinger, S.; Kornowski, A.; Lupton, J. M.; Rogach, A. L.; Benson, O.; Feldmann, J.; Weller, H. Nano Lett. 2003, 3, 1677–1681. (47) Wang, J.; Gudiksen, M. S.; Duan, X.; Cui, Y; Lieber, C. M. Science 2001, 293, 1455–1457. (48) Kan, S. H.; Mokari, T.; Rotheberg, T.; Bannin, U. Nat. Mater. 2003, 2, 155–158.
NL8033015
Nano Lett., Vol. 9, No. 2, 2009