ChemComm
Published on 13 February 2014. Downloaded by University of Windsor on 19/10/2014 12:36:43.
COMMUNICATION
Cite this: Chem. Commun., 2014, 50, 3480 Received 18th December 2013, Accepted 11th February 2014
View Article Online View Journal | View Issue
Low temperature synthesis and characterization of lanthanide-doped BaTiO3 nanocrystals† Sean P. Culver,a Viktor Stepanov,a Matthew Mecklenburg,b Susumu Takahashia and Richard L. Brutchey*a
DOI: 10.1039/c3cc49575b www.rsc.org/chemcomm
The vapor diffusion sol–gel (VDSG) method was employed for the room-temperature synthesis of B10 nm, aliovalently doped 0.4, 0.8, and 1.6 mol% La:BaTiO3 and 0.4, 0.6, and 1.2 mol% Dy:BaTiO3 nanocrystals. Maximum ensemble relative permittivities of 176 and 208 were observed in the 0.8 mol% La:BaTiO3 and the 1.2 mol% Dy:BaTiO3 nanocrystals, respectively, relative to 89 for undoped BaTiO3 (at 1 MHz, 25 8C) due to local disorder induced by aliovalent substitution.
Doping small concentrations of lanthanide (Ln) ions into bulk ABO3 perovskites (e.g., BaTiO3) has proven useful in modulating the associated dielectric,1,2 electrical,3 and optical properties4,5 in these materials. Three modes of aliovalent substitution within BaTiO3 are believed to be possible based on the ionic radius. Studies have demonstrated that large ions (e.g., La3+)6 substitute exclusively at the 12-coordinate A-site, while small ions (e.g., Yb3+)6 solely occupy the 6-coordinate B-site. Ions with intermediate ionic radii (e.g., Eu3+, Gd3+, Dy3+) are said to be amphoteric, whereby these intermediate sized Ln3+ can substitute at the A-site, the B-site, or both sites.7 Upon substitution, the charge imbalance introduced by the lattice defects must be compensated for in order to maintain charge neutrality. Thus, Ln3+ substitutions at the A-site can generate B-site vacancies and B-site substitutions can result in oxygen deficiency.8 Incorporation of La3+ within bulk BaTiO3 has been shown to occupy the A-site, induce B-site vacancies, and enhance the roomtemperature permittivity.9 Additionally, given the intermediate ionic radius of Dy3+ compared to the host lattice atoms (i.e., r(Ba2+) = 1.61 Å and r(Ti4+) = 0.61 Å for 12- and 6-coordinate environments, respectively6), it is believed to exhibit amphoteric character, while also elevating the associated room-temperature permittivity.10,11 a
Department of Chemistry, University of Southern California, Los Angeles, California 90089, USA. E-mail: [email protected] b Center for Electron Microscopy and Microanalysis, University of Southern California, Los Angeles, California 90089, USA † Electronic supplementary information (ESI) available: Experimental details for nanocrystal synthesis and materials characterization; TGA thermograms, TEM images, size distribution histograms, SAED patterns, and EPR spectra for Ln:BaTiO3 nanocrystals. See DOI: 10.1039/c3cc49575b
3480 | Chem. Commun., 2014, 50, 3480--3483
Generally, the aforementioned substitutions are achieved through high temperature (41000 1C) solid state synthesis. Though effective, such high temperature routes cause particle sintering and micron-sized grains.12 As electronic devices become smaller and smaller, the ability to achieve functional and processable nanocrystals with tunable properties is becoming more important.13 Therefore, the ability to aliovalently dope nanocrystals of BaTiO3 and related perovskites at low temperatures to prevent sintering is highly desirable. The vapor diffusion sol–gel method allows for the lowtemperature synthesis of functional metal oxide nanocrystals under ultra-benign conditions.14–16 The controlled flow of water vapor over an alcohol solution of metal alkoxide and metal acetylacetonate (acac) precursors induces their hydrolysis and polycondensation to nucleate and grow nanocrystals. Herein, the synthesis of xDy:BaTiO3 and xLa:BaTiO3 (0 r x o 2 mol%) occurs via kinetically controlled hydrolysis and cross polycondensation within alcohol solutions of BaTi(OR)6 (R = CH2CHCH3OCH3) and Ln(acac)3 precursors to yield the resulting nanocrystals at room temperature. Overall ceramic yields for the vapor diffusion sol–gel reactions were estimated by mass balance and thermal gravimetric analysis (TGA) to be 68, 73, and 76% for nominal 0.5, 1.0, and 2.0 mol% Dy:BaTiO3 and 69, 69, and 76% for nominal 0.5, 1.0, and 2.0 mol% La:BaTiO3, respectively. The organic content of the isolated nanocrystals was determined to be o8 wt% by TGA and can be attributed to unreacted surface alkoxy groups (ESI†). The crystallinity and phase purity of the resulting Dy:BaTiO3 and La:BaTiO3 nanocrystals were confirmed by powder X-ray diffraction (XRD). As shown in Fig. 1, all reflections can be indexed to the cubic perovskite structure with lattice constant a B 4.03 Å, belonging to the paraelectric Pm3% m space group (JCPDS no. 31-0174). Segregation of secondary crystalline carbonate phases, or Ln2Ti2O7 pyrochlore phases typically observed in Ln3+-doped BaTiO3 ceramics synthesized via high-temperature solid-state reactions,7 were not observed. Furthermore, no differences in the crystallinity or lattice constants were observed in any of the studied compositions, as indicated by the XRD patterns. Post-synthetic thermal treatment of the nanocrystals synthesized under ambient
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 13 February 2014. Downloaded by University of Windsor on 19/10/2014 12:36:43.
Communication
Fig. 1 Powder XRD patterns of xLn:BaTiO3 nanocrystals synthesized at room temperature.
conditions was not required to induce crystallization and incorporate the Ln3+ ions. We have recently demonstrated by synchrotron X-ray diffraction and total scattering that lanthanide dopants incorporate into the perovskite lattice (as opposed to forming core–shell structures) under these synthesis conditions, as evidenced by slight changes in the lattice parameter at low dopant concentrations.17 Herein, both inductively coupled plasma-optical emission spectroscopy (ICP-OES) and dielectric characterization (vide infra) suggest that the lanthanide is incorporated into the BaTiO3 nanocrystal host. Elemental analysis was performed with ICP-OES to quantify the La3+ and Dy3+ concentrations in the La:BaTiO3 and Dy:BaTiO3 nanocrystals, respectively. The La3+ and Dy3+ concentrations were found to be 0.4, 0.8, and 1.6 mol% and 0.4, 0.6, and 1.2 mol%, respectively, for nominal 0.5, 1.0, and 2.0 mol% Ln:BaTiO3 nanocrystals. Both of the lanthanide dopants exhibited less than unity incorporation into the host lattice against the nominal doping concentration, achieving doping efficiencies of 60–80% relative to nominal under these benign synthesis conditions. The morphology of the Dy:BaTiO3 and La:BaTiO3 nanocrystals was investigated by transmission electron microscopy (TEM). The nanocrystals possess a quasispherical shape with relatively monodisperse size distributions (s = 16–19%, N = 100) for nanocrystals synthesized at low temperature (Fig. 2, ESI†). The average diameters
Fig. 2 TEM images of (a) 0.6 mol% Dy:BaTiO3 and (b) 0.8 mol% La:BaTiO3 nanocrystals. High-resolution TEM images are provided in the insets; the corresponding lattice fringe d-spacing and lattice planes are indicated.
This journal is © The Royal Society of Chemistry 2014
ChemComm
were found to be 9.8 1.8, 10.1 1.6, and 10.2 1.9 nm for the 0.4, 0.6, and 1.2 mol% Dy:BaTiO3 nanocrystals and 9.8 1.9, 10.1 1.7, and 9.9 1.7 nm for the 0.4, 0.8, and 1.6 mol% La:BaTiO3 nanocrystals (ESI†). The lanthanide dopants did not have a significant effect on the size or shape of the resulting nanocrystals. The presence of well-defined lattice fringes corresponding to the (100) and/or (110) lattice planes in the highresolution TEM images suggest that the nanocrystals are composed of single crystalline domains (Fig. 2 insets). Additionally, selected area electron diffraction patterns were collected for all compositions and indexed to the cubic perovskite phase (ESI†), corroborating the XRD data. The ensemble dielectric properties of the Dy:BaTiO3 and La:BaTiO3 nanocrystals were measured from composite parallel plate capacitors obtained by pressing a mixture of the Ln:BaTiO3 nanocrystals with polyvinyl alcohol (PVA) into cylindrical pellets and coating both sides with silver contacts. Bruggeman’s effective medium model18 was employed to extrapolate the dielectric constant of the unsintered nanocrystal component: n X i¼1
ei eeff v ¼0 ei þ ðn 1Þeeff
(1)
The effective dielectric constant was modeled as a three component system (i.e., n = 3) consisting of volume fractions (v) of air, PVA, and Dy:BaTiO3 or La:BaTiO3 nanocrystals. The dielectric constants used for air and PVA were 1.00 and 1.95, respectively. It should be noted that high temperature annealing was not performed on the pellets in order to prevent nanocrystal sintering and to more accurately estimate the true ensemble dielectric constant of the B10 nm nanocrystals prepared at room temperature. Dielectric permittivity and loss tangent measurements on the Ln:BaTiO3 nanocrystals were conducted in a frequency range of 1 kHz to 2 MHz at 25 1C under nitrogen (Fig. 3). The undoped
Fig. 3 Relative permittivity (e) and dielectric loss tangent (tan d) of the Ln:BaTiO3 nanocrystals as a function of frequency. All measurements were conducted at 25 1C under a nitrogen atmosphere in a frequency range of 1 kHz to 2 MHz.
Chem. Commun., 2014, 50, 3480--3483 | 3481
View Article Online
Published on 13 February 2014. Downloaded by University of Windsor on 19/10/2014 12:36:43.
ChemComm
Communication
BaTiO3 nanocrystals exhibited a relative permittivity of 89 and a loss tangent of 0.017 at 1 MHz. Incorporation of 0.4 mol% La3+ did not significantly affect the nanocrystal permittivity, which exhibited a value of 103 at 1 MHz (Fig. 3a). The onset of the effect of La3+ concentration on the permittivity was found to occur at 0.8 mol%. Here, the permittivity is two times greater than undoped BaTiO3, with a relative permittivity of 176 at 1 MHz. As the concentration of La3+ is further increased to 1.2 mol%, the relative permittivity slightly decreases by 16% to 148. For the Dy:BaTiO3 nanocrystals, Dy3+ doping does not have a substantial effect on the relative permittivity below a concentration of 0.6 mol% (Fig. 3b), in agreement with previous reports on Dy:BaTiO3 bulk ceramics.10,11 The relative permittivity for the 0.4 and 0.6 mol% Dy:BaTiO3 compositions were 114 and 104, respectively, at 1 MHz. Upon increasing the Dy3+ concentration to 1.2 mol%, the relative permittivity increased to 208, which is nearly two and a half times greater than that of undoped BaTiO3. To test against the formation of amorphous Ln2O3 shells, or related species, being the cause of the observed dielectric properties, we subjected pre-formed, undoped BaTiO3 nanocrystals to a second vapor diffusion sol–gel reaction in the presence of Dy(acac)3 (nominal 2.0 mol%; 25 1C, 48 h). The dielectric constant of the BaTiO3 nanocrystals surface treated with Dy(acac)3 decreased by 18% relative to the untreated BaTiO3 nanocrystals, suggesting that the surface species resulting from such a reaction are not the cause of the observed dielectric effects. The dielectric loss tangents were between 0.02–0.05 for all Ln:BaTiO3 nanocrystals (i.e., greater than the undoped BaTiO3 nanocrystals for all compositions). The increase in permittivity may be attributed to local disorder promoted by aliovalent substitution. It has been computationally modeled for bulk ceramics that the titanium octahedra distort to electrostatically stabilize the lattice because of B-site vacancies, generating local polarizations that enhance the relative permittivity.19 As previously mentioned, La3+ is known to substitute at the A-site, whereas Dy3+ can substitute either the A- or B-site exclusively or both the A- and B-sites. Introducing trivalent ions into the A-site generates Ti4+ vacancies to accommodate the charge imbalance caused by the defect, according to:20 4LaBa - VTi 0 0 0 0
(2)
The Ti4+ vacancies distort the Ti–O bonding in the surrounding octahedra.19 Another possible charge compensation mechanism involves the reduction of Ti4+ to Ti3+. In order to elucidate which mechanism is occurring, electron paramagnetic resonance (EPR) spectra were collected at 78 K on undoped BaTiO3, 1.6 mol% La:BaTiO3, and 1.2 mol% Dy:BaTiO3 nanocrystals (ESI†). Beyond the observed sextet indicative of Mn2+, the absence of any paramagnetic centers in the g-value range of 1.907–1.974 belonging to Ti3+ suggests that the concentration of reduced titanium is below the EPR detection limit. Paramagnetic centers were observed at g-values of 2.004–2.005 for all three samples, which correspond to the presence of titanium vacancies (VTi).21,22 Thus, the formation of titanium vacancies upon A-site substitution is the dominant charge compensation mechanism occurring in this system. Regarding Dy3+ doping, a recent study by Rabuffetti et al. on
3482 | Chem. Commun., 2014, 50, 3480--3483
Eu3+ aliovalently doped into BaTiO3 nanocrystals reported that Eu3+ (r = 1.29 Å for a 12-coordinate environment; extrapolated value6) substitutes onto the A-site for concentrations r1 mol%, with a transition to A- and B-site substitution at 42 mol%.17 Due to the similar ionic radii of Eu3+ and Dy3+, these results suggest that Dy3+ may be preferentially occupying the A-site at the doping level investigated here; however, high resolution structural characterization is required to verify this hypothesis. With regards to the frequency dependent dielectric properties, the undoped BaTiO3 showed a permittivity of 103 at 1 kHz, which decreased to 91 by 100 kHz. Such permittivity frequency dependence results from interfacial polarization at lower frequencies.23 All compositions exhibited minimal frequency dependence, with reductions in permittivity ranging from 13–20% by 100 kHz. Following the low frequency dependence, the nanocrystals were stable in permittivity up to 2 MHz. Dielectric loss tangents between 0.05–0.08 were observed in the Dy:BaTiO3 and La:BaTiO3 nanocrystals at 1 kHz, which decreases to between 0.02–0.05 by 100 kHz. In summary, the VDSG method was employed to synthesize aliovalently doped La:BaTiO3 and Dy:BaTiO3 nanocrystals at room temperature under ultra-benign conditions. Dielectric properties and ICP-OES confirmed that the lanthanide ions could be incorporated at room temperature without the need for post-synthetic annealing. The size-controlled, quasispherical nanocrystals appear to be single crystalline, as indicated by high-resolution TEM. Maximum relative permittivities of 176 and 206 were obtained for the 0.8 mol% La:BaTiO3 and 1.2 mol% Dy:BaTiO3 nanocrystals, respectively, at 25 1C and 1 MHz. While we have previously demonstrated that the relative permittivity of BaTiO3 nanocrystals can be tuned by isovalent substitution of Ba2+ and Ti4+ with Sr2+ and Zr4+, respectively, these solid solutions require substitution of ca. 33 mol% Sr2+ for Ba2+ or 15 mol% Zr4+ for Ti4+ to achieve maximum relative permittivity.24–26 Therefore, much lower aliovalent substitution levels can achieve similar effects on the dielectric properties of BaTiO3 nanocrystals using the vapor-diffusion sol–gel route. This work is supported by the Department of Energy, Office of Basic Energy Sciences, under Grant No. DE-FG02-11ER46826. S.P.C. acknowledges the National Science Foundation for a Graduate Research Fellowship. We thank Prof. P. Qin for use of his EPR spectrometer (supported by NIH S10 RR028992). S.T. acknowledges financial support from the Searle Scholars Program.
Notes and references 1 A. R. West, T. B. Adams, F. D. Morrison and D. C. Sinclair, J. Eur. Ceram. Soc., 2004, 24, 1439. 2 A. J. Moulson and J. M. Herbert, Electroceramics, Chapman and Hall, London, 1990. 3 M. D. Glinchuk, I. P. Bykov, S. M. Kornienko, V. V. Laguta, A. M. Slipenyuk, A. G. Bilous, O. I. V’yunov and O. Z. Yanchevskii, J. Mater. Chem., 2000, 10, 941. 4 J. P. Vernon, N. Hobbs, Y. Cai, A. Lethbridge, P. Vukusic, D. D. Deheyn and K. H. Sandhage, J. Mater. Chem., 2012, 22, 10435. 5 F. A. Rabuffetti, J. S. Lee and R. L. Brutchey, Adv. Mater., 2012, 24, 1434. 6 R. D. Shannon, Acta Crystallogr., 1976, A32, 751. 7 D. Y. Lu, T. Koda, H. Suzuki and M. Toda, J. Ceram. Soc. Jpn., 2005, 113, 721.
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 13 February 2014. Downloaded by University of Windsor on 19/10/2014 12:36:43.
Communication 8 C. L. Freeman, J. A. Dawson, H. R. Chen, J. H. Harding, L. B. Ben and D. C. Sinclair, J. Mater. Chem., 2011, 21, 4861. 9 F. D. Morrison, D. C. Sinclair, J. M. S. Skakle and A. R. West, J. Am. Ceram. Soc., 1998, 81, 1957. 10 D. W. Kang, T. G. Park, J. W. Kim, J. S. Kim, H. S. Lee and H. Cho, Electron. Mater. Lett., 2010, 6, 145. 11 Y. Pu, W. Chen, S. Chen and H. T. Langhammer, Ceramica, 2005, 51, 214. 12 F. D. Morrison, D. C. Sinclair and A. R. West, J. Appl. Phys., 1999, 86, 6355. 13 A. Testino, Int. J. Appl. Ceram. Technol., 2013, 10, 723. 14 S. P. Culver, F. A. Rabuffetti, S. Zhou, M. Mecklenburg, Y. Song, B. C. Melot and R. L. Brutchey, Chem. Mater., 2013, 25, 4129. 15 F. A. Rabuffetti, J. S. Lee and R. L. Brutchey, Chem. Mater., 2012, 24, 3114. 16 F. A. Rabuffetti and R. L. Brutchey, Chem. Commun., 2012, 48, 1437.
This journal is © The Royal Society of Chemistry 2014
ChemComm 17 F. A. Rabuffetti, S. P. Culver, J. S. Lee and R. L. Brutchey, Nanoscale, 2014, 6, 2909. 18 W. R. Tinga, W. A. Voss and D. F. Blossey, J. Appl. Phys., 1973, 44, 3897. 19 C. L. Freeman, J. A. Dawson, J. H. Harding, L. B. Ben and D. C. Sinclair, Adv. Funct. Mater., 2012, 23, 491. 20 F. D. Morrison, D. C. Sinclair and A. R. West, J. Am. Ceram. Soc., 2001, 84, 531. 21 T. Kolodiazhnyi and A. Petric, J. Phys. Chem. Solids, 2003, 64, 953. 22 T. D. Dunbar, W. L. Warren, B. A. Tuttle, C. A. Randall and Y. Tsur, J. Phys. Chem. B, 2004, 108, 908. 23 R. Z. Hou, P. Ferreira and P. M. Vilarinho, Chem. Mater., 2009, 21, 3536. 24 C. W. Beier, M. A. Cuevas and R. L. Brutchey, J. Mater. Chem., 2010, 20, 5074. 25 F. A. Rabuffetti and R. L. Brutchey, Chem. Commun., 2012, 48, 1437. 26 F. A. Rabuffetti and R. L. Brutchey, ACS Nano, 2013, 7, 11435.
Chem. Commun., 2014, 50, 3480--3483 | 3483
Published on 13 February 2014. Downloaded by University of Windsor on 19/10/2014 12:36:43.
COMMUNICATION
Cite this: Chem. Commun., 2014, 50, 3480 Received 18th December 2013, Accepted 11th February 2014
View Article Online View Journal | View Issue
Low temperature synthesis and characterization of lanthanide-doped BaTiO3 nanocrystals† Sean P. Culver,a Viktor Stepanov,a Matthew Mecklenburg,b Susumu Takahashia and Richard L. Brutchey*a
DOI: 10.1039/c3cc49575b www.rsc.org/chemcomm
The vapor diffusion sol–gel (VDSG) method was employed for the room-temperature synthesis of B10 nm, aliovalently doped 0.4, 0.8, and 1.6 mol% La:BaTiO3 and 0.4, 0.6, and 1.2 mol% Dy:BaTiO3 nanocrystals. Maximum ensemble relative permittivities of 176 and 208 were observed in the 0.8 mol% La:BaTiO3 and the 1.2 mol% Dy:BaTiO3 nanocrystals, respectively, relative to 89 for undoped BaTiO3 (at 1 MHz, 25 8C) due to local disorder induced by aliovalent substitution.
Doping small concentrations of lanthanide (Ln) ions into bulk ABO3 perovskites (e.g., BaTiO3) has proven useful in modulating the associated dielectric,1,2 electrical,3 and optical properties4,5 in these materials. Three modes of aliovalent substitution within BaTiO3 are believed to be possible based on the ionic radius. Studies have demonstrated that large ions (e.g., La3+)6 substitute exclusively at the 12-coordinate A-site, while small ions (e.g., Yb3+)6 solely occupy the 6-coordinate B-site. Ions with intermediate ionic radii (e.g., Eu3+, Gd3+, Dy3+) are said to be amphoteric, whereby these intermediate sized Ln3+ can substitute at the A-site, the B-site, or both sites.7 Upon substitution, the charge imbalance introduced by the lattice defects must be compensated for in order to maintain charge neutrality. Thus, Ln3+ substitutions at the A-site can generate B-site vacancies and B-site substitutions can result in oxygen deficiency.8 Incorporation of La3+ within bulk BaTiO3 has been shown to occupy the A-site, induce B-site vacancies, and enhance the roomtemperature permittivity.9 Additionally, given the intermediate ionic radius of Dy3+ compared to the host lattice atoms (i.e., r(Ba2+) = 1.61 Å and r(Ti4+) = 0.61 Å for 12- and 6-coordinate environments, respectively6), it is believed to exhibit amphoteric character, while also elevating the associated room-temperature permittivity.10,11 a
Department of Chemistry, University of Southern California, Los Angeles, California 90089, USA. E-mail: [email protected] b Center for Electron Microscopy and Microanalysis, University of Southern California, Los Angeles, California 90089, USA † Electronic supplementary information (ESI) available: Experimental details for nanocrystal synthesis and materials characterization; TGA thermograms, TEM images, size distribution histograms, SAED patterns, and EPR spectra for Ln:BaTiO3 nanocrystals. See DOI: 10.1039/c3cc49575b
3480 | Chem. Commun., 2014, 50, 3480--3483
Generally, the aforementioned substitutions are achieved through high temperature (41000 1C) solid state synthesis. Though effective, such high temperature routes cause particle sintering and micron-sized grains.12 As electronic devices become smaller and smaller, the ability to achieve functional and processable nanocrystals with tunable properties is becoming more important.13 Therefore, the ability to aliovalently dope nanocrystals of BaTiO3 and related perovskites at low temperatures to prevent sintering is highly desirable. The vapor diffusion sol–gel method allows for the lowtemperature synthesis of functional metal oxide nanocrystals under ultra-benign conditions.14–16 The controlled flow of water vapor over an alcohol solution of metal alkoxide and metal acetylacetonate (acac) precursors induces their hydrolysis and polycondensation to nucleate and grow nanocrystals. Herein, the synthesis of xDy:BaTiO3 and xLa:BaTiO3 (0 r x o 2 mol%) occurs via kinetically controlled hydrolysis and cross polycondensation within alcohol solutions of BaTi(OR)6 (R = CH2CHCH3OCH3) and Ln(acac)3 precursors to yield the resulting nanocrystals at room temperature. Overall ceramic yields for the vapor diffusion sol–gel reactions were estimated by mass balance and thermal gravimetric analysis (TGA) to be 68, 73, and 76% for nominal 0.5, 1.0, and 2.0 mol% Dy:BaTiO3 and 69, 69, and 76% for nominal 0.5, 1.0, and 2.0 mol% La:BaTiO3, respectively. The organic content of the isolated nanocrystals was determined to be o8 wt% by TGA and can be attributed to unreacted surface alkoxy groups (ESI†). The crystallinity and phase purity of the resulting Dy:BaTiO3 and La:BaTiO3 nanocrystals were confirmed by powder X-ray diffraction (XRD). As shown in Fig. 1, all reflections can be indexed to the cubic perovskite structure with lattice constant a B 4.03 Å, belonging to the paraelectric Pm3% m space group (JCPDS no. 31-0174). Segregation of secondary crystalline carbonate phases, or Ln2Ti2O7 pyrochlore phases typically observed in Ln3+-doped BaTiO3 ceramics synthesized via high-temperature solid-state reactions,7 were not observed. Furthermore, no differences in the crystallinity or lattice constants were observed in any of the studied compositions, as indicated by the XRD patterns. Post-synthetic thermal treatment of the nanocrystals synthesized under ambient
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 13 February 2014. Downloaded by University of Windsor on 19/10/2014 12:36:43.
Communication
Fig. 1 Powder XRD patterns of xLn:BaTiO3 nanocrystals synthesized at room temperature.
conditions was not required to induce crystallization and incorporate the Ln3+ ions. We have recently demonstrated by synchrotron X-ray diffraction and total scattering that lanthanide dopants incorporate into the perovskite lattice (as opposed to forming core–shell structures) under these synthesis conditions, as evidenced by slight changes in the lattice parameter at low dopant concentrations.17 Herein, both inductively coupled plasma-optical emission spectroscopy (ICP-OES) and dielectric characterization (vide infra) suggest that the lanthanide is incorporated into the BaTiO3 nanocrystal host. Elemental analysis was performed with ICP-OES to quantify the La3+ and Dy3+ concentrations in the La:BaTiO3 and Dy:BaTiO3 nanocrystals, respectively. The La3+ and Dy3+ concentrations were found to be 0.4, 0.8, and 1.6 mol% and 0.4, 0.6, and 1.2 mol%, respectively, for nominal 0.5, 1.0, and 2.0 mol% Ln:BaTiO3 nanocrystals. Both of the lanthanide dopants exhibited less than unity incorporation into the host lattice against the nominal doping concentration, achieving doping efficiencies of 60–80% relative to nominal under these benign synthesis conditions. The morphology of the Dy:BaTiO3 and La:BaTiO3 nanocrystals was investigated by transmission electron microscopy (TEM). The nanocrystals possess a quasispherical shape with relatively monodisperse size distributions (s = 16–19%, N = 100) for nanocrystals synthesized at low temperature (Fig. 2, ESI†). The average diameters
Fig. 2 TEM images of (a) 0.6 mol% Dy:BaTiO3 and (b) 0.8 mol% La:BaTiO3 nanocrystals. High-resolution TEM images are provided in the insets; the corresponding lattice fringe d-spacing and lattice planes are indicated.
This journal is © The Royal Society of Chemistry 2014
ChemComm
were found to be 9.8 1.8, 10.1 1.6, and 10.2 1.9 nm for the 0.4, 0.6, and 1.2 mol% Dy:BaTiO3 nanocrystals and 9.8 1.9, 10.1 1.7, and 9.9 1.7 nm for the 0.4, 0.8, and 1.6 mol% La:BaTiO3 nanocrystals (ESI†). The lanthanide dopants did not have a significant effect on the size or shape of the resulting nanocrystals. The presence of well-defined lattice fringes corresponding to the (100) and/or (110) lattice planes in the highresolution TEM images suggest that the nanocrystals are composed of single crystalline domains (Fig. 2 insets). Additionally, selected area electron diffraction patterns were collected for all compositions and indexed to the cubic perovskite phase (ESI†), corroborating the XRD data. The ensemble dielectric properties of the Dy:BaTiO3 and La:BaTiO3 nanocrystals were measured from composite parallel plate capacitors obtained by pressing a mixture of the Ln:BaTiO3 nanocrystals with polyvinyl alcohol (PVA) into cylindrical pellets and coating both sides with silver contacts. Bruggeman’s effective medium model18 was employed to extrapolate the dielectric constant of the unsintered nanocrystal component: n X i¼1
ei eeff v ¼0 ei þ ðn 1Þeeff
(1)
The effective dielectric constant was modeled as a three component system (i.e., n = 3) consisting of volume fractions (v) of air, PVA, and Dy:BaTiO3 or La:BaTiO3 nanocrystals. The dielectric constants used for air and PVA were 1.00 and 1.95, respectively. It should be noted that high temperature annealing was not performed on the pellets in order to prevent nanocrystal sintering and to more accurately estimate the true ensemble dielectric constant of the B10 nm nanocrystals prepared at room temperature. Dielectric permittivity and loss tangent measurements on the Ln:BaTiO3 nanocrystals were conducted in a frequency range of 1 kHz to 2 MHz at 25 1C under nitrogen (Fig. 3). The undoped
Fig. 3 Relative permittivity (e) and dielectric loss tangent (tan d) of the Ln:BaTiO3 nanocrystals as a function of frequency. All measurements were conducted at 25 1C under a nitrogen atmosphere in a frequency range of 1 kHz to 2 MHz.
Chem. Commun., 2014, 50, 3480--3483 | 3481
View Article Online
Published on 13 February 2014. Downloaded by University of Windsor on 19/10/2014 12:36:43.
ChemComm
Communication
BaTiO3 nanocrystals exhibited a relative permittivity of 89 and a loss tangent of 0.017 at 1 MHz. Incorporation of 0.4 mol% La3+ did not significantly affect the nanocrystal permittivity, which exhibited a value of 103 at 1 MHz (Fig. 3a). The onset of the effect of La3+ concentration on the permittivity was found to occur at 0.8 mol%. Here, the permittivity is two times greater than undoped BaTiO3, with a relative permittivity of 176 at 1 MHz. As the concentration of La3+ is further increased to 1.2 mol%, the relative permittivity slightly decreases by 16% to 148. For the Dy:BaTiO3 nanocrystals, Dy3+ doping does not have a substantial effect on the relative permittivity below a concentration of 0.6 mol% (Fig. 3b), in agreement with previous reports on Dy:BaTiO3 bulk ceramics.10,11 The relative permittivity for the 0.4 and 0.6 mol% Dy:BaTiO3 compositions were 114 and 104, respectively, at 1 MHz. Upon increasing the Dy3+ concentration to 1.2 mol%, the relative permittivity increased to 208, which is nearly two and a half times greater than that of undoped BaTiO3. To test against the formation of amorphous Ln2O3 shells, or related species, being the cause of the observed dielectric properties, we subjected pre-formed, undoped BaTiO3 nanocrystals to a second vapor diffusion sol–gel reaction in the presence of Dy(acac)3 (nominal 2.0 mol%; 25 1C, 48 h). The dielectric constant of the BaTiO3 nanocrystals surface treated with Dy(acac)3 decreased by 18% relative to the untreated BaTiO3 nanocrystals, suggesting that the surface species resulting from such a reaction are not the cause of the observed dielectric effects. The dielectric loss tangents were between 0.02–0.05 for all Ln:BaTiO3 nanocrystals (i.e., greater than the undoped BaTiO3 nanocrystals for all compositions). The increase in permittivity may be attributed to local disorder promoted by aliovalent substitution. It has been computationally modeled for bulk ceramics that the titanium octahedra distort to electrostatically stabilize the lattice because of B-site vacancies, generating local polarizations that enhance the relative permittivity.19 As previously mentioned, La3+ is known to substitute at the A-site, whereas Dy3+ can substitute either the A- or B-site exclusively or both the A- and B-sites. Introducing trivalent ions into the A-site generates Ti4+ vacancies to accommodate the charge imbalance caused by the defect, according to:20 4LaBa - VTi 0 0 0 0
(2)
The Ti4+ vacancies distort the Ti–O bonding in the surrounding octahedra.19 Another possible charge compensation mechanism involves the reduction of Ti4+ to Ti3+. In order to elucidate which mechanism is occurring, electron paramagnetic resonance (EPR) spectra were collected at 78 K on undoped BaTiO3, 1.6 mol% La:BaTiO3, and 1.2 mol% Dy:BaTiO3 nanocrystals (ESI†). Beyond the observed sextet indicative of Mn2+, the absence of any paramagnetic centers in the g-value range of 1.907–1.974 belonging to Ti3+ suggests that the concentration of reduced titanium is below the EPR detection limit. Paramagnetic centers were observed at g-values of 2.004–2.005 for all three samples, which correspond to the presence of titanium vacancies (VTi).21,22 Thus, the formation of titanium vacancies upon A-site substitution is the dominant charge compensation mechanism occurring in this system. Regarding Dy3+ doping, a recent study by Rabuffetti et al. on
3482 | Chem. Commun., 2014, 50, 3480--3483
Eu3+ aliovalently doped into BaTiO3 nanocrystals reported that Eu3+ (r = 1.29 Å for a 12-coordinate environment; extrapolated value6) substitutes onto the A-site for concentrations r1 mol%, with a transition to A- and B-site substitution at 42 mol%.17 Due to the similar ionic radii of Eu3+ and Dy3+, these results suggest that Dy3+ may be preferentially occupying the A-site at the doping level investigated here; however, high resolution structural characterization is required to verify this hypothesis. With regards to the frequency dependent dielectric properties, the undoped BaTiO3 showed a permittivity of 103 at 1 kHz, which decreased to 91 by 100 kHz. Such permittivity frequency dependence results from interfacial polarization at lower frequencies.23 All compositions exhibited minimal frequency dependence, with reductions in permittivity ranging from 13–20% by 100 kHz. Following the low frequency dependence, the nanocrystals were stable in permittivity up to 2 MHz. Dielectric loss tangents between 0.05–0.08 were observed in the Dy:BaTiO3 and La:BaTiO3 nanocrystals at 1 kHz, which decreases to between 0.02–0.05 by 100 kHz. In summary, the VDSG method was employed to synthesize aliovalently doped La:BaTiO3 and Dy:BaTiO3 nanocrystals at room temperature under ultra-benign conditions. Dielectric properties and ICP-OES confirmed that the lanthanide ions could be incorporated at room temperature without the need for post-synthetic annealing. The size-controlled, quasispherical nanocrystals appear to be single crystalline, as indicated by high-resolution TEM. Maximum relative permittivities of 176 and 206 were obtained for the 0.8 mol% La:BaTiO3 and 1.2 mol% Dy:BaTiO3 nanocrystals, respectively, at 25 1C and 1 MHz. While we have previously demonstrated that the relative permittivity of BaTiO3 nanocrystals can be tuned by isovalent substitution of Ba2+ and Ti4+ with Sr2+ and Zr4+, respectively, these solid solutions require substitution of ca. 33 mol% Sr2+ for Ba2+ or 15 mol% Zr4+ for Ti4+ to achieve maximum relative permittivity.24–26 Therefore, much lower aliovalent substitution levels can achieve similar effects on the dielectric properties of BaTiO3 nanocrystals using the vapor-diffusion sol–gel route. This work is supported by the Department of Energy, Office of Basic Energy Sciences, under Grant No. DE-FG02-11ER46826. S.P.C. acknowledges the National Science Foundation for a Graduate Research Fellowship. We thank Prof. P. Qin for use of his EPR spectrometer (supported by NIH S10 RR028992). S.T. acknowledges financial support from the Searle Scholars Program.
Notes and references 1 A. R. West, T. B. Adams, F. D. Morrison and D. C. Sinclair, J. Eur. Ceram. Soc., 2004, 24, 1439. 2 A. J. Moulson and J. M. Herbert, Electroceramics, Chapman and Hall, London, 1990. 3 M. D. Glinchuk, I. P. Bykov, S. M. Kornienko, V. V. Laguta, A. M. Slipenyuk, A. G. Bilous, O. I. V’yunov and O. Z. Yanchevskii, J. Mater. Chem., 2000, 10, 941. 4 J. P. Vernon, N. Hobbs, Y. Cai, A. Lethbridge, P. Vukusic, D. D. Deheyn and K. H. Sandhage, J. Mater. Chem., 2012, 22, 10435. 5 F. A. Rabuffetti, J. S. Lee and R. L. Brutchey, Adv. Mater., 2012, 24, 1434. 6 R. D. Shannon, Acta Crystallogr., 1976, A32, 751. 7 D. Y. Lu, T. Koda, H. Suzuki and M. Toda, J. Ceram. Soc. Jpn., 2005, 113, 721.
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 13 February 2014. Downloaded by University of Windsor on 19/10/2014 12:36:43.
Communication 8 C. L. Freeman, J. A. Dawson, H. R. Chen, J. H. Harding, L. B. Ben and D. C. Sinclair, J. Mater. Chem., 2011, 21, 4861. 9 F. D. Morrison, D. C. Sinclair, J. M. S. Skakle and A. R. West, J. Am. Ceram. Soc., 1998, 81, 1957. 10 D. W. Kang, T. G. Park, J. W. Kim, J. S. Kim, H. S. Lee and H. Cho, Electron. Mater. Lett., 2010, 6, 145. 11 Y. Pu, W. Chen, S. Chen and H. T. Langhammer, Ceramica, 2005, 51, 214. 12 F. D. Morrison, D. C. Sinclair and A. R. West, J. Appl. Phys., 1999, 86, 6355. 13 A. Testino, Int. J. Appl. Ceram. Technol., 2013, 10, 723. 14 S. P. Culver, F. A. Rabuffetti, S. Zhou, M. Mecklenburg, Y. Song, B. C. Melot and R. L. Brutchey, Chem. Mater., 2013, 25, 4129. 15 F. A. Rabuffetti, J. S. Lee and R. L. Brutchey, Chem. Mater., 2012, 24, 3114. 16 F. A. Rabuffetti and R. L. Brutchey, Chem. Commun., 2012, 48, 1437.
This journal is © The Royal Society of Chemistry 2014
ChemComm 17 F. A. Rabuffetti, S. P. Culver, J. S. Lee and R. L. Brutchey, Nanoscale, 2014, 6, 2909. 18 W. R. Tinga, W. A. Voss and D. F. Blossey, J. Appl. Phys., 1973, 44, 3897. 19 C. L. Freeman, J. A. Dawson, J. H. Harding, L. B. Ben and D. C. Sinclair, Adv. Funct. Mater., 2012, 23, 491. 20 F. D. Morrison, D. C. Sinclair and A. R. West, J. Am. Ceram. Soc., 2001, 84, 531. 21 T. Kolodiazhnyi and A. Petric, J. Phys. Chem. Solids, 2003, 64, 953. 22 T. D. Dunbar, W. L. Warren, B. A. Tuttle, C. A. Randall and Y. Tsur, J. Phys. Chem. B, 2004, 108, 908. 23 R. Z. Hou, P. Ferreira and P. M. Vilarinho, Chem. Mater., 2009, 21, 3536. 24 C. W. Beier, M. A. Cuevas and R. L. Brutchey, J. Mater. Chem., 2010, 20, 5074. 25 F. A. Rabuffetti and R. L. Brutchey, Chem. Commun., 2012, 48, 1437. 26 F. A. Rabuffetti and R. L. Brutchey, ACS Nano, 2013, 7, 11435.
Chem. Commun., 2014, 50, 3480--3483 | 3483
