Fabrication of BaTiO 3 –Carbon Nanocomposite and Porous BaTiO 3

Published: August 19, 2011

r 2011 American Chemical Society 5215 dx.doi.org/10.1021/cg200031p | Cryst. Growth Des. 2011, 11, 5215–5220

ARTICLE

pubs.acs.org/crystal

Fabrication of BaTiO3�Carbon Nanocomposite and Porous BaTiO3

Ru Z. Hou, Paula Ferreira,* and Paula M. Vilarinho*

Department of Ceramics and Glass Engineering, CICECO, University of Aveiro, 3810-193 Aveiro, Portugal

’ INTRODUCTION

Morphology control on ferroelectrics at nanometer scale hasinspired a new round of materials exploration and described a seriesof prototype applications.1�11 Vortex polarization was recentlyreported in ultrasmall disks of conventional ferroelectrics,2 andthen nanorings were designed to stabilize the vortex state andprovide samples for investigation of polarization vortices.5,6

Ferroelectrics with one-dimensional morphology, that is, nanorods,nanowires, and nanotubes, were predicted to possess enhancedferroelectricity as compared to their bulk counterparts.7,8

Nanoislands of ferroelectrics are good candidates for the studyof fundamental issues such as size effects.9 Physical and chemical“bottom up” approaches to prepare ferroelectric nanostructureswere reviewed.10 The nanoscale design strategy was also em-ployed in the fabrication of porous incipient ferroelectrics films.11

Attention is moving frommacroscopy to the scale of hundreds oflattice parameters or even smaller, where the shape, surface area,curvature, and defects significantly affect the properties of thestudiedmaterials.Within this context, more findings are expectedwith the nanoscale morphology modifications of ferroelectrics.

Meanwhile, the combination of different materials at thenanoscale creates multifunctional nanocomposites, where novelproperties may result from the scale, interface, and interactions.Ferroelectrics-based nanocomposites with other functional com-ponents, such as polymer, carbon, metal, and ferromagneticphase, etc., have been reported for ultrahigh dielectric constant,power storage, or magnetoelectric applications.12�17 Recently,quantum-confinement effects of nanocrystals of BaTiO3, SrBi2-Ta2O9, and LiTaO3 in mesoporous silica (MCM-41)18�20 werereported with uniform pores of several nanometers in a two-dimensional p6mm hexagonal symmetry.21 Blue-shift of opticalband gap and superparaelectric behaviors were found in theconfined and separated nanocrystals due to their limited size andweak cooperation.

In this work, (i) the incorporation of BaTiO3 intomesoporouscarbon (CMK-3), which possesses the hexagonal symmetry to

form a new BaTiO3�carbon composite, and (ii) the removal thecarbon matrix resulting in BaTiO3 particles with unique nano-porous structures are presented. CMK-3 is a carbon replication ofmesoporous silica SBA-15, which has a larger pore size than thatof MCM-41 due to the utilization of a block copolymer asstructure direct reagent.22,23 The carbon matrix is much easier toremove (simply by calcination) as compared to silica, which hasto be dissolved by hydrofluoric acid or strong alkaline solutionthat is harmful to ferroelectric oxides.

This work also has its meaning in terms of nanocasting.Various mesoporous metal oxides replicated fromCMK-3 matrixhave been recently realized using nitrates as metal source.24�29

The replication of CMK-3 withmetal nitrates takes advantages ofthe low operating temperature (as low as 473 K) to convertnitrates into oxides. However, crystalline complex multimetalcompounds have not been involved in the replication of meso-porous carbon. One of the underlying difficulties to expand thistechnique to complex compounds is the unavailability of thenitrate salt of the metal element. In addition, to synthesize thecomplex compound, temperature much higher than the decom-position temperature of metal nitrate is required. Moreover,reaction between the carbon and metal compounds occursoccasionally. Here, the possibility of replicating mesoporouscarbon into typical ferroelectrics BaTiO3 is explored, and theproblems encountered during this process are reported anddiscussed.

’EXPERIMENTAL SECTION

Synthesis.Mesoporous carbon CMK-3 was prepared by replicationof mesoporous silica SBA-15 as reported in the literature.23 Thesynthesis of SBA-15 followed the method of preparation of mesoporous

Received: January 10, 2011Revised: August 15, 2011

ABSTRACT: A novel nanocomposite was fabricated by intro-ducing BaTiO3 into mesoporous carbon CMK-3 with hexago-nal porous structure and pore size around 3.4 nm. NanosizedBaTiO3 crystallites were fully dispersed and confined in thechannels of CMK-3. This is the first combination of conven-tional ferroelectrics with newly developed mesoporous carbonmaterials. Furthermore, the carbonmatrix was thermally removed in oxygen atmosphere to form unique nanoporous structures withthe framework of crystalline BaTiO3. The particle morphology evolution from CMK-3 to BaTiO3�CMK-3, and then to porousBaTiO3, was followed by high-resolution scanning electron microscopy (HRSEM). Although this work was aimed at preparation,the present nanocomposite of conductive amorphous carbon and insulate crystalline ferroelectrics, as well as the nanoporousBaTiO3, are expected to have interesting properties in areas such as power storage, microwave adsorbing, and gas sensor.

5216 dx.doi.org/10.1021/cg200031p |Cryst. Growth Des. 2011, 11, 5215–5220

Crystal Growth & Design ARTICLE

silica using triblock copolymer reported by Zhao et al.22 To form theBaTiO3�CMK-3 composite, CMK-3 was impregnated with BaTiO3

precursor solution, which was prepared by mixing 0.015 mol of tetra-butyl orthotitanate (97%, Fluka) stabilized by 0.03 mol of acetylacetone(99.3%, Fluka) and 0.015 mol of barium acetate (99%, Carlo Erba)dissolved in 30 mL of acetic acid (99.8%, Merck). Typically, 0.1 g ofCMK-3 was dispersed in 2 mL of BaTiO3 precursor solution. The mixturewas stirred for 2 h before the excess of solution was removed with aceramic filter. The wet powders were dried in an oven and then calcinedin flowing nitrogen at 700 �C (heating rate 5 �C/min) with various time tocrystallize BaTiO3 inside the porous carbon. Finally, some of the BaTiO3�CMK-3 powders were heat treated at 500 �C in air, or at 400 �C inoxygen (heating rate 1.5 �C/min, 30 min), to get porous BaTiO3.Characterization. Nitrogen adsorption�desorption measurements

on SBA-15 and CMK-3 powders were conducted with a Micromeritics

Gemini 2375 system at 77 K, based on which the Brunauer�Emmett�Teller (BET) surface area and Barrett�Joyner�Halenda (BJH) poresize distribution were obtained. A thermoanalyzer (Labsys SetaramTG-DTA/DSC, Caluire) was used for thermogravimetry (TG) anddifferential thermoanalysis (DTA) carried out in air. X-ray diffraction(XRD) patterns of the composite and porous BaTiO3 were collected ona diffractometer of Rigaku D/Max-B. Sherrer equation30 was applied tocalculate the crystallite mean size attending to the fact that instrumentalpeak width is 0.34 (for NIST standard 660a), close to the sample peak

Figure 1. Nitrogen adsorption�desorption isotherms of SBA-15 andCMK-3.

Figure 2. TG and DTA plots for CMK-3 impregnated with BaTiO3

precursor solution, performed in air.

Figure 3. XRD patterns for BaTiO3�carbon composites attained fromcalcination at 700 �C in nitrogen for various time. For comparison, thesolid lines depict the BaTiO3 (2) and BaCO3 (0) reflections accordingto the JCPDS card nos. 04-012-8129 and 01-085-0720, respectively.

Figure 4. TEM image of (a) CMK-3, (b) BaTiO3�CMK-3 composite;and (c) the HRTEM image of the nanocomposite showing fringes ofBaTiO3 nanocrystals confined in CMK-3 (the arrows indicate the areascorresponding to BaTiO3 and CMK-3).



chosen for calculation (39� 2θ, hkl 111). This peak is always a single peakin either BaTiO3 cubic or tetragonal crystallographic phases.

CMK-3, BaTiO3�CMK-3 composite, and porous BaTiO3 werestudied with transmission electron microscopy (TEM) (H9000-NA,Hitachi), and the high-resolution transmission electron microscopy(HRTEM) images were obtained using another instrument (JEM-2200FS, JEOL). Particle morphology was recorded with high-resolutionscanning electron microscopy (HRSEM) (FE-SEM SU-70, Hitachi).

’RESULTS AND DISCUSSION

Nitrogen adsorption�desorption isotherms of SBA-15 andCMK-3 used in this work are shown in Figure 1. Both of theisotherms exhibit type IV shape classified by IUPAC,31 provingthe mesoporous nature of the templates. SBA-15 had a BETsurface area of 671 m2/g and a narrow pore size distributionaround 6 nm, while the BET surface area and pore size of CMK-3were 1046 m2/g and 3.4 nm, respectively. These results con-firmed the successful synthesis of highly porous carbonmatrix forthe next step of impregnation with BaTiO3.

Figure 2 represents the thermal analysis (carried out in air) ofCMK-3 that has been impregnated with BaTiO3 precursorsolution. The TG plot mainly contains two sharp mass lossstages. The one in the range from room temperature to 100 �C

corresponds to the evaporation of residual water. The other onecovering 100�350 �C should be related to the decomposition ofthe organic components, burning out of carbon, and the elimina-tion of generated water from the decomposition reaction.According to the DTA plot, the formation temperature of BaTiO3

was determined as a bit lower than 700 �C. However, the calcina-tion in air at this temperature resulted in considerable amounts ofBaCO3 byproduct, and it was proved that the nitrogen atmo-sphere was effective to prevent BaCO3 formation. Accordingly,mainly BaTiO3�CMK-3 samples attained from calcination at700 �C in flowing nitrogen, to guarantee the crystallization ofBaTiO3 in the porous carbon matrix, were studied, althoughsamples calcined at lower temperatures showed occasionally the

Figure 5. XRD patterns for powders after thermal removal of CMK-3matrix in air and in oxygen. For comparison, the solid lines depict theBaTiO3 (2) and BaCO3 (0) reflections according to the JCPDS cardnos. 04-012-8129 and 01-085-0720, respectively.

Figure 6. Typical TEM micrograph of the resulted porous BaTiO3

particle after thermal removal of CMK-3 matrix in flowing oxygen. TheSAED is shown as the inset.

Figure 7. (a) Typical HRSEM micrograph of the porous BaTiO3

particles. (b) Magnified view of the selected area in (a), and (c)magnified view of the selected area in (b).



existence of crystalline BaTiO3 as well. XRD patterns for BaTiO3�CMK-3 composites prepared by calcination at 700 �C in nitrogenfor various times (10 min, 30 min and 1 h) are shown in Figure 3.All of the patterns indicate single crystalline BaTiO3 phasewithout carbonate, and the carbon matrix is amorphous. Applica-tion of Scherrer’s equation indicates that the crystallite sizeslightly increases with the calcination time, from 11.2 nm(10 min) to 11.9 nm (30 min), then to 13.3 nm (1 h). Theaverage crystallite size of more than 10 nm does not contradictthe 3.4 nm cross-section diameter of CMK-3 channel becausethere is no limitation for crystallite to grow along the channels.These BaTiO3�CMK-3 composites had largely lowered specificBET surface area, down to the level around 400 m2/g, suggestingthe loading of BaTiO3 into the carbon matrix. The repeatedimpregnation was also tried with the aim to further improve theloading, but it was realized that the second process intensified theundesired deposition of precursor onto the external surface ofmesoporous carbon particles. This proves that during the firstimpregnation, the template pores were fully filled by the pre-cursors solution. During the heat treatment to achieve crystal-lization of BaTiO3, the precursors partially decompose withrelease of massive gas byproduct and consequent formation ofvoids. In this way, the mesostructure cannot be easily replicatedafter the removal of the hard template. It is expected that theBaTiO3 particles migrate inside the channels and aggregate insmall areas of the template to allow the formation of smalldomains of well replicated structure.

TEM and HRTEM characterizations were employed to con-firm the incorporation of BaTiO3 into CMK-3 (Figure 4). Longparallel channels (vertical to the electron beam) of as-preparedmesoporous carbon CMK-3 are clearly seen, as shown inFigure 4a. After impregnation and crystallization processes, TEMrevealed heterogeneous dispersion of nanosized BaTiO3 particlesinside the carbon matrix (Figure 4b). Figure 4c shows the latticefringes of BaTiO3 in the environment of amorphous carbon. Thewidth of single BaTiO3 crystallite is limited by the cross-sectiondiameter of CMK-3 channel, while it can grow longer along thedirection parallel to the channels, as depicted in the HRTEMimage. These nanoparticles were isolated but not aggregated inthe pore system. Such structure provides a composite of insulatecrystalline nanoparticles with high dielectric constant fully dis-persed in a conductive amorphous matrix.

The carbonmatrix canbe removedby thermal combustion to attaina porous structure. Here, the calcination of the BaTiO3�CMK-3powders in air and oxygen, respectively, was tried. Whitepowders were obtained in both cases. The XRD pattern of thepowders calcined in air indicates a large amount of BaCO3

(Figure 5), which is attributed to the reaction between generatedCO2 and residual barium species. Calcination in flowing oxygen

effectively prohibits the formation of carbonate, and the XRDpattern suggests BaTiO3 single phase in the produced powderswith improved crystallinity (crystal size∼20.4 nm estimated withScherrer’s equation) as compared to that of the confined crystal-lites in CMK-3. The function of flowing oxygen is not only theimprovement of oxidization of carbon at relatively lower tem-perature, but may be also in the dilution of the generated CO2 inthe atmosphere during heating.

Figure 6 is a TEM image of the resulted BaTiO3 after thecarbon matrix was burnt out, showing a porous particle com-posed of grown crystallites, consistent with the XRD data.Without confinement of the carbon matrix, the nanosizedBaTiO3 crystallites were quite easy to grow during the heattreatment, which broke the ordering of the porous structure. Inaddition, different from the decomposition of metal nitrate intosimple oxide, which occurs at low temperatures of 200�300 �C,the formation of BaTiO3 by calcination requires a high tempera-ture above 600 �C that makes it hard to keep the crystallites assmall as several nanometers. Similar heat treatment at hightemperature before carbon removal was reported to result inthe crystal growth and loss of the porosity ordering in the finalproduct of porous MgO using nitrate as precursor.29 Thus, thecrystallite growth that occurs during both the BaTiO3 formationand the carbon removal steps becomes a fatal obstacle to replicatethe ordered porous structure of CMK-3. The low BET specificsurface area calculated for the BaTiO3 powders (∼15 m2/g)further confirms the loss of porosity probably due to crystalgrowth.

Anyway, it is still interesting to study the present porousstructures at nanoscale with high-resolution SEM, which ispowerful to give more information and display the aspects thatTEM cannot reveal. Typical HRSEM micrographs of the porousBaTiO3 particles with different magnification are shown inFigure 7, giving a detailed view on how the porous structure isorganized. The big trunk-like particle (Figure 7a) is a cluster offibers. With higher magnification, it is revealed that the fibers areporous constructed by nanosized BaTiO3 crystallites (Figure 7band c). This can be seen as a kind of hierarchical porous structure,taking the gaps among the fibers as first rank porosity. BaTiO3

particles with smaller size were also observed but with similarporous structure. Figure 8 illustrates the particle morphologyrelation among CMK-3, BaTiO3�CMK-3 composite, andBaTiO3. Both big trunk-like particles and clusters of rods werefound in CMK-3 powders, and there was a faithful particlemorphology succession from CMK-3 to BaTiO3�CMK-3 com-posite, and then to the porous BaTiO3. This explains why porousBaTiO3 particles are of various size and shape.

It is known that mesoporous silica SBA-15 can be synthesizedin various forms such as fiber-, rope-, sphere-like, etc.32 Because

Figure 8. HRSEMmicrographs showing the particle morphology succession from CMK3 to BaTiO3�CMK-3 composite, and then to porous BaTiO3.



the BaTiO3 particle morphology is templated from the CMK-3,and CMK-3 is replicated from mesoporous silica SBA-15, it ispossible to modify the porous BaTiO3 morphology by changingthe morphology of SBA-15. In this work, SBA-15 with sphericalparticle shape to prepare CMK-3 was tried as well. In turn,spherical porous BaTiO3 particles were attained, as shown inFigure 9. In this case, the porous structure showed a certaindegree of ordering. Grains of BaTiO3 were clearly aligned,reflecting the direction of the parallel channels of the removedCMK-3. If the growth of the crystallites can be avoided, BaTiO3

with hexagonally well-ordered porous structure is then expected.

’SUMMARY

BaTiO3 nanocrystallites were introduced into the channels ofmesoporous carbon via the impregnation of precursor solutioninto CMK-3 matrix followed by calcination under nitrogen.Thus, a novel nanocomposite was fabricated, described ascrystalline ferroelectrics particles fully dispersed and confinedin a conductive porous carbon matrix with high BET surface area.Further, unique hierarchical porous structures with BaTiO3

framework were attained after thermal removal of the carbonmatrix. It was found that the oxygen atmosphere was helpful toprevent the generation of carbonate at this step. Particle mor-phology faithfully succeeded from CMK-3 to BaTiO3�CMK-3and then to porous BaTiO3. By changing the particle morphol-ogy of the matrix, the particle morphology of the producedporous BaTiO3 can be varied. A certain degree of ordering wasachieved in the porous BaTiO3, while the main problem to getwell-ordered porous structure lies in the control of crystallite sizeduring BaTiO3 formation and carbon removal process. Thepresent ferroelectrics�carbon composite may be promising inthe areas of power storage or microwave adsorbing, and thenanoporous BaTiO3 may find its applications in areas such as gassensor.

’AUTHOR INFORMATION

Corresponding Author*Email: [email protected] (P.M.V.); [email protected] (P.F.).

’ACKNOWLEDGMENT

We are thankful to Dr. Marc-Georg Willinger for HRTEMcharacterizations. We thank FCT and FEDER (QREN �COMPETE) for funding the project PTDC/CTM/098130/2008. FCT is also thanked for the Postdoctoral fellowshipSFRH/BPD/26711/2006.

’REFERENCES

(1) Scott, J. F. Science 2007, 315, 954.(2) Naumov, I. I.; Bellaiche, L.; Fu, H. Nature 2004, 432, 737.(3) Naumov, I.; Fu, H. Phys. Rev. Lett. 2007, 98, 077603.(4) Naumov, I.; Bratkovsky, A.M. Phys. Rev. Lett. 2008, 101, 107601.(5) Zhu, X. H.; Evans, P. R.; Byrne, D.; Schilling, A.; Douglas, C.;

Pollard, R. J.; Bowman, R. M.; Gregg, J. M.; Morrison, F. D.; Scott, J. F.Appl. Phys. Lett. 2006, 89, 122913.

(6) Byrne, D.; Schilling, A.; Scott, J. F.; Gregg, J. M. Nanotechnology2008, 19, 165608.

(7) Morozovska, A. N.; Eliseev, E. A.; Glinchuk, M. D. Phys. Rev. B2006, 73, 214106.

(8) Morozovska, A. N.; Glinchuk, M. D.; Eliseev, E. A. PhaseTransitions 2007, 80, 71.

(9) Szafraniak, I.; Harnagea, C.; Scholz, R.; Bhattacharyya, S.; Hesse,D.; Alexe, M. Appl. Phys. Lett. 2003, 83, 2211.

(10) Alexe, M.; Hesse, D. J. Mater. Sci. 2006, 41, 1.(11) Grosso, D.; Boissi�ere, C.; Smarsly, B.; Brezesinski, T.; Pinna,

N.; Albouy, P. A.; Amenitsch, H.; Antonietti, M.; Sanchez, C.Nat. Mater.2004, 3, 787.

(12) Lee, J. Y.; Lee, J. H.; Hong, S. H.; Lee, Y. K.; Choi, J. Y. Adv.Mater. 2003, 15, 1655.

(13) Pecharrom�an, C.; Esteban-Beteg�on, F.; Bartolom�e, J. F.; L�opez-Esteban, S.; Moya, J. S. Adv. Mater. 2001, 13, 1541.

(14) Huang, Q.; Gao, L.; Liu, Y.; Sun, J. J. Mater. Chem. 2005,15, 1995.

(15) Nagao, Y.; Nakayama, Y.; Oda, H.; Ishikawa, M. J. Power Sources2007, 166, 595.

(16) Zheng, H.; Wang, J.; Lofland, S. E.; Ma, Z.; Mohaddes-Ardabili,L.; Zhao, T.; Salamanca-Riba, L.; Shinde, S. R.; Ogale, S. B.; Bai, F.;Vieland, D.; Jia, Y.; Schlom, D. G.; Wuttig, M.; Roytburd, A.; Ramesh, R.Science 2004, 303, 661.

(17) Nan, C. W.; Bichurin, M. I.; Dong, S.; Viehland, D.; Srinivasan,G. J. Appl. Phys. 2008, 103, 031101.

(18) Kohiki, S.; Nogami, S.; Kawakami, S.; Takada, S.; Shimooka, H.;Deguchi, H. Appl. Phys. Lett. 2003, 82, 4134.

(19) Kohiki, S.; Takada, S.; Shimizu, A.; Yamada, K.; Higashijima,H.; Mitome, M. J. Appl. Phys. 2000, 87, 474.

(20) Kinka, M.; Banys, J.; B€ohlmann, W.; Bierwirth, E.; Hartmann,M.; Michel, D.; V€olkel, G.; P€oppl, A. IEEE Trans. Ultrason. Ferroelectr.Freq. Control 2006, 53, 2305.

(21) Beck, J. S.; Vartuli, J. C.; Roth, W. J.; Leonowicz, M. E.; Kresge,C. T.; Schmitt, K. D.; Chu, C. T-W.; Olson, D. H.; Sheppard, E. W.;McCullen, S. B.; Higgins, J. B.; Schlenker, J. L. J. Am. Chem. Soc. 1992,114, 10834.

(22) (a) Zhao, D. Y.; Feng, J. L.; Huo, Q. S.; Melosh, N.; Fredrick-son, G. H.; Chmelka, B. F.; Stucky, G. D. Science 1998, 279, 548. (b)Zhao, D.; Huo, Q.; Feng, J.; Chmelka, B. F.; Stucky, G. D. J. Am. Chem.Soc. 1998, 120, 6024.

(23) Jun, S.; Joo, S. H.; Ryoo, R.; Kruk, M.; Jaroniec, M.; Liu, Z.;Ohsuna, T.; Terasaki, O. J. Am. Chem. Soc. 2000, 122, 10712.

(24) Tiemann, M. Chem. Mater. 2008, 20, 961.(25) Lai, X.; Li, X.; Geng, W.; Tu, J.; Li, J.; Qiu, S. Angew. Chem., Int.

Ed. 2007, 46, 738.

Figure 9. (a) Porous BaTiO3 particles derived from spherical CMK-3particles. (b) Magnified view of the selected area in (a).



(26) Roggenbuck, J.; Sch€afer, H.; Tsoncheva, T.; Minchev, C.;Hanss, J.; Tiemann, M. Microporous Mesoporous Mater. 2007, 101, 335.(27) Waitz, T.; Tiemann, M.; Klar, P. J.; Sann, J.; Stehr, J.; Meyer,

B. K. Appl. Phys. Lett. 2007, 90, 123108.(28) Roggenbuck, J.; Waitz, T.; Tiemann, M. Microporous Mesopor-

ous Mater. 2008, 113, 575.(29) Roggenbuck, J.; Koch, G.; Tiemann, M. Chem. Mater. 2006,

18, 4151.(30) Patterson, A. Phys. Rev. 1939, 56, 978.(31) Gregg, S. J.; Sing, K. S. W. Adsorption, Surface Area, and Porosity,

2nd ed.; Academic Press: London, 1982.(32) Zhao, D.; Sun, J.; Li, Q.; Stucky, G. D. Chem. Mater. 2000,

12, 275.

Fabrication of BaTiO 3 –Carbon Nanocomposite and Porous BaTiO 3

Documents