Experimental Evidence of Self-Limited Growth of Nanocrystals in

Europe Nanoport, FEI Company. Info icon. Your current ... For the first time, we do experimentally prove this concept of self-limited growth on the na...
0 downloads 0 Views 5MB Size
NANO LETTERS

Experimental Evidence of Self-Limited Growth of Nanocrystals in Glass

2009 Vol. 9, No. 6 2493-2496

Somnath Bhattacharyya,†,⊥ Christian Bocker,‡ Tobias Heil,§ Jo¨rg R. Jinschek,| Thomas Ho¨che,*,† Christian Ru¨ssel,‡ and Helmut Kohl§ Leibniz-Institut fu¨r Oberfla¨chenmodifizierung e.V., Permoserstrasse 15, 04318 Leipzig, Germany, Otto-Schott-Institut, UniVersita¨t Jena, Fraunhoferstrasse 6, 07743 Jena, Germany, Physikalisches Institut, Westfa¨lische Wilhelms-UniVersita¨t, Wilhelm-Klemm-Strasse 10, 48149 Mu¨nster, Germany, and Europe Nanoport, FEI Company, Achtseweg Noord 5, Bldg, 5651 GG EindhoVen, The Netherlands Received April 22, 2009; Revised Manuscript Received April 30, 2009

ABSTRACT Growth of nanocrystals precipitated in glasses with specific compositions can be effectively limited by diffusion barriers forming around crystallites. For the first time, we do experimentally prove this concept of self-limited growth on the nanoscale for a SiO2/Al2O3/Na2O/K2O/BaF2 glass in which BaF2 nanocrystals are formed. As shown by advanced analytical transmission electron microscopy techniques, the growth of these BaF2 crystals, having great potential for photonic applications, is inherently limited by the formation of a ca. 1 nm wide SiO2 shell.

Nanoglass ceramics, a novel yet largely ignored class of nanomaterials are about to play a major role for various optical applications. In particular, so-called ultratransparent glass-ceramics containing metal fluoride nanocrystals sized between 5 and 100 nm possess a huge application potential since rare-earth-doped metal fluoride nanocrystals feature significantly enhanced fluorescence,1 luminescence,2,3 and upconversion.4,5 But how can the stringent requirements on size, size distribution, and volume concentration of nanocrystalline precipitates be met? And, in particular, how can coarsening of particle size, e.g., by Ostwald ripening, be efficiently suppressed? One approach was proposed by Ru¨ssel et al. in a detailed study of the crystallization behavior in the glass systems Na2O/K2O/CaO/CaF2/Al2O3/SiO26 and Na2O/K2O/ CaO/BaF2/Al2O3/SiO2.7 It was shown that, despite an increasing volume fraction of the crystalline phase with increasing annealing time, the mean crystallite size remained constant. It was further proved that the glass transition temperature Tg of partially crystallized glasses increases with annealing temperature and time and finally approaches a value equal to the temperature at which the glass ceramics were annealed. In light of these findings, the hypothesis was developed that during nucleation and subsequent growth of fluoride crystals, a silica-enriched layer is formed around the †

Leibniz-Institut fu¨r Oberfla¨chenmodifizierung e.V. Otto-Schott-Institut, Universita¨t Jena. § Physikalisches Institut, Westfa¨lische Wilhelms-Universita¨t. | Europe Nanoport, FEI Company. ⊥ Present address: Department of Condensed Matter Physics and Materials Science, Tata Institute of Fundamental Research, Mumbai 400005, India. ‡

10.1021/nl901283r CCC: $40.75 Published on Web 05/07/2009

 2009 American Chemical Society

growing nanocrystal. Via an increased viscosity, the developing core-shell structure acts as a diffusion barrier at the interface between crystal and glass matrix. Eventually, this barrier limits further crystal growth. In the present contribution, using energy filtering transmission electron microscopy (EFTEM) as well as electron energy-loss spectrometry (EELS) spot analyses, we provide experimental evidence of exactly such a silica-enriched shell surrounding BaF2 nanocrystals in a Na2O/K2O/BaF2/Al2O3/ SiO2 glass. Glasses with the composition 69.6% SiO2, 7.52% Al2O3, 15.04% K2O, 1.88% Na2O, 6% BaF2 (in mole %) were melted from reagent-grade Na2CO3, K2CO3, BaF2, Al(OH)3, and SiO2 (quartz) in batches of 200 g in a platinum crucible. The melt was kept for 1 h at 1590 °C. Then it was cast on a copper block and the glass was placed in a furnace preheated to 450 °C and finally allowed to slowly cool down to room temperature. The glass-ceramics were prepared by annealing at 700 °C for 2 h. TEM samples were prepared by cutting slices, plane-parallel grinding, dimpling to a residual thickness of 10-15 µm, and ion-beam thinning using Ar+ ions. By means of double-sided ion-beam etching at small angles (