Session W32: Glassy and Amorphous Systems

First-principles investigation of pressure-induced amorphization in zeolites

Crystalline zeolites can be transformed into amorphous structures by application of pressure, without ever forming a liquid. Upon release of the applied pressure, some zeolites transform back to their crystalline structure while others do not. Thus, zeolites are very interesting from the point of view of the theories trying to explain {\sl reversible} amorphization by pressure [Cohen et al., JNCS 307-310, 602 (2002)]. On the other hand, recent studies of the amorphization process in zeolites have led to the identification of co-existing phases of the same composition but markedly different densities and degrees of disorder [Greaves et al., Nat. Mats. 2, 622 (2003)]. Further, it has been argued that upon application of pressure (or, equivalently, temperature) zeolites render a low-entropy, low-density, amorphous phase that could constitute a new type of glass, with physical properties that might differ considerably from those of {\sl typical} glasses obtained by slow cooling from the melt. In this talk we will report a first-principles investigation of the structural changes induced by pressure in zeolites. More precisely, we will show results for three zeolites with the so-called LTA structure (Na-ZK4, Na-A, and an idealized SiO$_2$ system with the ZK4 structure). We will discuss the implications regarding the various amorphous phases experimentally found and the reversibility of the amorphization.   

Glass transition and viscosity of simple glasses and liquids

The theoretical understanding of liquids and glasses at an atomistic level lags well behind that of crystalline materials, even though they are important in many fields including biology and the medical sciences.~ We present a simple microscopic model for the glass transition based on topological fluctuations in the bonding network. The model makes predictions for important parameters of the glassy state, such as the glass transition temperature, $T_{g}$, and the liquid fragility coefficient, $m$, based on microscopic variables. Excellent agreement with a number of experimental observations from metallic glasses is demonstrated. A key to this success is to focus on the dependence on Poisson's ratio, following the work of Novikov and Sokolov,$^{1}$ that characterizes the interaction between local density and shear fluctuations. To our knowledge, this is the first model to~predict $T_{g}$ and $m$ quantitatively from microscopic variables. It presents a simple conceptual framework that should provide the basis for a more general microscopic understanding of liquids and glasses, including molecular systems. \newline 1. V. N. Novikov and A. P. Sokolov, \textit{Nature}, \textbf{431}, 961-963 (2004).   

Temperature induced density anomaly in Te rich liquid Germanium Tellurides : p versus sp3 bonding ?

The density anomaly of liquid Ge$_{0.15}$Te$_{0.85}$ measured between 633K and 733K is investigated with ab initio Molecular Dynamics calculations at four temperatures and at the corresponding experimental densities. For box sizes ranging from 56 to 112 atoms, an 8 k-points sampling of the Brillouin zone is necessary to obtain reliable results. Contrary to other Ge chalcogenides, no sp3 hybridization of the Ge bonding is observed. As a consequence the negative thermal expansion of the liquid is not related to a tetrahedral bonding as in the case of water or silica. We show that it results from the symmetry recovery of the local environment of Ge atoms that is distorted at low temperature by a Peierls-like mechanism acting in the liquid state in the same way as in the parent solid phases.   

Giant photocontraction effects in obliquely-deposited chalcogenide glass thin-films*

Ge$_{x}$Se$_{1-x}$ thin-films at several obliqueness angles $\alpha $ (= 0, 20, 45, 60, 80) and compositions x ( = 0.15, 0.20, 0.23, 0.25 and 0.33) were vapor-deposited, and examined in Raman scattering and SEM measurements both in the pristine and illuminated state. The films, placed in an inert ambient, were exposed to Hg lamp radiation, and photo-contraction of the films established using a profilometer. Raman scattering of the pristine and exposed films were studied as a function of depth using a confocal microscope attachment. Our results show (i) Raman scattering of the normally deposited ($\alpha $ = 0) films in the pristine state are similar to those of corresponding bulk glasses, (ii) obliquely deposited films at x = 1/3 reveal Raman lineshapes that change qualitatively with $\alpha $, suggestive of nanoscale phase separation of the films, while those at x = 0.23 show Raman lineshapes that are largely independent of $\alpha $, (iii) the photocontraction effect maximizes in the 0.20$<$ x $<$ 0.25 range, confirming the earlier finding (ref1) (iv) light illumination partially undoes effects associated with nanoscale phase separation. Possible interpretation of these results in relation to origin of photocontraction effects will be presented. *Supported by NSF grant DMR 04-56472. 1.Bhanwar Singh, S. Rajagopalan, P. K. Bhat, D. K. Pandya and K. L. Chopra, Solid State Communications, Vol. 29, pp. 167-169 (1979)   

Non-perturbative Renormalization Group Study of A Kinetically Constrained Model for Glasses

We study a dynamic field theory, which is based on the dynamic heterogeneity, with the non-perturbative renormalization group (NRG) method. Dynamic heterogeneity has been observed both experimentally, and in numerical simulations. It may play an important role in the physics of glass transition. Dynamic heterogeneity means that the slow dynamics of glass formers is dominated by the spatial fluctuations. And the length scales of dynamically correlated regions increase when the system approaches the glass transition, together with the increase of time scale. This is similar to the conventional dynamic critical phenomena. People have derived a dynamical field theory for a kinetically constrained model based on the Fredrickson-Anderson (FA) model, which puts the dynamic heterogeneity at its core. Here we study this field theory with the NRG method. The NRG method is not restricted to small parameters, and thus can be applied to more general cases. The critical exponents are calculated, and the phase structure is given. Especially the 1D spatial dimension case, which is difficult for the usual perturbative calculation, is also studied. Other possible models are also mentioned.   

Self-organization of glass networks in a model with equilibration

Recent experimental results suggest the existence of an intermediate phase in covalent glasses attributed to a self-organization of the glass network minimizing its internal stress. While a number of models have been proposed recently to explain this phenomenon, a full understanding of the network self-organization and the intermediate phase is lacking. We modify a previously studied model,$^(*)$ in which a network is grown in a way that keeps it stressless, by allowing continuous equilibration as the mean coordination is increased while still avoiding stress. In our model, an unusual intermediate phase appears, in which both rigid and floppy networks have a chance to occur, a result similar to that obtained by Barr\'{e} {\it et al.} for less realistic Bethe lattices. We discuss various structural properties of the resulting self-organized networks, as well as some results for the entropy cost of self-organization. \newline \vskip 2pt \noindent $^(*)$ M.F. Thorpe, D.J. Jacobs, M.V. Chubynsky, and J.C. Phillips, {\it J. Non-Cryst. Solids} {\bf 266-269}, 859 (2000)   

The Impact of Liquid Structure and Long Range Diffusion on Glass Formation and Nanoscale Devitrification

Recently, synchrotron X-ray diffraction measurements of electrostatically levitated samples revealed a growing icosahedral order in many undercooled metallic liquids and alloys. In a TiZrNi liquid, this icosahedral order catalyzed the nucleation of a metastable icosahedral phase, instead of the stable C14 tetrahedral Laves phase, confirming a half-century-old hypothesis made by Frank that connects the crystal nucleation barrier with the local structure of the liquid. This is one example of a growing number of cases of multiple order parameter coupling in nucleation, which are not readily described within the framework of the commonly used classical theory of nucleation. The implications for glass formation, nanoscale devitrification and the role of microalloying of such coupled nucleation processes and of a new coupled flux kinetic model for nucleation are discussed.   

Thermally Activated Processes in Polymer Glasses

A derivation is given for the Vogel-Fulcher-Tammann thermal activation law for the glassy state of a bulk polymer. Our microscopic considerations involve the entropy of closed polymer molecular chains (i.e. polymer closed strings). For thin film polymer glasses, one obtains open polymer strings in that the boundary surfaces serve as possible string endpoint locations. The Vogel-Fulcher-Tammann thermal activation law thereby holds true for bulk polymer glass but does not hold true in the ultra thin film limit of polymer glass.   

Transient nucleation in a Zr-Ti-Cu-Ni-Al metallic glass

Upon annealing Zr-Ti-Cu-Ni-Al bulk metallic glasses (BMG) crystallize to a nano-quasicrystal/amorphous composite. To probe the nucleation processes that underlie this microstructure formation we have determined the nucleation rate as a function of temperature, employing the two-step annealing method commonly used in silicate glasses. Samples were first annealed at temperatures where the nucleation rates were high, but the growth rates were low to produce a population of nuclei. These were subsequently grown to an observable size for transmission electron microscopy study by annealing at a higher temperature where the nucleation rate is small. We present the first quantitative time-dependent nucleation data obtained for a metallic glass by this method. The data are analyzed in terms of the classical theory of nucleation and an extended kinetic model that includes the coupling between the interfacial attachment and long-range diffusion fluxes.   

The role of Ti in the formation of Zr-Ti-Cu-Ni-Al glasses

It has been widely reported that glass formation improves in Zr$_{62}$Cu$_{20}$Ni$_{8}$Al$_{10 }$alloys when small amounts of Ti are substituted for Zr. Glasses containing greater than 3 at.{\%} Ti crystallize to a metastable icosahedral phase, suggesting that Ti enhances icosahedral short range order (ISRO) in the liquid/glass, making crystallization more difficult during cooling. Based on \textit{in-situ} high-energy synchrotron diffraction studies of electrostatically levitated (ESL) supercooled liquids and rapidly quenched amorphous alloys, we demonstrate ISRO in all cases irrespective of the Ti concentration. Further, our ESL solidification studies show that Ti inhibits surface crystallization, but does not improve glass formation.   

High-Energy X-ray Diffraction Study of Liquid Structure of Zr-based Binary Alloys

High-energy (E = 113 keV) synchrotron x-ray diffraction experiments were performed for metallic glass-forming Zr-Cu alloys in the liquid state at high temperatures. Accurate structure information of highly reactive melts has been obtained by applying conical nozzle levitation technique as a containerless method. The total structure factor extracted for the liquid Zr$_{50}$Cu$_{50}$ alloy near its melting temperature shows a particular shoulder at the high-$Q$ side on the second peak as well as the liquid Zr$_{70}$Cu$_{30}$ alloy. This feature of structure factor is similar to those of structure factors observed in deeply undercooled metallic liquids or metallic glasses, in which local icosahedral short range ordering was found to exist. With the use of reverse Monte Carlo simulation analysis, it was demonstrated that short-range ordered clusters exist even in the equilibrium liquid state of Zr$_{70}$Cu$_{30}$ alloy.   

Elasticity and conductivity thresholds in solid electrolyte glasses

The solid electrolyte glass, AgI, possesses a low mean coordination number [1], and when alloyed in the base oxide glass, AgPO$_{3}$, steadily lowers the connectivity of the alloyed glass, (AgI)x(AgPO3)1-x, as reflected in reduction of glass transition temperatures T$_{g}$(x). Non-reversing enthalpy associated with T$_{g}$s vanish in the 0.10 $<$ x $<$ 0.35 range, the reversibility window, which we identify with the Intermediate elastic phase [2]. Glasses at x $<$ 0.10 belong to the Stressed-Rigid while those at x $>$ 0.35 to the Floppy elastic phase. Electrical conductivity, $\sigma $(x), reveal a mild increase near x = 0.10 as glasses become unstressed, and a pronounced increase near x = 0.35, when glasses become floppy. The correlation between $\sigma $(x) and the elastic phases opens a new paradigm in understanding electrical transport in glasses. Chains of PO$_{4}$ tetrahedra ( Q$^{2})$ present in the pristine oxide are steadily cut and eventually transformed into rings as networks become less connected with increasing x, as revealed by Raman and P$^{31}$ NMR measurements. \newline [1] P. Boolchand and W.J.Bresser, Nature \textbf{410}, 1070 (2001). \newline [2] S.Chakravarty, D.Georgiev, P.Boolchand {\&} M.M.Micoulant, JPCM \textbf{17}, L1-L7 (2005).   

Chemical alloying induced collapse of reversibility windows in ternary As-S-I glasses*

Thermally reversing windows represent glass compositions across which glass transitions are thermally reversing in character. These windows have been observed in several chalcogenide glasses, and are identified$^{1}$ with self-organized phases of glassy networks. Upon alloying halogen (iodine) in base chalcogenide glasses (Ge-Se, Ge-S), the reversibility windows collapse$^{2}$ about the mean-field rigidity transition. We attempt to understand this behavior better. We have now synthesized ternary glass compositions of the type, (AsI$_{3})_{x}$(As$_{0.30}$S$_{0.70})_{1-x}$ and (AsI$_{3})_{y}$ (As$_{2}$S$_{3})_{1-y}$ over wide composition ranges of x and y, and have examined them systematically in Raman scattering and MDSC experiments. Along with earlier results$^{3}$ on binary As$_{z}$S$_{1-z}$ glasses, the present results permit mapping the reversibility window over the glass forming range of the present As-S-I ternary. The results show the window region to be of nearly triangular shape, with a base extending in the 0.20 $<$ z $<$ 0.27 range and a vertex located near y = 0.28. A possible interpretation of the results will be presented. * Supported by NSF grant DMR-04 -56472 1. P.Boolchand et al.Phil. Mag \underline {85},3823 (2005). 2. Y. Wang et al. Phys. Rev. Lett. \underline {87}, 18, 5503 (2001) 3. D.G. Georgiev, Ph.D. Thesis , Univ. of Cincinnati (2003) unpublished   

Glass structure and electrical conductivity in (As$_2$S$_3$ $_{1-x}$ (Ag$_{2}$S)$_{x}$

We have synthesized titled glasses in the 0 $<$ x $<$ 0.16 range, and have examined them in modulated DSC experiments. The starting materials, As$_{2}$S$_{3}$ and Ag$_{2}$S lumps, were reacted in evacuated fused quartz tubings, and glasses synthesized by water-quench of homogenized melts. Thermal measurements used a TA instruments model 2920 operated at 3\r{ }C/min scan rate and 1\r{ }C/100s modulation rate. Preliminary results reveal a single glass transition in the 0 $<$ x $<$ 0.05 range, which steadily decrease from a value of 210\r{ }C at x = 0 to 182\r{ } C near x = 0.05. In contrast, bimodal glass transitions are observed at x $>$ 0.09, with one T$_{g}$(1) near 167\r{ }C and the second, T$_{g}$(2) near 186\r{ }C, and with the endotherm associated with T$_{g}$(1) steadily increasing with x. Non-reversing enthalpies associated with T$_{g}$s are found to steadily decrease in the 0 $<$ x $<$ 0.09 range, to nearly vanish in the 0.10 $<$ x $<$ 0.12 range and to increase thereafter ( x $>$ 0.12).These findings suggest that glasses at low x ( $<$ 0.09) are Stressed- rigid, those at x $>$ 0.12 Floppy while those in between in the Intermediate phase$^{1}$. The present results correlate well with earlier$^{2}$ electrical conductivity results in suggesting the possibility of an elastic origin to the conductivity thresholds in solid electrolyte glasses. 1. P. Boolchand, D.Georgiev and B. Goodman, J.Opto {\&} Adm. Mater. 3, 703 (2001). 2. E.A. Kazakova and Z.U.Borisova, Fiz. Khim.Stekla \textbf{6}, 424(1980).   

New glasses in the alumina-calcia-monazite (LaPO$_{4})$ system: structural evidence from NMR, Raman scattering and thermal properties

A new group of glasses has been synthesized from the well-known compounds CaAl$_{2}$O$_{4}$ and (CaO)$_{12}$(Al$_{2}$O$_{3})_{7}$, melted together with monazite (LaPO$_{4})$ in compositions containing 2 to $>$75{\%} of the latter. Raman and $^{31}$P NMR spectra in the solid state show that PO$_{4}$ groups do not share bridging oxygens, i.e. that the materials are orthophosphates. Thermal properties and $^{27}$Al NMR in both liquid and solid states indicate the presence of a strong aluminate network, based upon AlO$_{4}$ tetrahedra sharing corners. P/Al TRAPDOR NMR measurements show that P and Al are in close proximity in the glasses, most likely sharing P-O-Al linkages. However, clear and unambiguous signatures of the aluminate network are still sought. Models for the structures of these glasses, drawn from experiment, will be presented.