Search

We construct a thermodynamic potential that can guide training of a generative model defined on a set of binary degrees of freedom. We argue that upon reduction in description, so as to make the generative model computationally-manageable, the potential develops multiple minima. This is mirrored by the emergence of multiple minima in the free energy proper of the generative model itself. The variety of training samples that employ N binary degrees of freedom is ordinarily much lower than the size 2^N of the full phase space. The non-represented configurations, we argue, should be thought of as comprising a high-temperature phase separated by an extensive energy gap from the configurations composing the training set. Thus, training amounts to sampling a free energy surface in the form of a library of distinct bound states, each of which breaks ergodicity. The ergodicity breaking prevents escape into the near continuum of states comprising the high-temperature phase; thus it is necessary for proper functionality. It may however have the side effect of limiting access to patterns that were underrepresented in the training set. At the same time, the ergodicity breaking within the library complicates both learning and retrieval. As a remedy, one may concurrently employ multiple generative models -- up to one model per free energy minimum., Comment: 51 pages, 12 figures, accepted to Neural Computation

We argue that the dominant charge carrier in glassy semiconducting alloys is a compound particle in the form of an electron or hole bound to an intimate pair of topological lattice defects; the particle is similar to the polaron solution of the Su-Schrieffer-Heeger Hamiltonian. The spatial component of the density of states for these special polarons is determined by the length scale of spatial modulation of electronegativity caused by a separate set of standalone topological defects. The latter length scale is fixed by the cooperativity size for structural relaxation; the size is largely independent of temperature in the glass but above melting, it decreases with temperature. Thus we predict that the temperature dependence of the electrical conductivity should exhibit a jump in the slope near the glass transition; the size of the jump is predicted to increase with the fragility of the melt. The predicted values of the jump and of the conductivity itself are consistent with experiment., Comment: 12 pages, 8 figures, 4 tables

Angle-resolved photoemission spectra of alkali metals exhibit a puzzling, non-dispersing peak in the apparent density of states near the Fermi energy. We argue that the holes left behind a significant fraction of photoejected electrons are not wavepacket-like objects used to describe excitations of an equilibrium Fermi liquid but, instead, are relatively localized entities resulting from a photon-induced cavitation in the electron fluid. At the same time, these special localized holes can be thought of as vacancies in a transient Wigner solid. The corresponding contribution to the photoemission current is non-dispersive and is tied to the Fermi level; it exhibits certain similarities to photoemission from localized core orbitals such as the presence of recoil currents. Calculated spectra are consistent with experiment. We briefly discuss the present findings in the context of quantum measurement.

We argue that one can associate a pseudo-time with sequences of configurations generated in the course of classical Monte Carlo simulations for a single-minimum bound state, if the sampling is optimal. Hereby the sampling rates can be, under special circumstances, calibrated against the relaxation rate and frequency of motion of an actual physical system. The latter possibility is linked to the optimal sampling regime being a universal crossover separating two distinct suboptimal sampling regimes analogous to the physical phenomena of diffusion and effusion, respectively. Bound states break symmetry; one may thus regard the pseudo-time as a quantity emerging together with the bound state. Conversely, when transport among distinct bound states takes place -- thus restoring symmetry -- a pseudo-time can no longer be defined. One can still quantify activation barriers, if the latter barriers are smooth, but the simulation becomes impractically slow and pertains to overdamped transport only. Specially designed Monte Carlo moves that bypass activation barriers -- so as to accelerate sampling of the thermodynamics -- amount to effusive transport and lead to severe under-sampling of transition-state configurations that separate distinct bound states while destroying the said universality. Implications of the present findings for simulations of glassy liquids are discussed.

We present a microscopic picture rationalizing the surprisingly steep decrease of the band gap with temperature in insulators, crystalline or otherwise. The gap narrowing largely results from fluctuations of long-wavelength optical phonons---when the latter are present---or their disordered analogs, if the material is amorphous. We elaborate on this notion to show that possibly with the exception of weakly bound solids made of closed-shell atoms, the existence of an insulating gap or pseudo-gap in a periodic solid implies that optical phonons must be present, too. This means that in an insulating solid, the primitive cell must have at least two atoms and/or that a charge density wave is present. As a corollary, a (periodic) elemental solid whose primitive unit contains only one atom will ordinarily be a metal, possibly unless the element belongs to group 18 of the periodic table, consistent with observation. Some implications of the present results for quantum solids are briefly discussed, such as that the ground state of the Wigner crystal must be anti-ferromagnetic. A simple field theory of the metal-insulator transition is constructed that ties long-wavelength optical vibrations with fluctuations of an order parameter for the metal-insulator transition; symmetry-breaking aspects of the latter transition are thereby highlighted., Comment: submitted to J Chem Phys

Amorphous solids manifest puzzling effects of mysterious degrees of freedom that give rise to a heat capacity and phonon scattering in great excess over what would be expected for a solid that has a unique vibrational ground state. Of particular conceptual importance is the apparent near universality of phonon scattering in amorphous solids made by quenching a liquid. To rationalise this universality, scale-free scenarios have been proposed that either hinge on there being long-range interactions between bare structural degrees of freedom or that invoke long-range criticality stemming from the emergence of marginally stable vibrational modes. In a contrasting, local scenario, the puzzling low-temperature degrees of freedom are, instead, weakly-interacting, strongly anharmonic degrees of freedom each of which involves the motion of a few hundred particles. In this scenario, the universality of phonon scattering comes about because the characteristic energy scale of the local anharmonic resonances and the strength of their interaction with phonons are both set by the glass transition temperature $T_g$, while their concentration is set by the cooperativity size $\xi$ for dynamics at $T_g$. The nanoscopic length $\xi$ is manifested in vibrational excitations of the spatial boundary of the resonances, which underlie the so-called Boson peak, and very deep, topological midgap electronic states in glassy semiconductors, which are implicated in a number of strange optoelectronic phenomena in amorphous chalcogenides. I discuss the merits of the above scenarios when confronted with experimental data., Comment: Adv. Phys. X, 2018, vol. 3, 1510296. https://doi.org/10.1080/23746149.2018.1510296

We show systematically that a steady-state ensemble of mesoscopic inclusions of a solute-rich fluid can emerge in liquid solutions well outside the region of stability of the solute-rich phase. Unanticipated by conventional treatments, this type of reversible aggregation nonetheless can take place if the solute molecules bind transiently with each other to form long-lived complexes. The binding causes kinetic stabilization of inclusions of the solute-rich phase---within a substantial size range---so as to render the critical size for nucleation of the inclusions finite. Individual droplets nucleate and grow until they become mechanically unstable because of a concomitant drop in the internal pressure, the latter drop caused by the thermodynamic metastability of the solute-rich phase. At the same time, the {\em ensemble} of the droplets is steady-state on long times. In a freshly prepared solution, the ensemble is predicted to evolve similarly to the conventional Ostwald ripening, during which larger droplets grow at the expense of smaller droplets. The present mechanism is proposed to underlie the puzzling mesoscopic clusters observed in solutions of proteins and other molecules., Comment: submitted to Nature Materials

We determine the electronic density of states for computationally-generated bulk samples of amorphous chalcogenide alloys As$_{x}$Se$_{100-x}$. The samples were generated using a structure-building algorithm reported recently by us ({J. Chem. Phys.} ${\bf 147}$, 114505). Several key features of the calculated density of states are in good agreement with experiment: The trend of the mobility gap with arsenic content is reproduced. The sample-to-sample variation in the energies of states near the mobility gap is quantitatively consistent with the width of the Urbach tail in the optical edge observed in experiment. Most importantly, our samples consistently exhibit very deep-lying midgap electronic states that are delocalized significantly more than what would be expected for a deep impurity or defect state; the delocalization is highly anisotropic. These properties are consistent with those of the topological midgap electronic states that have been proposed by Zhugayevych and Lubchenko as an explanation for several puzzling opto-electronic anomalies observed in the chalcogenides, including light-induced midgap absorption and ESR signal, and anomalous photoluminescence. In a complement to the traditional view of the Urbach states as a generic consequence of disorder in atomic positions, the present results suggest these states can be also thought of as intimate pairs of topological midgap states that cannot recombine because of disorder. Finally, samples with an odd number of electrons exhibit neutral, spin $1/2$ midgap states as well as polaron-like configurations that consist of a charge carrier bound to an intimate pair of midgap states; the polaron's identity---electron or hole---depends on the preparation protocol of the sample., Comment: submitted to J Phys Chem B

Apart from not having crystallized, supercooled liquids can be considered as being properly equilibrated and thus can be described by a few thermodynamic control variables. In contrast, glasses and other amorphous solids can be arbitrarily far away from equilibrium and require a description of the history of the conditions under which they formed. In this paper we describe how the locality of interactions intrinsic to finite-dimensional systems affects the stability of amorphous solids far off equilibrium. Our analysis encompasses both structural glasses formed by cooling and colloidal assemblies formed by compression. A diagram outlining regions of marginal stability can be adduced which bears some resemblance to the quasi-equilibrium replica meanfield theory phase diagram of hard sphere glasses in high dimensions but is distinct from that construct in that the diagram describes not true phase transitions but kinetic transitions that depend on the preparation protocol. The diagram exhibits two distinct sectors. One sector corresponds to amorphous states with relatively open structures, the other to high density, more closely-packed ones. The former transform rapidly owing to there being motions with no free energy barriers; these motions are string-like locally. In the dense region, amorphous systems age via compact activated reconfigurations. The two regimes correspond, in equilibrium, to the collisional or uniform liquid and the so called landscape regime, respectively. These are separated by a spinodal line of dynamical crossovers. Owing to the rigidity of the surrounding matrix in the landscape, high-density part of the diagram, a sufficiently rapid pressure quench adds compressive energy which also leads to an instability toward string-like motions with near vanishing barriers. (SEE REST OF ABSTRACT IN THE ARTICLE.), Comment: submitted to J Phys Chem B

We first argue that the covalent bond and the various closed-shell interactions can be thought of as symmetry broken versions of one and the same interaction, viz., the multi-center bond. We use specially chosen molecular units to show that the symmetry breaking is controlled by density and electronegativity variation. We show that the bond order changes with bond deformation but in a step-like fashion, regions of near constancy separated by electronic localization transitions. These will often cause displacive transitions as well so that the bond strength, order, and length are established self-consistently. We further argue on the inherent relation of the covalent, closed-shell, and multi-center interactions with ionic and metallic bonding. All of these interactions can be viewed as distinct sectors on a phase diagram with density and electronegativity variation as control variables; the ionic and covalent/secondary sectors are associated with on-site and bond-order charge density wave respectively, the metallic sectorwith an electronic fluid. While displaying a contiguity at low densities, the metallic and ionic interactions represent distinct phases separated by discontinuous transitions at sufficiently high densities. Multi-center interactions emerge as a hybrid of the metallic and ionic bond that results from spatial coexistence of delocalized and localized electrons. In the present description, the issue of the stability of a compound is that of mutual miscibility of electronic fluids with distinct degrees of electron localization, supra-atomic ordering in complex inorganic compounds comes about naturally. The notions of electronic localization advanced hereby suggest a high throughput, automated procedure for screening candidate compounds and structures with regard to stability, without the need for computationally costly geometric optimization., Comment: 21 pages and 25 figures in main text with a 16 page/37 figure supplement containing method details. New revision corrects minor typographical errors in text on pages 3 and 7, and captions for figures 5, 6 and 11

We set up the problem of finding the transition state for phase nucleation in multi-component fluid mixtures, within the Landau-Ginzburg density functional. We establish an expression for the coordinate-dependent local pressure that applies to mixtures, arbitrary geometries, and certain non-equilibrium configurations. The expression allows one to explicitly evaluate the pressure in spherical geometry, \`a la van der Waals. Pascal's law is recovered within the Landau-Ginzburg density functional theory, formally analogously to how conservation of energy is recovered in the Lagrangian formulation of mechanics. We establish proper boundary conditions for certain singular functional forms of the bulk free energy density that allow one to obtain droplet solutions with thick walls in essentially closed form. The hydrodynamic modes responsible for mixing near the interface are explicitly identified in the treatment; the composition at the interface is found to depend only weakly on the droplet size. Next we develop a Landau-Ginzburg treatment of the effects of amphiphiles on the surface tension; the amphiphilic action is seen as a violation of Pascal's law. We explicitly obtain the binding potential for the detergent at the interface and the dependence of the down-renormalization of the surface tension on the activity of the detergent. Finally, we argue that the renormalization of the activation barrier for escape from long-lived structures in glassy liquids can be viewed as an action of uniformly seeded, randomly oriented amphiphilic molecules on the interface separating two dissimilar aperiodic structures. This renormalization is also considered as a "wetting" of the interface. The resulting conclusions are consistent with the random first order transition (RFOT) theory.

The random first-order transition (RFOT) theory of the structural glass transition is reviewed in a pedagogical fashion. The rigidity that emerges in crystals and glassy liquids is of the same fundamental origin. In both cases, it corresponds with a breaking of the translational symmetry; analogies with freezing transitions in spin systems can also be made. The common aspect of these seemingly distinct phenomena is a spontaneous emergence of the molecular field, a venerable and well-understood concept. In crucial distinction from periodic crystallisation, the free energy landscape of a glassy liquid is vastly degenerate, which gives rise to new length and time scales while rendering the emergence of rigidity gradual. We obviate the standard notion that to be mechanically stable a structure must be essentially unique; instead, we show that bulk degeneracy is perfectly allowed but should not exceed a certain value. The present microscopic description thus explains both crystallisation and the emergence of the landscape regime followed by vitrification in a unified, thermodynamics-rooted fashion. The article contains a self-contained exposition of the basics of the classical density functional theory and liquid theory, which are subsequently used to quantitatively estimate, without using adjustable parameters, the key attributes of glassy liquids, viz., the relaxation barriers, glass transition temperature, and cooperativity size. These results are then used to quantitatively discuss many diverse glassy phenomena, including: the intrinsic connection between the excess liquid entropy and relaxation rates, the non-Arrhenius temperature dependence of $\alpha$-relaxation, the dynamic heterogeneity, ... (see Comments for the remainder of Abstract.), Comment: violations of the fluctuation-dissipation theorem, glass ageing and rejuvenation, rheological and mechanical anomalies, super-stable glasses, enhanced crystallisation near the glass transition, the excess heat capacity and phonon scattering at cryogenic temperatures, the Boson peak and plateau in thermal conductivity, and the puzzling midgap electronic states in amorphous chalcogenides

We explore several potential issues that have been raised over the years regarding the "entropic droplet" scenario of activated transport in liquids, due to Wolynes and coworkers, with the aim of clarifying the status of various approximations of the random first order transition theory (RFOT) of the structural glass transition. In doing so, we estimate the mismatch penalty between alternative aperiodic structures, above the glass transition; the penalty is equal to the typical magnitude of free energy fluctuations in the liquid. The resulting expressions for the activation barrier and the cooperativity length contain exclusively bulk, static properties; in their simplest form they contains only the bulk modulus and the configurational entropy per unit volume. The expressions are universal in that they do not depend explicitly on the molecular detail. The predicted values for the barrier and cooperativity length and, in particular, the temperature dependence of the barrier are in satisfactory agreement with observation. We thus confirm that the entropic droplet picture is indeed not only internally-consistent but is also fully constructive, consistent with the apparent success of its many quantitative predictions. A simple view of a glassy liquid as a locally metastable, degenerate pattern of frozen-in stress emerges in the present description. Finally, we derive testable relationships between the bulk modulus and several characteristics of glassy liquids and peculiarities in low-temperature glasses., Comment: minor corrections made, appeared in J. Phys. Chem. B, vol. 118, pp. 13744-13759

We show that the vibrational response of a glassy liquid at finite frequencies can be described by continuum mechanics despite the vast degeneracy of the vibrational ground state; standard continuum elasticity assumes a unique ground state. The effective elastic constants are determined by the bare elastic constants of individual free energy minima of the liquid, the magnitude of built-in stress, and temperature, analogously to how the dielectric response of a polar liquid is determined by the dipole moment of the constituent molecules and temperature. In contrast with the dielectric constant---which is enhanced by adding polar molecules to the system---the elastic constants are down-renormalized by the relaxation of the built-in stress. The renormalization flow of the elastic constants has three fixed points, two of which are trivial and correspond to the uniform liquid state and an infinitely compressible solid respectively. There is also a nontrivial fixed point at the Poisson ratio equal to 1/5, which corresponds to an isospin-like degeneracy between shear and uniform deformation. The present description predicts a discontinuous jump in the (finite frequency) shear modulus at the crossover from collisional to activated transport, consistent with the RFOT theory., Comment: 24 pages, 9 figures, revised version, to appear in The Journal of Chemical Physics

We compute the temperature-dependent barrier for alpha-relaxations in several liquids, without adjustable parameters, using experimentally determined elastic, structural, and calorimetric data. We employ the random first order transition(RFOT) theory, in which relaxation occurs via activated reconfigurations between distinct, aperiodic minima of the free energy. Two different approximations for the mismatch penalty between the distinct aperiodic states are compared, one due to Xia and Wolynes, which scales universally with temperature as for hard spheres, and one due to Rabochiy and Lubchenko, which employs measured elastic and structural data for individual substances. The agreement between the predictions and experiment is satisfactory, given the uncertainty in the measured experimental inputs. The explicitly computed barriers are used to calculate the glass transition temperature for each substance---a kinetic quantity---from the static input data alone. The temperature dependence of both the elastic and structural constants enters the temperature dependence of the barrier over an extended range to a degree that varies from substance to substance. The lowering of the configurational entropy, however, seems to be the dominant contributor to the barrier increase near the laboratory glass transition, consistent with previous experimental tests of the RFOT theory using the XW approximation. In addition, we compute the temperature dependence of the dynamical correlation length, also without using adjustable parameters. These agree well with experimental estimates obtained using the Berthier et al. procedure. Finally, we find the temperature dependence of the complexity of a rearranging region is consistent with the picture based on the RFOT theory but is in conflict with the assumptions of the Adam-Gibbs and "shoving" scenarios for the viscous slowing down in supercooled liquids., Comment: submitted to J Phys Chem B

We propose a structural model that treats in a unified fashion both the atomic motions and electronic excitations in quenched melts of pnictide and chalcogenide semiconductors. In Part I (submitted to J. Chem. Phys.), we argued these quenched melts represent aperiodic $pp\sigma$-networks that are highly stable and, at the same time, structurally degenerate. These networks are characterized by a continuous range of coordination. Here we present a systematic way to classify these types of coordination in terms of discrete coordination defects in a parent structure defined on a simple cubic lattice. We identify the lowest energy coordination defects with the intrinsic midgap electronic states in semiconductor glasses, which were argued earlier to cause many of the unique optoelectronic anomalies in these materials. In addition, these coordination defects are mobile and correspond to the transition state configurations during the activated transport above the glass transition. The presence of the coordination defects may account for the puzzling discrepancy between the kinetic and thermodynamic fragility in chalcogenides. Finally, the proposed model recovers as limiting cases several popular types of bonding patterns proposed earlier, including: valence-alternation pairs, hypervalent configurations, and homopolar bonds in heteropolar compounds., Comment: 17 pages, 15 figures, revised version, final version to appear in J. Chem. Phys

Semiconductor glasses exhibit many unique optical and electronic anomalies. We have put forth a semi-phenomenological scenario (J. Chem. Phys. 132, 044508 (2010)) in which several of these anomalies arise from deep midgap electronic states residing on high-strain regions intrinsic to the activated transport above the glass transition. Here we demonstrate at the molecular level how this scenario is realized in an important class of semiconductor glasses, namely chalcogen and pnictogen containing alloys. Both the glass itself and the intrinsic electronic midgap states emerge as a result of the formation of a network composed of $\sigma$-bonded atomic $p$-orbitals that are only weakly hybridized. Despite a large number of weak bonds, these $pp\sigma$-networks are stable with respect to competing types of bonding, while exhibiting a high degree of structural degeneracy. The stability is rationalized with the help of a hereby proposed structural model, by which $pp\sigma$-networks are symmetry-broken and distorted versions of a high symmetry structure. The latter structure exhibits exact octahedral coordination and is fully covalently-bonded. The present approach provides a microscopic route to a fully consistent description of the electronic and structural excitations in vitreous semiconductors., Comment: 22 pages, 17 figures, revised version, final version to appear in J. Chem. Phys

We formulate a minimal ansatz for local stress distribution in a solid that includes the possibility of strongly anharmonic short-length motions. We discover a broken-symmetry metastable phase that exhibits an aperiodic, frozen-in stress distribution. This aperiodic metastable phase is characterized by many distinct, nearly degenerate configurations. The activated transitions between the configurations are mapped onto the dynamics of a long range classical Heisenberg model with 6-component spins and anisotropic couplings. We argue the metastable phase corresponds to a deeply supercooled non-polymeric, non-metallic liquid, and further establish an order parameter for the glass-to-crystal transition. The spin model itself exhibits a continuous range of behaviors between two limits corresponding to frozen-in shear and uniform compression/dilation respectively. The two regimes are separated by a continuous transition controlled by the anisotropy in the spin-spin interaction, which is directly related to the Poisson ratio $\sigma$ of the material. The latter ratio and the ultra-violet cutoff of the theory determine the liquid configurational entropy. Our results suggest that liquid's fragility depends on the Poisson ratio in a non-monotonic way. The present ansatz provides a microscopic framework for computing the configurational entropy and relaxational spectrum of specific substances., Comment: 11 pages, 5 figures, Final version published in J Phys Chem B

The steady state values of the viscosity and the intrinsic ionic-conductivity of quenched melts are computed, in terms of independently measurable quantities. The frequency dependence of the ac dielectric response is estimated. The discrepancy between the corresponding characteristic relaxation times is only apparent; it does not imply distinct mechanisms, but stems from the intrinsic barrier distribution for $\alpha$-relaxation in supercooled fluids and glasses. This type of intrinsic ``decoupling'' is argued not to exceed four orders in magnitude, for known glassformers. We explain the origin of the discrepancy between the stretching exponent $\beta$, as extracted from $\epsilon(\omega)$ and the dielectric modulus data. The actual width of the barrier distribution always grows with lowering the temperature. The contrary is an artifact of the large contribution of the dc-conductivity component to the modulus data. The methodology allows one to single out other contributions to the conductivity, as in ``superionic'' liquids or when charge carriers are delocalized, implying that in those systems, charge transfer does not require structural reconfiguration., Comment: submitted to J Chem Phys

We review the Random First Order Transition Theory of the glass transition, emphasizing the experimental tests of the theory. Many distinct phenomena are quantitatively predicted or explained by the theory, both above and below the glass transition temperature $T_g$. These include: the viscosity catastrophe and heat capacity jump at $T_g$, and their connection; the non-exponentiality of relaxations and their correlation with the fragility; dynamic heterogeneity in supercooled liquids owing to the mosaic structure; deviations from the Vogel-Fulcher law, connected with strings or fractral cooperative rearrangements; deviations from the Stokes-Einstein relation close to $T_g$; aging, and its correlation with fragility; the excess density of states at cryogenic temperatures due to two level tunneling systems and the Boson Peak., Comment: submitted to Ann. Rev. Phys. Chem

Melting is analyzed dynamically as a problem of localization at a liquid-solid interface. A Lindemann-like criterion of melting is derived in terms of particular vibrational amplitudes, which turn out to equal a universal quotient (about one-tenth) of the molecular spacing, at the interface. The near universality of the Lindemann ratio apparently arises owing to strongly overdamped dynamics near melting, and despite the anharmonic interactions' being system-specific. A similar criterion is derived for structural displacements in the bulk of the {\em solid}, in particular the premelted layer; the criterion is no longer strictly universal, but still depends only on the harmonic properties of the solid. We further compute the dependence of the magnitude of the elemental molecular translations, in deeply supercooled fluids, on the temperature and the high frequency elastic constants. We show explicitly that the surface tension between distinct liquid states, near the glass transition of a supercooled liquid, is nearly evenly split between entropic and energetic contributions., Comment: to appear in J. Phys. Chem. B

Amorphous solids exhibit intrinsic, local structural transitions, that give rise to the well known quantum-mechanical two-level systems at low temperatures. We explain the microscopic origin of the electric dipole moment of these two-level systems: The dipole emerges as a result of polarization fluctuations between near degenerate local configurations, which have nearly frozen in at the glass transition. An estimate of the dipole's magnitude, based on the random first order transition theory, is obtained and is found to be consistent with experiment. The interaction between the dipoles is estimated and is shown to contribute significantly to the Gr\"{u}neisen parameter anomaly in low $T$ glasses. In completely amorphous media, the dipole moments are expected to be modest in size despite their collective origin. In partially crystalline materials, however, very large dipoles may arise, possibly explaining the findings of Bauer and Kador, J. Chem. Phys. {\bf 118}, 9069 (2003)., Comment: Submitted for publication; April 27, 2005 version

The quantum excitations in glasses have long presented a set of puzzles for condensed matter physicists. A common view is that they are largely disordered analogs of elementary excitations in crystals, supplemented by two level systems which are chemically local entities coming from disorder. A radical revision of this picture argues that the excitations in low temperature glasses are deeply connected to the energy landscape of the glass when it vitrifies: the excitations are not low excited states built on a single ground state but locally defined resonances, high in the energy spectrum of a solid. According to a semiclassical analysis, the two level systems involve resonant collective tunneling motions of around two hundred molecular units which are relics of the mosaic of cooperative motions at the glass transition temperature $T_g$. The density of states of the TLS is determined by $T_g$ and the mosaic's length scale, which is a weak function of the cooling rate. The universality of phonon scattering in insulating glasses is explained. The Boson Peak and the plateau in thermal conductivity, observed at higher temperatures, are also quantitatively understood within the picture as arising from the same cooperative motions, but now accompanied by thermal activation of the mosaic's vibrational modes. The dynamics of some of the local structural transitions have significant quantum corrections to the semiclassical picture. These corrections lead to a deviation of the heat capacity and conductivity from the standard tunneling model results and explain the anomalous time dependence of the heat capacity. Interaction between tunneling centers contributes to the large and negative value of the Gr\"{u}neisen parameter often observed in glasses., Comment: submitted for publication, Dec. 17, 2004 version

The novel isotope and magnetic effects on two level system echoes in glasses observed by Nagel et al.(Phys.Rev.Lett. 92, 245511 (2004)) provide unique experimental insight into the nature of tunneling systems in amorphous solids. We show their observations are consistent with tunneling systems being extended systems having of order 100 moving units, as predicted by theory (Phys.Rev.Lett. 87, 195901 (2001)). The numerous nuclear quadrupole interactions within a tunneling center explain the magnitude of the magnetic and frequency dependence observed in deuterated glycerol. The extended nature of two level systems explains why a small residual effect can still be observed in regular glycerol even with its low abundance of quadrupolar nuclei., Comment: Submitted to Phys. Rev. Lett

The random first order transition theory of the dynamics of supercooled liquids is extended to treat aging phenomena in nonequilibrium structural glasses. A reformulation of the idea of ``entropic droplets'' in terms of libraries of local energy landscapes is introduced which treats in a uniform way the supercooled liquid (reproducing earlier results) and glassy regimes. The resulting microscopic theory of aging makes contact with the Nayaranaswamy-Moynihan-Tool nonlinear relaxation formalism and the Hodge-Scherer extrapolation of the Adam-Gibbs formula, but deviations from both approaches are predicted and shown to be consistent with experiment. The nonlinearity of glassy relaxation is shown to quantitatively correlate with liquid fragility. The residual nonArrhenius temperature dependence of relaxation observed in quenched glasses is explained. The broadening of relaxation spectra in the nonequilibrium glass with decreasing temperature is quantitatively predicted. The theory leads to the prediction of spatially fluctuating fictive temperatures in the long-aged glassy state, which have non-Gaussian statistics. This can give rise to ``ultra-slow'' relaxations in systems after deep quenches., Comment: Submitted to J. Chem. Phys

According to the Random First Order Transition (RFOT) theory of glasses, the barriers for activated dynamics in supercooled liquids vanish as the temperature of a viscous liquid approaches the dynamical transition temperature from below. This occurs due to a decrease of the surface tension between local meta-stable molecular arrangements much like at a spinodal. The dynamical transition thus represents a crossover from the low $T$ activated bevavior to a collisional transport regime at high $T$. This barrier softening explains the deviation of the relaxation times, as a function of temperature, from the simple $\log \tau \propto 1/s_c$ dependence at the high viscosity to a mode-mode coupling dominated result at lower viscosity. By calculating the barrier softening effects, the RFOT theory provides a {\em unified} microscopic way to interpret structural relaxation data for many distinct classes of structural glass formers over the measured temperature range. The theory also provides an unambiguous procedure to determine the size of dynamically cooperative regions in the presence of barrier renormalization effects using the experimental temperature dependence of the relaxation times and the configurational entropy data. We use the RFOT theory framework to discuss data for tri-naphthyl benzene, salol, propanol and silica as representative systems., Comment: Submitted to J. Chem. Phys

We argue that the intrinsic glassy degrees of freedom in amorphous solids giving rise to the thermal conductivity plateau and the ``boson peak'' in the heat capacity at moderately low temperatures are directly connected to those motions giving rise to the two-level like excitations seen at still lower temperatures. These degrees of freedom can be thought of as strongly anharmonic transitions between the local minima of the glassy energy landscape that are accompanied by ripplon-like domain wall motions of the glassy mosaic structure predicted to occur at $T_g$ by the random first order transition theory. The energy spectrum of the vibrations of the mosaic depends on the glass transition temperature, the Debye frequency and the molecular length scale. The resulting spectrum reproduces the experimental low temperature Boson peak. The ``non-universality'' of the thermal conductivity plateau depends on $k_B T_g/\hbar \omega_D$ and arises from calculable interactions with the phonons., Comment: 4 pages, submitted to PRL

Several puzzling regularities concerning the low temperature excitations of glasses are quantitatively explained by quantizing domain wall motions of the random first order glass transition theory. The density of excitations agrees with experiment and scales with the size of a dynamically coherent region at $T_g$, being about 200 molecules. The phonon coupling depends on the Lindemann ratio for vitrification yielding the observed universal relation $l/\lambda \simeq 150$ between phonon wavelength $\lambda$ and mean free path $l$. Multilevel behavior is predicted to occur in the temperature range of the thermal conductivity plateau., Comment: 4 pages, submitted to PRL