Silver sulfide based glasses (I). Glass forming regions,structure and ionic conduction of glasses in GeS2Ag2S and GeS2Ag2SAgI systems

Ag₂S forms with GeS₂ stable glasses over a wide range of compositions (0–55% Ag₂S mol%). In the same system, more complex glasses obtained by dissolving silver iodide have been synthesized with up to 50 mol% AgI.Raman spectra are presented and a vibrational assignment in terms of bridging and non-bridging sulfur has been made. The electrical conductivity of these glasses has been measured over a temperature range (?50°C? + 50°C) and for various compositions by the complex impedance diagram method. At 25°C, the conductivity reached a maximum value of 6 × 10^?3 Ω^?1 cm^?1. Whatever the glass used, the same limit value of conductivity $\text{(σ ? 10 su?2 Ω}{}^{\mathrm{?1}}\text{cm}{}^{\mathrm{?1}}$) and activation energy ($\text{E}{}_{\mathrm{\sigma }}\text{? 0.25}\text{eV}$) are obtained for the highest content of silver iodide. A conduction mechanism is proposed.     

Energy transfer in Eu3+-Yb3+ and Tb3+-Yb3+ system in glasses

The behavior of the phonon-assisted energy transfer between trivalent rare-earth ions in glasses was investigated. The ions Eu³⁺ and Tb³⁺ as energy donors and Yb³⁺ as acceptor were selected. The energy gap between the levels of the donor and acceptor was estimated on the basis of the energy diagram of each ion determined from absorption and emission spectra. The probability for the transfers of (Eu, ⁵D₀-⁷F₆): (Yb, ${}^2\text{F}{}_{\mathrm{<mtext>7</mtext><mtext>2</mtext>}}\text{-}{}^2\text{F}{}_{\mathrm{<mtext>5</mtext><mtext>2</mtext>}}$) and (Tb, ⁵D₄-⁷F₀): (Yb, ${}^2\text{F}{}_{\mathrm{<mtext>7</mtext><mtext>2</mtext>}}\text{-}{}^2\text{F}{}_{\mathrm{<mtext>5</mtext><mtext>2</mtext>}}$) in silicate, borosilicate, phosphate and germanate glasses was measured in the temperature range of liquid-nitrogen temperature - 650K. The probability of transfer was the smallest in phosphate glass and B₂O₃ had the effect of increasing it. In germanate glass the dependence of the probability of the energy gap was relatively weak. These results were correlated to the difference in the phonon energy and the strength of the electron-lattice coupling in each glass.     

Nucleation kinetics of sodium disilicate

The kinetics of crystal nucleation in Na₂O · 2SiO₂ have been determined over the range of undercoolings between 173 and 373°C. The plot of log(I_v?) versus $\text{1}\text{ΔT}{}^2{}_{\mathrm{r}}\text{T}{}^3{}_{\mathrm{r}}$ is a straight line of negative slope over some 13 orders of magnitude in I_v. The slope of this relation indicates a nucleation barrier of about 45 kT at ΔT_r = 0.2, and the intercept at $\text{1}\text{ΔT}{}^2{}_{\mathrm{r}}\text{T}{}^3{}_{\mathrm{r}}\text{= 0}$ is 10²⁶ cm^-3 sec^-1. poise. The results are in good agreement with predictions of the theory of homogeneous nucleation, even in the pre-exponential factor.     

Retarded elasticity in B2O3GeO2 glasses

The retarded elasticity was investigated for B₂O₃GeO₂ glasses having network structure in the glass transition range by using a compressive method. The compliance (J_∞) determined at the final stage of each measurement displayed a maximum for roughly constant viscosity (η ? 10¹⁴ P) in all the B₂O₃GeO₂ glasses and was simulated by the same equation applied for AsS glasses reported in a previous paper [1].

$\text{J}{}_{\mathrm{\infty }}\text{=(1?k}{}_2\text{keta;}{}_{\mathrm{G}}\text{keta;}\text{)[?}{}_1\text{(}\text{k}{}_1\text{keta;}\text{)+?}{}_2\text{(}\text{nk}{}_1\text{keta;}\text{)]}$

, where $\text{K}{}_1\text{, k}{}_2\text{, ?}{}_1\text{, ?}{}_2\text{and}\text{n}$ are parameters and η_G is the viscosity related to the retarded elasticity. The terms (k₁/η) and (nk₁/η) are assumed to be equal to one for all their values exceeding one. For B₂O₃GeO₂ glasses, the deformation due to the retarded elasticity could be alloted to two structural elements: the first element related to the term $\text{?}{}_1\text{(k}{}_1\text{/η)}$ and the second element related to the term $\text{?}{}_2\text{(nk}{}_1\text{/η)}$. The values of $\text{?}{}_1$ showed almost no variance with the glass composition, but $\text{?}{}_2$ had a minimum at the composition of 50 mol% GeO₂. These data suggest that the contribution of the second element is the smallest at B₂O₃/GeO₂ = 1. The values of k₂ were close to that of As₂S₃ glass having the network structure. k₁ and n (or nk₁) were almost constant regardless of the composition, respectively. These data suggest that the inhibition due to the viscosity starts at an approximately constant viscosity in B₂O₃GeO₂ glasses.     

Study of amorphous Ge–SiO2–P+Si tunnel junctions

A new method to determine ac conductivity of amorphous Ge using Al-amorphous Ge–SiO₂–P⁺Si tunnel junctions is presented. Frequency dependence of ac conductivity is found to satisfy the power law in the frequency range between 1 and 50 kHz and the density of localized states at the Fermi level is estimated to be ～ 1.7 × 10²⁰ cm^?3 eV^?1 which decreases to ～ 4.5 × 10¹⁹ cm^?3 eV^?1 after annealing at 175°C.Temperature dependence of tunneling conductance of Al-amorphous Ge–SiO₂–P⁺Si junctions is appreciable only near zero bias. Zero bias conductance of the junctions obeys the $\text{T}{}^{\mathrm{<mtext>?</mtext><mtext>1</mtext><mtext>4</mtext>}}$ law of Mott; the density of localized states obtained from the $\text{T}{}^{\mathrm{<mtext>?</mtext><mtext>1</mtext><mtext>4</mtext>}}$ law is one order of magnitude smaller than that obtained by ac conductivity measurements, being insensitive to annealing. This behavior of the tunnel junctions differes in many respects from those of Al–Al₂O₃-amorphous Ge tunnel juntions.     

A new method of stirring for LPE growth

This paper describes a method of stirring a liquid during crystal growth in which each point on the growth face moves round a circle of radius b. If the face has a radius a and completes n revolutions per second, then the stirring is roughly equivalent to rotating the disc about a single axis at a rate $\text{2πc}{}^2\text{b}{}^2\text{n}\text{a}{}^2$ where c is a constant (c ∽ 2) but the system behaves as if only the liquid up to a distance $\text{(}\text{2v}\text{n}\text{)}{}^{\mathrm{<mtext>1</mtext><mtext>2</mtext>}}$ from the face is stirred. Uniform stirring occurs over the whole face except for an annulus of width 2b at the periphery. The necessary design criteria are discussed and a simple dynamically balanced apparatus and the results obtained with it whilst growing iron garnet films are described.     

Electrical,structural and optical properties of amorphous carbon

The planar and transverse electrical resistivity of amorphous carbon (a-C) films getter-sputtered at low temperature (77–95 K) is well-fitted by the expression $\text{? = ?}{}_0\text{exp}\text{(T}{}_0\text{/T)}{}^{\mathrm{<mtext>1</mtext><mtext>4</mtext>}}$ The exponent T₀ being approximately the same in both cases (≈ 7 × 10⁷ K) suggests that the amorphous films are isotropic. Films thinner than 600 Å display a two-dimensional hopping conductivity from which one deduces a density of states N(E_F) at the Fermi level of 10¹⁸ eV^?1 cm^?3 and a radius of the localized wave functions (a) of 12 Å. Tunneling experiments and optical absorption measurements are consistent with a pseudogap of approximately 0.8 eV. Electron diffraction experiments indicate that a-C films consist of a mixture of diamond and graphite bonds; this fact taken in the light of the other experiments would suggest that the graphite bonds act as the localized conduction states.     

Electric field ionization from deep trapping levels in a-Selenium

The deep trapping levels in a-selenium have been measured by the electric field stimulated current method. It is found that the enhancement of detrapping from the trapping levels is proportional to EXP ($\text{α/}\text{kTxE}{}^{\mathrm{<mtext>1</mtext><mtext>2</mtext>}}$) where $\text{α/}\text{kT}\text{=0.012 (}\text{m/V}\text{)}{}^{\mathrm{<mtext>1</mtext><mtext>2</mtext>}}$) and E is electric field. The distribution of trapping time is also obtained.     

Solarization mechanism of glass containing Ce3+ and As5+

The solarization mechanism in a glass containing both Ce³⁺ and As⁵⁺, 16Na₂O·11CaO·73SiO₂:0.15AsO_x·0.015CeO_x (in mol.%), is newly proposed by elucidating the valence and coordination structure of arsenic after the photochemical reaction, the mechanism being traditionally expressed as

$\text{2}\text{Ce}{}^{\mathrm{3+}}\text{+}\text{As}{}^{\mathrm{5+}}\text{→}\text{hv}\text{2}\text{Ce}{}^{\mathrm{4+}}\text{+}\text{As}{}^{\mathrm{3+}}$

ESR hyperfine quartets due to an As⁴⁺ ranging from 0.1 to 0.5 T built up on UV-irradiation and their line shape varied with the duration of the irradiation. The line shape analysis of the ESR spectra employing a computer simulation technique has led to the following conclusions; (1) As⁵⁺ is reduced to As⁴⁺ in the solarization process. (2) The geometry around the As⁴⁺ in the solarized glass is tetrahedral during the early stage and trigonal-pyramidal during the latter stage of the reaction.     

Electrical conductivity and photoconductivity of As2S3PbS glasses

Measurements of dc electrical conductivity and photoconductivity of various glassy compositions (x = 0.1?0.625) in (As₂S₃)_1?x(PbS)_x have been made. Experimental results of the temperature dependence of dc conductivity from room temperature to 200°C (which includes the glass transition temperature) are reported. All the compositions exhibit intrinsic conduction in the measured temperature range. Thermal activation energy, glass transition temperature and σ₀ for the compositions studied, were determined from the experimental data. The low value of $\text{σ}{}_0\text{(10?10}{}^{\mathrm{?2}}\text{Ω}{}^{\mathrm{?1}}\text{cm}{}^{\mathrm{?1}}$) in these semiconducting glasses is attributed to the greater participation of localized states in the conduction process.In the measurements of photoconductivity, the variation of photocurrent with temperature, photon energy, light intensity and electric field is observed. The recombination model has been involved to explain the results of photoconductivity. Both electrical and photoconductivity data support the presence of higher density of localized states in the x = 0.1 composition than in others.     

Electrical properties of glassy AsxGe10Te90−x alloys

This paper analyses the electrical properties of glassy alloys of As_xGe₁₀Te_90?x, while reporting the conductivity and dielectric constant of As₅Ge₁₀Te₈₅ and As₁₅Ge₁₀Te₇₅ compositions in the temperature range 77–383 K and the frequency range from dc to 5 MHz. The dc conductivity has been shown to be of the form

The ratio σ₀₁/σ₀₂ is of the order of 10⁶. ΔE₁, the higher temperature activation energy, is dependent on the composition, while ΔE₂, the lower temperature activation energy, is less dependent on the composition. The dielectric constant has been found to be independent of temperature and frequency up to about 253 K. However, at higher temperatures, it becomes activated and proportional to log ω.Some common features of As_xGe₁₀Te_90?x are a kink in dc conductivity, a ω^0.8 relationship for ac conductivity, no evidence of variable-range hopping at low temperatures, field-dependent conductivity and memory switching. The data can be interpreted in terms of the dangling-bond theory of Mott and his collaborators. A high density of states of the order of 10²⁰eV^?1 cm^?3 near the Fermi level may be expected.     

Theory of molten zone shape and stability

The Laplace-Young capillary equation for the shape of an axisymmetric liquid-vapor interface has been solved numerically for boundary conditions relevant to a model of the floating zone process. The stability of these solutions with respect to axisymmetric and asymmetric perturbations which conserve volume has been determined via the conjugate point criterion of the calculus of variations. The liquid zone shape is governed by five dimensionless parameters: $\text{R}{}_{\mathrm{m}}\text{R}{}_{\mathrm{f}}$, $\text{L}\text{Rf}$, $\text{V}\text{φR}{}^2{}_{\mathrm{f}}\text{L}$, $\text{? = ρg}\text{R}{}^2{}_{\mathrm{f}}\text{γ}$, and $\text{?}{}_{\mathrm{R}}\text{= ρΩ}{}^2\text{R}{}^3{}_{\mathrm{f}}\text{γ}$, where R_m and R_f are the radii of the melting and freezing solids, respectively, L is zone length , V is the zone volume, ρ is the density difference between liquid vapor, g is the gravitational acceleration, γ is the liquid-vapor surface tension, and Ω is the constant angular velocity of the uniformly rotating zone. For growth of constant diameter crystals, the angle ø_f, measured between the meniscus and the growth axis at the freezing interface, is constant. For R_m = R_f, ?_R = 0, and ø_f = 0, the maxi mum value of ? for which a stable liquid zone exists has been calculated for various values of L/R_f. For some values of ?, two different stable liquid zones with different volumes (but all other parameters identical) give the same value of ø_f.     

Observation and simulation of step motion on evaporating crystal surfaces

The motion of interacting parallel monatomic 〈10〉 steps on free-evaporating (100) NaCl cleavage surfaces has been investigated electron-microscopically and simulated numerically in the temperature range from 260 to 320°C. To make visible the advance of steps due to annealing a special matched-face decoration technique was employed. By comparison of simple step configurations before and after their isothermal activation the process-determining parameters could be obtained: velocity of isolated monatomic straight steps $\text{(υ}{}_{\mathrm{<mtext>isol</mtext>}}\text{(T=1.25 × 10}{}^{23}\text{exp}\text{(}\text{-2.5eV}\text{kT}\text{)}\text{A}\text{?}\text{min}\text{)}$, surface self-diffusion length $\text{(x}{}_{\mathrm{s}}\text{(T)=5.5}\text{exp}\text{(}\text{0.31 eV;}\text{kT)}\text{A}\text{?}\text{)}$, and retardation factor (ζ(T)=1). Moreover, any symmetry in diffusion fields at a straight step was found to be negligible in the present case. By means of these parameters of motion it was possible to stimulate mutual interactions of steps observed within finite trains of parallel 〈10〉 steps. The confrontation of experimental and numerical results has shown that the applied step model which bases on the extended BCF theory describes the propagation of monatomic straight steps on evaporating NaCl cleavage surfaces in detail.     

Relation between electrical and optical gaps of amorphous semiconductors in the SiAsTe system

Electrical and optical properties of semiconducting SiAsTe glasses have been investigated. Compositional dependences of the properties in the Si_xAs_yTe_z system are examined as a function of atomic percentage x (or y, z) of one element with parameters of constant atomic ratio y/z (or x/z, x/y) of the other two elements. A pre-exponential factor σ₀ in the dc conductivity formula is estimated to be $\text{(2.1 ± 0.6) × 10}{}^4\text{(Ω ·}\text{cm}\text{)}{}^{\mathrm{?1}}$, inependently of the compositions. A systematic relationship between the compositional changes in the electrical gap E_g(el) and optical gap E_g(op) has been found. The energy gaps increase linearly with increasing Si content and decreasing Te content, but are almost independent of As content. The relation between E_g(el) and E_g(op) is expressed by E_g(el) = 1.60 E_g(op) ? 0.15 in eV. On the other hand, the optical absorption coefficient α(Ω) near the band edge follows the empirical formula, $\text{α(Ω) = α}{}_0$ exp ($\text{h}\text{?}\text{Ω/E}{}_{\mathrm{<mtext>s</mtext>}}$). The experimentally determined factor E_s increases linearly with E_g(op) and is closely related to the energy difference between the two gaps. A tentative model to explain these experimental results is proposed by taking into account of the effect of the potential fluctuations in such disordered materials.     

Effect of thermal history on conductivity and electrical relaxation in alkali silicate glasses

Electrical conductivity σ₀ and electric field relaxation measurements have been carried out as a function of thermal history for two alkali silicate glasses, Na₂O3SiO₂ and K₂O3SiO₂. Specimens of each glass with three different thermal histories, two of the anneal-and-quench type and one of the rate-cool type, were studied. The average structural or fictive temperature T_f of each of the specimens was characterized by measuring their indices of refraction. Effects of thermal history on σ₀ and its activation enthalpy $\text{H}{}_{\mathrm{\sigma }}{}^1$ were in accord with results of previous investigators. That is, for a given type of thermal history σ₀ was lower and $\text{H}{}_{\mathrm{\sigma }}{}^1$ higher the lower T_f. In addition it was found that for two specimens with the same T_f or index of refraction but different thermal histories the rate-cooled specimen exhibited a lower conductivity than the annealed-and-quenched specimen, in accord with the results of Ritland. The distribution of relaxation times τ_σ for decay of the electric field due to ionic migration was found to be due primarily to a distribution in the pre-exponential term ln $\text{τ}{}_{\mathrm{\sigma }}{}^1$ in the equation $\text{ln}\text{τ}{}_{\mathrm{\sigma }}\text{=}\text{ln}\text{τ}{}_{\mathrm{\sigma }}{}^1\text{+ H}{}^1\text{/RT}$; the distribution in $\text{H}{}^1$ was extremely narrow. Differences in thermal history caused small differences in the distribution of τ_σ, but no difference in the average activation enthalpy $\text{〈H}{}^1\text{〉}$ for τ_σ. From this result it appeared that the dependence of the conductivity activation enthalpy $\text{H}{}_{\mathrm{\sigma }}{}^1$ on thermal history was due to the effect of thermal history on the temperature dependence of the distribution in τ_σ.     

X-ray photoemission study of fluorozirconate glass and related crystals

Silicate glasses with the composition (in per cent by weight) of 74 SiO₂, 20 Na₂O, 6 CaO and different Fe₂O₃ concentrations were subjected to X-ray irradiation and the behaviour of the iron as well as that of the radiation-induced defects was investigated systematically using ESR. The defects which are abundant in iron-free glasses (they produce a strong signal at g = 2.01) and which lead to colouring of the glass, decrease with rising Fe₂O₃ concentration and cease to exist at 2% Fe₂O₃. These glasses are radiation-resistant.Because of the X-rays a valence change of the iron takes place. This process is initiated by a radiation-induced redox interaction between Fe³⁺ and Fe²⁺ according to:

$\text{Fe}{}^{\mathrm{2+}}\text{+}\text{Fe}{}^{\mathrm{3+}}\text{→}\text{hv}\text{Fe}{}^{\mathrm{3+1}}\text{+}\text{Fe}{}^{\mathrm{2+1}}$

Because of the different structures (coordinations) of Fe²⁺ and Fe³⁺ the process leads to a decrease in Fe³⁺ at g = 4.2 in glasses with low Fe₂O₃ concentration. This decrease is reduced with increasing concentration of Fe₂O₃.     

Small polaron conduction in V2O5P2O5 glasses

Dc conductivity measurements have been made between 90 and 520 K on three bulk samples of V₂O₅P₂O₅ glass. Heat treatment is found to result in a reduction of the activation energy at a given temperature and this is most noticeable at low temperatures. The behaviour at low temperatures can be described using Mott's variable range hopping arguments, and at high temperatures by non-adiabatic small polaron hopping between nearest neighbours. At intermediate temperatures a simple model is used in which excitations by optical and acoustic phonons are considered to make independent contributions to the jump frequency. Mott's theory is extended to the polaron case for $\text{T>}\text{1}\text{4}\text{?}$ and is shown to be in good agreement with results. Values for $\text{r}{}_{\mathrm{<mtext>p</mtext>}}\text{(～2.8}\text{A}\text{?}\text{)}$ the polaron radius and $\text{α(～3.5}\text{A}\text{?}{}^{\mathrm{?1}}\text{)}$ the electron decay constant are shown to be consistent with the model for small polarons. A method is suggested for obtaining α and N(E_F) from the ac conductivity and the slope of 1nσ versus $\text{1}\text{T}{}^{\mathrm{<mtext>1</mtext><mtext>4</mtext>}}$ at low temperatures. Values of N(E) are obtained which correlate with those obtained by the previous analysis. This implies that the disorder energy separating adjacent sites Δ₀ is large (～0.4 eV) in these materials.     

Electrical and optical properties of amorphous boron and amorphous concept for ß-rhombohedral boron

Measurements of the conductivity (σ), thermoelectric power (S) and thermal conductivity (κ) of amorphous boron are made over wide temperature ranges (T = 77–850 K for σ, T = 300–850 K for S and T = 80–1100 K for κ). The room temperature spectral dependencies of the reflection (R) and absorption (α) coefficients are determined for the wavelength intervals 2–25 μm and 1.3–25 μm respectively. The I–V characteristics are also studied and shown to be consistent with the Poole-Frenkel law.The value obtained for the thermal energy gap of amorphous boron (1.3 eV) is slightly smaller than that of crystal ß-rhombohedral boron (1.4 eV). The temperature dependence of the electrical conductivity can be satisfactorily described by the Mott law $\text{ln}\text{σ ≈ ?(T}{}_0\text{/T)}{}^{\mathrm{<mtext>1</mtext><mtext>4</mtext>}}$, where $\text{T}{}_0\text{? 10}{}^8\text{K}$. This gives an estimate, N ≈ 10¹⁸ cm^?3, for the concentration of trapping levels responsible for the hopping conduction. The value $\text{?}{}_0\text{? 9}$ is found from the spectral dependence of R while α has Urbach-like character $\text{? α ≈}\text{exp}\text{(}\text{h}\text{?}\text{ω/Δ)}$, where $\text{Δ ? 0.19}\text{eV}$.A comparison is made between amorphous boron and crystalline ß-rhombohedral boron. Because of the very complex crystal structure and the large dimensions of the unit cell of ß-boron, some of its physical properties could be qualitatively described on the basis of the so-called ‘amorphous concept’.     

Spinodal decomposition during continuous cooling of a PbOB2O3Al2O3 glass

Spinodal decomposition during continuous cooling of the PbOB₂O₃Al₂O₃ quasi-binary glass system was analysed by numerical integration of Cook's differential equation (which includes the contribution of random density fluctuations) for small angle X-ray scattering (SAXS) intensity. The SAXS curves derived from the calculations have a wide range of k-Fourier components (0 < k < k_c) for which a positive amplification factor occurs and they show a “crossover” point at $\text{k}{}_{\mathrm{c}}\text{= 0.155}\text{A}\text{?}{}^{\mathrm{?1}}$. The wavenumber which receives maximal amplification, k_m, increases with the cooling rate, Q, as $\text{Q}{}^{\mathrm{<mtext>1</mtext><mtext>n</mtext>}}$, with n = 10.9. This Q dependence of k_m is similar to that predicted by Huston et al., however our results show a higher value of n. The dependence on Q and k_m of the SAXS intensity I(k_m) was also deduced. The measurements of SAXS curves were performed on glass samples prepared by the splat-cooling technique. Because of the difficulties which arise in the determination of the cooling rate of the samples, the only experimental results that could be compared with the theory are the k_m dependence of I and the value of k_c. These results are satisfactorily understood in terms of the present analysis.     

Kinetics of ion exchange in glasses

The kinetics of $\text{K}{}^{\mathrm{+}}\text{?}\text{Na}{}^{\mathrm{+}}$ exchange in two glass systems, 20Na₂O·(60?x)B₂O₃· (20 + x)Si₂ (where x = 0, 15, 30 and 45 mol%) and Na₂O·3SiO₂, were studied as a function of glass composition, salt bath composition, exchange temperature and time The distribution of K in the glass specimens after exchange in molten KNO₃ was determined with an electron probe. Stresses in these speciments were measured photoelastically. The interdiffusion coefficient $\text{D}$ for ion exchange was calculated as a function of local composition in the glass using the Boltzmann-Matano method. The strong variation of $\text{D}$ in any particular glass approximated that predicted by a mixed alkali model (as advanced by Lacharme), where the glass in the ion-exchanged region approximates a composite of stacked layers of mixed alkali glasses with a gradually varying alkali ratio. The small discrepancy between the experiment and the mixed alkali model was partly, but not fully, reconciled by considering the strains in the glasses. The observation which remained unexplained was that the calculated stress profiles did not show perfect agreement, both in magnitude and in shape, with the experimentally measured stress profiles. It appeared that the kinetics of ion exchange in the glasses were also influenced by a network relaxation process which may have occurred well below the glass transition temperature.