Nuclear magnetic resonance studies of alkaline earth phosphosilicate and aluminoborosilicate glasses

The ¹¹B, ²⁷Al, ²⁹Si and ³¹P magic angle spinning (MAS) NMR spectra of MO–P₂O₅, MO–SiO₂–P₂O₅ and MO(M^′₂O)–SiO₂–Al₂O₃–B₂O₃ (M=Mg, Ca, Sr and Ba, M^′=Na) glasses were examined. In binary MO–P₂O₅ (M=Ca and Mg) glasses, the distributions of the phosphate sites, P(Q_n), can be expressed by a theoretical prediction that P₂O₅ reacts quantitatively with MO. In the ternary 0.30MO–0.05SiO₂–0.65P₂O₅ glasses, the 6-coordinated silicon sites were detected, whose population increases in the order of MgOxCaO–0.05SiO₂–(0.95−x)P₂O₅ glasses, its population increases with an increase in f (=([P₂O₅]−[MO]−[B₂O₃]−[Na₂O])/[SiO₂]) and has maximum at f=9. The signal due to the 5-coordinated silicon atoms is also observed when x is smaller than 0.45. When three network-forming oxides such as SiO₂, Al₂O₃ and B₂O₃ coexist, Al₂O₃ reacts preferably with MO. The populations of 4-coordinated boron atoms, N₄, are expressed well with r/(1−r), where r=([Na₂O]−[Al₂O₃])/([Na₂O]−[Al₂O₃]+[B₂O₃]). The correlation of the Raman signal at 1210 and 1350 cm⁻¹ with the NMR signal of Si(Q₆) at −215 ppm is also seen.     

Properties and structure of RO---Na2O---Al2O3---P2O5 (R = Mg, Ca, Sr, Ba) glasses

The properties and structure of (45 - x)RO · xNa₂O · 2.5Al₂O₃ · 52.5P₂O₅ (R = Mg, Ca, Sr, Ba, 0 x 31 mol%) glasses were investigated. The variation in the molar volumes of glasses in the MgO series is closely related to the formation of the end groups in the glasses with the substitution of Na⁺ ions for Mg²⁺ ions, resulting in a variation of the density and refractive index of the glasses. The properties of glasses containing CaO in terms of Na₂O substitution depend mainly on the low field strength of Na⁺ ions substituting for CaO even though the end groups occurring in the glasses increased. The variation in properties of the glasses containing SrO and BaO, some of which were substituted by Na₂O, could be explained by differences in masses, field strength and polarizability between the Na⁺ ions and the alkaline-earth ions due to a small variation in the structure of the glasses despite Na₂O substitution.     

Li conductivity in lithium niobate: silica glasses

Glasses of composition 65 mol% LiNbO₃:: 35 mol% SiO₂ have been shown to be Li⁺ ion conductors with a conductivity at 200°C > 1 × 10⁻⁵ (η cm)⁻¹ and an activation energy of 0.54 eV. The addition of approximately 0.1 mol% Fe₂O₃ leads to an enhancement of conductivity to ≈10⁻³ (η cm)⁻¹ at 200°C and an activation energy of 0.67 eV. The effect of Fe is shown to be in the control of microstructure in the glass, with Fe₂O₃ concentrations < 1 mol% acting as a grain growth inhibitors and larger concentrations acting as a nucleating agents. A model for this process based on the expected stoichiometry of the melt and the effect of Fe²⁺ and Fe³⁺ in charge compensation is in excellent agreement with experimental data from electron spin resonance.     

Study of electrical properties of MoO3–Fe2O3–P2O5 and SrO–Fe2O3–P2O5 glasses by impedance spectroscopy. II

The electrical and dielectric properties for three series of MoO₃–Fe₂O₃–P₂O₅ and one series of SrO–Fe₂O₃–P₂O₅ glasses were measured by impedance spectroscopy in the frequency range from 0.01 Hz to 3 MHz and over the temperature range from 303 to 473 K. It was shown in Part I that the MoO₃ is incorporated into phosphate network and the structure/properties are strongly influenced by the overall O/P ratio. The Fe₂O₃ content and Fe(II)/Fe_tot ratio in these glasses have significant effects on the electrical conductivity and dielectric permittivity. With decreasing Fe₂O₃ content in MoO₃–Fe₂O₃–P₂O₅ glasses with O/P at 3.5 the dc conductivity, σ_dc(ω) decreases for two orders of magnitude, which indicates that the conductivity for these glasses depends on Fe₂O₃ and is independent of the MoO₃ content. Also, the dielectric properties such as ^′(ω), ^″(ω) and σ_ac(ω) and their variation with frequency and temperature indicates a decrease in relaxation intensity with increase in the concentration of MoO₃. On the other hand, the dc conductivity for MoO₃–Fe₂O₃–P₂O₅ glasses with O/P > 3.5 increases with the substitution of MoO₃ which has been explained by an increase in the number of non-bridging oxygens and formation of Fe–O–P bonds that are responsible for formation of small polarons. The increase in the dielectric permittivity, ^′(ω) with increasing MoO₃ content is attributed to the increase in the deformation of glass network with increasing bonding defects. For SrO–Fe₂O₃–P₂O₅ glasses the conductivity and dielectric permittivity remained constant with increasing SrO.     

Mössbauer studies on some glasses and crystals in the Na2O---Fe2O3---SiO2 system

Mössbauer absorption measurements have been made at room temperature on ⁵⁷Fe in iron sodium silicate glasses containing 3–15 mol% Fe₂O₃ and various iron alkali silicate crystals in order to study the state of iron in these glasses. The spectra of all the glasses gave one doublet with a quadrupole splitting varying from 0.73–0.78 mm s⁻¹, while those of Na₂O · Fe₂O₃ · 4 SiO₂ and 5 Na₂O · Fe₂O₃ · 8 SiO₂ crystals showed much smaller quadrupole splitting, 0.28 mm s⁻¹ and 0.10 mm s⁻¹, respectively, and an asymmetrical doublet of much narrower linewidth. When sodium was replaced by other alkali metals of larger size, such as K and Cs, in MFeSi₂O₆ and MFeSi₃O₈ crystals, the quadrupole splitting became wider and approached to 0.73 mm s⁻¹. Such a variation was not observed for glasses. These results suggest that a larger number of non-identical sites exist in iron sodium silicate glasses than in the corresponding crystals.     

The effect of MgO on the thermodynamics of the Fe/Fe-redox equilibrium and the incorporation of iron in soda-magnesia-aluminosilicate melts

Melts with the basic compositions 10Na₂O · 10MgO · xAl₂O₃ · (80−x)SiO₂ (x=0, 5, 10, 15 and 20), 10Na₂O · xMgO · 10Al₂O₃ · (80−x)SiO₂ (x=5, 10, 15 and 20) and xNa₂O · 10MgO · 10Al₂O₃ · (80−x)SiO₂ (x=5, 10 and 15) all doped with 0.25 mol% Fe₂O₃ were studied using square-wave voltammetry. The temperatures applied were in the range of 1000–1600 °C. The square-wave voltammograms recorded show peaks caused by the reduction of Fe³⁺ to Fe²⁺. The attributed peak potentials measured decreased linearly with decreasing temperatures. Increasing the MgO-concentration led to more negative peak potentials. Introducing alumina in the melt first resulted in less negative peak potentials. If the molar Al₂O₃-concentration is equal to that of Na₂O (=10 mol%) the peak potentials are least negative. Further increase of the Al₂O₃-concentration led to more negative peak potentials. The variation of the Na₂O-concentration led to a maximum in the peak potentials at an Na₂O-concentration of 10 mol%. An empirical formula which allows the calculation of standard potentials from the chemical composition is proposed. Furthermore, a structural explanation for the effect of the chemical composition is given. Especially, the incorporation of Al₂O₃ as AlO₄⁻-tetrahedra at [Al₂O₃] < [Na₂O] and as network modifier at larger concentrations was structurally explained by the similarities of Fe²⁺ and Mg²⁺, with respect to cation radii and metal–oxygen bond lengths.     

The effect of hydrostatic pressure on the dielectric constants, and their temperature dependences, of phosphate and tellurite glasses

Measurements of the pressure dependence of the static dielectric constant of tellurite (pure TeO₂ and 67%TeO₂ + 33%WO₃) and samarium phosphate (5%Sm₂O₃ + 95%P₂O₅ and 15%Sm₂O₃ + 85%P₂O₅) glasses at elevated pressures (0–70 kbar) for a range of temperatures (77–380 K) are reported. The electrical properties under pressure have been determined from the low-frequency complex plane analysis of glass discs contained within a Bridgman opposed anvil cell. The most notable observation is that the pressure dependence of the static dielectric constant, of all glasses studied, is positive, for example for vitreous TeO₂ ln ε/P is equal to 4.41 × 10⁻¹¹ (Pa⁻¹) at 293 K. Behaviour of this type is common to a number of materials (plastics and chalcogenide glasses) for which it is not possible to define any long-range order. It is in direct contrast with the behaviour of crystalline insulators, for which ε/P is usually negative. The effect of pressure on the dielectric constant has been analysed using two different approaches based on the macroscopic Clausius-Mossotti equation. The effects of high pressure on the dielectric constant have been correlated with the temperature dependence of the dielectric constant at atmospheric pressure.     

An infrared study of crystallization in sodium-disilicate glasses containing iron oxides

The infrared absorption spectra of sodium-disilicate glasses containing various amounts of Fe₂O₃ ([Na₂O · 2 SiO₂]_1−x [Fe₂O₃]_x, where X = 0.05, 0.1 and 0.2) were investigated in the wavenumber range from 200–2000 cm⁻¹. The addition of Fe₂O₃ to the sodium-disilicate glass does not seem to introduce any new absorption band as compared with the spectrum of a pure sodium-disilicate glass; nevertheless, a general shift of the existing absorption bands toward lower wavenumbers is observed. The amount of shift is, in fact, proportional to the content of Fe₂O₃ in the glass. This observation is consistent with the recently proposed structural model for the bonding of Fe³⁺ ions in the iron-sodium-silicate glass system.

Annealing of 20 mol% iron oxide glasses at 550 and 580°C produced an extra sharp infrared absorption peak at about 610 cm⁻¹ wavenumber. This new peak is believed to be related to the crystallized particles of the glass as concluded from both a scanning electron micrograph and an electron diffraction pattern.     

Microstructure and growth rate of chemical vapour deposited TiN films in a vertical cold wall reactor

TiN films were grown on SUS304 substrates heated by an induction furnace in a vertical cold wall reactor. Scanning electron microscopy, transmission electron microscopy and X-ray diffraction were used to characterize the microstructures of films obtained at different deposition conditions (temperature, gas flow rate and gas composition). Film structures obtained in the present vertical reactor had the following features compared with those in the tubular reactor: (1) Abnormally grown “star-shaped” crystals were observed on the surfaces of films deposited in the following ranges of total gas flow rate (Q_T), temperature (T) and partial pressures (P): 9.0×10⁻⁶ ≤ Q_T ≤ 1.6×10⁻⁵ m³ s⁻¹, 1223 ≤ T ≤ 1273 K, 0.92 ≤ P_TiCl4 ≤ 6.18 kPa, P_H2 = P_N2. The matrix grains were responsible for (211) preferred orientation. (2) Surface morphologies did not vary so much with P_TiCl4. On the other hand, a drastic change was brought about by adding HCl to the source gas, i.e., plate-shaped crystals dominated and the large “star-shaped” crystals were no longer present. (3) The apparent activation energy for deposition reaction was 230 kJ/mol (1173 ≤ T ≤ 1273 K) and 76.5 kJ/mol (1273 ≤ T ≤ 1373 K) at P_TiCl4 = 2.43 kPa and P_H2 = P_N2 = 49.45 kPa.     

Structural distribution and rigidity of the glass network determined by EPR of Cu

The linewidth-broadening of the EPR spectra of Cu²⁺ in silicate, borate and phosphate glasses was analyzed in terms of the distribution of g_| and A_| (δ_| and δA_|) and related to the distribution of the rigidity of the network structure. X- and K-band spectra were measured for the glasses doped with ⁶³Cu²⁺ (93% abundance). The linewidth of the HFS shoulders with parallel orientation to H increased linearly with increasing m or microwave frequency. δg_| and δA_| showed a marked dependence on glass composition. For example, in Na₂O---B₂O₃ glasses, on going from x (mol% of Na₂O) being small through intermediate to large, δg_| varied from small through large to negligibly small. In contrast to these glasses δg_| was extremely large for 75PbO · 25B₂O₃ glass. The large δg_| for the Na₂O---B₂O₃ glassesof intermediate x was attributed to the coexistence of various borate groups competitively coordinating to Cu²⁺. Negligibly small δg_| for 70Na₂O · 30B₂O₃ glass and extremely large δg_| for 75PbO ·25B₂O₃ glass, both with a narrower structural distribution, reflect regidity of the glass network. The Pb---O bonding is strong enough to distort the coordination of Cu²⁺-complex. The situation is the reverse in Na₂O---B₂O₃ glasses.     

The effect of mixed alkaline earths on the diffusivity and the incorporation of iron in 5Na2O · xMgO · (15−x)CaO · yAl2O3 · (80−y)SiO2 melts

Diffusion coefficients of iron were measured in glass melts with the basic compositions 5Na₂O · xMgO · (15−x)CaO · yAl₂O₃ · (80−y)SiO₂ with x=5, 10 and y=0, 5, 7.5, 10 and 15. The melts were doped with 0.25 mol% Fe₂O₃ and studied in the temperature range from 1000 to 1600 °C using square-wave voltammetry. The voltammograms exhibited distinct peaks attributed to the reduction of Fe³⁺ to Fe²⁺, from which peak currents mixed diffusion coefficients of iron were calculated. Diffusion coefficients in all melt compositions which did not show crystallization could be fitted to Arrhenius equation. The diffusivities measured in different melt compositions were related to the same viscosity, i.e. not the same temperature. Increasing the alumina concentration from 5 to 10 mol% resulted in an increase of the viscosity corrected diffusivities. At further increasing alumina concentrations, the diffusivities get smaller again. This can be explained by the stabilizing effect of Na⁺ and Ca²⁺ on FeO⁻₄ and AlO⁻₄-tetrahedra, which strengthens the incorporation of Fe³⁺ into the glass structure.     

Internal friction in vanadium-phosphate glasses doped with Na2O

The internal friction of _xNa₂O·(0.5−x)V₂O₅·O.5P₂O₅(x = 0.025–0.3) glasses was studied using the low-frequency torsion pendulum technique. The temperature spectrum of internal friction reveals three maxima. Maximum 1, the so-called “electron” maximum, is the same as observed in binary vanadium-phosphate glasses. The origin of maximum 2 can be attributed to ion migration. Maximum 3 appears for glasses containing more than 10 mol.% Na₂O and is probably connected with sodium-proton interactions.     

Alkali diffusion and electrical conductivity in sodium borate glasses

The diffusion coefficient of sodium, ²²Na, and silver, ¹¹⁰Ag, and electrical conductivity for sodium borate glasses (4–24 mol% Na₂O) has been measured from 100°C to slightly below the glass transition temperature, T_g. Unlike silicate glasses where the self-diffusion coefficient is much larger than the impurity diffusion coefficient, D_Na and D_Ag had close to the same magnitude in the sodium borate glasses. In some glasses, D_Ag was slightly larger than D_Na. Na₂O-Ag₂O-B₂O₃ glasses show a relatively small mixed alkali effect, despite the significant mass difference between Ag(≈108) and Na(≈23). It is concluded that the mixed alkali effect is more dependent upon the difference in ionic radii rather than differences in mass.     

Electrical conductivities of mixed cation glasses

The electrical conductivities of (1−x) Li₂O · x BaO · 2 SiO₂, (1−x) Na₂O · x MgO ·2 SiO₂, (1−x) Na₂O · x CaO · SiO₂ and (1−x) Na₂O · x BaO · 2SiO₂ glasses were measured at temperature ranging from room temperature to 450°C. The transport numbers for Na⁺ ion in (1−x) Na₂O · x BaO · 2 SiO₂ glasses were measured. It was found that the alkali ion carried a significant part of the current in these glasses except one that had no alkali ions, and the conductivity decreased markedly as the alkali oxide was substituted by an alkaline earth oxide. The results of conductivity measurements combined with the data hitherto reported on mixed alkali glasses led to the proposal that the so-called mixed alkali effect could be explained on the basis of the independent path model in which it is assumed that cations can move only through vacant sites left by those of the same type.     

Oscillator strength of the infrared absorption band near 1080 cm in SiO2 films

It has been well known that the absorption maximum of the peak near 1080 cm⁻¹ in amorphous SiO₂ films shifts continuously with variation of thickness and properties such as stress. This is a first report on the oscillator strength of the absorption against frequency at the absorption maximum. SiO₂ films on silicon wafers were prepared by thermal growth in either dry O₂ or an O₂/H₂ mixture or liquid-phase deposition in HF saturated with silica gel. The oscillator strength continuously decreased from 1×10⁻⁴ down to 1×10⁻⁵ with the frequency shift from 1099 to 1063 cm⁻¹.     

SIMS study of the concentration of dopants in LPCVD silicon thin films

Measurements of solid phase dopant concentration (S) of LPCVD Si thin films as a function of substrate temperature (T_s = 500−640 ° C) and gas phase doping ratio (R = 1 × 10⁻⁵ −4 × 10⁻²) by SIMS indicate different behaviors of P and B in the films. A linear relation S = b(T)R is observed for B-doped film with b(T) varying from 4 to 50 depending on T_s. Boron-doped microcrystalline film has a higher doping efficiency than that of P-doped ones.     

Raman spectroscopic study on the structure of ZnCl2---ZnX2 and ZnCl2---KX (X = Br, I) glasses

Raman spectra have been measured for ZnCl₂---ZnX₂ and ZnCl₂---KX (X = Br, I) glasses to investigate the structure of the glasses with varying composition. The assignment of each band was made, and the change of the spectra with composition was explained in terms of the bridging and non-bridging states of halide ions and the change of the tetrahedral units, ZnX_nCl_4−n²⁻ (n = 0–4), formed in the glasses. As the content of ZnX₂ in ZnCl₂---ZnX₂ glasses increases (20 → 80 mol%), the peak frequency of the Zn---Cl stretching mode increases (238 → 248 cm⁻¹ in X = I glasses, 238 → 259 cm⁻¹ in X = Br glasses) while the Zn---I and Zn---Br stretching frequencies decrease (173 → 120 cm⁻¹ for Zn---I, 196 → 157 cm⁻¹ for Zn---Br). The decrease of the Zn---I and Zn---Br band frequencies was attributed to the increase of the number n of the ZnX_nCl_4−n²⁻ tetrahedra. The increase of the Zn---Cl frequency suggests the existence of the bonding state of Cl⁻ ions which is intermediate between the bridging and the non-bridging states. In ZnCl₂---KX glasses, the Zn---Cl_non-bridging band at about 300 cm⁻¹ was observed in addition to the bands observed in ZnCl₂---ZnX₂ glasses. The addition of KX produces non-bridging anions while the tetrahedral units, ZnX_nCl_4−n²⁻ are also formed.     

Determination of diffusion coefficient of titanium ion in TiO2---SiO2 wet gel prepared from metal alkoxides during leaching

It was found that the radial gradient refractive index (r-GRIN) glass rod could be produced by the leaching of wet gel prepared by a sol-gel method from metal alkoxides and heat-treatment of the gel. The concentration gradient of the doped component resulting from its diffusion in the wet gel was studied. In order to estimate the exact diffusion coefficient of the titanium ion in the binary TiO₂---SiO₂ wet gel, its value was measured by using two methods. The observed diffusion coefficient by colorimetric analysis was about 1.1 × 10⁻⁶ cm²/s, while the value obtained from analysis of EPMA was about 1.2 × 10⁻⁶ cm²/s, and both values were nearly equal to each other. The activation energy of titanium ion diffusion in the wet gel was estimated to be about 11 kcal/mol.     

EXAFS and XANES joint analyses for semiconducting vanadium phosphate glasses

Vanadium EXAFS spectra of 50V₂O₅-50P₂O₅ glasses with different C = V⁺⁴/V_tot have been measured. The V-O distances increase by ΔR₁ = (0.03 ± 0.02) Å to ΔR₂ = (0.07 ± 0.03) Å going from a glass with C = 0.64 to C = 0.84 and from C = 0.51 to C = 0.84, respectively. The EXAFS data show a basically similar structure of the vanadium sites for both the V⁴⁺ and V⁵⁺ ionic states. The density of the glasses increases with C from the value d₁ = 2.81 g/cm³ for C = 0.51 to d₂ = 2.92 g/cm³ for C = 0.84 indicating a more random packing of glass units.     

Ionic transport and defect structure of vitreous beryllium fluoride

Vitreous BeF₂ was prepared by two techniques; (1) remelting of a technical grade material, and (2) vacuum distillation/fluoridation. Infrared spectroscopy studies have established that the first material contains about 0.5 wt.% hydroxyl, predicted to be coherently incorporated into the vitreous network as edge-linked [Be(OH)₄]²⁻ units. The distilled BeF₂ is water-free. The dc electrical conductivity of the remelted BeF₂ was measured as σ = (7.9 × 10³/T) exp(−24500 cal/mol/RT) ω⁻¹ cm⁻¹ and for the distilled BeF₂ as σ = (3.0 × 10⁵/T) exp(−36700 cal/mol/RT ω⁻¹ cm⁻¹ at temperatures to 280°C. Ionic transport studies utilizing a dc electrolysis polarization technique with N₂−F₂ and H₂−HF gas electrodes have demonstrated that the fluorine ion is the transport species. A general model for fluorine transport is proposed based upon a modified anti-Frenkel defect model. The difference in the fluorine transport process for the undistilled grade of BeF₂ is seen as a consequence of the anti-Frenkel defect pair interaction with the [Be(OH)₄[²⁻ groupings.