Search

Proton concentration of simple binary metaphosphate glasses under a pseudo fuel cell anode reaction was investigated at 350 °C by utilizing an in-situ Fourier transform infrared (FTIR) spectroscopy. Proton concentrations of both MO·P2O5 (M = Ca, Sr) glasses increase under humidified hydrogen atmosphere and applying a direct current voltage between platinum electrodes, suggesting that a proton infiltration occurs into these glasses. Compared with CaO·P2O5 glass, SrO·P2O5 glass shows higher proton concentration at the same period of time. Structure of glass network was investigated using synchrotron radiation X-ray diffraction, 31P magic-angle spinning nuclear magnetic resonance (MAS NMR) and FTIR (KBr method) spectroscopies. Strength of the cross-linking between phosphate chains via M2+ metal cations is correlated for the proton infiltration into phosphate glasses.

Thermal, structural and ionic conduction characteristics of lithium niobophosphate glass (mol% 50Li2O–45P2O5–5Nb2O5) are reported and compared with the characteristics of lithium metaphosphate glass (mol% 50Li2O–50P2O5). The presence of the second glass former (Nb2O5) not only improves the chemical durability and thermal stability but also enhances the conduction characteristics of lithium metaphosphate glass. The ionic conductivity measurements are reported in the extended temperature range i.e. above the glass transition temperature (Tg) where a non-Arrhenius temperature dependence of the ionic conductivity has been observed. The ion conduction process below Tg is governed by a well-defined activation energy whereas the conduction process above Tg is activated by at least two different phenomena discussed in the light of the structural relaxation and crystallization.

The thermodynamic and kinetics of the dehydration and hydration of CsH_(2)PO_(4) is investigated by thermogravimetric, differential scanning calorimetry and X-ray diffraction analysis in the temperature range of 200 to 400 °C, water partial pressure range of 0.06 to 0.90 atm. SiO_(2) powder was added to CsH_(2)PO_(4) to accelerate both dehydration and hydration. When dehydrated in the presence of SiO_(2), CsH_(2)PO_(4) has a fine microstructure (~200 nm feature size), and no reaction between CsH2PO4 and SiO_(2) is observed. By making use of the enhanced kinetics afforded by SiO_(2), the phase boundary between CsH_(2)PO_(4), CsPO_(3) and dehydrated liquid was precisely determined. The triple point connecting these phases is located at pH2O = 0.35 ± 0.2 atm and T = 267.5 ± 1.0 °C. The stability of CsH2PO4 and the liquid dehydrate, CsH_(2(1−x))PO_(4−x)(l), were confirmed by the complete reversal of dehydration to recover these phases in the appropriate temperature and water partial pressure ranges. Rehydration and conversion of CsPO3(s) to CsH2PO4(s) occur over a period of several hours, depending on temperature, water partial pressure, and morphology of the metaphosphate. High pH2O and small particles favor rapid dehydration, whereas the temperature dependence of the rehydration kinetics is non-monotonic, reaching its fastest rate in the vicinity of the superprotonic transition.

The ionic conductivity of CdI 2 doped silver ultraphosphate, metaphosphate and polyphosphate glasses has been reported in this paper. We have observed that the conductivity of metaphosphate and polyphosphate glasses increases and that of the ultraphosphate glasses decreases with the increase of CdI 2 content in the compositions. We have observed that insertion of CdI 2 expands the host glass network of all compositions leading to increased migration pathway for Ag + ions. However, the expansion of the glass network is not enough to influence the dynamics of Ag + ions in ultraphosphate glasses. We have explained the composition dependence of the conductivity of these glasses on the basis of coordination exchange between mobile Ag + and immobile Cd 2+ ions. Due to this exchange effect, Cd 2+ and Ag + ions coordinate to the oxygen and iodine ions respectively. The magnitude of the exchange is different for different phosphate networks and the conductivity can either increase or decrease depending on the magnitude of the exchange. The activation energy for the Ag + migration depends on the degree of Cd–Ag exchange as well as on the network structure of the glass compositions.

We report a TEM microstructural and compositional study of proton-conducting (La,Ca)P3O9 (Al-free) and (La,Ca,Al)P3O9 (Al-containing) glass-ceramics, prepared by the crystallization of quenched glass melts. Bright-field Transmission Electron Microscopy imaging, annular dark-field Scanning Transmission Electron Microscopy, and energy dispersive x-ray compositional analysis were used to characterize the structure and chemistry of the glass and crystalline phases of the samples that developed following the treatments at various temperatures, heating-rate and dwell-time. TEM chemical characterization revealed that the crystallized domains were mainly Ca-doped LaP3O9, while the remaining phase was a Ca-rich phosphate glass. Crystallization improved conductivity first (up to a conductivity of nearly 10− 4.5 S/cm at 550 °C in air). However, the conductivity dropped when the samples were treated at higher temperature or for longer dwell-times, due to a continued Ca-departure from the La-rich crystalline into glass phase. Possible proton transfer mechanisms and strategy to improve conductivity in this doped rare earth metaphosphate glass-ceramic are discussed.

The electrical properties of [La(1 − x)Mx]–P3O9 metaphosphate glasses, where M is Ba, Sr, Ca, and 0 ≤ x ≤ 0.8 were investigated in the 300–500 °C range. The protonic conductivity increases two orders of magnitude from the unsubstituted to the 60% substituted glass reaching a maximum conductivity of about 10− 5 S/cm at 450 °C. The activation energy is found to be approximately independent of concentration though dependent on the type of modifying substitutional cation, and decreases from about 1.04 eV for Ca and Sr to ∼ 0.96 eV for Ba. AC impedance measurements were used to find the DC ionic conductivities. Mean ion travel distances at the frequencies, νΖ* or νΜ*, associated with the maximum in Z″ or M″ were found to be up to 30 times larger than the average oxygen spacing in the glasses. While diffusion coefficients derived from conductivities, Dσ, were vastly lower than an uncorrelated diffusion coefficient computed from first principles, the travel distance at νZ* or νZ* derived from a simple random walk expression using DΓ, was surprisingly close to those derived from the more rigorous statistical treatments. The ratio ξ = Dσ/DΓ ≪ 1 may be interpreted as indicative of highly correlated ion motion and strongly reduced mean attempt frequencies, together with a mean activation energy for motion increased by ΔE = Eσ − EΓ over that of the uncorrelated intertetrahedral motion of an ideal crystal characterized by DΓ.

Al 2 O 3 was added to a 2CaO–La 2 O 3 –5P 2 O 5 metaphosphate, to replace 10% of the Ca 2+ ions by Al 3+ , forming a phosphate with the nominal composition 1.8CaO–0.1Al 2 O 3 –La 2 O 3 –5P 2 O 5 . The effect of Al 2 O 3 addition and heat treatment on the microstructure and conductivity of the resulting glass–ceramics was investigated by XRD, SEM, TEM, and AC impedance spectroscopy. Upon transformation from glass to glass–ceramic, conductivities increased significantly. The glasses were isochronally transformed at 700 and at 800 °C for 1 h or 5 h, in air, following heating at 3 or 10 °C/min. With Al 2 O 3 addition, after a heat treatment at 700 °C, 100–300 nm nano-domains of LaP 3 O 9 crystallized from the glass matrix. Annealing at 800 °C produced a further order of magnitude conductivity increase for the Al-free glass, but less so for the Al-containing glass.

The influence of the preparation conditions and metaphosphate secondary phase on the proton conduction and chemical stability of the single and polycrystalline CsH 2 PO 4 samples was studied by the ac admittance technique. The relation between frequency dependences of the admittance plots and adsorbed water as well as a content of the metaphosphate phase was determined. It was shown that in ambient conditions the thermal decomposition reactions start on the sample's surface at temperatures above specific point, T 1 = 425 K, and the rate of the decomposition reactions on the interfaces is strongly influenced by the sample's quality. It is suggested that this reaction is initiated by the melting of the thin layer of CsH 5 (PO 4 ) 2 on the surface of CsH 2 PO 4 . The variation of the contributions of the intergrain and electrode–proton conductor boundaries to the total admittance with the temperature was studied for polycrystalline CDP samples and obtained data were compared with those for nominally pure CDP single crystal samples.

The glass forming regions of MIn–AgPO3 (M=Li, Na, K, Rb, Cs and Tl for n=1, M=Cd for n=2 and M=Bi for n=3) systems were determined. The glass transition temperature, density and ionic conductivity for these glasses were investigated. The increase of ionic conductivity was obtained with increasing iodide content in all the glasses examined in the present investigation. Ionic conductivity was distinguished by the difference of valence of M ion. It was found that isothermal conductivity and activation energy for MIn–AgPO3 glasses are mainly affected by iodine ion concentration. It is considered that ion exchange between M and Ag ions occurs in the glass and then Ag ions contiguous with iodine ions tend to move in the gateway formed by iodine ions and phosphate chains.

dc Biasing with a pulse train enabled us to follow the process of solid-state anodization of aluminum on silver metaphosphate glass, a silver-ion conductor. Anodization induced formation of a dielectric oxide in which the capacitance (around 10 −7 F/cm 2 ) decreased with increasing bias field, caused by an increase in oxide thickness. The oxide formed would not be the genuine alumina but contains phosphorus (and possibly silver).

The structures of ion conducting mixed-alkali metaphosphate glasses (A 2 O) x –(B 2 O) 1− x –P 2 O 5 (A,B=Li,Na,Rb) have been examined by neutron diffraction and Raman spectroscopy. The results show that the short range order of the phosphorous–oxygen network is almost independent of the composition. The length-scale of the intermediate range order correlations increases with increasing size of the cation. For all the investigated compositions the nearest Li–O, Na–O and Rb–O distances have been estimated from the atomic pair correlation functions to be 2.0±0.1, 2.35±0.1 (Na–O I), 2.6±0.1 (Na–O II) and 2.9±0.1 A, and the corresponding average coordination numbers are approximately 4, 3 (I)+2.5 (II) and 5, respectively. Thus, the cation environments are distinctly different for Li + , Na + and Rb + and each of them remain the same for all compositions within experimental errors. The alkali ions are quasi-uniformly distributed and we were not able to observe any significant preference for either similar or dissimilar pairs. This implies that the mixed-alkali effect can be qualitatively understood on a purely structural basis. The A sites are energetically unfavourable to the B + ions (and vice versa) and jumps to dissimilar sites are then unlikely to occur. Thus, the A sites will have the tendency of blocking the pathways for the B + ions and vice versa.

The influence of the structure of the glass network as well as the influence of gradual substitution of CuI by CuCl in the systems (40 − x)CuI-xCuCl-30Cu2O-30P2O5 (metaphosphate) and (40 − x)CuI-xCuCl-45Cu 2O-15P2O5 (orthophosphate), respectively, on the properties of Cu+-ion-conducting glasses were investigated. The glass formation was observed in the complete range of substitution of CuI by CuCl in both systems under study. The conductivities of the metaphosphate glasses reach orders of 10−3 S cm−1, the conductivities of orthophosphate glasses are one order of magnitude lower. Moreover, substitution of CuI by CuCl in the metaphosphate system causes a positive deviation of the conductivity from the additivity rule, with a maximum deviation for the molar ratio Cl(Cl +I) = 0.750.

Cuprous ion conductive glasses were prepared in the systems (40 − x)CuIxCuBr30Cu2O30P2O5 and (40 − x)CuI xCuCl30Cu2O30P2O5. Glass-formation was observed in the complete range of substitution of CuBr or CuCl for Cul. The glass transition temperatures of prepared glasses ranged from 355 K up to 407 K. Infrared studies confirmed that the glasses contain mainly metaphosphate chain groups. The conductivity of the glasses at room temperature is of the order of 10−4 S cm−1 to 10−3 S cm−1 and shows a positive deviation from the additivity rule for both systems, with a maximum deviation for the molar ratio X(X + I) = 0.750.

The structure of glasses prepared in the Na2OGa2O3P2O5 system was investigated by vibrational (IR, Raman) and X-ray absorption spectroscopies. For the orthophosphate compositions, gallium is found in four-fold coordination, whereas it is found in six-fold coordination in a metaphosphate glass. For pyro-, tri and tetraphosphate glasses, gallium atoms are distributed on tetrahedral and octahedral sites. Gallium atoms found on octahedral sites are likely chelated by pyrophosphate groups. The ionic conductivity of the glasses depends mainly on their sodium content.

Raman spectroscopy and X-ray photoelectron spectroscopy (XPS) are used to investigate the 0.5 Li 2 O:0.5[(1− x )P 2 O 5 : x WO 3 ] glassy system. Raman spectroscopy shows that besides the expected metaphosphate tetrahedra and tungstate tetrahedra, pyrophosphate tetrahedra and tungstate octahedra are also present. A reaction in the glass melt is proposed: 2 LiPO 3 +2 Li 2 WO 4 ⇌Li 4 P 2 O 7 +Li 2 W 2 O 7 . The proportions of the various structural groupings are determine d from the associated Raman peak areas. Structural groupings characteristic of the highly conducting Li 4 P 2 O 7 and Li 2 W 2 O 7 phases are the most abundant at around x =0.6. This offers a plausible explanation for the conductivity maximum previously observed near this composition. From the XPS spectra, binding energies of the Li 1s, P 2p, W 4f and O 1s core-levels are determined. The O 1s spectrum is found to consist of two peaks which are attributed to the P=O, P-O-P, P-O, W-O-W and W-O oxygen species. The proportion of the P-O-P species decreases as x increases. Good agreement is found between the XPS data and the Raman data.

Seeking possible divalent cation conducting glasses, we have investigated glass formation and conductivity in Pb ( PO 3 ) 2 + PbX 2 (X = halide) glasses. Glass formation is found to extend to 40 mol% PbF 2 , 60 mol% PbCl 2 , 80 mol% PbBr 2 but no glasses could be obtained with PbI 2 . Comparison of PbBr 2 - and PbCl 2 -containing glasses suggests the observed conductivity is anionic. In this case it seems that 40 Pb(PO 3 ) 2 ·60 PbCl 2 , with σ 200°C = 7.08 × 10 −6 Ω −1 cm , may be the best anion-conducting glass yet reported. The PbBr 2 -containing glasses have the lower activation energies, however, which would be consistent with cationic conduction. t + data are need to clarify.

Glass formation and electrical conductivity have been studied in the system LiF-Li2O-Al(PO3)3. Compositions have been specified by the ratio, N, of Li ions to the sum of glass-formers, Al + P, and by the ratio, X, of fluorine atoms to phosphorus atoms. Clear glasses were obtained with N=1 to 2.3 and X=0.5 to 2.5. Limited substitutions of Br for F and of BO 3 2 for 1 4 Al(PO3)3 were found possible. A correlation was found between the activation energies for conduction, Ea, and the pre-exponential constants, A. Both Ea and A were found to decrease with increasing N and to increase with increasing X. Anderson and Stuart's method of calculating activation energy is criticized, and a simple electrostatic calculation of Ea is introduced in which Ea depends on the Li+ ion jump distance calculated from the mean Li+ ion concentration. However this model cannot explain the large observed variations of A and its correlation with E. It is suggested that large variations of jump distance with small variations of composition occur due to a change of mechanism from single to multiple-ion jumps.