Search NASA⌕ Search

SEARCH · Search NASA

Results for “diffractive networks”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 235 records · Page 13

Deep learning of structural morphology imaged by scanning X-ray diffraction microscopy

Scanning X-ray nanodiffraction microscopy is a powerful technique for spatially resolving nanoscale structural morphologies by diffraction contrast. One of the critical challenges in experimental nanodiffraction data analysis is posed by the convergence angle of nanoscale focusing optics which creates simultaneous dependency of the far-field scattering data on three independent components of the local strain tensor-corresponding to dilation and two potential rigid body rotations of the unit cell. All three components are in principle resolvable through a spatially mapped sample tilt series; however, traditional data analysis is computationally expensive and prone to artifacts. In this study, we implement NanobeamNN, a convolutional neural network specifically tailored to the analysis of scanning probe X-ray microscopy data. NanobeamNN learns lattice strain and rotation angles from simulated diffraction of a focused X-ray nanobeam by an epitaxial thin film and can directly make reasonable predictions on experimental data without the need for additional fine-tuning. We demonstrate that this approach represents a significant advancement in computational speed over conventional methods, as well as a potential improvement in accuracy over the current standard.

Luo, Aileen [Cornell Univ., Ithaca, NY (United Sta↗

Structure of jadeite melt at high pressures up to 4.9;#8201;GPa

The structure of jadeite (NaAlSi{sub 2}O{sub 6}) melts has been studied using multiple-angle energy-dispersive x-ray diffraction up to 4.9 GPa and 1923 K. The first sharp diffraction peak in the structure factor shifts toward higher momentum transfer as pressure increases, indicating the shrinkage of the intermediate network in the melt. The radial distribution function shows a monotonous decrease in average T-T length and T-O-T angle with increasing pressure, but displays no detectable change in the average bond length between tetrahedrally coordinated cations and oxygen (T-O length, where T = Si{sup 4+}, Al{sup 3+}). Our observations indicate that the dominant structural changes occur in the intermediate range order at pressures up to 4.9 GPa. The changes in T-O length, T-T length, and T-O-T angle appear to correlate with the viscosity anomaly in this pressure range.

36 MATERIALS SCIENCE↗

A new glass-free LTCC microwave ceramic – (1 − x) Li{sub 2.08}TiO{sub 3} + xLiF

Graphical abstract: - Highlights: • Sintering temperature of Li{sub 2}TiO{sub 3} was reduced to 900 °C by doping with LiF and excessive lithium. • A new glass-free LTCC microwave ceramic – (1 − x) Li{sub 2.08}TiO{sub 3} + xLiF was obtained. • The obtained LTCC material exhibited excellent microwave dielectric properties. - Abstract: The structure and microstructure of the (1 − x) Li{sub 2.08}TiO{sub 3} + xLiF (0.11 ≤ x ≤ 0.14) ceramics were studied by X-ray powder diffraction (XRD), thermal dilatometry and scanning electron microscopy (SEM). The microwave dielectric properties of the ceramics were studied with a network analyzer. All samples exhibited single monoclinic phase. The sintering temperature was successfully reduced to about 900 °C/2 h for the LiF-doped Li{sub 2.08}TiO{sub 3} specimens. The dielectric permittivity slightly decreased, whereas the Q × f value changed little, with the increase in LiF dopant. The temperature coefficient of resonant of frequency (τ{sub f}) value changed from positive to negative with increasing x, and a near zero value was obtained at x = 0.13 composition. Optimized microwave dielectric properties with ε{sub r} = ∼22.8, Q × f = ∼63,000 GHz and τ{sub f} = ∼1.0 ppm/°C for x = 0.13 composition sintered at 900 °C/2 h. The XRD and back scattering SEM analysis showed that the ceramic was compatible with Ag powders after sintering at 900 °C/2 h.

36 MATERIALS SCIENCE↗

Local structural investigation of non-crystalline materials at high pressure: the case of GeO 2 glass

Abstract Local structures play a crucial role in the structural polyamorphism and novel electronic properties of amorphous materials, but their accurate measurement at high pressure remains a formidable challenge. In this article, we use the local structure of network-forming GeO 2 glass as an example, to present our recent approaches and advances in high-energy x-ray diffraction, high-pressure x-ray absorption fine structure, and ab initio first-principles density functional theory calculations and simulations. Although GeO 2 glass is one of the best studied materials in the field of high pressure research due to its importance in glass theory and geophysical significance, there are still some long-standing puzzles, such as the existence of appreciable distinct fivefold [5] Ge coordination at low pressure and the sixfold-plus [6+] Ge coordination at ultrahigh pressure. Our work sheds light on the origin of pressure-induced polyamorphism of GeO 2 glass, and the [5] Ge polyhedral units may be the dominant species in the densification mechanism of network-forming glasses from tetrahedral to octahedral amorphous structures.

Physics↗

Tri-molybdenum phosphide (Mo 3 P) and multi-walled carbon nanotube junctions for volatile organic compounds (VOCs) detection

Detection and analysis of volatile organic compounds' (VOCs) biomarkers lead to improvement in healthcare diagnosis and other applications such as chemical threat detection and food quality control. Here, we report on tri-molybdenum phosphide (Mo3P) and multiwalled carbon nanotube (MWCNT) junction-based vapor quantum resistive sensors (vQRSs), which exhibit more than one order of magnitude higher sensitivity and superior selectivity for biomarkers in comparison to pristine MWCNT junctions based vQRSs. Transmission electron microscope/scanning tunneling electron microscope with energy dispersive x-ray spectroscopy, x-ray diffraction, and x-ray photoelectron spectroscopy studies reveal the crystallinity and the presence of Mo and P elements in the network. The presence of Mo3P clearly enhanced the performance of vQRS as evidenced in sensitivity and selectivity studies. The vQRSs are stable over extended periods of time and are reproducible, making them a potential candidate for sensing related applications.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Sparse Representations to Replace TOFD Images in Non-destructive Testing of Materials

This paper describes a proposed method for the selection of relevant samples of ultrasonic signals during automatic material inspection. Instead of the well-known time of flight diffraction (TOFD) images, data are stored as a sparse matrix in which the elements only indicate whether a defect has been detected. This technique avoids storage of useless signals received during probe displacement in cases of low and high signal-to-noise ratios that correspond to coarse-grained and fine-grained materials, respectively. The approach is based on comparing the positions of maximum amplitudes, which are randomly located when signals only consist of noise but are in the same signal range when a defect is detected. The matrix elements are then applied as inputs to a self organizing map by neural networks to produce a normalized sparse matrix as the output, with a constant number of elements. This approach will be beneficial to enable the use of selected data in intelligent systems requiring a fixed number of inputs to characterize defects.

42 ENGINEERING↗

Navigating high-dimensional process-structure–property relations in nanocrystalline Pt-Au alloys with machine learning

For decades, materials scientists have relied on the process-structure–property paradigm to guide investigations into material behaviors. Traditional studies often examine a limited number of process-structure–property variables, striving to elucidate mechanisms governing material response. However, this approach is time consuming and can limit exploration, as well as the discovery of process-structure–property relations in novel materials. In this paper, we combined combinatorial sputter deposition and multi-modal high-throughput materials characterization with feedforward neural networks to establish high-dimensional process-structure–property relations in Pt-Au alloys, yielding nanocrystalline alloys with high hardness and low resistivity relevant to electrical contact switch applications. We mapped three indicators of process conditions (composition and two atomic deposition characteristics) onto four indicators of material structure (X-ray diffraction, film thickness, density, and surface roughness) and two indicators of material properties (hardness and resistivity), resulting in 784 unique combinations evaluated over a 13-dimensional space. The neural networks predicted Pt-Au alloys with 18–24 at.% Au, when deposited at specific conditions, to have a nanoindentation hardness up to 7.2 GPa. This high hardness value, comparable to some steels, represents a 3-fold improvement in hardness over “hard gold”, a commonly used electrical contact alloy, while maintaining requisite electrical conductivity. The neural network models provide an avenue to identify expected process windows capable of maximizing material performance.

Electrical contact materials↗

Physics constrained unsupervised deep learning for rapid, high resolution scanning coherent diffraction reconstruction

By circumventing the resolution limitations of optics, coherent diffractive imaging (CDI) and ptychography are making their way into scientific fields ranging from X-ray imaging to astronomy. Yet, the need for time consuming iterative phase recovery hampers real-time imaging. While supervised deep learning strategies have increased reconstruction speed, they sacrifice image quality. Furthermore, these methods’ demand for extensive labeled training data is experimentally burdensome. Here, we propose an unsupervised physics-informed neural network reconstruction method, PtychoPINN, that retains the factor of 100-to-1000 speedup of deep learning-based reconstruction while improving reconstruction quality by combining the diffraction forward map with real-space constraints from overlapping measurements. In particular, PtychoPINN gains a factor of 4 in linear resolution and an 8 dB improvement in PSNR while also accruing improvements in generalizability and robustness. This blend of performance and computational efficiency offers exciting prospects for high-resolution real-time imaging in high-throughput environments such as X-ray free electron lasers (XFELs) and diffraction-limited light sources.

97 MATHEMATICS AND COMPUTING↗

Visualizing Crystallization Dynamics and Transformation Pathways of Disordered Rocksalt Oxides During Thermally Activated Sol–Gel Synthesis

Sol–gel synthesis is a wet-chemical processing route for fabricating functional materials with control over composition and microstructure at relatively low temperatures compared to conventional solid-state synthesis. While sol–gel process initiates with intermixed molecular precursors, the early-stage nucleation pathways are insufficiently understood. Here, in this study, the chemical and structural transformation of ion disordered rocksalt (DRX) Li 1.2 Mn 0.4 Ti 0.4 O 2 (LMTO), a promising cathode material for lithium batteries, is studied by multiscale characterizations. In situ heating transmission electron microscopy (TEM) using a liquid cell visualizes and identifies crystallization pathways at the nanoscale. While some regions follow a classical multi-step transition through thermodynamically stable intermediates, others exhibit a kinetic shortcut via a localized amorphous matrix to directly form the DRX structure. Macroscale Fourier transform infrared spectroscopy corroborates the findings and reveals that transition metal ions are more strongly incorporated into the acetate-coordinated network than lithium. Although in situ heating TEM captures diverse local transformation pathways, in situ synchrotron X-ray diffraction indicates that the macroscopic transformation proceeds predominantly through spinel LMTO and lithium titanates toward DRX-LMTO. The findings uncover the spatiotemporal chemical and structural transformations in sol–gel derived DRX-LMTO materials, and call for fine-tuning of such sol–gel chemistries to manipulate the crystallization pathways and achieve target material homogeneity more efficiently.

cathode material↗

Ultrafast radiographic imaging and tracking: An overview of instruments, methods, data, and applications

Ultrafast radiographic imaging and tracking (U-RadIT) use state-of-the-art ionizing particle and light sources to experimentally study sub-nanosecond transients or dynamic processes in physics, chemistry, biology, geology, materials science and other fields. These processes are fundamental to modern technologies and applications, such as nuclear fusion energy, advanced manufacturing, communication, and green transportation, which often involve one mole or more atoms and elementary particles, and thus are challenging to compute by using the first principles of quantum physics or other forward models. One of the central problems in U-RadIT is to optimize information yield through, e.g. high-luminosity X-ray and particle sources, efficient imaging and tracking detectors, novel methods to collect data, and large-bandwidth online and offline data processing, regulated by the underlying physics, statistics, and computing power. We review and highlight recent progress in: (a.) Detectors such as high-speed complementary metal-oxide semiconductor (CMOS) cameras, hybrid pixelated array detectors integrated with Timepix4 and other application-specific integrated circuits (ASICs), and digital photon detectors; (b.) U-RadIT modalities such as dynamic phase contrast imaging, dynamic diffractive imaging, and four-dimensional (4D) particle tracking; (c.) U-RadIT data and algorithms such as neural networks and machine learning, and (d.) Applications in ultrafast dynamic material science using XFELs, synchrotrons and laser-driven sources. Hardware-centric approaches to U-RadIT optimization are constrained by detector material properties, low signal-to-noise ratio, high cost and long development cycles of critical hardware components such as ASICs. Interpretation of experimental data, including comparisons with forward models, is frequently hindered by sparse measurements, model and measurement uncertainties, and noise. Alternatively, U-RadIT make increasing use of data science and machine learning algorithms, including experimental implementations of compressed sensing. Machine learning and artificial intelligence approaches, refined by physics and materials information, may also contribute significantly to data interpretation, uncertainty quantification and U-RadIT optimization.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Nondestructive characterization of polycrystalline 3D microstructure with time-domain Brillouin scattering

In this paper, we employ gigahertz frequency ultrasonic waves to non-invasively image grain microstructure in polycrystalline ceria, an important electrolyte material for solid oxide fuel cells. This technique, termed time domain Brillouin scattering (TDBS), is a pump-probe technique that launches and monitors ultrasonics waves as they propagate from the surface into the depth of a sample. We demonstrate the capability of TDBS to image surface and subsurface grain structure by measuring mode and depth dependent velocity of the ultrasonic wave. A neural network tool is adopted to determine grain orientation from measured velocities. The TDBS results compare closely with electron backscatter diffraction (EBSD) measurements. This work expands the application of picosecond ultrasonics to three-dimensional imaging and offers a new route for the nondestructive microstructure characterization and in situ examination of polycrystalline materials.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

NASA Tech Briefs, November 2003

Topics covered include: Computer Program Recognizes Patterns in Time-Series Data; Program for User-Friendly Management of Input and Output Data Sets; Noncoherent Tracking of a Source of a Data-Modulated Signal; Software for Acquiring Image Data for PIV; Detecting Edges in Images by Use of Fuzzy Reasoning; A Timer for Synchronous Digital Systems; Prototype Parts of a Digital Beam-Forming Wide-Band Receiver; High-Voltage Droplet Dispenser; Network Extender for MIL-STD-1553 Bus; MMIC HEMT Power Amplifier for 140 to 170 GHz; Piezoelectric Diffraction-Based Optical Switches; Numerical Modeling of Nanoelectronic Devices; Organizing Diverse, Distributed Project Information; Eigensolver for a Sparse, Large Hermitian Matrix; Modified Polar-Format Software for Processing SAR Data; e-Stars Template Builder; Software for Acoustic Rendering; Functionally Graded Nanophase Beryllium/Carbon Composites; Thin Thermal-Insulation Blankets for Very High Temperatures; Prolonging Microgravity on Parabolic Airplane Flights; Device for Locking a Control Knob; Cable-Dispensing Cart; Foam Sensor Structures Would be Self-Deployable and Survive Hard Landings; Real-Gas Effects on Binary Mixing Layers; Earth-Space Link Attenuation Estimation via Ground Radar Kdp; Wedge Heat-Flux Indicators for Flash Thermography; Measuring Diffusion of Liquids by Common-Path Interferometry; Zero-Shear, Low-Disturbance Optical Delay Line; Whispering-Gallery Mode-Locked Lasers; Spatial Light Modulators as Optical Crossbar Switches; Update on EMD and Hilbert-Spectra Analysis of Time Series; Quad-Tree Visual-Calculus Analysis of Satellite Coverage; Dyakonov-Perel Effect on Spin Dephasing in n-Type GaAs; Update on Area Production in Mixing of Supercritical Fluids; and Quasi-Sun-Pointing of Spacecraft Using Radiation Pressure.

Source record↗

Facile one-pot synthesis and characterization of nickel supported on hierarchically porous carbon

Highlights: • Novel, inexpensive, one-pot, synthesis method for Ni on hierarchically porous carbon. • Disappearance of surfactant mesopores seen with incorporation of nickel. • Distribution of Ni nanoparticles on the hierarchically porous carbon support was studied by SEM. • Nickel nanoparticles were dispersed on macropore walls and not within carbon. - Abstract: Described is a novel, facile route for the synthesis of nickel supported on hierarchically porous carbon (Ni/HPC) using a one-pot co-gelation sol–gel method. Ni/HPC with varying nickel loadings (0.5, 1, 2.5 and 5 wt% Ni) were synthesized and the materials characterized by nitrogen physisorption, X-ray diffraction (XRD), scanning electron microscopy (SEM), and Fourier transform infrared (FTIR) and Raman spectroscopies. The results show a three-dimensional network of disordered carbon with fine nickel nanoparticles of sizes ranging from 8 nm to 13 nm at 0.5 wt% Ni loading which gradually increased with increase in the Ni loading. The carbon structure was retained at the macropore level, but not at the mesoscale where the ordered mesopores were lost on nickel addition. The nickel nanoparticles were observed to grow on the surface of the ligaments. This may make them particularly suitable for low pressure Ni-catalyzed organic transformations e.g., hydrogenations, C–C coupling, C-heteroatom coupling, etc.

36 MATERIALS SCIENCE↗

Different dimensional coordination polymers with 4,4'-oxybis(benzoate): Syntheses, structures and properties

Five transition-metal coordination polymers, namely, [Zn{sub 7}Cl{sub 6}(oba){sub 4}]{sub n} (1), [Cd{sub 7}Cl{sub 6}(oba){sub 4}]{sub n} (2), [Zn(oba)(H{sub 2}O)]{sub n} (3), [Ag{sub 2}(oba)]{sub n} (4) and [Co(oba)(H{sub 2}O){sub 2}]{sub n} (5) (H{sub 2}oba=4,4′-oxydibenzoic acid), have been achieved under hydrothermal conditions and structurally characterized by IR, elemental analyses, X-ray single-crystal diffraction and TGA. The X-ray single-crystal diffraction reveals that compounds 1 and 2 are isomorphism, featuring pillared-layer 3D motifs, in which the 2D inorganic layers (Zn{sub 6}Cl{sub 7}){sub n} (or (Cd{sub 6}Cl{sub 7}){sub n}) are connected by oba{sup 2−} pillars. Compound 3 exhibits 1D stair-like chain and extends to a 3D network by two different interchain O–H–O hydrogen bonding interactions while compound 4 features wave chains and stretches to 2D layer by interchain Ag–O weak contacts. Compound 5 shows 2D network in which Co-chains are pillared by oba{sup 2−} ligand and then forms a 3D network by four different O–H–O hydrogen bonding interactions. Furthermore, 1–4 exhibit luminescent properties at a solid state and 5 shows antiferromagnetic behavior. - Graphical abstract: Five new transition-metal coordination complexes 1–5 have been synthesized and characterized by single-crystal X-ray diffractions, IR spectra, elemental analyses, thermogravimetric analyses (TGA), photoluminescent spectra and magnetic measurement. - Highlights: • Compound 1 exhibits a pillared-layer 3D network. • The photoluminescent properties of 1–4 have been measured. • Compound 5 exhibits antiferromagnetic behavior.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Changes in mobility of plastic crystal ethanol during its transformation into the monoclinic crystal state

Transformation of deuterated ethanol from the plastic crystal phase into the monoclinic one is investigated by means of a singular setup combining simultaneously dielectric spectroscopy with neutron diffraction. We postulate that a dynamic transition from plastic crystal to supercooled liquid-like configuration through a deep reorganization of the hydrogen-bonding network must take place as a previous step of the crystallization process. Once these precursor regions are formed, subsequent crystalline nucleation and growth develop with time.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Selective substitution of Cr in CaFe{sub 4}As{sub 3} and its effect on the spin density wave.

Single crystals of CaCr{sub 0.84}Fe{sub 3.16}As{sub 3}, a Cr substituted analog of CaFe{sub 4}As{sub 3}, were grown from Sn flux and characterized with single crystal neutron diffraction. CaCr{sub 0.84}Fe{sub 3.16}As{sub 3} crystallizes in the orthorhombic space group Pnma with a three-dimensional framework, where Fe, Cr, and As form a covalent channel-like network with Ca{sup 2+} cations residing in the channels. CaCr{sub 0.84}Fe{sub 3.16}As{sub 3} has a unit cell of a = 12.057(4) {angstrom}, b = 3.7374(13) {angstrom}, and c = 11.694(3) {angstrom}, as determined by room temperature single crystal neutron diffraction (R{sub 1} = 0.0747, wR{sub 2} = 0.1825). Structural data was also collected at 10 K. The single crystal neutron data showed that Cr selectively occupies a particular metal site, Fe(4). The antiferromagnetic transition associated with spin density wave (SDW) in the parent compound is preserved and shifts from 96 to 103 K with the selective Cr doping. Moessbauer, magnetic, and electrical resistivity measurements are reported.

36 MATERIALS SCIENCE↗

Organic carboxylate anions effect on the structures of a series of Mn(II) complexes based on 2-phenylimidazo[4,5-f]1,10-phenanthroline ligand

In our efforts to tune the structures of Mn(II) complexes by selection of organic carboxylic acid ligands, six new complexes [Mn(PIP){sub 2}Cl{sub 2}] (1), [Mn(PIP){sub 2}(4,4'-bpdc)(H{sub 2}O)].2H{sub 2}O (2), [Mn(PIP){sub 2}(1,4-bdc)] (3), [Mn(PIP)(1,3-bdc)] (4), [Mn(PIP){sub 2}(2,6-napdc)].H{sub 2}O (5), and [Mn(PIP)(1,4-napdc)].H{sub 2}O (6) were obtained, where PIP=2-phenylimidazo[4,5-f]1,10-phenanthroline, 4,4'-H{sub 2}bpdc=biphenyl-4,4'-dicarboxylic acid, 1,4-H{sub 2}bdc=benzene-1,4-dicarboxylic acid, 1,3-H{sub 2}bdc=benzene-1,3-dicarboxylic acid, 2,6-H{sub 2}napdc=2,6-naphthalenedicarboxylic acid, 1,4-H{sub 2}napdc=1,4-naphthalenedicarboxylic acid. All complexes have been structurally characterized by IR, elemental analyses, and single crystal X-ray diffraction. Structural analyses show that complexes 1 and 2 possess mononuclear structures, complexes 3, 4, and 5 feature chain structures, and complex 6 exhibits a 2D (4,4) network. The structural difference of 1-6 indicates that organic carboxylate anions play important roles in the formation of such coordination architectures. Furthermore, the thermal properties of complexes 1-6 and the magnetic property of 4 have been investigated. - Graphical Abstract: Through selecting organic carboxylate anions, six Mn(II) complexes have been synthesized under hydrothermal conditions and characterized by single crystal X-ray diffraction.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Three Pb{sup II} coordination polymers based on 2-(pyridin-2-yl)-1H-imidazole-4,5-dicarboxylic acid: Syntheses, crystal structures, and fluorescent properties

Three lead coordination polymers, [PbCl(C{sub 10}H{sub 6}N{sub 3}O{sub 4})(H{sub 2}O)·H{sub 2}O]{sub n} (1), [Pb(C{sub 10}H{sub 6}N{sub 3}O{sub 4}){sub 2}(H{sub 2}O)]{sub n} (2) and [Pb{sub 3}(C{sub 10}H{sub 5}N{sub 3}O{sub 4}){sub 3}]{sub n} (3) (C{sub 10}H{sub 7}N{sub 3}O{sub 4}=2-(pyridin-2-yl)-1H-imidazole-4,5-dicarboxylic acid), have been hydrothermally synthesized and structurally characterized by single-crystal X-ray diffraction. In 1, Cl anions connected neighboring wave-like 2D layers, which are constructed with left- and right-handed helical chains, into a 3D network structure with a (6{sup 3})(6{sup 5}·8) topology. In 2, Pb cations are linked into a 3D 6{sup 6} network with left- and right-handed helixes by μ{sub 2}-bridging C{sub 10}H{sub 6}N{sub 3}O{sub 4}{sup −} ligands. In 3, C{sub 10}H{sub 5}N{sub 3}O{sub 4}{sup 2−} ligands link Pb{sub 6}O{sub 12} clusters into a 3D (4{sup 12}·6{sup 3}) network. Their fluorescent properties were also investigated. - Graphical abstract: Three 3D lead compounds based on 2-(pyridin-2-yl)-1H-imidazole-4,5-dicarboxylic acid have been hydrothermally synthesized. Four new coordination modes of the organic ligand are first reported. Display Omitted - Highlights: • Three new Pb(II) complexes have been synthesized and characterized. • Left- and right-handed helical chains can be found in the 3D networks of 1 and 2. • Pb{sub 6}O{sub 12} clusters are connected into (4{sup 12}·6{sup 3}) network in 3.

36 MATERIALS SCIENCE↗