Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “diffusion length”

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 667 records · Page 37

Flux Equations for Osmotically Moderated Sorption–Diffusion Transport in Rigid Microporous Membranes

The development of non-aqueous reverse osmosis separations via the use of rigid microporous membrane materials suggests that flux equations originally developed for swellable polymeric membranes should be revisited. This paper demonstrates that the gradient of fractional occupancy of penetrant molecules within the micropores of the membrane is the driving force for permeation without requiring assumptions about pressure within the membrane. Flux equations are derived using both Fickian and Maxwell-Stefan approaches, and different behavior in the permeate flux versus upstream hydraulic pressure relationship is shown to arise as a result of differences in the loading dependence of the single component Maxwell-Stefan diffusivity. Molecular modeling results available in the literature and experimental data obtained from carbon molecular sieve (CMS) membranes showcase that these loading-dependent changes in the Maxwell-Stefan diffusivity are possible. This loading dependence is separated into three regimes: so-called “weak confinement” diffusion and “strong confinement” diffusion, both of which have been discussed at length in the literature, and a new “hybrid confinement” diffusion, which is introduced here. Furthermore, the separation mechanism of solvent molecules through a rigid bimodally microporous membrane is studied using xylene molecules passing through CMS membranes fabricated under different conditions as examples. Altogether, this study provides fundamental insight and guidance into osmotically-moderated sorption-diffusion transport of solvent molecules through rigid microporous membranes.

36 MATERIALS SCIENCE↗

Kinetics of oxygen interstitial injection and lattice exchange in rutile TiO{sub 2}

The existence of a facile surface pathway for generation of O interstitials (O{sub i}) in rutile that can facilitate annihilation of O undesirable vacancies has been demonstrated recently. Through isotopic self-diffusion experiments, the present work determines a value of approximately 1.8 eV for the activation energy of O{sub i} injection from TiO{sub 2} (110). The mean path length for O{sub i} diffusion decreases by nearly an order of magnitude upon adsorption of 0.1 monolayer of sulfur. Sulfur apparently inhibits the surface annihilation rate of Ti interstitials, lowering their bulk concentration and the corresponding catalytic effect they seem to exert upon O{sub i} exchange with the lattice.

36 MATERIALS SCIENCE↗

Effect of inflow density on ion diffusion region of magnetic reconnection: Particle-in-cell simulations

We perform a systematic study of the effect of inflow density on reconnection diffusion regions using a 2.5-D particle-in-cell code. The diffusion region structures are analyzed at times when all simulations have reconnected the same amount of magnetic flux. We find that reducing the inflow density from 1 to 1/100th of the current sheet density dramatically increases the diffusion region physical size and the reconnection rate. The width of the diffusion region scales with the upstream ion inertial length systematically. Consistent with the presence of counter-streaming inflowing ion beams near the x-line, the ion meandering width in the diffusion region also scales with the ion inertial length. The aspect ratio of the ion diffusion region remains a constant, independent of the inflow density. The quadrupole Hall magnetic field is reduced. The upstream magnetic field deviates from its asymptotic value by {approx}50% at the lowest simulated inflow density. The downstream ion outflow velocity scales linearly with the upstream Alfven speed with a multiplication factor {approx}0.4 < 1. When applied to magnetic reconnection in the Earth's magnetotail, this factor of 0.4 is a possible explanation as to why bulk flow velocities in the magnetotail are typically on the order of 500 km/s, while the Alfven speeds of inflowing plasmas can exceed 2000 km/s.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Phase transitions in a reaction-diffusion model on a line with boundaries

A one-dimensional model on a line of length L is investigated, which involves particle diffusion as well as single particle annihilation. There are also creation and annihilation at the boundaries. The static and dynamical behaviors of the system are studied. It is seen that the system could exhibit a dynamical phase transition. For small drift velocities, the relaxation time does not depend on the absorption rates at the boundaries. This is the fast phase. For large velocities, the smaller of the absorption rates at boundaries enter the relaxation rate and makes it longer. This is the slow phase. Finally, the effect of a random particle creation in the bulk is also investigated.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

The importance of parameter variances, correlations lengths, and cross-correlations in reactive transport models: key considerations for assessing the need for microscale information

A process-oriented modeling approach is implemented to examine the importance of parameter variances, correlation lengths, and especially cross-correlations in contaminant transport predictions over large scales. It is shown that the most important consideration is the correlation between flow rates and retardation processes (e.g., sorption, matrix diffusion) in the system. lf flow rates are negatively correlated with retardation factors in systems containing multiple flow pathways, then characterizing these negative correlation(s) may have more impact on reactive transport modeling than microscale information. Such negative correlations are expected in porous-media systems where permeability is negatively correlated with clay content and rock alteration (which are usually associated with increased sorption). Likewise, negative correlations are expected in fractured rocks where permeability is positively correlated with fracture apertures, which in turn are negatively correlated with sorption and matrix diffusion. Parameter variances and correlation lengths are also shown to have important effects on reactive transport predictions, but they are less important than parameter cross-correlations. Microscale information pertaining to contaminant transport has become more readily available as characterization methods and spectroscopic instrumentation have achieved lower detection limits, greater resolution, and better precision. Obtaining detailed mechanistic insights into contaminant-rock-water interactions is becoming a routine practice in characterizing reactive transport processes in groundwater systems (almost necessary for high-profile publications). Unfortunately, a quantitative link between microscale information and flow and transport parameter distributions or cross-correlations has not yet been established. One reason for this is that quantitative microscale information is difficult to obtain in complex, heterogeneous systems. So simple systems that lack the complexity and heterogeneity of real aquifer materials are often studied. Another is that instrumentation used to obtain microscale information often probes only one variable or family of variables at a time, so linkages to other variables must be inferred by indirect means from other lines of evidence. Despite these limitations, microscale information can be useful in the development and validation of reactive transport models. For example, knowledge of mineral phases that have strong affinities for contaminants can help in the development of cross-correlations between flow and sorption parameters via characterization of permeability and mineral distributions in aquifers. Likewise, microscale information on pore structures in low-permeability zones and contaminant penetration distances into these zones from higher-permeability zones (e.g., fractures) can provide valuable constraints on the representation of diffusive mass transfer processes between flowing porosity and secondary porosity. The prioritization of obtaining microscale information in any groundwater system can be informed by modeling exercises such as those conducted for this study.

54 ENVIRONMENTAL SCIENCES↗

Using shaped pulses to probe energy deposition during laser-induced damage of SiO2 surfaces

Laser-induced damage initiation in silica has been shown to follow a power-law behavior with respect to pulse-length. Models based on thermal diffusion physics can successfully predict this scaling and the effect of pulse shape for pulses between about 3ns and 10ns. In this work we use sophisticated new measurement techniques and novel pulse shape experiments to test the limits of this scaling. We show that simple pulse length scaling fails for pulses below about 3ns. Furthermore, double pulse initiation experiments suggest that energy absorbed by the first pulse is lost on time scales much shorter than would be predicted for thermal diffusion. This time scale for energy loss can be strongly modulated by maintaining a small but non-zero intensity between the pulses. By producing damage with various pulse shapes and pulse trains it is demonstrated that the properties of any hypothetical thermal absorber become highly constrained.

42 ENGINEERING↗

Diagnosing isopycnal diffusivity in an eddying, idealized midlatitude ocean basin via Lagrangian, in Situ, Global, High-Performance Particle Tracking (LIGHT)

Isopycnal diffusivity due to stirring by mesoscale eddies in an idealized, wind-forced, eddying, midlatitude ocean basin is computed using Lagrangian, in Situ, Global, High-Performance Particle Tracking (LIGHT). Simulation is performed via LIGHT within the Model for Prediction across Scales Ocean (MPAS-O). Simulations are performed at 4-, 8-, 16-, and 32-km resolution, where the first Rossby radius of deformation (RRD) is approximately 30 km. Scalar and tensor diffusivities are estimated at each resolution based on 30 ensemble members using particle cluster statistics. Each ensemble member is composed of 303 665 particles distributed across five potential density surfaces. Diffusivity dependence upon model resolution, velocity spatial scale, and buoyancy surface is quantified and compared with mixing length theory. The spatial structure of diffusivity ranges over approximately two orders of magnitude with values of O(10 5 ) m 2 s –1 in the region of western boundary current separation to O(10 3 ) m 2 s –1 in the eastern region of the basin. Dominant mixing occurs at scales twice the size of the first RRD. Model resolution at scales finer than the RRD is necessary to obtain sufficient model fidelity at scales between one and four RRD to accurately represent mixing. Mixing length scaling with eddy kinetic energy and the Lagrangian time scale yield mixing efficiencies that typically range between 0.4 and 0.8. In conclusion, a reduced mixing length in the eastern region of the domain relative to the west suggests there are different mixing regimes outside the baroclinic jet region.

42 ENGINEERING↗

Characterization of diffusion flames for synthesis of single-walled carbon nanotubes

Recent studies have shown that Fe/Si/O catalysts on the fuel side of an oxygen-enriched inverse diffusion flame produce micron-length single-walled carbon nanotubes at rapid rates (>100{mu}m/s). Despite the favorable catalyst/flame interaction for nanotube nucleation and growth, the catalyst lifetimes are only a few milliseconds. To increase catalyst lifetime and hence, carbon nanotube length, it is necessary to know how the local environment changes as the catalyst moves through the flame. A 2-D computational fluid dynamics model with detailed chemistry is employed to investigate the nature of the flame environment along various catalyst trajectories. The results indicate that temperature and species concentrations do not change significantly along individual catalyst trajectories, although not all trajectories experience the same environment due to the steep gradients in the radial direction. On the other hand, analysis of catalyst particle composition before and after nanotube growth shows that catalyst oxygen content decreases significantly during nanotube growth. This change in catalyst composition could affect the relative rates of carbon supply versus removal from the catalyst surface, such that carbon encapsulation and thus poisoning of the catalyst is favored after sufficient time. The results of this work indicate that catalyst deactivation, not a changing catalyst environment, is responsible for rapid encapsulation of the catalyst by amorphous carbon and thus, the short catalyst lifetimes observed in oxygen-enriched diffusion flames. (author)

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

TURBULENCE IN THE SOLAR WIND MEASURED WITH COMET TAIL TEST PARTICLES

By analyzing the motions of test particles observed remotely in the tail of Comet Encke, we demonstrate that the solar wind undergoes turbulent processing enroute from the Sun to the Earth and that the kinetic energy entrained in the large-scale turbulence is sufficient to explain the well-known anomalous heating of the solar wind. Using the heliospheric imaging (HI-1) camera on board NASA's STEREO-A spacecraft, we have observed an ensemble of compact features in the comet tail as they became entrained in the solar wind near 0.4 AU. We find that the features are useful as test particles, via mean-motion analysis and a forward model of pickup dynamics. Using population analysis of the ensemble's relative motion, we find a regime of random-walk diffusion in the solar wind, followed, on larger scales, by a surprising regime of semiconfinement that we attribute to turbulent eddies in the solar wind. The entrained kinetic energy of the turbulent motions represents a sufficient energy reservoir to heat the solar wind to observed temperatures at 1 AU. We determine the Lagrangian-frame diffusion coefficient in the diffusive regime, derive upper limits for the small scale coherence length of solar wind turbulence, compare our results to existing Eulerian-frame measurements, and compare the turbulent velocity with the size of the observed eddies extrapolated to 1 AU. We conclude that the slow solar wind is fully mixed by turbulence on scales corresponding to a 1–2 hr crossing time at Earth; and that solar wind variability on timescales shorter than 1–2 hr is therefore dominated by turbulent processing rather than by direct solar effects.

79 ASTRONOMY AND ASTROPHYSICS↗

Incorporation of lead and titanium into a porous alumina membrane during Pb(Zr{sub 0.52}Ti{sub 0.48})O{sub 3} nanotube synthesis

We report the identification and characteristics of lead aluminate (Pb{sub 2}Al{sub 2}O{sub 5} or PbAl{sub 12}O{sub 19})/aluminum titanate (Al{sub 2}TiO{sub 5}) formed as a result of the interaction between a Pb(Zr{sub 0.52}Ti{sub 0.48})O{sub 3} (PZT) precursor and a porous alumina membrane (PAM) during PZT nanotube synthesis. The surface morphology and composition of PZT nanotubes were examined as a function of the etching time and annealing temperature, and the diffusion of elements over nanometer length scales during PZT nanotube formation was investigated based on analysis using field emission transmission electron microscopy, scanning transmission electron microscopy, and energy dispersive x-ray microscopy. As the annealing temperature increased, the volume of lead aluminate/aluminum titanate increased, suggesting that considerable diffusion of Pb and Ti occurred and, in turn, dramatically affected the nanotube characteristics. The diffusion of Pb and Ti was clearly evident, resulting in the formation of a distinct lead aluminate/aluminum titanate at the PZT nanotube/PAM interface.

77 NANOSCIENCE AND NANOTECHNOLOGY↗

PLASMA DYNAMICS AT THE PROMINENCE-CORONA INTERFACE

The interface between the cool and dense plasma typical of a prominence and its tenuous and hot surrounding coronal plasma is poorly understood. We study the plasma dynamics at this interface using a three-dimensional particle-in-cell code, which enables us to carry out simulations on spatial and temporal scales of the order of the Debye length and plasma period, respectively. The results show that anomalous Bohm diffusion across magnetic field lines occurs at the interface, leading to mixing of the two plasmas. It is also shown that collisions with neutral hydrogen within the prominence plasma are of little importance for the plasma dynamics in the prominence-corona transition region. In particular, the temperature of the prominence plasma crossing the interface into the corona can become anisotropic due to preferential heating by instabilities originating from unstable velocity distributions. Our results pertain to spatial scales significantly smaller than scales commonly used in magnetohydrodynamic simulations, and they shed light on processes that are very likely to be present at the interface.

79 ASTRONOMY AND ASTROPHYSICS↗

Manipulating surface diffusion and elastic interactions to obtain quantum dot multilayer arrangements over different length scales

An innovative multilayer growth of InAs quantum dots on GaAs(100) is demonstrated to lead to self-aggregation of correlated quantum dot chains over mesoscopic distances. The fundamental idea is that at critical growth conditions is possible to drive the dot nucleation only at precise locations corresponding to the local minima of the Indium chemical potential. Differently from the known dot multilayers, where nucleation of new dots on top of the buried ones is driven by the surface strain originating from the dots below, here the spatial correlations and nucleation of additional dots are mostly dictated by a self-engineering of the surface occurring during the growth, close to the critical conditions for dot formation under the fixed oblique direction of the incoming As flux, that drives the In surface diffusion.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

A multifidelity method for a nonlocal diffusion model

Nonlocal models feature a finite length scale, referred to as the horizon, such that points separated by a distance smaller than the horizon interact with each other. Such models have proven to be useful in a variety of settings. However, due to the reduced sparsity of discretizations, they are also generally computationally more expensive compared to their local differential equation counterparts. In this work, we introduce a multifidelity Monte Carlo method that combines the high-fidelity nonlocal model of interest with surrogate models that use coarser grids and/or smaller horizons and thus have lower fidelities and lower costs. Using the multifidelity method, the overall computational cost of uncertainty quantification is reduced without compromising accuracy. It is shown for a one-dimensional nonlocal diffusion example that speedups of up to two orders of magnitude can be achieved using the multifidelity method to estimate the expectation of an output of interest.

97 MATHEMATICS AND COMPUTING↗

Hydrogen Bonding Stiffens Peptide Amphiphile Supramolecular Filaments by Aza-Glycine Residues

Peptide amphiphiles (PAs) are a class of molecules comprised of short amino acid sequences conjugated to hydrophobic moieties that may exhibit self-assembly in water into supramolecular structures. Here, we investigate here how mechanical properties of hydrogels formed by PA supramolecular nanofibers are affected by hydrogen bond densities within their internal structure by substituting glycine for aza-glycine (azaG) residues. We found that increasing the number of PA molecules that contain azaG up to 5 mol% in PA supramolecular nanofibers increases their persistence length fivefold and decreases their diffusion coefficients as measured by fluorescence recovery after photobleaching. When these PAs are used to create hydrogels, their bulk storage modulus (G') was found to increase as azaG PA content in the supramolecular assemblies increases up to a value of 10 mol% and beyond this value a decrease was observed, likely due to diminished levels of nanofiber entanglement in the hydrogels as a direct result of increased supramolecular rigidity. Interestingly, we found that the bioactivity of the scaffolds toward dopaminergic neurons derived from induced pluripotent stem cells can be enhanced directly by persistence length independently of storage modulus. We hypothesize that this is due to interactions between the cells and the extracellular environment across different size scales: from filopodia adhering to individual nanofiber bundles to cell adhesion sites that interact with the hydrogel as a bulk substrate. Fine tuning of hydrogen bond density in self-assembling peptide biomaterials such as PAs provides an approach to control nanoscale stiffness as part of an overall strategy to optimize bioactivity in these supramolecular systems. supramolecular biomaterials.

36 MATERIALS SCIENCE↗

Are oxygen limitations under recognized regulators of organic carbon turnover in upland soils?

Understanding the processes controlling organic matter (OM) stocks in upland soils, and the ability to management them, is crucial for maintaining soil fertility and carbon (C) storage as well as projecting change with time. OM inputs are balanced by the mineralization (oxidation) rate, with the difference determining whether the system is aggrading, degrading or at equilibrium with reference to its C storage. In upland soils, it is well recognized that the rate and extent of OM mineralization is affected by climatic factors (particularly temperature and rainfall) in combination with OM chemistry, mineral–organic associations, and physical protection. Here we examine evidence for the existence of persistent anaerobic microsites in upland soils and their effect on microbially mediated OM mineralization rates. Further, we corroborate long-standing assumptions that residence times of OM tend to be greater in soil domains with limited oxygen supply (aggregates or peds). Moreover, the particularly long residence times of reduced organic compounds (e.g., aliphatics) are consistent with thermodynamic constraints on their oxidation under anaerobic conditions. Incorporating (i) pore length and connectivity governing oxygen diffusion rates (and thus oxygen supply) with (ii) ‘hot spots’ of microbial OM decomposition (and thus oxygen consumption), and (iii) kinetic and thermodynamic constraints on OM metabolism under anaerobic conditions will thus improve conceptual and numerical models of C cycling in upland soils. We conclude that constraints on microbial metabolism induced by oxygen limitations act as a largely unrecognized and greatly underestimated control on overall rates of C oxidation in upland soils.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Activated penetrant dynamics in glass forming liquids: size effects, decoupling, slaving, collective elasticity and correlation with matrix compressibility

In this work, we employ the microscopic self-consistent cooperative hopping theory of penetrant activated dynamics in glass forming viscous liquids and colloidal suspensions to address new questions over a wide range of high matrix packing fractions and penetrant-to-matrix particle size ratios. The focus is on the mean activated relaxation time of smaller tracers in a hard sphere fluid of larger particle matrices. This quantity also determines the penetrant diffusion constant and connects directly with the structural relaxation time probed in an incoherent dynamic structure factor measurement. The timescale of the non-activated fast dissipative process is also studied and is predicted to follow power laws with the contact value of the penetrant–matrix pair correlation function and the penetrant–matrix size ratio. For long time penetrant relaxation, in the relatively lower packing fraction metastable regime the local cage barriers are dominant and matrix collective elasticity effects unimportant. As packing fraction and/or penetrant size grows, much higher barriers emerge and the collective elasticity associated with the correlated matrix dynamic displacement that facilitates penetrant hopping becomes important. This results in a non-monotonic variation with packing fraction of the degree of decoupling between the matrix and penetrant alpha relaxation times. The conditions required for penetrant hopping to become slaved to the matrix alpha process are determined, which depend mainly on the penetrant to matrix particle size ratio. By analyzing the absolute and relative importance of the cage and elastic barriers we establish a mechanistic understanding of the origin of the predicted exponential growth of the penetrant hopping time with size ratio predicted at very high packing fractions. A dynamics-thermodynamics power law connection between the penetrant activation barrier and the matrix dimensionless compressibility is established as a prediction of theory, with different scaling exponents depending on whether matrix collective elasticity effects are important. Quantitative comparisons with simulations of the penetrant relaxation time, diffusion constant, and transient localization length of tracers in dense colloidal suspensions and cold viscous liquids reveal good agreements. Multiple new predictions are made that are testable via future experiments and simulations. Extension of the theoretical approach to more complex systems of high experimental interest (nonspherical molecules, semiflexible polymers, crosslinked networks) interacting via variable hard or soft repulsions and/or short range attractions is possible, including under external deformation.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Dynamics of the plumes produced by ultrafast laser ablation of metals

We have analyzed ultrafast laser ablation of a metallic target (Nickel) in high vacuum addressing both expansion dynamics of the various plume components (ionic and nanoparticle) and basic properties of the ultrafast laser ablation process. While the ion temporal profile and ion angular distribution were analyzed by means of Langmuir ion probe technique, the angular distribution of the nanoparticulate component was characterized by measuring the thickness map of deposition on a transparent substrate. The amount of ablated material per pulse was found by applying scanning white light interferometry to craters produced on a stationary target. We have also compared the angular distribution of both the ionic and nanoparticle components with the Anisimov model. While the agreement for the ion angular distribution is very good at any laser fluence (from ablation threshold up to {approx_equal}1 J/cm{sup 2}), some discrepancies of nanoparticle plume angular distribution at fluencies above {approx_equal}0.4 J/cm{sup 2} are interpreted in terms of the influence of the pressure exerted by the nascent atomic plasma plume on the initial hydrodynamic evolution of the nanoparticle component. Finally, analyses of the fluence threshold and maximum ablation depth were also carried out, and compared to predictions of theoretical models. Our results indicate that the absorbed energy is spread over a length comparable with the electron diffusion depth L{sub c} ({approx_equal}30 nm) of Ni on the timescale of electron-phonon equilibration and that a logarithmic dependence is well-suited for the description of the variation in the ablation depth on laser fluence in the investigated range.

36 MATERIALS SCIENCE↗

Constraining primordial magnetic fields with distortions of the black-body spectrum of the cosmic microwave background: pre- and post-decoupling contributions

Primordial magnetic fields that exist before the photon-baryon decoupling epoch are damped on length scales below the photon diffusion and free-streaming scales. The energy injected into the plasma by dissipation of magnetosonic and Alfv and apos;en waves heats photons, creating a y-type distortion of the black-body spectrum of the cosmic microwave background. This y-type distortion is converted into a μ-type distortion when elastic Compton scattering is efficient. Therefore, we can use observational limits on y- and μ-type distortions to constrain properties of magnetic fields in the early universe. Assuming a Gaussian, random, and non-helical field, we calculate μ and y as a function of the present-day strength of the field, B{sub 0}, smoothed over a certain Gaussian width, k{sub c}{sup −1}, as well as of the spectral index of the power spectrum of fields, n{sub B}, defined by P{sub B}(k)∝k{sup n{sub B}}. For a nearly scale-invariant spectrum with n{sub B} = −2.9 and a Gaussian smoothing width of k{sub c}{sup −1} = 1Mpc, the existing COBE/FIRAS limit on μ yields B{sub 0} < 40 nG, whereas the projected PIXIE limit on μ would yield B{sub 0} < 0.8 nG. For non-scale-invariant spectra, constraints can be stronger. For example, for B{sub 0} = 1 nG with k{sub c}{sup −1} = 1Mpc, the COBE/FIRAS limit on μ excludes a wide range of spectral indices given by n{sub B} > −2.6. After decoupling, energy dissipation is due to ambipolar diffusion and decaying MHD turbulence, creating a y-type distortion. The distortion is completely dominated by decaying MHD turbulence, and is of order y ≈ 10{sup −7} for a few nG field smoothed over the damping scale at the decoupling epoch, k{sub d,} {sub dec} ≈ 290(B{sub 0}/1nG){sup −1}Mpc{sup −1}. The projected PIXIE limit on y would exclude B{sub 0} > 1.0 and 0.6 nG for n{sub B} = −2.9 and -2.3, respectively, and B{sub 0} > 0.6 nG for n{sub B} ≥ 2. Finally, we find that the current limits on the optical depth to Thomson scattering restrict the predicted y-type distortion to be y∼<10{sup −8}.

79 ASTRONOMY AND ASTROPHYSICS↗