scispace - formally typeset
Search or ask a question

Showing papers in "Philosophical Magazine Part B in 1992"


Journal ArticleDOI
TL;DR: In this article, the authors reviewed the experience with the growth of a wide variety of intermetallic compounds from molten fluxes and discussed common problems associated with this method of sample growth.
Abstract: Our experience with the growth of a wide variety of intermetallic compounds from molten fluxes is reviewed. Common problems associated with this method of sample growth are discussed, as are problems and advantages of particular fluxes.

619 citations


Journal ArticleDOI
L. B. Schein1
TL;DR: In this paper, the current status of these and other theories in explaining the data is reviewed and a review of the current state of the polaron theory and disorder theory is given.
Abstract: It is well known that charge carrier mobilities μ in molecularly doped polymers, determined by transient photoconductivity experiments, depend on the temperature T, the mean distance p between dopant molecules and the electric field E. Recently introduced deconvolution procedures have made possible the systematic analysis of the dependence of μ on T, p and E. Two theories which have been proposed to account for the data are the polaron theory and the disorder theory of Bassler and co-workers. The current status of these and other theories in explaining the data is reviewed.

184 citations


Journal ArticleDOI
TL;DR: In this paper, the electronic structure of amorphous Si-C (a-Si1-xCx) and hydrogenated Si-H alloys has been calculated as a function of composition, chemical ordering, C coordination and H configuration using the tight binding method.
Abstract: The electronic structure of amorphous Si—C (a-Si1-xCx) and hydrogenated amorphous (a-Si1-x Cx: H) alloys has been calculated as a function of composition, chemical ordering, C coordination and H configuration using the tight-binding method. The maximum in the optical bandgap observed in a-Si1-x Cx: H at around x = 0·6 is not due to chemical ordering but to a change in band edge character, from Si-Si bond states for x 0·6. The gap is controlled by the degree of clustering of sp2 sites for x > 0·6. Interpretation of the optical gap, photoemission and X-ray emission data suggests that a moderate degree of chemical ordering exists in a-Si1-xCx: H and is higher in a-Si1-xCx. Hydrogenation widens the gap over the entire composition range, in Si-rich alloys by a recession of the valence band and in C-rich alloys by reducing the cluster sizes. The position of defect states due to Si and C dangling bonds is calculated and combined into a band model for the alloys. Experimen...

165 citations


Journal ArticleDOI
TL;DR: In this paper, it was shown that crystalline Si (c-Si) nucleates within the Al layers and penetrates the Al as the c-Si grows, and the overall activation energy of the reaction, determined by calorimetry, is 1·2 ± 1 eV.
Abstract: The crystallization of amorphous Si (a-Si) in Al/a-Si multilayer thin films has been investigated by ex situ and in situ transmission electron microscopy (TEM), X-ray diffraction and calorimetry. The a-Si crystallizes at about 200°C, a significantly lower temperature than for the pure elemental state, with a heat of crystallization of about 12 kJ (mol Si)−1. We show that crystalline Si (c-Si) nucleates within the Al layers and penetrates the Al as the c-Si grows. The speed of the growth of c-Si observed by in situ TEM was a few angstroms per second at 220°C. Al grains are separated and the layered structure is destroyed, while the Al(111) film texture is enhanced. The overall activation energy of the reaction, determined by calorimetry, is 1·2 ± 01 eV. We propose a model in which diffusion of Si through the Al grains and rearrangement of the Al grains occur simultaneously.

131 citations


Journal ArticleDOI
TL;DR: In this paper, a theoretical framework for the description and calculation of electronic transport characteristics in granular metals, which are composite materials consisting of a random mixture of nanometer-sized metal and insulator grains, is presented.
Abstract: A brief review is given of the theoretical framework for the description and calculation of electronic transport characteristics in granular metals, which are composite materials consisting of a random mixture of nanometer-sized metal and insulator grains. In the metal-rich regime, electrical conduction is by electron percolation through connected metallic networks. The formulation of an effective-medium theory is described for the calculation of this percolative aspect of the electrical transport. Even in the metallic regime, however, metallic conductivity behaviour is violated at low temperatures due to the electron localization effects caused by random scattering. A quantum percolation model is used to take into account the quantum-wave interference effects and to simulate both the temperature and magnetic-field dependence of the low-temperature electrical conductivity in thin granular metal films. The same theoretical model is also used to explore the mesoscopic transport behaviour of small g...

118 citations


Journal ArticleDOI
TL;DR: In this paper, the Debye-Waller parameter of nanocrystalline Pd was found to be due to increased static displacement of atoms from their equilibrium sites compared with the static displacements in coarse-grained material.
Abstract: Quantitative X-ray diffraction measurements made over a temperature range 16–300K on a Pd sample of 8·3nm median grain size were compared with data acquired from a coarse-grained Pd reference sample. The larger Debye-Waller parameter of nanocrystalline Pd was found to be due to increased static displacements of atoms from their equilibrium sites compared with the static displacements in coarse-grained material. Consistent with this behaviour, the strain distribution determined from the width of intensity peaks was significantly broader in the nanocrystalline sample than in the coarse-grained sample. No grain-size-correlated differences in either thermal vibrational amplitude or lattice parameter were observed. In contradiction to previous reports of greatly increased thermal expansion coefficients in nanocrystalline metals, the change in lattice parameter of Pd from 16 to 300 K was observed to be independent of grain size. The results indicate that some regions of the nanocrystalline sample are i...

88 citations


Journal ArticleDOI
TL;DR: In this paper, two distinct orthorhombic phases form in the Al65Cu20Fe10Cr5 and Al63Cu17.5Si2 alloys and represent an important class of approximant structures of the decagonal phase.
Abstract: Two distinct orthorhombic phases form in the Al65Cu20Fe10Cr5 and Al63Cu17.5Co17.5Si2 alloys and represent an important class of approximant structures of the decagonal phase. Their structural units (convex pentagon, concave pentagon and rhombus) as revealed by high resolution electron microscopy (HREM) images can also be used to construct the Penrose tiling. We propose that one single orthorhombic phase may not be able to account for the transformation towards the high-temperature decagonal phase. For this transformation to be possible, the microcrystalline structure must satisfy a delicate balance among building blocks that is required to achieve a Penrose tiling. Two concepts are distinguished: the approximant phase and the approximation state. An analysis of the orientation relationships between a CsCl type of structure and orthorhombic phases suggests that such orthorhombic phases are three-dimensional superstruc-tures based on the CsCl unit cell. Furthermore, this relationship leads to the d...

69 citations


Journal ArticleDOI
M. Abkowitz1
TL;DR: In this paper, it has been shown that a diverse assortment of disordered molecular materials in fact share a common pattern of features in their transport properties, such as a convoluted field and temperature dependence which is characterized by In μ proportional to E 1/2 where the slope of this relation typically diminishes with increasing temperature.
Abstract: Over the last decade it has become increasingly clear, through broad application of time-of-flight drift mobility techniques, that a diverse assortment of disordered molecular materials in fact share, in remarkable detail, a common pattern of features in their transport properties. In this paper the latter will be clearly illustrated using as the principal examples certain of the polysilylenes and polygermylenes and poly-tetraphenylbenzidine, a polymer incorporating a transport-active aromatic amine on its main chain. The features demonstrated are as follows. Firstly the drift mobility exhibits a convoluted field and temperature dependence which is characterized by In μ proportional to E1/2 where the slope of this relation typically diminishes with increasing temperature and can even undergo sign reversal and an activated type of temperature dependence which is itself field dependent. In addition, some studies appear to indicate a sensitivity of transport activation to changes in microscopic pola...

67 citations


Journal ArticleDOI
TL;DR: The Si-Na cage structures undergo an insulator-to-metal transition when doped with alkali metals and so have some structural and chemical similarities to the C60 fulleren as discussed by the authors.
Abstract: The Si-Na ‘clathrate’ phases are cage structures that undergo an insulator-to-metal transition when doped with alkali metals and so have some structural and chemical resemblance to the C60 fulleren...

59 citations


Journal ArticleDOI
TL;DR: The sp2 sites form mainly sixfold rings within the clusters and the presence of a small fraction of adjacent fivefold and sevenfold rings washes out the third-neighbour peak at 2 % 4 A in the radial distribution function of sputtered a-C as discussed by the authors.
Abstract: Total energies favour a segregation of the sp2 sites in amorphous carbon (a-C) and hydrogenated amorphous carbon (a-C: H) into sp2-bonded clusters embedded in a sp3-bonded matrix. The size of the clusters determines the local bandgap. The presence of clusters forces the mobility edges further into the bands. Clustering also introduces strong symmetrical fluctuations of the band edges. This causes intracluster transitions between localized and localized states to become allowed. The effect of such clustering on the luminescence properties, the resonant Raman spectra, the Urbach tail and the subgap defect optical absorption of a-C: H is discussed. The sp2 sites form mainly sixfold rings within the clusters. The presence of a small fraction of adjacent fivefold and sevenfold rings washes out the third-neighbour peak at 2 % 4 A in the radial distribution function of sputtered a-C.

59 citations


Journal ArticleDOI
TL;DR: In this article, it was shown that the superlattice should be energetically preferred on (0001) basal faces below about 30 K and on (0110) prism face below about 70 K. The rate of such a reconstruction is probably slow because of the low temperature involved, though molecular interchange with the vapour may facilitate it.
Abstract: It is generally accepted that the faces of ice crystals in equilibrium with the vapour undergo reconstruction to a quasi-liquid transition film at temperatures above about–20°C, though experimental evidence is equivocal. At much lower temperatures it appears that the free energy can be reduced by a less drastic surface reconstruction involving the orientation of surface dipoles to a superlattice. Calculations suggest that this superlattice should be energetically preferred on (0001) basal faces below about 30 K and on (0110) prism faces below about 70 K. On (0001) faces, twin boundaries have very low energy and may exist in profusion. The rate of such a reconstruction is probably slow because of the low temperature involved, though molecular interchange with the vapour may facilitate it. It is possible that this structure could be observed by electron diffraction or by fine-focus infrared spectroscopy. A more radical reconstruction is also discussed, but considered to be energetically unlikely.

Journal ArticleDOI
TL;DR: In this article, a Penrose-tiling approximant of the Al-Mn decagonal quasicrystal was derived from the structure of two sets of Penrose type tiles.
Abstract: A new Al-Mn orthorhombic crystalline phase isostructural with Al60 Mn11 Ni4, Al20 Mn3 Cu2, and the T3 AlMnZn phase has been found in coexistence with the decagonal quasicrystal in a rapidly solidified Al-Mn alloy. From the structure of this Penrose-tiling approximant two sets of Penrose-type tiles were derived. A structural model of the Al-Mn decagonal quasicrystal based on these tiles was proposed. This model agrees well with the recent experimental X-ray structural determination of the Al-Mn decagonal quasicrystal.

Journal ArticleDOI
TL;DR: In this paper, the authors developed a computational simulation of the two-dimensional soap froth, including finite Plateau borders, and showed that the structure is that of a dense random packing of hard discs with contact number equal to four.
Abstract: We have developed a computational simulation of the two-dimensional soap froth, including finite Plateau borders. As the gas fraction фis decreased, there is a loss of rigidity at approximately ф=0·84, at which point the structure is that of a dense random packing of hard discs with contact number equal to four. This result and the general properties of the system are further illuminated by various elementary models and analogies.

Journal ArticleDOI
TL;DR: In this article, a detailed investigation of hydrogen bonding and optical absorption in the 0.6-2 eV range for four series of hydrogenated amorphous silicon films (a-Si:H) deposited under four different plasma conditions and for substrate temperatures ranging from 100 to 300°C.
Abstract: We present the results of a detailed investigation of hydrogen bonding and optical absorption in the 0.6–2 eV range for four series of hydrogenated amorphous silicon films (a-Si:H) deposited under four different plasma conditions and for substrate temperatures ranging from 100 to 300°C. We have found that, for each plasma condition, the subgap absorption shows a sharp decrease at a well defined temperature, which allows us to denne, for each series of samples, an equilibrium temperature T ED for defects above which the subgap absorption remains constant. A similar trend is observed for the structural properties of the films (hydrogen content, hydrogen bonding, optical gap and refractive index). However, for each series, the equilibrium temperature TED for the defects is about 50°C lower than the equilibrium temperature T ED for the hydrogen-related properties. Our results clearly show that the defect structure and the hydrogen-related structure of these a-Si: H samples do not follow the same subs...

Journal ArticleDOI
TL;DR: In this paper, a multipole theory of light propagation in matter is used to describe magnetic effects due to electric quadrupoles and magnetic dipoles induced by light wave fields in non-absorbing antiferromagnetic crystals of the uniaxial and cubic systems.
Abstract: A multipole theory of light propagation in matter is used to describe magnetic effects due to electric quadrupoles and magnetic dipoles induced by light wave fields in non-absorbing antiferromagnetic crystals of the uniaxial and cubic systems. A new effect predicted by the theory is linear birefringence, of electric quadrupole origin, for light travelling along the edge directions in certain cubic crystals. Also predicted in many classes of uniaxial crystal, when propagation is along a crystallographic axis perpendicular to the axis of highest symmetry, is the existence of skew or S-rays. These are characterized by a divergence of the Poynting vector from the wave normal. A previously known effect, non-reciprocal or gyrotropic birefringence, is reconsidered and shown to be a property of media which exhibit Jones birefringence. Furthermore, the theory shows that non-reciprocal birefringence is not in general gyrotropic and that, where effects have previously been described, the results obtained by...

Journal ArticleDOI
TL;DR: In this article, the resistivity in the cross-over region between the low-temperature Fermi liquid and the high temperature perturbative regimes was calculated via an extension of the numerical renormalization group method.
Abstract: One of the most difficult quantities to calculate for the non-degenerate Anderson and Kondo models has been the resistivity in the cross-over region between the low-temperature Fermi liquid and the high-temperature perturbative regimes. Here we show that this quantity can be calculated via an extension of the numerical renormalization group method. We present results deduced from the spectral density as a function of temperature over the full temperature range. The approach can be used to explore the full parameter space of the Anderson model, including additional interactions such as conduction-electron screening.

Journal ArticleDOI
TL;DR: In this article, the existence of the following five structures has been recognized in these alloys by electron microscopy observations: traditional F-type icosahedral phase, an F − type icosahedra phase with a new superlattice ordering, a third F −type icoshedral phase with highly dense distribution of reflections, and two kinds of approximant structures.
Abstract: Al–Pd–Mn–Si alloys with nominal compositions (Al0.710Pdo.205Mno.085)1−x Six and 0 ≤ x ≤; 0.03 were annealed at various temperatures between 1075 and 779 K. The existence of the following five structures has been recognized in these alloys by electron microscopy observations: traditional F–type icosahedral phase, an F–type icosahedral phase with a new superlattice ordering, a third F–type icosahedral phase with a highly dense distribution of reflections, and two kinds of approximant structures. The main feature of the second icosahedral phase was interpreted by a doubled superlattice of the original six-dimensional cubic lattice; a 6D = 2.58nm. This ordering occurred in the alloys annealed at lower temperatures: 876 and 779 K for x=0 alloys and 1025, 980 and 878K for x = 0.01 alloys. A drastic structural improvement by the addition of silicon was discovered in the third F–type icosahedral phase, which was formed in the alloys annealed below 1025 K. The occurrence of these two highly ordered phases...

Journal ArticleDOI
TL;DR: In this paper, negative capacitance has been observed in a wide range of other systems which can now also be understood within the theoretical framework presented in this paper. But the application of equivalent-circuit analysis to such systems has not been discussed.
Abstract: Impedance analysis of ZnS metal–insulator–metal structures has revealed strong negative capacitance behaviour Under low-field conditions, the devices are highly insulating with capacitance consistent with their geometry and dielectric constant However, at high fields they become conducting and the capacitance falls to zero and then becomes negative It may exceed the geometrie value by over five orders of magnitude This frequency-domain behaviour is explained in terms of the unusual time-domain behaviour of the devices The origins of negative capacitance are shown to He in the modulation of the differential conductance by the ac measurement voltage, combined with the interpretation of the response in terms of an electrical equivalent circuit The application of equivalent-circuit analysis to such systems is discussed Negative capacitance has been observed in a wide range of other systems which can now also be understood within the theoretical framework presented

Journal ArticleDOI
TL;DR: In this article, an electron-irradiation-induced crystalline-to-amorphous (CA) transition in α-SiC has been studied by high-resolution electron microscopy (HREM).
Abstract: An electron-irradiation-induced crystalline-to-amorphous (CA) transition in α-SiC has been studied by high-resolution electron microscopy (HREM). The irradiation-produced damage structure was examined as a function of dose of electrons by taking high-resolution maps extending from the unirradiated crystalline region to the completely amorphized region. In the intermediate region between those two regions, that is in the CA transition region, the damage structure was essentially a mixture of crystalline and amorphous phases. The volume fraction of the amorphous phase was found to increase with increasing dose of electrons and no discrete crystalline-amorphous interface was observed in CA transition region. These facts indicate the heterogeneous and gradual nature of the CA transition. In the transition region close to the unirradiated crystalline region, a sort of fragmentation of the crystal lattice was observed to occur; crystallites with slightly different orientations with respect to the paren...

Journal ArticleDOI
TL;DR: In this article, Monte Carlo calculations have been performed in order to study the kinetics of carrier trapping in sparsely populated systems and possible consequences for thermally stimulated conductivity and thermoluminescence performed on microcrystalline or two-dimensional solids are discussed.
Abstract: Monte Carlo calculations have been performed in order to study the kinetics of carrier trapping in sparsely populated systems. The method of calculation overcomes the problem of a ‘very large number’ of carriers participating simultaneously in the relaxation process. The algorithm of the simulation is described. The calculations were performed for the case of a linearly varying temperature. Possible consequences for thermally stimulated conductivity and thermoluminescence performed on microcrystalline or two-dimensional solids are discussed.

Journal ArticleDOI
TL;DR: In this paper, a few new possible models for the observed behaviour are considered, including correlated hopping in a Coulomb gap, and the best of those that they examined is based on correlated hopping.
Abstract: The conductivity in cermets and in granular films is known to vary as exp [-(To/T)1/2], but the existing theories for such a behaviour have been subjected to criticism. A few new possible models for the observed behaviour are considered here. Most of them cannot explain the observations. The best of those that we examined is based on correlated hopping in a Coulomb gap.

Journal ArticleDOI
TL;DR: In this paper, a review of charge transport in molecular solids is discussed, with focus on the effects of static and dynamic disorder, and the theoretical development that led to the now accepted idea that the carriers are polaronic as a result of strong interaction with librations is described.
Abstract: Recent controversial issues in charge transport in molecular solids are discussed, with focus on the effects of static and dynamic disorder. The situation in ordered solids such as aromatic hydrocarbon crystals is reviewed with attention to the nature of charge carriers, and the theoretical development that led to the now accepted idea that the carriers are polaronic as a result of strong interaction with librations is described. Next, recent experiments on disordered solids (specifically molecularly doped polymers), are reviewed, and our theoretical work directed at their understanding is presented and compared with alternatives existing in the literature. Comparative comments are made on the three available approaches: the ordered polaronic, the disordered polaronic and the disordered non-polaronic. It is argued that none is in a satisfactory state at the moment. Key problems that require theoretical attention are identified.

Journal ArticleDOI
TL;DR: Luminescence spectra of Eu3+ in sol-gel-derived silica (SiO2) samples are reported and discussed as a function Eu 3+ concentration and of heating as mentioned in this paper.
Abstract: Luminescence spectra of Eu3+ in sol-gel-derived silica (SiO2) samples are reported and discussed as a function Eu3+ concentration and of heating. Lifetimes, linewidths and Stark splittings of the transitions are used for monitoring structure evolutions of the gel network; the data in the interval 25–250°C allow us to describe the general behaviour of the system during SiO, gel→txerogel transition.

Journal ArticleDOI
TL;DR: In this article, experimental results on reentrant metallic systems are discussed, which are interpreted on a model for the magnetic structure based on the "transverse spin component freezing" approach.
Abstract: We discuss experimental results on re-entrant metallic systems, which we interpret on a model for the magnetic structure based on the ‘transverse spin component freezing’ approach.

Journal ArticleDOI
TL;DR: In this paper, two different methods of promoting microcrystal formation in the amorphous network of hydrogenated silicon carbide (Sic: H) have been investigated: glow-discharge plasma-enhanced chemical vapour deposition (PECVD) with high hydrogen dilution and high power density, and annealing of device-quality SiC: H films.
Abstract: In the present work, two different methods of promoting microcrystal formation in the amorphous network of hydrogenated silicon carbide (Sic: H) have been investigated: glow-discharge plasma-enhanced chemical vapour deposition (PECVD) with high hydrogen dilution and high power density, and annealing of device-quality amorphous SiC: H films. The microcrystalline SiC: H films directly produced by PECVD have a high electrical conductivity (10−5ω-1cm−1), a large energy gap (2–2.1eV) and they are hydrogenated. The annealed films have a high electrical conductivity (7× 10−4ω−1cm−1) and a smaller optical bandgap and they do not contain hydrogen.

Journal ArticleDOI
TL;DR: In this article, the melting and freezing temperatures of small In particles embedded in an Fe matrix have been measured as a function of particie size using in situ heating in an electron microscope.
Abstract: The melting and freezing temperatures of small In particles embedded in an Fe matrix have been measured as a function of particie size using in situ heating in an electron microscope. The melting temperature of small particles of In is lower than that of bulk material and decreases with decreasing radius of the particles (r). The freezing temperature also decreases with decreasing r. This behaviour of the melting of In particles embedded in an Fe matrix contrasts sharply with that of In particles embedded in an Al matrix, for which the melting temperature increases with decreasing r. The difference is discussed in terms of the interfacial energies between Al or Fe matrix and liquid or solid In.

Journal ArticleDOI
TL;DR: A survey of experimental and theoretical work on the Coulomb interactions in a Mott-Anderson insulator can be found in this article, with a focus on recent developments in the literature.
Abstract: A survey is presented of experimental and theoretical work on the problem of Coulomb interactions in a Mott-Anderson insulator, since its inception, but emphasizing recent developments The paper also attempts to explain the basic physical considerations, to point out some existing disagreements and to assess to what degree these may have been resolved by recent work

Journal ArticleDOI
TL;DR: In this article, a scheme for the explicit treatment of charge-transfer processes in molecular-dynamics simulations of condensed phases is described, and the method is applied to pure amorphous silica and to silica to which small numbers of dilithium oxide impurities have been added.
Abstract: A scheme is described for the explicit treatment of charge-transfer processes in molecular-dynamics simulations of condensed phases. Technical details of the simulations are discussed, and it is shown that it is possible to implement the method without serious cost in computing time relative to standard molecular-dynamics calculations for charged systems. The scheme used ensures that the charge-transfer forces are dynamically conservative, and no numerical instabilities have been observed in the integration of the equations of motion. The method is applied to pure amorphous silica and to silica to which small numbers of dilithium oxide impurities have been added. The distributions of atomic charges have their expected form in each case. In particular, addition of the cation impurities leads to the spontaneous generation of an almost equal number of oxygen defects, and it is shown that it is possible to distinguish unambiguously between bridging and dangling oxygens on electrostatic rather than ge...

Journal ArticleDOI
TL;DR: In this article, structural relations of the two approximant phases to the icosahedral phase are discussed, in the framework of the projection scheme, on the basis of X-ray diffraction data.
Abstract: It has been found in a Mg Ga-Al-Zn alloy that an icosahedral phase produced by melt-spinning transforms to the 1/1 cubic approximant of the icosahedral phase at 653 K and successively to the 2/1 cubic approximant at around 668 K upon heating. Structural relations of the two approximant phases to the icosahedral phase are discussed, in the framework of the projection scheme, on the basis of X-ray diffraction data. Electrical resistivities have been measured from 4·2 K to 280 K for single-phase samples of the icosahedral and approximant phases. The icosahedral phase and the 2/1 approximant one show similar behaviour, in which the resisitivity is ∼ 120 μω2 cm at 4-2 K and the temperature coefficient is negative. In contrast, the 1/1 approximant phase shows a relatively smali resistivity value of ∼ 60 μω cm at 4-2 K and a positive temperature coefficient. Changes in the resistivity accompanied by the phase transformations have been confirmed by measuring the resistivity upon heating the melt-spun sam...

Journal ArticleDOI
TL;DR: In this article, the ratios of magnetic to charge form factors have been measured for 11 reflections in ferromagnetic α-iron by diffraction of synchrotron radiation, and a white-beam single-crystal approach has produced data of a comparable auality to early polarized neutron work, and it seems likely that this techniaue will become a valuable new tool for studying ferromagnets.
Abstract: The ratios of magnetic to charge form factors have been measured for 11 reflections in ferromagnetic α-iron by diffraction of synchrotron radiation. A novel white-beam single-crystal approach has produced data of a comparable auality to early polarized neutron work, and it seems likely that this techniaue will become a valuable new tool for studying ferromagnets.