jap banner
  • Volume/Page
  • Keyword
  • DOI
  • Citation
  • Advanced
   
 
 
 

Flickr Twitter iResearch App Facebook

BROWSE VOLUMES

Year Range: 
Search Issue | RSS Feeds RSS
Previous Issue Next Issue

1 Oct 1996

Volume 80, Issue 7, pp. 3619-4239

Page 2 of 4 Pages Previous Page Next Page | Jump to Page

On the unusual electromigration behavior of copper interconnects

E. Glickman and M. Nathan

J. Appl. Phys. 80, 3782 (1996); http://dx.doi.org/10.1063/1.363330 (10 pages) | Cited 33 times

Full Text: | Download PDF

Show Abstract
The present uncertainty in the ability of Cu to substitute for Al as the conductor material in very large scale integration arises from the perplexing electromigration behavior of Cu interconnects: the electromigration activation energy in multigrained lines is often about two times lower than for grain‐boundary diffusion, while the pre‐exponential factor in the electromigration rate expression is several orders of magnitude smaller than that characteristic of electromigration along grain boundaries. Using literature data, in particular those of drift velocity experiments, we show that regardless of these unusual facts, grain boundaries are still most likely the major electromigration diffusion pathways in Cu interconnects. Based upon recent progress in the theory of grain‐boundary grooving with an arbitrary grain‐boundary flux [Klinger et al. J. Appl. Phys. 78, 3833 (1995)], and the specific model applying the general theory to electromigration [Glickman, Phys. Low‐Dim. Struct. 11/12, 69 (1994)], we explain the major features of electromigration in Cu in terms of the extension of slit‐like grooves along the interconnect line, followed by their merging. Fitting the electromigration activation energy reported for pure Cu into the model suggests that surface diffusion along freshly created groove walls is slow, with an activation energy above 2 eV. Most likely, this is due to trace surface contaminations. Having this new key element in the grain‐boundary grooving model, with surface diffusion acting in effect as a ‘‘healing’’ mechanism rather than as an independent pathway parallel to grain‐boundary diffusion, and using empirical surface diffusion parameters for Cu, enable us to rationalize the major features of Cu electromigration behavior reported in the literature: the values of the electromigration pre‐exponents, the island‐like morphology of the electromigration displacement region, the current density exponent, and the origin of the Sn effect on electromigration parameters. Our main conclusion is that the full advantages of Cu as the most promising electromigration resistant interconnect material can be realized, provided that trace contaminations responsible for suppressing diffusion along freshly created surfaces are identified, controlled and eliminated. This will allow unalloyed, multigrained Cu interconnects to have an electromigration resistance four orders of magnitude larger than Al(2% Cu) at 100 °C. © 1996 American Institute of Physics.
Show PACS
66.30.Qa Electromigration
85.40.Ls Metallization, contacts, interconnects; device isolation

An integral equation approach to electromigration under arbitrary time‐dependent stress

V. M. Dwyer, M. J. Kearney, and P. C. Bressloff

J. Appl. Phys. 80, 3792 (1996); http://dx.doi.org/10.1063/1.363331 (6 pages) | Cited 2 times

Full Text: | Download PDF

Show Abstract
We present an integral equation reformulation of the drift–diffusion equation for modeling void buildup in integrated circuit interconnects due to the electromigration of aluminum atoms. The method can deal with arbitrary time‐dependent current stress and offers a number of advantages over the conventional finite‐difference approach. In particular, the boundary conditions are included in an intuitive and straightforward fashion. By way of illustration we identify the conditions under which the time‐to‐failure scales as r−2 for unidirectional square‐wave stress of duty cycle r. © 1996 American Institute of Physics.
Show PACS
66.30.Qa Electromigration
85.40.Ls Metallization, contacts, interconnects; device isolation

Competitive evolution of the fine contrast modulation and CuPt ordering in InGaP/GaAs layers

A. Diéguez, F. Peiró, A. Cornet, J. R. Morante, F. Alsina, and J. Pascual

J. Appl. Phys. 80, 3798 (1996); http://dx.doi.org/10.1063/1.363332 (6 pages) | Cited 7 times

Full Text: | Download PDF

Show Abstract
We use transmission electron microscopy to characterize the morphology of InGaP epitaxial layers grown by metal‐organic vapor‐phase epitaxy over misoriented GaAs (001) substrates, with a cutoff angle in a range from 0° to 25°. The occurrence of phase separation and CuPt‐type ordered superstructures has been observed. The most ordered configuration has been found to appear in layers grown on 2° off substrates, and the strength of order decreases with increasing the misorientation angle beyond α=2°. Conversely, whereas the phase separation is less evident in the layer grown at 2°, the sample grown with a misorientation of 25° exhibits the most phase separated configuration. The completion between these two phenomena is discussed depending on the misorientation angle. © 1996 American Institute of Physics.
Show PACS
68.55.-a Thin film structure and morphology
68.37.Hk Scanning electron microscopy (SEM) (including EBIC)
68.37.Lp Transmission electron microscopy (TEM)

Structural characterization of ordered SiGe films grown on Ge(100) and Si(100) substrates

T. Araki, N. Fujimura, T. Ito, A. Wakahara, and A. Sasaki

J. Appl. Phys. 80, 3804 (1996); http://dx.doi.org/10.1063/1.363333 (4 pages) | Cited 2 times

Full Text: | Download PDF

Show Abstract
We observe the formation of an ordered structure in Si1−xGex films grown on Ge(100) substrates, as well as on Si(100) substrates, by molecular beam epitaxy. The structural characterization of these ordered films is performed. The degree of order in the films is quantitatively measured using x‐ray diffraction. The dependence of the degree of order on Ge composition is similar between films on Ge(100) and Si(100) substrates. By careful x‐ray diffraction analysis, we find that the degree of order is not equivalent in variants. © 1996 American Institute of Physics.
Show PACS
68.55.-a Thin film structure and morphology
81.15.Hi Molecular, atomic, ion, and chemical beam epitaxy

Scaling laws for the reduction of threading dislocation densities in homogeneous buffer layers

J. S. Speck, M. A. Brewer, G. Beltz, A. E. Romanov, and W. Pompe

J. Appl. Phys. 80, 3808 (1996); http://dx.doi.org/10.1063/1.363334 (9 pages) | Cited 48 times

Full Text: | Download PDF

Show Abstract
In the heteroepitaxial growth of films with large misfit with the underlying substrate (linear mismatch strains in excess of 1%–2%) the generation of misfit dislocations and threading dislocations (TDs) is ubiquitous for thicknesses well in excess of the equilibrium critical thickness. Experimental data suggest that the TD density in relaxed homogeneous buffer layers can be divided into three regimes: (i) an entanglement region near the film/substrate interface corresponding to TD densities of ∼1010–1012 cm−2; (ii) a falloff in TD density that is inversely proportional to the film thickness h, applicable to densities in the range ∼107–109 cm−2; and (iii) saturation or weak decay of the TD density with further increase in film thickness. Typical saturation densities are on the order of ∼106–107 cm−2. In this article, we show that the TD reduction may be described in terms of effective lateral motion of TDs with increasing film thickness. An analytic model is developed that successfully predicts both the 1/h scaling behavior and the saturation of TD densities. Long‐range fluctuations in the net Burgers vector content of the local TDs is a cause for saturation behavior. These models are supported by computer simulations. © 1996 American Institute of Physics.
Show PACS
61.72.Lk Linear defects: dislocations, disclinations
61.72.Bb Theories and models of crystal defects
68.60.Bs Mechanical and acoustical properties

Band lineup of lattice mismatched InSe/GaSe quantum well structures prepared by van der Waals epitaxy: Absence of interfacial dipoles

O. Lang, A. Klein, C. Pettenkofer, W. Jaegermann, and A. Chevy

J. Appl. Phys. 80, 3817 (1996); http://dx.doi.org/10.1063/1.363335 (5 pages) | Cited 19 times

Full Text: | Download PDF

Show Abstract
Epitaxial growth of the strongly lattice mismatched (6.5%) layered chalcogenides InSe and GaSe on each other is obtained with the concept of van der Waals epitaxy as proven by low‐energy electron diffraction and scanning tunnel microscope. InSe/GaSe/InSe and GaSe/InSe/GaSe quantum well structures were prepared by molecular beam epitaxy and their interface properties were characterized by soft x‐ray photoelectron spectroscopy. Valence and conduction band offsets are determined to be 0.1 and 0.9 eV, respectively, and do not depend on deposition sequence (commutativity). As determined from the measured work functions the interface dipole is 0.05 eV; the band lineup between the two materials is correctly predicted by the Anderson model (electron affinity rule). © 1996 American Institute of Physics.
Show PACS
68.65.-k Low-dimensional, mesoscopic, nanoscale and other related systems: structure and nonelectronic properties
81.15.Hi Molecular, atomic, ion, and chemical beam epitaxy
68.37.Hk Scanning electron microscopy (SEM) (including EBIC)
68.37.Lp Transmission electron microscopy (TEM)

Morphology of wetting reaction of eutectic SnPb solder on Au foils

P. G. Kim and K. N. Tu

J. Appl. Phys. 80, 3822 (1996); http://dx.doi.org/10.1063/1.363336 (6 pages) | Cited 38 times

Full Text: | Download PDF

Show Abstract
We have studied the wetting behaviors of eutectic SnPb solder caps on pure Au foils at 200 °C. Surface morphology, wetting angle, and wetting tip stability were examined by scanning electron microscopy and energy dispersion x‐ray analysis. In addition, interfacial reaction in the bulk diffusion couple of the solder and Au was studied. Intermetallic compound (AuSn4) was observed on the surface of the solder cap, often called the cold joint, as early as 2 s of reflow at 200 °C. The wetting angle decreased with reflow time but remained constant after 180 s of reflow. However, the side‐band width and the diameter of the solder cap continuously increase with the reflow time to 5 min when the entire eutectic SnPb solder cap is fully consumed by intermetallic compound formation. In the bulk diffusion couple of eutectic SnPb and Au, the growth and morphology of the intermetallic compound (AuSn4) was influenced by Au dissolution. Since Au, showing excellent wetting behavior, forms the platelet‐type intermetallic compound, we postulate that the wetting rate may be improved with a system containing platelet‐type intermetallic compound formation for the eutectic SnPb solder. © 1996 American Institute of Physics.
Show PACS
68.08.Bc Wetting
81.30.Bx Phase diagrams of metals, alloys, and oxides
82.80.Ej X-ray, Mössbauer, and other γ-ray spectroscopic analysis methods

C+‐energy‐dependent residual ion damage in GaAs:C grown by the low‐energy ion‐beam doping method

Tsutomu Iida, Yunosuke Makita, Takayuki Shima, Shinji Kimura, Joachim Horn, Hans L. Hartnagel, and Shin‐ichiro Uekusa

J. Appl. Phys. 80, 3828 (1996); http://dx.doi.org/10.1063/1.363306 (6 pages) | Cited 1 time

Full Text: | Download PDF

Show Abstract
The effects of residual ion damage in low‐energy (30 eV–30 keV) C+‐doped GaAs were investigated with regard to the electrical and optical activation of C as a function of C+ ion acceleration energy EC+. Systematic variation of EC+ demonstrated that, in the energy range of EC+<170 eV, the net hole concentration (∣NA−ND∣) slightly increases as EC+ increases and the highest ∣NA−ND∣ was obtained at EC+=170 eV under the constant C+ ion‐beam current density. For EC+≳170 eV, an increase in EC+ gave rise to an abrupt decrease of ∣NA−ND∣ down to two orders of magnitude smaller than that obtained at EC+=170 eV. In low‐temperature (2 K) photoluminescence spectra for as‐grown samples with EC+=240 and 350 eV, a novel emission ascribable to residual ion damage was observed instead of an essential acceptor–acceptor emission of [g−g]β. However, subsequent annealing at 850 °C made this novel emission disappear and the proper [g−g]β emission was merely observed. An activation process observed for EC+=5, 10, and 30 keV samples was very similar to that by high‐energy ion implantation, indicating low activation rate of 10%–15%. Minority‐carrier lifetime measurements using scanning tunneling microscope stimulated time‐resolved luminescence demonstrated the presence of residual ion damage in as‐grown samples at EC+=240 and 350 eV and annealed ones at EC+=5, 10, and 30 keV while no ion damage was observed in as‐grown sample at EC+=30 eV. The incorporation and activation behaviors of C atoms that take the form of low‐energy ions were found to be considerably affected by changes in ion–surface interaction with increasing EC+ and by the presence of residual ion damage in the layer. © 1996 American Institute of Physics.
Show PACS
61.80.Jh Ion radiation effects
81.15.Jj Ion and electron beam-assisted deposition; ion plating
79.20.Rf Atomic, molecular, and ion beam impact and interactions with surfaces
68.55.Ln Defects and impurities: doping, implantation, distribution, concentration, etc.

Electrical behavior of implanted carbon impurities in fluorine co‐implanted GaAs

Marco Peroni, Patrizio D’Eustacchio, Nazzareno Di Virginio, Roberto Graffitti, and Andrea Gasparotto

J. Appl. Phys. 80, 3834 (1996); http://dx.doi.org/10.1063/1.363337 (6 pages) | Cited 1 time

Full Text: | Download PDF

Show Abstract
The effect of fluorine co‐implantation in carbon‐implanted GaAs has been investigated by secondary‐ion‐mass spectrometry (SIMS), Hall, C–V and deep level transient spectroscopy (DLTS) techniques. The variation of carbon activation with different co‐implanted doses of fluorine has been analyzed on the basis of the concentration and mobility versus depth profiles of the samples, obtained via the differential Hall technique. The acceptor activation is found to increase with the co‐implanted F dose only in the surface region, whereas in the deeper part of the carbon‐implanted layer the co‐implantation process is less effective, if not counterproductive. The presence of a midgap hole trap level, with an emission energy ET=0.48 eV, has been observed by means of the DLTS technique. The trap concentration is found to be higher in the samples with the lower C activation level (corresponding to lower co‐implanted F dose); furthermore, by comparing C–V measurements (at 77 K and room temperature) and SIMS profiles, the trap concentration profile is found to be almost coincident with the unactivated C concentration: In other words, the difference between the implanted C and the activated acceptor concentration corresponds to the hole trap concentration. This result enables a better understanding of the anomalous behavior observed when implanted carbon is used to compensate donor impurities, with almost 100% efficiency, while as an acceptor its activation is much lower. All these effects cannot be ascribed to any chemical bond of the co‐implanted F since from SIMS analysis it results that it completely outdiffuses after the rapid thermal anneal process. © 1996 American Institute of Physics.
Show PACS
71.55.Eq III-V semiconductors
61.72.S- Impurities in crystals
61.72.uj III-V and II-VI semiconductors

Transient picosecond/subpicosecond Raman scattering studies of nonequilibrium electron distributions and phonons in CdTe

E. Grann, Y. Chen, K. T. Tsen, D. K. Ferry, T. Almeida, Y. P. Chen, J. P. Faurie, and S. Sivananthan

J. Appl. Phys. 80, 3840 (1996); http://dx.doi.org/10.1063/1.363338 (4 pages) | Cited 7 times

Full Text: | Download PDF

Show Abstract
Nonequilibrium electron distributions and phonons in CdTe have been studied by transient picosecond/subpicosecond Raman spectroscopy at T=300 K. Our experimental results show that for photoexcited electron–hole density of n≂1018 cm−3, the electron distributions can be reasonably well described by Fermi–Dirac distribution functions with effective electron temperature substantially higher than the lattice temperature. From an ensemble Monte Carlo analysis of the nonequilibrium phonon population as a function of photoexcited electron–hole pair density, the LO phonon lifetime in CdTe has been deduced to be τ≂0.75±0.25 ps. © 1996 American Institute of Physics.
Show PACS
78.30.Fs III-V and II-VI semiconductors
78.47.-p Spectroscopy of solid state dynamics
63.20.D- Phonon states and bands, normal modes, and phonon dispersion

Majority and minority electron and hole mobilities in heavily doped gallium aluminum arsenide

Herbert S. Bennett

J. Appl. Phys. 80, 3844 (1996); http://dx.doi.org/10.1063/1.363339 (10 pages) | Cited 7 times

Full Text: | Download PDF

Show Abstract
The majority electron and minority hole mobilities were calculated in Ga1−yAlyAs for donor densities between 1016 and 1019 cm−3. Similarly, the majority hole and minority electron mobilities were calculated for acceptor densities between 1016 and 1020 cm−3. The mole fraction of AlAs, y, varies between 0.0 and 0.3 in these calculations. All the important scattering mechanisms have been included. The ionized impurity and carrier–carrier scattering processes were treated with a quantum‐mechanical, phase‐shift analysis. These calculations are the first to use a phase‐shift analysis for minority carriers scattering from majority carriers in ternary compounds such as Ga1−yAlyAs. The results are in good agreement with experiment for majority mobilities and predict that at high dopant densities minority mobilities should increase with increasing dopant density for a short range of densities. This effect occurs because of the reduction of plasmon scattering and the removal of carriers from carrier–carrier scattering due to the Pauli exclusion principle. These calculations do not treat the density‐of‐state modifications due to heavy doping, which should have only a small effect on the mobility at room temperature. The results are important for device modeling because of the need to have physically reasonable values for minority mobilities when simulating the electrical behavior of heterojunction bipolar transistors.
Show PACS
72.20.Dp General theory, scattering mechanisms
72.80.Ey III-V and II-VI semiconductors

Small‐signal capacitance and conductance of biased a‐Si structures

J. Furlan, I. Skubic, F. Smole, P. Popović, and M. Topič

J. Appl. Phys. 80, 3854 (1996); http://dx.doi.org/10.1063/1.363340 (6 pages) | Cited 1 time

Full Text: | Download PDF

Show Abstract
Small‐signal capacitance and conductance of experimental samples of a‐Si n‐i‐n structures were measured in a wide frequency range under various bias conditions. The measured capacitance at low frequencies greatly exceeds the expected value derived from the ΔQ/ΔV ratio, where ΔQ is a change of the trapped charge corresponding to a change ΔV of the applied voltage. This capacitance increases with the steady‐state bias and decreases with the frequency of the measuring signal. The measured low‐frequency small‐signal conductance equals the differential conductance obtained from the steady‐state current–voltage characteristics, but it increases with the rising frequency of the measuring signal. A small‐signal analytical model of an a‐Si n‐i‐n structure is developed which agrees well with the experimental results. With this model, the high capacitive effect of the n‐i‐n device at low frequencies is explained on the basis of a phase shift which arises from the delayed capture–emission mechanism of carriers in the localized states. Using this model, it is shown that an increasing frequency of small‐signal excitation moves the energy region of gap states engaged in the delaying action toward the conduction band, resulting in a decreasing capacitance and an increasing conductance of the a‐Si n‐i‐n structure. © 1996 American Institute of Physics.
Show PACS
73.61.Jc Amorphous semiconductors; glasses
73.40.Lq Other semiconductor-to-semiconductor contacts, p-n junctions, and heterojunctions

Electrical resistivity of TiB2 at elevated pressures and temperatures

Xiaoyuan Li, Murli H. Manghnani, Li‐Chung Ming, and Dennis E. Grady

J. Appl. Phys. 80, 3860 (1996); http://dx.doi.org/10.1063/1.363341 (3 pages) | Cited 9 times

Full Text: | Download PDF

Show Abstract
The electrical resistivity of TiB2 has been measured using a DIA‐6 cubic anvil apparatus at pressures up to 8 GPa and temperatures up to 800 K. The ambient‐condition resistivity is determined to be 13.3 (±0.9) μΩ cm. The resistivity decreases with increasing pressure. At pressures above 2 GPa, the pressure dependence of the resistivity is about −0.36 μΩ cm/GPa. On heating, the resistivity increases linearly with temperature. The measurements at simultaneously high pressure (3.2 GPa) and high temperatures yield a temperature dependence of 46 (±5) nΩ cm/K for the resistivity. © 1996 American Institute of Physics.
Show PACS
72.15.Eb Electrical and thermal conduction in crystalline metals and alloys
81.40.Vw Pressure treatment

Polar optical oscillations in coupled quantum wells: The electron‐phonon interaction and scattering

Jun‐jie Shi and Shao‐hua Pan

J. Appl. Phys. 80, 3863 (1996); http://dx.doi.org/10.1063/1.363342 (13 pages) | Cited 21 times

Full Text: | Download PDF

Show Abstract
Within the framework of the dielectric continuum model, optical phonon modes and their interaction and scattering with electrons in general coupled quantum wells (GCQWs) are studied. The dispersion relation of interface phonons and the electron‐interface‐phonon coupling strengths as functions of coordinate z and wave‐number k are derived and illustrated. We find that the forbidden‐frequency behavior of asymmetric heterostructures exists in GCQWs, which may cause an obvious polarization and interaction with electrons. The scattering rates as functions of quantum well parameters are calculated and numerical examples are given graphically. It is shown that the behaviors of intersubband scattering rates are much more complicated than those of intrasubband scattering ones. The results are useful for subsequent theoretical modeling of optical or transport experiments. © 1996 American Institute of Physics.
Show PACS
72.10.Di Scattering by phonons, magnons, and other nonlocalized excitations
63.20.D- Phonon states and bands, normal modes, and phonon dispersion
63.20.K- Phonon interactions

On the conductance and the conductivity of disordered quantum wires

A. Haque and A. N. Khondker

J. Appl. Phys. 80, 3876 (1996); http://dx.doi.org/10.1063/1.363343 (5 pages) | Cited 2 times

Full Text: | Download PDF

Show Abstract
We present a model, based on the Keldysh formalism, to study the transport properties of disordered quantum wires of finite lengths. Unlike the phenomenological models, we estimate the electron in‐ and out‐scattering rates using the local density of states of various transverse modes that allow electrons to change their directions. The formulation, therefore, guarantees the conservation of both the charge and the current at any cross section of the device. Using the model we calculate the average two‐probe conductances of quantum wires that are terminated at nonideal contacts. © 1996 American Institute of Physics.
Show PACS
72.10.Bg General formulation of transport theory
72.20.Dp General theory, scattering mechanisms
73.50.Bk General theory, scattering mechanisms

Analysis of magnetic field dependent Hall data in narrow bandgap Hg1−xCdxTe grown by molecular beam epitaxy

J. Antoszewski and L. Faraone

J. Appl. Phys. 80, 3881 (1996); http://dx.doi.org/10.1063/1.363344 (12 pages) | Cited 15 times

Full Text: | Download PDF

Show Abstract
The analysis of magnetic field dependent Hall data is presented for three representative Hg1−xCdxTe layers grown by Molecular Beam Epitaxy with x in the range 0.193 to 0.244. These samples exhibit ‘‘anomalous’’ Hall characteristics which are analyzed using a hybrid approach consisting of mobility spectrum (MS) analysis followed by a multi‐carrier fitting (MCF) procedure. This hybrid approach is able to readily separate contributions to the total conductivity arising from extrinsic carriers, thermally activated intrinsic electrons, and two‐dimensional electron layers. The extracted transport parameters for thermally activated intrinsic electrons are shown to be in excellent agreement with established physical models for narrow bandgap HgCdTe. The two‐dimensional electron layers are found to be only weekly temperature dependent with an electron mobility in the range of 2 to 4×104 cm2/V s and a sheet density in the range of 1011 to 1012 cm−2. Of particular interest, is the fact that the ‘‘anomalous’’ Hall characteristics exhibited by all three samples are shown not to be indicative of poor‐quality material. This anomalous behavior is found to be due to comparable contributions to the total conductivity from either bulk majority carrier holes and intrinsic minority electrons, or bulk majority carriers and a two‐dimensional electron population. The practical aspects of implementing the MS+MCF procedure are discussed, with particular emphasis on the range of magnetic fields required for unambiguous and accurate parameter extraction. © 1996 American Institute of Physics.
Show PACS
73.50.Jt Galvanomagnetic and other magnetotransport effects (including thermomagnetic effects)
73.61.Ga II-VI semiconductors

Magneto‐optics in coupled quantum dots

Ryuichi Ugajin

J. Appl. Phys. 80, 3893 (1996); http://dx.doi.org/10.1063/1.363345 (4 pages)

Full Text: | Download PDF

Show Abstract
We calculate the optical transition coefficient of two electrons confined in a pair of coupled quantum dots under a magnetic field. When the distance between quantum dots is small, the applied magnetic field causes considerable change in the optical‐transition coefficients and induces new types of optical transition. On the other hand, when the distance is large, the Mott insulating states of a pair of electrons are less sensitive to an external magnetic field because of the small size of each quantum dot. © 1996 American Institute of Physics.
Show PACS
73.21.-b Electron states and collective excitations in multilayers, quantum wells, mesoscopic, and nanoscale systems
78.20.Ls Magneto-optical effects

Novel method for minority‐carrier mobility measurement using photoconductance decay with chemically passivated and plasma damaged surfaces

A. W. Stephens and M. A. Green

J. Appl. Phys. 80, 3897 (1996); http://dx.doi.org/10.1063/1.363346 (7 pages) | Cited 1 time

Full Text: | Download PDF

Show Abstract
A method for measuring minority‐carrier mobility using microwave‐detected photoconductance decay without requiring bulk lifetime, estimates is presented. Three different measurements on a single sample yield values for surface recombination velocity, bulk lifetime, and diffusivity. For each measurement the surface conditions of the sample are changed, allowing extraction of different parameters. The usefulness of 0.08 molar ethanol/iodine solution as a means of achieving such good surface passivation is demonstrated. The following procedure was used to achieve high surface recombination. A CF4 plasma surface etch was shown to achieve the same level of surface damage as mechanical abrasion. The advantage of the new method is that it completely eliminates the chance of breaking samples during the abrasion process, which is of particular advantage for thin samples. The new experimental method for minority‐carrier mobility measurement is evaluated using carrier lifetime measurements made on a commercially available Leo Giken ‘‘Wafer‐τ’’ lifetime tester. © 1996 American Institute of Physics.
Show PACS
72.20.Fr Low-field transport and mobility; piezoresistance
72.40.+w Photoconduction and photovoltaic effects
73.25.+i Surface conductivity and carrier phenomena

Microstructural and electrical investigations of Pd/Ge/Ti/Au ohmic contact to n‐type GaAs

J. S. Kwak, H. N. Kim, H. K. Baik, J.‐L. Lee, D. W. Shin, C. G. Park, H. Kim, and K.‐E. Pyun

J. Appl. Phys. 80, 3904 (1996); http://dx.doi.org/10.1063/1.363347 (6 pages) | Cited 4 times

Full Text: | Download PDF

Show Abstract
Interfacial microstructures of Pd/Ge/Ti/Au ohmic contact to n‐type GaAs have been investigated using cross‐sectional transmission electron microscopy, and the results are used to interpret the electrical properties. Annealing at 300 °C yields a contact resistance of 0.62 Ω mm and the layer structure is changed to GaAs/PdGe/Au4Ti/TiO. The ohmic contact is formed through a solid phase regrowth of GaAs heavily doped with Ge below the PdGe layer. At 380 °C, the lowest contact resistance of 0.43 Ω mm is obtained. The layer structure is changed to GaAs/(Ge–Ti)/PdGe/TiO. Spikes composed of Au and AuGa are found at the grain boundaries of the PdGe compound. The formation of AuGa at 380 °C reduces the contact resistance through the creation of more Ga vacancies at the interface of GaAs/PdGe, and the incorporation of elemental Ge. © 1996 American Institute of Physics.
Show PACS
68.55.Nq Composition and phase identification
73.40.Cg Contact resistance, contact potential
68.37.Hk Scanning electron microscopy (SEM) (including EBIC)
68.37.Lp Transmission electron microscopy (TEM)

Transmission electron microscopy study of rapid thermally annealed Pd/Ge contacts on In0.53Ga0.47As

P. Ressel, W. Österle, I. Urban, I. Dörfel, A. Klein, K. Vogel, and H. Kräutle

J. Appl. Phys. 80, 3910 (1996); http://dx.doi.org/10.1063/1.363348 (5 pages) | Cited 3 times

Full Text: | Download PDF

Show Abstract
Phase formation in rapid thermally annealed Pd/Ge contacts on In0.53Ga0.47As has been investigated by means of cross‐sectional transmission electron microscopy, convergent‐beam electron diffraction, and energy‐dispersive x‐ray analysis. Solid‐phase regrowth is observed to occur similarly as in Pd/Ge contacts on GaAs or InP. The reaction starts at low temperatures with the formation of an amorphous Pd–In–Ga–As layer, which crystallizes at elevated temperatures yielding hexagonal Pd4In0.53Ga0.47As being first described in this work. At temperatures ≳250 °C, this phase decomposes due to epitaxial solid‐phase regrowth of In0.53Ga0.47As and formation of Pd–Ge phases. The stable composition is reached at temperatures ≳350 °C with excess Ge diffused through top Pd–Ge to the contact interface and growing epitaxially on the semiconductor. © 1996 American Institute of Physics.
Show PACS
68.37.Hk Scanning electron microscopy (SEM) (including EBIC)
68.37.Lp Transmission electron microscopy (TEM)
68.55.Nq Composition and phase identification
61.05.jm Convergent-beam electron diffraction, selected-area electron diffraction, nanodiffraction

Conductance measurements on Pb centers at the (111) Si:SiO2 interface

M. J. Uren, J. H. Stathis, and E. Cartier

J. Appl. Phys. 80, 3915 (1996); http://dx.doi.org/10.1063/1.363349 (8 pages) | Cited 38 times

Full Text: | Download PDF

Show Abstract
The electrical properties of the Pb center have been measured using the conductance technique over the temperature range 130–290 K. A high concentration of Pb centers was created by vacuum annealing of 28‐nm‐thick thermal oxides on (111) silicon surfaces. Fitting the conductance data allowed the contribution of the (0/−) Pb level to be separated from the U‐shaped background states. The (0/−) peak in the density of states was found to be asymmetrical with a broad shoulder on the conduction band side. The Pb levels were found to show a capture cross section which fell toward the band edges and which could be fitted by assuming an activated cross section with an activation energy which increased toward the band edges. By contrast, the background states showed a cross section which was temperature and band bending independent.
Show PACS
73.40.Qv Metal-insulator-semiconductor structures (including semiconductor-to-insulator)
73.20.At Surface states, band structure, electron density of states

Quasiparticle injection‐detection experiments in niobium

J. G. Gijsbertsen and J. Flokstra

J. Appl. Phys. 80, 3923 (1996); http://dx.doi.org/10.1063/1.363350 (11 pages) | Cited 2 times

Full Text: | Download PDF

Show Abstract
We present the results of injection‐detection experiments that have been performed to study diffusion and losses of quasiparticles in Nb strips employing a series array of Nb/Al junctions. The I–V curve of the detector junction was measured as a function of the spacing between the (high‐quality) detector junction and the bias point of the injector. For our sample geometry, the quasiparticle gas is described in terms of a one‐dimensional diffusion model. The experiment shows that quasiparticle trapping losses at the interface between the Nb base electrode and the anodized Nb2O5 dominate. In addition, a very low quasiparticle diffusion constant at 1 K, D=1.2×10−4 m2 s−1, has been found. We estimate that this corresponds to an electron mean free path of l≊6.5 nm. Due to the difference between the band gaps of the base and top electrode and to the peak in the density of states at the gap edge, the excess detector current shows a characteristic voltage dependence for bias voltages below 0.5 mV. Theoretical lines for the excess I–V curves are calculated on the basis of a microscopic proximity effect model. All the features observed in the excess I–V curves are well described. © 1996 American Institute of Physics.
Show PACS
74.45.+c Proximity effects; Andreev reflection; SN and SNS junctions
74.70.Ad Metals; alloys and binary compounds (including A15, MgB2, etc.)

Dynamics of a ring with three Josephson junctions

U. Geigenmüller

J. Appl. Phys. 80, 3934 (1996); http://dx.doi.org/10.1063/1.363351 (7 pages) | Cited 4 times

Full Text: | Download PDF

Show Abstract
The dynamics of a current biased superconducting ring with three Josephson junctions is investigated theoretically. For suitable magnetic frustration, the current–voltage (I–V) characteristic shows a surprising series of cusps. This feature, as well as the critical current, are discussed in detail, and explained using an extension of the familiar ‘‘tilted washboard’’ model. The stability of the new structure in the I–V curves against changes of various parameters is studied. An estimate shows that experimental observation should be feasible. © 1996 American Institute of Physics.
Show PACS
85.25.Cp Josephson devices
74.50.+r Tunneling phenomena; Josephson effects

Microstructure and magnetism in FeTaN films deposited in the nanocrystalline state

B. Viala, M. K. Minor, and J. A. Barnard

J. Appl. Phys. 80, 3941 (1996); http://dx.doi.org/10.1063/1.363352 (16 pages) | Cited 73 times

Full Text: | Download PDF

Show Abstract
As‐deposited, magnetically soft nanocrystalline FeTaN films are successfully grown by dc‐magnetron reactive sputtering. Growth conditions are instrumental in extending the solubility limit of Ta in the bcc FeTa alloy. Nitrogen incorporation in FeTa films is found to be much higher than in Fe films and can be explained in terms of thermochemistry using a large Ta‐N interaction coefficient. The influence of different alloying elements is discussed theoretically, with regard to the metal‐nitrogen affinity. A ‘‘typical columnar microstructure’’ associated with the sputtering process is identified and its evolution versus the extent of nitrogenation is described in detail. Stress, magnetostriction, resistivity, and magnetic properties are respectively described as a function of both Ta and N contents of the films. The magnetic behavior of as‐deposited nanocrystalline FeTaN is found to be very sensitive to both the dimension of the grains, their morphology and the nature of the grain boundary material which represents a non‐negligible volume fraction in nanocrystalline films. It is proposed that the columnar structure plays the key role in promoting a large perpendicular anisotropy component (K⊥) and controls a ‘‘Stripe Domain’’‐like behavior observed at high N contents, which cannot be explained in terms of film stress in this material. The contribution of the magnetoelastic anisotropy is also described. In summary, by breaking the columnar structure, the incorporation of nitrogen first decreases K⊥ below the critical limit for formation of stripe domains. In these conditions, N acts as a ‘‘grain refiner’’ and excellent soft magnetic behavior is reported and explained in terms of ‘‘vanishing magnetocrystalline anisotropy.’’ The good thermal stability of such soft films is confirmed. By contrast, higher nitrogen incorporation increases K⊥ above the critical limit, leading to a stable stripe domain‐like structure which does not allow for soft magnetic properties. This phenomenon has been found to be reversible at low temperature where a complete restoration of the soft magnetic behavior has been observed. This anomalous result is explained by the transformation of the grain boundary material into a ‘‘low Curie temperature phase’’ for large N contents. © 1996 American Institute of Physics.
Show PACS
75.70.Ak Magnetic properties of monolayers and thin films
68.55.-a Thin film structure and morphology
75.80.+q Magnetomechanical effects, magnetostriction

Kerr rotation enhancement and correlation between perpendicular anisotropy and Kerr effect in PtCu(Al,Ag)/Co multilayers

Y. J. Wang, Z. H. Guo, D. K. Zhu, and C. H. Shang

J. Appl. Phys. 80, 3957 (1996); http://dx.doi.org/10.1063/1.363353 (5 pages) | Cited 7 times

Full Text: | Download PDF

Show Abstract
The effects of doping Cu, Al, and Ag in Pt spacer layers on the perpendicular anisotropy Ku and the polar Kerr rotation Θk in the wavelength range of 400–800 nm have been investigated for sputtered 0.8 nm Pt/0.3 nm Co multilayers. Ku and Θk measured below 633 nm decrease with the increase of the concentrations of Cu, Al, and Ag in the Pt spacer layers. It is found that the variation of Ku and Θk with the doping concentrations follows a quadratic equation of Ku=aΘ2k+b (a, b are constants here). This suggests that both Ku and Θk originate from a common micromechanism, i.e., spin‐orbit coupling. An obvious enhancement in the peak of the polar Kerr rotation appears at 770 nm for Cu and Ag dopings and at 680 nm for Al doping. Moreover, the polar Kerr rotation Θk∗ at this enhanced peak shows an oscillation behavior with the increasing doping concentrations in the Pt spacer layers. © 1996 American Institute of Physics.
Show PACS
75.70.Cn Magnetic properties of interfaces (multilayers, superlattices, heterostructures)
75.30.Gw Magnetic anisotropy
78.20.Ls Magneto-optical effects
Page 2 of 4 Pages Previous Page Next Page | Jump to Page
Close

Google Calendar

ADVERTISEMENT

Go to AIP Home   © AIP Publishing LLC
close