Library

Notes: Abstract It is well known that during volitional sinusoidal tracking the long-latency reflex modulates in parallel with the volitional EMG activity. In this study, a series of experiments are reported demonstrating several conditions in which an uncoupling of reflex from volitional activity occurs. The paradigm consists of a visually guided task in which the subject tracked a sinusoid with the wrist. The movement was perturbed by constant torque or controlled velocity perturbations at 45° intervals of the tracking phase. Volitional and reflex-evoked EMG and wrist displacement as functions of the tracking phase were recorded. The relationship of both short-latency (30–60 ms) and longer-latency (60–100 ms) reflex components to the volitional EMG was evaluated. In reflex tracking, the peak reflex amplitude occurs at phases of tracking which correspond to a maximum of wrist joint angular velocity in the direction of homonymous muscle shortening and a minimum of wrist compliance. Uncoupling of the reflex and volitional EMG was observed in three situations. First, during passive movement of the wrist through the sinusoidal tracking cycle perturbation-evoked long-latency stretch reflex peak is modulated as for normal, volitional tracking. However, with passive joint movement the volitional EMG modulation is undetectable. Second, a subset of subjects demonstrate a normally modulated and positioned long-latency reflex with a single peak. However, these subjects have distinct bimodal peaks of volitional EMG. Third, the imposition of an anti-elastic load (positive position feedback) shifts the volitional EMG envelope by as much as 180° along the tracking phase when compared with conventional elastic loading. Yet the long-latency reflex peak remains at its usual phase in the tracking cycle, corresponding to the maximal velocity in the direction of muscle shortening. Furthermore, comparison of the results from elastic and anti-elastic loads reveals a dissociation of short- and long-latency reflex activity, with the short-latency reflex shifting with the volitional EMG envelope. Comparable results were also obtained for controlled velocity perturbations used to control for changes in joint compliance. The uncoupling of the reflex and volitional EMG activity in the present series of experiments points to a flexible relationship between reflex and volitional control systems, altered by peripheral input and external load.

Notes: A dependence of the strength of the antiferromagnetic coupling across Cu on the Co layer thickness has been observed. The Co thickness dependence displays two clear peaks consistent with the recently predicted oscillation period of 6.2 A(ring) Co. Apart from the two peaks also several small peaks are visible on a scale of about 1 monolayer Co. Free-electron calculations indicate that these rapid variations in strength may result from slight differences between the slopes and starting points of the two Co wedges that were involved in the experiment.

Notes: We have used the magneto-optical Kerr effect to study the thickness dependence of the perpendicular magnetic anisotropy, coercivity, nucleation field and remanence in an ultrathin Co wedge (0–9 monolayers) grown epitaxially on Pt(111) and capped by a thin Pt overlayer. In addition, the spin-reversal mechanism in the Co layers has been investigated using Kerr domain imaging. The sample showed perpendicular magnetization up to an extrapolated Co thickness of 15 monolayers (ML), with derived volume and interface anisotropy constants of −0.77 MJ/m3 and 1.15 mJ/m2, respectively. The Kerr rotation and Kerr ellipticity demonstrated linear dependences on the Co thickness (tCo), with sizeable extrapolated offsets at tCo=0, attributed to polarization of the Pt by the Co. The coercivity (Hc) and nucleation field both rose to peak values at tCo=1.5 ML, with a subsequent monotonic fall-off for higher Co thicknesses. In the case of Hc, this fall-off did not demonstrate the tCo−5/2 dependence shown by a similar Pd(111)/Co/Pd sample investigated earlier.

Notes: A unique sample was prepared on a Cu(100) single crystal, consisting of three Co layers separated by two Cu layers in the form of wedges oriented perpendicular to each other: Cu(100)/80 A(ring) Co/Cu wedge A/30 A(ring) Co/Cu wedge B/30 A(ring) Co/7 A(ring) Cu/30 A(ring) Au. Position-sensitive magneto-optical Kerr effect measurements along Cu wedge B, at a fixed position on Cu wedge A corresponding to maximum antiferromagnetic (AF) coupling, enabled us to investigate not only the AF but also the ferromagnetic (F) coupling between the two 30 A(ring) Co layers as a function of the Cu thickness. The measurements confirmed both long and short period coupling oscillations in the AF regime, and revealed the predicted extension of the short period through the F regime.

Notes: The encoding of direction and speed in the discharge of dorsal premotor (PMd) and primary motor (MI) neurons was studied during two-dimensional visually-instructed pursuit arm movements in which eight directions and four constant speeds were independently manipulated. Each trial consisted of equal durations of visual observation of target movement without hand movement (cue) and visual pursuit-tracking of the target with the hand (track). A total of 240 neurons was recorded from PMd and MI in two Macaca mulatta monkeys. Two classes of regression analyses were used to relate neuronal firing during the cue and track periods to direction and speed. First, the average firing from each period was fitted to target direction or speed. Period-averaged firing significantly correlated with direction more frequently in the track than in the cue period. Conversely, correlations with speed (with or without direction) were more common in the cue than in the track period. Secondly, a binwise regression evaluated the temporal evolution of firing correlations with direction and speed. Supporting the period-based results, significant binwise correlations of the discharge with speed occurred preferentially during the cue period when there was no hand movement. Prior to movement, correlations of the firing with direction became significant and continued through the movement. Both analyses demonstrated a distinct tendency for neurons to be modulated by speed information early and by direction information later. This temporal parcellation reflects both the sequential demands of the task and constraints placed on the neural computations. The early representation of target speed is hypothesized to reflect the need to calculate a ‘go signal’ for the initiation of movement.

Notes: Photon-emission experiments on silicon-rich hydrogenated amorphous silicon-nitride metal–semiconductor–metal diodes, have shown the existence of hot electrons under applied field strengths of approximately 106 V/cm. The effective temperatures and mean free path between collision for the electrons were estimated from the spectra. It is shown that, in general, asymmetrical changes in the electrical characteristics of the devices occur after prolonged dc stressing at high fields. Two drift mechanisms can be distinguished. The first is called "cathodic'' drift and is driven by recombination between band-tail carriers in the semiconductor. The other is called "anodic'' drift, and results from the effects of hot electrons at the anode. The spatial and time dependence of these drift mechanism is explained using a simple model. © 1996 American Institute of Physics.

Notes: Magnetic anisotropies of ultrathin transition-metal films are inherently related to their structural properties. In ultrathin films the large fraction of atoms located at the film interface generates strong interface anisotropies, whereas elastic strain fields caused by the forced registry of atoms at the substrate/film interface induce magnetoelastic anisotropy contributions. So far the experimental confirmation of the transition from these thin-film properties to bulk anisotropy properties, characterized by a dominating magnetocrystalline anisotropy, has not yet been presented. Magnetic anisotropies reflect, depending on their origin, both the crystallographic symmetry and the symmetry of the film geometry. For a clear separation between magnetoelastic, magnetocrystalline and Néel-type interface anisotropy contributions, the film symmetry and thickness must be chosen such that the respective different anisotropy contributions appear with different symmetries and film thickness dependencies. This is the case for (110)-oriented fcc Co films. In the present study we use the Brillouin light-scattering technique for the determination of the anisotropy contributions. An analysis of the spin-wave frequency measured as a function of the in-plane direction of the external field and the film thickness yields information about all relevant anisotropies. The samples used were molecular-beam-epitaxy grown in ultrahigh vacuum. Onto a Cu (110) single-crystal substrate a wedge-type sample and two staircase-shaped samples with distinct thicknesses in the range of 8–110 A(ring) were grown.To obtain symmetric Co/Cu interfaces the Co layers were covered with a 12 A(ring) Cu layer. Finally, a 25-A(ring)-thick Au protective layer was deposited. Low-energy electron-diffraction studies were used to obtain the structural data of the films. All relevant anisotropy contributions—the magnetocrystalline anisotropy, and the uniaxial in-plane and out-of-plane anisotropy contributions—were determined. Three different anisotropy regimes are observed as a function of the Co layer thickness dCo. This thickness regime up to 13 A(ring) is dominated by the magnetoelastic anisotropy contributions as a result of the pseudomorphic film growth of the Co layer. For Co layer thicknesses larger than 13 A(ring) we find a reduction of the magnetoelastic anisotropy contributions. This is structurally correlated to an anisotropic relaxation of the in-plane Co lattice constant. In the regime of dCo(approximately-greater-than)50 A(ring) we observe a thickness-independent value for the magnetocrystalline anisotropy contribution K1=−8.5×105 erg/cm3. This anisotropy contribution is largely suppressed for dCo〈50 A(ring). This finding might either indicate a breakdown of the usually postulated linear superposition principle of magnetic anisotropy contributions to the free anisotropy energy, or it might point to a subtle modification of the electronic band structure. At the onset of the magnetocrystalline anisotropy we find a change in the easy magnetization direction from 〈001〉 for thin Co films to 〈111〉 for thicker ones. For a more detailed discussion see Ref. .

Notes: We discuss the fabrication and microstructuring techniques of pillar structures made of high vacuum sputtered Fe/Cr multilayers and of molecular beam epitaxy evaporated Co/Cu multilayers, for which we measured the giant magnetoresistance effect with the current perpendicular to the multilayer plane from 4 K to 300 K. Using optical lithography and reactive ion etching techniques we obtained structures with a typical height of 0.5 μm and a width ranging between 3 and 10 μm. For both Fe/Cr and Co/Cu multilayers we find an enhanced magnetoresistance with respect to the in-plane case. The perpendicular magnetoresistance of the Fe/Cr pillars strongly decreases with temperature, while for the Co/Cu samples the temperature dependence is weaker, indicating electron-magnon scattering processes of different strength.

Notes: Magnetic anisotropies and misfit strain relaxations have been investigated in Cu/Ni-wedge/Cu (100) and (111) sandwiches deposited by molecular beam epitaxy on single-crystal Cu substrates. Our results reveal a clear distinction in the nature of the measured anisotropy at Ni thicknesses below and above the critical value tc, where the growth becomes incoherent. Below tc, coherent lattice strain modifies only the volume anisotropy, while interface anisotropy is Néel type; above tc, magnetoelastic effects are found to contribute to the interface anisotropy.