The

Abstract:We report on the surprisingly strong, broadband emission of coherent terahertz pulses from ultrathin layers of semiconductors such as amorphous silicon, germanium and polycrystalline cuprous oxide deposited on gold, upon illumination with femtosecond laser pulses. The strength of the emission is surprising because the materials are considered to be bad (amorphous silicon and polycrystalline cuprous oxide) or fair (amorphous germanium) terahertz emitters at best. We show that the strength of the emission is partly explained by cavity-enhanced optical absorption. This forces most of the light to be absorbed in the depletion region of the semiconductor/metal interface where terahertz generation occurs. For an excitation wavelength of 800 nm, the strongest terahertz emission is found for a 25 nm thick layer of amorphous germanium, a 40 nm thick layer of amorphous silicon and a 420 nm thick layer of cuprous oxide, all on gold. The emission from cuprous oxide is similar in strength to that obtained with optical rectification from a 300 μm thick gallium phosphide crystal. As an application of our findings we demonstrate how such thin films can be used to turn standard optical components, such as paraboloidal mirrors, into self-focusing terahertz emitters.

Abstract:We report on direct measurements of the magnetic near-field of metamaterial split ring resonators at terahertz frequencies using a magnetic field sensitive material. Specifically, planar split ring resonators are fabricated on a single magneto-optically active terbium gallium garnet crystal. Normally incident terahertz radiation couples to the resonator inducing a magnetic dipole oscillating perpendicular to the crystal surface. Faraday rotation of the polarisation of a near-infrared probe beam directly measures the magnetic near-field with 100 femtosecond temporal resolution and (λ /200) spatial resolution. Numerical simulations suggest that the magnetic field can be enhanced in the plane of the resonator by as much as a factor of 200 compared to the incident field strength. Our results provide a route towards hybrid devices for dynamic magneto-active control of light such as isolators, and highlight the utility of split ring resonators as compact probes of magnetic phenomena in condensed matter. Planken, "Advanced terahertz electric near-field measurements at sub-wavelength diameter metallic apertures," Opt. Express 16, 7407-7417 (2008). 19. A. Sell, A. Leitenstorfer, and R. Huber, "Phase-locked generation and field-resolved detection of widely tunable terahertz pulses with amplitudes exceeding 100 MV/cm," Opt. Lett. 33, 2767-2769 (2008).

When (nanostructured) metals, such as gold and silver, are illuminated with laser pulses having a duration in the femtosecond range, they can emit pulses of THz light. Most of these experiments have been performed using amplified lasers, giving rise to energy densities on the sample on the order of mJ cm −2. The results of the different experiments are surprisingly inconsistent in both the measurements of the THz fluence as a function of laser fluence and in the interpretation of the results. This paper reviews the current state of affairs of this interesting topic and discusses some effects related to surface preparation that may influence the emission THz light on metals, particularly silver and copper. We also show results of measurements on nanostructured metals using unamplified laser pulses, which emphasize the role played by plasmons in the generation of THz light. When increasing the optical energy density on a specially nanostructured sample, we observe a transition from a 'classical' second-order non-linear optical process to a higher-order process as the source of the THz radiation. This supports recent results on a differently structured metal by Polyushkin et al (2014 Phys. Rev. B 89 125426), who also observe two different power regimes when decreasing the intensity coming from the high energy density side.

Semisupervised clustering algorithms partition a given data set using limited supervision from the user. The success of these algorithms depends on the type of supervision and also on the kind of dissimilarity measure used while creating partitions of the space. This paper proposes a clustering algorithm that uses supervision in terms of relative comparisons, viz., x is closer to y than to z. The proposed clustering algorithm simultaneously learns the underlying dissimilarity measure while finding compact clusters in the given data set using relative comparisons. Through our experimental studies on high-dimensional textual data sets, we demonstrate that the proposed algorithm achieves higher accuracy and is more robust than similar algorithms using pairwise constraints for supervision.