Search

International audience; Ultrafast imaging requires accurate synchronization of pump and probe pulses, as well as the removal of anypulse front tilt of the probe pulse. Indeed, any pulse front tilt is highly detrimental to the accuracy of the imagingtechnique. In addition, the ”zero delay” between pump and probe is only true for a specific propagation distance.This is because along the path of pump and probe beams, the dispersion strongly modifies the retardance betweenthe pulses. As an example, for pump and probe beams in a collinear configuration, the propagation through 100 µmthick glass slide is enough to induce a shift by more than 37 fs between 800 nm and 400 nm pulses because of thedifference in terms of refractive index. However, it is extremely difficult to characterize weak probe pulses.

The analogy existing between spatial and temporal optics has motivated many studies to interpret spatial phenomena in the domain of ultrafast optics [1] . Indeed, dispersion and one-dimensional diffraction may share the same mathematical formalism. Temporal analogues of common optical systems have been proposed such as lenses, imaging systems, diffraction gratings, which opens up a whole range of new possibilities for ultrafast photonics.

International audience; The linear or nonlinear optical propagation behavior lengthwise of micro- and nano-waveguides in operation is difficult to evaluate experimentally, although this information is crucial for their optimal development. Previously, apart from the destructive cut-back method, far-field qualitative imaging of Rayleigh scattering (RS) or optical backscatter reflectometry (OBR) technique in infrared wavelength (the state-of-art technique in terms of spatial resolution) have been used for rapid analysis of the propagation of light in optical waveguides (optical nanofiber ONF in Refs. [1] , [2] ). In the present work, we demonstrate a novel method, nondestructive and non- invasive, with a high spatial (lower than 5 µm) and spectral resolution, to analyze the light propagation in an optical waveguide in operation in linear or nonlinear regime. By measuring with a confocal spectrometer the RS radiated out of an ONF in operation, we show that our method can spatially characterize the entire waveguide in the linear regime with a micrometric spatial resolution. This approach provides direct observation of the higher-order modes (HOMs) leaking from the ONF surface along the taper region and the transition from core–cladding to cladding– air guidance. This method can also be used to characterize the linear loss and provides the localization of defects along an optical waveguide.

Fiber lasers emitting at wavelengths around 2 μm in continuous wave (CW) regime are promising sources of laser radiation for both civil (surgery, spectroscopy, lidar) and military applications as infrared countermeasures or laser weapons [1] . Optimizing the size, weight and power of lasers devoted to military activities is a major concern. The need of high power compact and robust laser sources leads us towards all-fiber solutions by using a single-oscillator architecture. In this contribution, we present a high power monolithic fiber laser source based on a Tm 3+ , Ho 3+ -codoped silica fiber emitting at around 2 µm in CW regime.

International audience; The interplay among the effects of dispersion, nonlinearity and gain/loss in optical fibres is a powerful tool to generate a broad range of pulse shapes with tuneable properties. Here we propose a method to optimise the systems parameters for a given pulse target. By reducing the system complexity and applying machine-learning strategies, we show that it is possible to efficiently identify the sets of parameters of interest. Two configurations are numerically investigated: pulse shaping in a passive normally dispersive fibre and pulse generation in a dual-pump nonlinear-amplifying-loop-mirror mode-locked fibre laser.

International audience; High-repetition-rate optical frequency-comb sources emitting picosecond pulses play important roles in variousscientific researches and industrial applications. Such ultrafast pulse sources are mostly generated in opticalcavities or microresonators. By means of the wavelength-conversion techniques, it is possible to transfer thecavity-based near-IR robust and compact sources to the mid-IR or to the visible wavelength regions [1-2], forwhich there is an increasing demand, for biophotonics and other applications. Here we demonstrate the generationof high-repetition-rate picosecond pulses in the visible wavelength range by using a fully optical cavity-lessconfiguration. First, we developed a tunable C-band picosecond pulse generator whose high-repetition-rate is alsotunable from 20 to 40 GHz. This near-IR optical cavity-less system makes use of high reliability componentsdeveloped for telecommunications [3]. Next, it was used to directly pump a nonlinear periodically poled ridgeLiNbO3 (PPLN) waveguide fabricated on silicon substrate [4]. This highly efficient χ(2)-type nonlinear material,which performs ultrafast-response of second harmonic generation (SHG), will offer us the opportunities to developfiber-coupled frequency doubling modules for fiber-integrated picosecond pulse sources in the visible. Figure 1(a)presents the experimental setup, where the generation of stable 40-GHz pulse trains is obtained through thenonlinear compression of an initial beat-signal in a cavity-less optical-fiber-based device. The initial sinusoidalbeating is generated by using a commercial LiNbO3 intensity modulator driven by a half repetition-rate externalRF clock and then amplified by means of Er-doped fiber amplifier. Moreover, we imposed a RF phase modulationto suppress Brillouin backscattering into the 2.2-km-long compression fiber. High-quality 6-ps Gaussian pulses ata repetition rate of 40 GHz are then obtained with an average power of 400 mW at the fiber output as shown inFig. 1(c1-1). The corresponding spectrum is reported in Fig. 1(b1) with a FWHM bandwidth about 100 GHz. Thispulse source is then injected into the fundamental mode of our PPLN waveguide by means of a lensed fiber. Thecenter wavelength and state of polarization of the near-IR pulse train is set to match the optimum SHG conversionof the PPLN waveguide whose temperature is stabilized near room temperature. The 20mm long SHG waveguidegives a normalized conversion coefficient of 40%/W. After beam collimation at the waveguide output, an opticalprism and an optical diaphragm are used to efficiently filter out the SHG signal and reject the residual pump andspectra generated by other nonlinear processes. The SHG signal is then collected into a multimode fiber andanalyzed by means of an optical spectrum analyzer (~15 GHz resolution) and our 45-GHz photodiode –oscilloscope system. As shown in Fig. 1(b2), the SHG spectrum exhibits a 40-GHz frequency-comb profilecentered at ~771 nm whose bandwidth is ~100 GHz. The SHG average power is measured to be about 30 mW.The temporal profile of our 40-GHz pulse train converted into the visible is shown in Fig. 1(c2) and exhibits a 40-GHz sinusoidal waveform due to the limitations of our detection system. The same behaviour and pulse durationis obtained when applying the same detection device to the measurement of our initial near-IR picosecond pulsetrain (see Fig. 1(c1-2)).

Recent experiments have shown that nonlinear wave propagation in multimode optical fibers leads to complex spatio-temporal phenomena. In this talk, we introduce new approaches for the control and optimization of nonlinear beam reshaping in the spatial, temporal and spectral dimensions. The first approach applies to spatial beam self-cleaning the technique of transverse wavefront shaping, which permits to launch an optimized input mode combination, that results in the stable generation of a whole nonlinear mode alphabet at the fiber output. The second approach introduces a longitudinal tapering of the core diameter of multimode active and passive fibers, which permits to generate ultra-wideband and high brightness supercontinuum light, featuring high spatial beam quality.

We record a sub-wavelength THz image of a caster sugar grain by performing optical rectification of a femtosecond laser pulse in the grain and scanning the grain surface. We obtain a spatial resolution equal to 50 μm, i.e. ~ λ THz / 12.

We report on our recent experimental and theoretical results on the use of a sinusoidally time-varying phase to suppress undesirable self-phase modulation of optical pulses propagating in fibre-optic systems.

Over the past years, an increasing amount of techniques and methods for control and measure the polarization state of the THz electric field has been developed [1], [2]. These efforts were mostly motivated by the growing number of applications in imaging and materials characterization for security implementations. The systems found in the literature are mainly based on classical THz time domain spectrometer (THz-TDS) with modifications in their setting up. Some solutions to control and detect the polarization state of the THz signal are achieved using metamaterials [2], two laser pumping [3], or electro-optic modulator. These methods present the disadvantage of using physical devices along the THz path or moving parts. These techniques are complex and produce experimental errors.

Using laser ablation method, we taper photonic crystal fiber (PCF) with keeping its inner structure well and then we achieved the TPG where the phase matching point PW is 512nm and THG is 1536nm with effective refractive index is 1.1314. Finally we calculate the evolution of the conversion efficiency as a function of the fiber length.

We overview recent advances in the spatio-temporal nonlinear dynamics of optical pulses propagating in multimode optical fibers. The Kerr effect leads to spatial beam self-cleaning in a graded-index multimode optical fiber, followed by sideband series generation spanning multiple octaves. Effectively single mode supercontinuum spanning from the visible to the mid-infrared was also demonstrated. Enhancement of Kerr beam self-cleaning was observed in active fiber with quasi-step index profile. Moreover, mutual self-cleaning was recently reported for both the fundamental and the second harmonic beams in optically poled multimode fibers with cubic and quadratic nonlinearity.

The scattering process occurring when a complex medium is illuminated with a monochromatic light is well known to generate polarization effects on the scattered field which are directly related to the scattering origins, and the scattering regime.

International audience

Unipolar forms of orthogonal frequency division multiplexing (OFDM) such as DC biased optical OFDM (DCO-OFDM) and asymmetrically clipped optical OFDM (ACO-OFDM) are widely used in intensity-modulated and direct detection (IM/DD) systems. At low data rate, ACO-OFDM is more efficient in terms of optical power but it suffers from spectral inefficiency. For large constellations, DCO-OFDM with large DC bias has more optical power efficiency, since it requires lower constellation size than ACO-OFDM. However, the required DC bias to obtain an acceptable clipping noise degrades the optical power efficiency. To overcome these drawbacks, we propose an optimized DCO-OFDM technique capable to transmit large constellations with a moderate DC bias. To reduce the clipping impact, a linear companding function is used in order to compress the negative part of the bipolar signal and therefore to reduce the amount of the clipped peaks and the loss of information. We demonstrate that the proposed technique has better bit error rate and optical power performances than conventional DCO-OFDM and ACO-OFDM. For a bit rate/normalized bandwidth of 6, a gain of 4dB in optical power is reached.

This paper revisits the way of performing depolarization measurements using a novel concept of field orthogonality breaking, providing a direct characterization of the depolarization strength of a material from a single measurement. This technique, inspired by a microwave photonics approach, uses a superposition of two orthogonally polarized modes with a slight frequency detuning Δν to enlighten the sample. Using such illumination, it can be shown that the depolarizing strength of a sample can be directly retrieved from a single measurement of the beat note intensity, appearing at the frequency Δν. In other words, the depolarization strength of a given material can be recovered from the way this material breaks the orthogonality between two incident orthogonal states. In addition, it must be noted that this technique is well suited for endoscopic remote sensing, being inherently insensitive to fiber/endoscope birefringences which preserve polarimetric orthogonality.

Summary form only given. The generation of optical supercontinua in the mid-infrared region and especially their expansion beyond the intrinsic limit dictated by fused silica is currently a subject of high interest. Tellurite and chalcogenide glasses have serious advantages because of their wide transmittance window which can reach more than 10 μm while the Kerr nonlinearity can be 500 times stronger than fused silica. These different features make them serious candidates for broad mid-infrared supercontinuum generation. For example, supercontinuum as broad as 4000-nm bandwidth has been generated in a sub-cm long Tellurite microstructured fiber by Domachuk et al. in ref. [1] by means of a femtosecond regime pumping. In this work, two kinds of soft-glasses, Tellurite and As2S3 chalcogenide suspended-core fibers with nearly 3-μm core diameter have been designed and pumped in their anomalous dispersion regime by means of an optical parametric oscillator delivering 200-fs pulses between 1700 and 2500 nm [2]. Figure 1a shows the resulting supercontinua obtained at the output of a 40-cm long sample of our tellurite fiber characterized by a 3.4-μm triangular core size. The nonlinear coefficient was evaluated to γ = 175 W-1.km-1, linear losses to 1.5 dB/m at 1550 nm and the zero dispersion wavelength (ZDW) was measured near 1660 nm. The pump wavelength was fixed to 1745 nm and resulting spectra have been recorded as a function of injected power. The broadest 112-mW supercontinuum is characterized by a 2000-nm bandwidth, corresponding to almost 2 octaves and is remarkably flat (a 1900-nm span contained in a -20 dB range). Figure 1b presents the corresponding numerical simulations, which reproduce quite well the dynamics highlighted during our experiments [2]. Figure 2a illustrates the experimental results obtained from a 45-mm long sample of our 3.2-μm suspended-core chalcogenide fiber. The nonlinear Kerr coefficient was estimated to U = 1175 W-1.km-1, linear losses to 1 dB/m at pump wavelength and a ZDW close to 2330 nm. Figure 2a shows the resulting spectrum for a pump wavelength of 2300 nm and an injected power of 70 mW. The supercontinuum extends from 1200 nm to 3200 nm in the -20 dB range. Note however that large residual OH absorption bands are clearly visible around 2.9 μm and beyond 3.2 μm and thus limiting supercontinuum expansion. Taking into account these extra losses, the corresponding numerical simulations illustrated in Fig. 2b are in good agreement with our experimental recordings.

Summary form only given. Frequency-shifted feedback was recently used to synchronize the beat note between the two modes of a dual-frequency laser with the radiofrequency emitted by a local microwave oscillator. This technique allows one to obtain an optically-carried RF clock signal that can find applications in microwave photonics. Two regimes of synchronization are then found [1]. When the frequency difference Δν between the two oscillators is smaller than a frequency fA determined by the feedback amount, the two oscillators are phase-locked, leading to a constant relative phase φ. Then, a regime of frequency locking without phase locking appears. This regime, also called bounded phase [2,3], extends the frequency synchronization beyond fA up to a frequency fB. Finally, when Δν is larger than fB, the two oscillators are not synchronized, resulting in an unbounded drifting phase. Here, we experimentally characterize the three regimes and compare the results to numerical simulations. Our dual-frequency Nd:YAG laser emitting at 1064 nm, is detailed in [1]. The laser cavity contains two quarter-wave plates in order to enforce dual-frequency operation with an adjustable beating frequency Qy.Qx between two linearly polarized x and y modes. We choose Qy. Qx around 200 MHz. The feedback arm contains a Bragg cell driven at fAO and a quarter-wave plate that flips the x polarization into a y polarization after reflection on the feedback mirror. The optical frequency of the reinjected beam is thus shifted by 2fAO. Hence the laser dynamics is governed by a third frequency Qx + 2fAO, and 'Q is equal to Qy - Qx - 2fAO. The figure shows the so-called phasor plots [3] and the corresponding spectra. It evidences the three regimes, i.e., locked (in which the relative phase I between the beat note and the local oscillator is fixed), bounded (in which I oscillates in a bounded interval) and unbounded (in which I grows indefinitely). By using a Lang-Kobayashi-type model, the phase evolution can be studied mathematically. In particular, using asymptotic methods, the transition between the locked and the bounded regime is found to be associated to a bifurcation. Conversely, the transition from the bounded to the unbounded regime is not a bifurcation, as formerly stated [2]. This is highlighted by the power spectra (e) and (f) which are similar. However, by measuring the phase noise spectra, we have found experimentally that the bounded regime has the same quality, in terms of spectral purity as the phase locked regime. In both cases the long-term phase stability of the local oscillator is reported on the beat note. Conversely, in the unbounded regime, the phase stability of the beat note is essentially that of the freerunning laser. This is confirmed by numerical simulations and demonstrates that bounded-phase regime truly enlarges the locking range. We are now extending our studies to injected semiconductor lasers, where boundedphase is also observed [3].

We developed an approach to design radiation-hardened rare earth -doped fibers and amplifiers. This methodology combines testing experiments on these devices with particle swarm optimization (PSO) calculations. The composition of Er/Yb-doped phosphosilicate fibers was improved by introducing Cerium inside their cores. Such composition strongly reduces the amplifier radiation sensitivity, limiting its degradation: we observed a gain decreasing from 19 dB to 18 dB after 50 krad whereas previous studies reported higher degradations up to 0°dB at such doses. PSO calculations, taking only into account the radiation effects on the absorption efficiency around the pump and emission wavelengths, correctly reproduce the general trends of experimental results. This calculation tool has been used to study the influence of the amplifier design on its radiation response. The fiber length used to ensure the optimal amplification before irradiation may be rather defined and adjusted to optimize the amplifier performance over the whole space mission profile rather than before integration in the harsh environments. Both forward and backward pumping schemes lead to the same kind of degradation with our active fibers. By using this promising coupled approach, radiation-hardened amplifiers nearly insensitive to radiations may be designed in the future.

This paper reviews important parameters for the two-photon absorption (TPA) laser stimulation technique and presents results concerning the characterization of the TPA effective spot size along lateral and axial directions. TPA scans are compared with classical photoelectric stimulation images to investigate TPA capabilities for failure analysis and design debug.

International audience; The depth of the Single-event burnout sensitive volume of power MOSFETs is investigated using TCAD simulation and TPA laser testing approaches.

This paper first describes the AM-FM demodulation of an arterial pressure signal. Although it is known to be efficient on signals modulated by breathing, we demonstrate that in case of lack of respiratory modulation (real or simulated central sleep apnea), the AM-FM algorithm doesn’t perform well in heart rate extraction. We introduce then a new algorithm based on Singular Spectrum Analysis eigen values which performs better cardiac frequency estimation in this context. Respiratory estimation is possible but is beyond the scope of this paper. The error for cardiac frequency estimation is around 0.2 BPM (Beats Per Minute) versus 5.5 BPM for the AM-FM demodulation. Further experimentations will be performed (with this time both cardiac and respiratory assessments) and will deal with real sleep apnea cases.

The impact of device temperature variations on the modulation response of an injection-locked Quantum-Dash laser is analyzed. Lower slave operating temperatures result in a large reduction in the undesirable pre-resonance dip in the modulation resp onse.

We report a theoretical analysis and experimental demonstration of a polarization attraction process at telecommunication wavelengths in isotropic optical fibers. The combined effects of polarization attraction and Raman amplification are also presented.

We experimentally demonstrate the use of an hybrid configuration to generate parabolic pulses. We combine dispersion decrease with distributed gain. This leads to several benefits on the parabolic generated pulses compared with a passive configuration.

To generate ultrashort pulses with higher energy in passively mode-locked fiber lasers, several groups have recently focused on the development of laser cavities operating in normal path-averaged dispersion, with a view to achieve wave-breaking-free operation. The mode-locking mechanism in fiber laser play a key role in providing scalability limitation: for instance, nonlinear polarization evolution can be overdriven. Furthermore, the differential gain of the doped fiber decreases along with the traveled distance. From all of these limitations, it follows that power scalability is always bounded for virtually any mode-locked laser cavity design. In general, the consequence is the emergence of multiple pulsing operation for high pumping powers. Interactions between these pulses could lead to the formation of stable bound pulses, although the pulses are highly chirped and energetic. The present study follows the experimental report of "parabolic bound pulses" in a high-power ytterbium-doped fiber laser operating in the normal dispersion regime. We explained through detailed numerical simulations how such bound pulses can be observed in a regime where wave-breaking-free operation was initially expected. The vectorial laser model takes explicitly into account gain saturation, finite amplification bandwidth, and nonlinear polarization evolution, in addition to self-phase modulation, group velocity dispersion and linear birefringence. We show that, for a given pumping power, the propagation regime can indeed appear like a "wave-breaking-free" regime, with only one minimum of the pulse duration per roundtrip. However, increasing the pumping power, pulse splitting finally occurs and leads to the stable binding of two highly-chirped, energetic pulses. The influence of gain filtering and total averaged cavity dispersion is analyzed in detail.

Summary form only given. Polymer-based photonic crystals have some advantages over their high refractive index counterparts in terms of: out-of-plane losses, low fabrication cost and optical non-linearities. Some of its disadvantages are the stability of the polymers and the unlikely development of a full gap. We present the status of our experimental and or theoretical work on waveguides, resonators and WDM elements, the latter based on ultrarefractive phenomena.