In supercontinuum generation, various propagation effects combine to produce a dramatic spectral broadening of intense ultrashort optical pulses. With a host of applications, supercontinuum sources are often required to possess a range of properties such as spectral coverage from the ultraviolet across the visible and into the infrared, shot-to-shot repeatability, high spectral energy density and an absence of complicated pulse splitting. Here we present an all-in-one solution, the first supercontinuum in a bulk homogeneous material extending from 450 nm into the mid-infrared. The spectrum spans 3.3 octaves and carries high spectral energy density (2 pJ nm−1–10 nJ nm−1), and the generation process has high shot-to-shot reproducibility and preserves the carrier-to-envelope phase. Our method, based on filamentation of femtosecond mid-infrared pulses in the anomalous dispersion regime, allows for compact new supercontinuum sources.

Attosecond light pulses in the extreme ultraviolet have drawn a great deal of attention due to their ability to interrogate electronic dynamics in real time. Nevertheless, to follow charge dynamics and excitations in materials, element selectivity is a prerequisite, which demands such pulses in the soft X-ray region, above 200 eV, to simultaneously cover several fundamental absorption edges of the constituents of the materials. Here, we experimentally demonstrate the exploitation of a transient phase matching regime to generate carrier envelope controlled soft X-ray supercontinua with pulse energies up to 2.9±0.1 pJ and a flux of (7.3±0.1) × 107 photons per second across the entire water window and attosecond pulses with 13 as transform limit. Our results herald attosecond science at the fundamental absorption edges of matter by bridging the gap between ultrafast temporal resolution and element specific probing.

We report on the first X-ray absorption fine structure (XAFS) and near edge X-ray absorption fine structure (NEXAFS) spectroscopy are well established methods for retrieving structural information about the composition of solid state materials and soft matter. The water window spectral range between 284 eV and 543 eV is of special interest as it contains the K-shell absorption edges of the biological building blocks: carbon (284 eV), nitrogen (410 eV) and oxygen (543 eV). Up until recently only facility scale light sources have been capable of generating coherent water window radiation: synchrotrons with a high degree of spatial coherence and hundreds of femtoseconds pulse durations, and X-ray free electron lasers with a high degree of spatial coherence and femtosecond temporal resolution [1]. High harmonic generation (HHG) [2, 3] offers an attractive alternative approach since it is realizable on a small table-top scale and is capable of generating fully coherent radiation, i.e. femto-to atto-second and possibly even zeptosecond pulse durations. The ability to generate coherent water window radiation from HHG is extremely exciting as it would bring ultra-short time resolution to structural probing with a table top method. HHG is most commonly driven by Ti:sapphire sources at 800 nm with the highest achievable photon energy, the so called cutoff, scaling linearly with the laser intensity and quadratically with the driving wavelength [4]. While the water window range is reachable with such sources via nonphase-matched HHG [5], the contradicting requirements of increasing the cutoff with higher laser intensity while avoiding excessive ionisation, severely limits the achievable flux in the water window. A solution to this dilemma is to use a source with a similar peak intensity and pulse duration, but at much longer emission wavelengths in order to exploit the quadratic wavelength scaling of the HHG cutoff. A drawback of such an approach is the unfavourable single atom response scaling of harmonic yield with λ −9 [6] which can however be mitigated, to a large extent, through high gaspressure phase matching [7]. This concept was demonstrated by reaching a 1.6 keV cutoff when driving with a mid-IR laser system [8]. Despite this cutting-edge result, the 20 Hz repetition rate and stability of the system have thus far proved insufficient for applications, thereby underlining the need for significant improvements of the laser parameters.We find that while high X-ray flux can be achieved through phase-matched HHG driven by kHz or higher repetition rate long-wavelength sources, achieving sufficient intensity and carrier to envelope phase (CEP) stability of the driver laser is an essential key both for producing attosecond pulses and for generating reproducible X-ray spectra from each laser pulse and throughhout an X-ray measurement.Currently at the kHz level and with long wavelength drivers, the lower end of the water window at 300 eV was reached using a Ti:sapphire pumped optical parametric amplifier (OPA) at 1.5 μm [9]...

We generate self-carrier-to-envelope phase-stable, 630 μJ pulses, centered at 2.1 μm, with 42 fs (6 cycle) duration based on collinear optical parametric amplification in BiB(3)O(6) at 3 kHz. These pulses are generated through a traveling wave amplifier scheme, and the bandwidth supports 28 fs (4 cycle) pulse duration. Carrier-to-envelope phase stability was measured to be 410 mrad over 10 min or 260 mrad over 35 s.

No abstract