X-rays are produced when laser-wakefield accelerated electrons oscillate in the transverse electrostatic field of the accelerating structure. The measured characteristics of these betatron x-rays follow scaling laws relating them to the electron energy, charge, plasma density, and other observables. Here we report on the x-rays produced by electrons accelerated to energies >1 GeV and investigate the scaling laws for photon number, critical energy, and beam divergence.

Previous experiments and modeling examining the impact of an underdense, pre-formed plasma in laser-plasma interactions have shown that the fast electrons are generated with energies higher than predicted by ponderomotive scaling [4, 3-14]. We report on experiments using the Texas Petawatt high intensity (150 fs, 1.5 x 10(20) W cm(-2)) laser pulse, which were conducted to examine the mechanism for accelerating these high energy electrons. These experiments gauge the impact a controlled low density pre-formed plasma has on electron generation with a shorter time scale than previous experiments, 150-180 fs. Electron temperatures measured via magnetic spectrometer on experiment were found to be independent of preformed plasma. Supplemental computational results using 1D PIC simulations predict that super-ponderomotive electrons are generated inside a potential well in the pre-plasma [1]. However, while the potential well is established around 150 fs, the electrons require at least an additional 50 fs to be trapped and heated inside it.

We have completed a pulse contrast upgrade on the Texas Petawatt Laser. This improvement enables the use of thin and reduced mass targets for ion acceleration, and reduces pre-plasma effects on all experiments. The new design starts with two BBO-based OPCPA stages pumped by an optically synchronized 8-ps laser. These stages amplify slightly chirped few ps pulses by six orders of magnitude. Next there are two LBO-based OPCPA stages that are pumped by 4 ns pulses. With much less gain than before, parametric fluorescence has been reduced by about three orders of magnitude. Prior to the upgrade, lenses caused pencil beam prepulses. Since tilting or wedging lenses was not a viable option, we replaced all lenses in the glass amplifiers with off-axis parabolic mirrors. There are still weak prepulses that we attribute to surface scattering. We eliminated thin transmissive optics to avoid post pulses that would result in prepulses by nonlinear (B-integral) conversion. This required us to reduce from eight to four passes in the 64-mm glass amplifier and to add a two-pass 25-mm ``booster amplifier.'' As a final upgrade we added an Acousto-Optic Programmable Dispersive-Filter (AOPDF) to improve higher order dispersion and steepen the rising edge of the compressed pulse.

In this work, we present a new and general method for measuring the astrophysical S factor of nuclear reactions in laser-induced plasmas and we apply it to H-2(d, n)He-3. The experiment was performed with the Texas Petawatt Laser, which delivered 150-270 fs pulses of energy ranging from 90 to 180 J to D-2 or CD4 molecular clusters (where D denotes H-2). After removing the background noise, we used the measured time-of-flight data of energetic deuterium ions to obtain their energy distribution. We derive the S factor using the measured energy distribution of the ions, the measured volume of the fusion plasma, and the measured fusion yields. This method is model independent in the sense that no assumption on the state of the system is required, but it requires an accurate measurement of the ion energy distribution, especially at high energies, and of the relevant fusion yields. In the H-2(d, n) He-3 and He-3(d, p)He-4 cases discussed here, it is very important to apply the background subtraction for the energetic ions and to measure the fusion yields with high precision. While the available data on both ion distribution and fusion yields allow us to determine with good precision the S factor in the d + d case (lower Gamow energies), for the d + He-3 case the data are not precise enough to obtain the S factor using this method. Our results agree with other experiments within the experimental error, even though smaller values of the S factor were obtained. This might be due to the plasma environment differing from the beam target conditions in a conventional accelerator experiment.

The irradiation of few-nm-thick targets by a finite-contrast high-intensity short-pulse laser results in a strong pre-expansion of these targets at the arrival time of the main pulse. The targets decompress to near and lower than critical densities with plasmas extending over few micrometers, i.e. multiple wavelengths. The interaction of the main pulse with such a highly localized but inhomogeneous target leads to the generation of a short channel and further self-focusing of the laser beam. Experiments at the Glass Hybrid OPCPA Scaled Test-bed (GHOST) laser system at University of Texas, Austin using such targets measured non-Maxwellian, peaked electron distribution with large bunch charge and high electron density in the laser propagation direction. These results are reproduced in 2D PIC simulations using the EPOCH code, identifying direct laser acceleration (DLA) [1] as the responsible mechanism. This is the first time that DLA has been observed to produce peaked spectra as opposed to broad, Maxwellian spectra observed in earlier experiments [2]. This high-density electrons have potential applications as injector beams for a further wakefield acceleration stage as well as for pump-probe applications. Copyright (C) 2016 Science and Technology Information Center, China Academy of Engineering Physics. Production and hosting by Elsevier B.V.

We generate a collimated (< 1mrad) beam of gamma rays that penetrate several centimeters of lead by retro-reflecting a petawatt laser pulse onto trailing similar to 2 GeV electrons with a plasma mirror after driving a laser-plasma accelerator.

We report on experiments irradiating isolated plastic spheres with a peak laser intensity of 2-3 x 10(20) W cm(-2). With a laser focal spot size of 10 mu m full width half maximum (FWHM) the sphere diameter was varied between 520 nm and 19.3 mu m. Maximum proton energies of similar to 25 MeV are achieved for targets matching the focal spot size of 10 mu m in diameter or being slightly smaller. For smaller spheres the kinetic energy distributions of protons become nonmonotonic, indicating a change in the accelerating mechanism from ambipolar expansion towards a regime dominated by effects caused by Coulomb repulsion of ions. The energy conversion efficiency from laser energy to proton kinetic energy is optimized when the target diameter matches the laser focal spot size with efficiencies reaching the percent level. The change of proton acceleration efficiency with target size can be attributed to the reduced cross-sectional overlap of subfocus targets with the laser. Reported experimental observations are in line with 3D3V particle in cell simulations. They make use of well-defined targets and point out pathways for future applications and experiments.

A system of two collinear probe beams with different wavelengths and pulse durations was used to capture simultaneously snapshot interferograms and streaked interferograms of laser produced plasmas. The snapshots measured the two dimensional, path-integrated, electron density on a charge-coupled device while the radial temporal evolution of a one dimensional plasma slice was recorded by a streak camera. This dual-probe combination allowed us to select plasmas that were uniform and axisymmetric along the laser direction suitable for retrieving the continuous evolution of the radial electron density of homogeneous plasmas. Demonstration of this double probe system was done by measuring rapidly evolving plasmas on time scales less than 1 ns produced by the interaction of femtosecond, high intensity, laser pulses with argon gas clusters. Experiments aimed at studying homogeneous plasmas from high intensity laser-gas or laser-cluster interaction could benefit from the use of this probing scheme. (C) 2016 AIP Publishing LLC.

In this work, we explore the possibility that the motion of the deuterium ions emitted from Coulomb cluster explosions is highly disordered enough to resemble thermalization. We analyze the process of nuclear fusion reactions driven by laser-cluster interactions in experiments conducted at the Texas Petawatt laser facility using a mixture of D-2+He-3 and CD4+He-3 cluster targets. When clusters explode by Coulomb repulsion, the emission of the energetic ions is ``nearly'' isotropic. In the framework of cluster Coulomb explosions, we analyze the energy distributions of the ions using a Maxwell-Boltzmann (MB) distribution, a shifted MB distribution (sMB), and the energy distribution derived from a log-normal (LN) size distribution of clusters. We show that the first two distributions reproduce well the experimentally measured ion energy distributions and the number of fusions from d-d and d-He-3 reactions. The LN distribution is a good representation of the ion kinetic energy distribution well up to high momenta where the noise becomes dominant, but overestimates both the neutron and the proton yields. If the parameters of the LN distributions are chosen to reproduce the fusion yields correctly, the experimentally measured high energy ion spectrum is not well represented. We conclude that the ion kinetic energy distribution is highly disordered and practically not distinguishable from a thermalized one.