We consider the acceleration of charged particles near ultrarelativistic shocks, with Lorentz factor . We present simulations of the acceleration process and compare these with results from semi‐analytical calculations. We show that the spectrum that results from acceleration near ultrarelativistic shocks is a power law, , with a nearly universal value for the slope of this power law. We confirm that the ultrarelativistic equivalent of the Fermi acceleration at a shock differs from its non‐relativistic counterpart by the occurrence of large anisotropies in the distribution of the accelerated particles near the shock. In the rest frame of the upstream fluid, particles can only outrun the shock when their direction of motion lies within a small loss cone of opening angle around the shock normal. We also show that all physically plausible deflection or scattering mechanisms can change the upstream flight direction of relativistic particles originating from downstream by only a small amount: . This limits the energy change per shock crossing cycle to , except for the first cycle where particles originate upstream. In that case the upstream energy is boosted by a factor for those particles that are scattered back across the shock into the upstream region.

Prolific electron-positron pair production is possible at laser intensities approaching 10;{24} W cm;{-2} at a wavelength of 1 mum. An analysis of electron trajectories and interactions at the nodes (B=0) of two counterpropagating, circularly polarized laser beams shows that a cascade of gamma rays and pairs develops. The geometry is generalized qualitatively to linear polarization and laser beams incident on a solid target.

We extend the eigenfunction method of computing the power-law spectrum of particles accelerated at a relativistic shock front to apply to shocks of arbitrarily high Lorentz factor. In agreement with the findings of Monte-Carlo simulations, we find the index of the power-law distribution of accelerated particles which undergo isotropic diffusion in angle at an ultrarelativistic, unmagnetized shock is s = 4.23 ± 0.01 (where s = −d ln f /d ln p with f the Lorentz invariant phase-space density and p the momentum). This corresponds to a synchrotron index for uncooled electrons of α = 0.62 (taking cooling into account α = 1.12), where α = −d ln F ν /d ln ν, F ν is the radiation flux and ν the frequency. We also present an approximate analytic expression for the angular distribution of accelerated particles, which displays the effect of particle trapping by the shock: compared with the non-relativistic case the angular distribution is weighted more towards the plane of the shock and away from its normal. We investigate the sensitivity of our results to the transport properties of the particles and the presence of a magnetic field. Shocks in which the parameter σ (the ratio of Poynting to kinetic energy flux) upstream is not small are less compressive and lead to larger values of s.

It is generally thought that most of the spin-down power of a pulsar is carried away in an MHD wind dominated by Poynting Ñux. In the case of an oblique rotator, a signiÐcant part of this energy can be considered to be in a low-frequency wave, consisting of stripes of a toroidal magnetic Ðeld of alternating polarity propagating in a region around the equatorial plane. Magnetic reconnection in such a structure has been proposed as a mechanism for transforming the Poynting Ñux into particle energy in the pulsar wind. We have reexamined this process and conclude that the wind accelerates signiÐcantly in the course of reconnection. This dilates the timescale over which the reconnection process operates so that the wind requires a much larger distance than was previously thought in order to convert the Poynting Ñux to particle Ñux. In the case of the Crab pulsar, the wind is still Poynting-dominated at the radius at which a standing shock is inferred from observation. An estimate of the radius of the termination shock for other pulsars implies that all except the millisecond pulsars have Poynting ÑuxÈdominated winds all the way out to the shock front.

In simulations of a 10PW laser striking a solid we demonstrate the possibility of producing a pure electron-positron plasma by the same processes as those thought to operate in high-energy astrophysical environments. A maximum positron density of 10 26 m −3 can be achieved, seven orders of magnitude greater than achieved in previous experiments. Additionally, 35% of the laser energy is converted to a burst of gamma-rays of intensity 10 22 Wcm −2 , potentially the most intense gammaray source available in the laboratory. This absorption results in a strong feedback between both pair and γ-ray production and classical plasma physics in the new 'QED-plasma' regime.Electron-positron (e − e + ) plasmas are a prominent feature of the winds from pulsars and black holes [1,2]. They result from the presence of electromagnetic fields strong enough to cause non-linear quantum electrodynamics (QED) reactions [3] in these environments leading to a cascade of e − e + pair production [4]. These fields can be much lower than the Schwinger field for vacuum breakdown [5] if they interact with highly relativistic electrons (γ >> 1) [3]. Non-linear QED has been probed experimentally with lasers in two complementary ways:(1) with a particle accelerator accelerating electrons to the necessary γ and a laser supplying the fields [6-8]; or (2) with a laser accelerating the electrons and goldnuclei supplying the fields [9][10][11]. An alternative configuration, using next-generation high-intensity lasers to provide both the acceleration and the fields [12], has the potential to generate dense e − e + plasmas. Analytical calculations and simulations exploring this configuration have shown that an overdense e − e + plasma can be generated from a single electron by counter-propagating 100PW lasers [12][13][14][15]. Here we will show that such a plasma can be generated with an order of magnitude less laser power by firing the laser at a solid target, putting such experiments in reach of next-generation 10PW lasers [16].The dominant non-linear QED effects in 10PW laserplasma interactions are: synchrotron gamma-ray photon (γ h ) emission from electrons in the laser's electromagnetic fields; and pair-production by the multiphoton Breit-Wheeler process, γ h + nγ l → e − + e + , where γ l is a laser photon [3,17,18]. Each reaction is a strongly multiphoton process, the former process being non-linear Compton scattering, e − + mγ l → e − + γ h [19,20], in the limit m → ∞. Therefore, these reactions only become important at the ultra-high intensities reached in 10PW laser-plasma interactions. The importance of synchrotron emission is determined by the parameter η. This depends on the ratio of the electric and magnetic fields in the plasma to the Schwinger field [5] (E s = 1.3 × 10 18 Vm −1 ). For ultra-relativistic particles 17,18]. γ is the Lorentz factor of the emitting electron or positron, β is the corresponding velocity normalised to c and E ⊥ is the electric field perpendicular to its motion. As η approaches unity each emitted photon takes a ...