Exploring novel target schemes for laser wakefield accelerators is essential to address the challenge of increasing repetition rates while ensuring the stability and quality of the produced electron beams. This paper introduces and discusses the prototyping of a two-chamber gas target, integrated into the beamline and operating under continuous gas flow, in the framework of ionization injection. We present the numerical fluid modeling employed to assist the density profile shaping, with a focus on gas mixing and dopant confinement. The importance of localized high-Z gas for ionization injection is demonstrated through particle-in-cell simulations using the simulated gas profiles. We describe the test bench used for prototype evaluation, specifically addressing the plasma electron density and the longitudinal distribution of species relevant to ionization injection. The lifetime of the target at 10 Hz and 60 mJ is measured for different materials, and its effect on the resulting electron beam is assessed using particle-in-cell simulations. Finally, we outline perspectives on high-power operation.

Designing a high-quality plasma injector electron source driven by a laser beam relies on numerical parametric studies using particle-in-cell codes. The common input parameters to explore are laser characteristics, plasma species and density profiles produced by computational fluid dynamics studies. We demonstrate the construction of surrogate models using machine learning techniques for a laser-plasma injector (LPI) based on more than $3000$ particle-in-cell simulations of laser wakefield acceleration performed for sparsely spaced input parameters published by Drobniak et al.[Phys. Rev. Accel. Beams, 26, 091302, (2023)]. The surrogate model developed in this article approximates a nonlinear mapping $\mathbb{R}^4 \to \mathbb{R}^4$. Surrogate models are highly relevant to the design and optimisation of LPI systems, as they enable the rapid mapping of a defined parameter hyperspace at a computational cost that is negligible compared to iterative PIC simulations. Their speed enables more efficient design studies by allowing extensive exploration of the input-output relationship without significant computational cost. We develop and compare the performance of three surrogate models, namely, multilayer perceptron (MLP), decision trees (DT) and Gaussian processes (GP). We show that using a simple and frugal MLP-based model trained on a reasonably-sized random scan data set of 500 particles in cell simulations, we can predict beam parameters with a coefficient of determination score $R^2=0.93$ . The best surrogate model is used to quickly find optimal working points and stability regions and to achieve targeted electron beam energy, charge, energy spread and emittance using different methods, namely random search, Bayesian optimisation and multi-objective Bayesian optimisation. This simple approach can serve more global design study of an LPI in a start-to-end linear laser-driven accelerator.

The present article reports on the generation of stable 50 pC low-divergence electron beams above 150 MeV from laser-driven wakefield acceleration using a continuous-flow gas target prototype tested at the 60 TW Salle Jaune facility at LOA. The gas target design is meant to be easily transported and integrated as an element of the beamline with a differential pumping system offering some 10⁻⁴ mbar pressure in the rest of the line. A dedicated gas injection system allows for the control of the gas mixture concentration and gas pressure in two different regions of the target within the frame of controlled ionisation injection schemes. The measured electron beam parameters show the importance of gas density profiles and longitudinal gas mixture confinement.