We investigate the self-consistent particle acceleration physics associated with the development of the kink instability (KI) in nonrelativistic, electron-ion plasma jets. Using 3D fully kinetic particle-in-cell simulations, we show that the KI efficiently converts the initial toroidal magnetic field energy into energetic ions. The accelerated ions form a nonthermal power-law tail in the energy spectrum, containing 10% of the initial magnetic field energy, with the maximum ion energy extending to the confinement energy of the jet. We find that the ions are efficiently accelerated by the concerted action of the motional electric field and the highly tangled magnetic field that develop in the nonlinear phase of the KI: fast curvature drift motions of ions across magnetic field lines enable their acceleration along the electric field. We further investigate the role of Coulomb collisions in the ion acceleration efficiency and identify the collisional threshold above which nonthermal ion acceleration is suppressed. Our results reveal how energetic ions may result from unstable nonrelativistic plasma jets in space and astrophysics and provide constraints on the plasma conditions required to reproduce this acceleration mechanism in laboratory experiments.

Jets of plasma threaded by magnetic fields are ubiquitous in space and astrophysical environments. Throughout their evolution, these jets can be subject to current-driven magnetohydrodynamic (MHD) instabilities, with the most common being the sausage (m = 0) and kink (m = 1) modes.1 The development of such instabilities is believed to be a candidate mechanism to explain the conversion of the jet's magnetic energy into energetic particles and radiation in a wide range of space physics and astrophysics scenarios. These include the flaring activity and energetic particle generation in jets emanating from the solar atmosphere,2–4 the synchrotron emission from bright knots, and cosmic ray acceleration in jets from active galaxies,5–7 among others. These instabilities are also associated with violent disruptions and particle acceleration in laboratory fusion devices, such as Tokamaks8 and Z-pinches.9–11 Yet, the detailed physics underlying how the development of such instabilities in jets accelerates particles remains poorly understood.

MHD simulations play an important role in understanding where and when sausage and kink modes develop over the course of the formation and propagation of astrophysical6,12 and laboratory13 plasma jets and how their nonlinear development impacts the structural integrity of jets.14 However, the kinetic physics that underpins the acceleration of particles is not captured by MHD models. Simulations of test particle dynamics in the fields computed by MHD simulations have shown nonthermal acceleration15 but do not provide a self-consistent picture of the injection mechanisms or feedback of energetic particles on the plasma fields. Fully kinetic simulations are necessary to address this central problem in a self-consistent manner. Using 3D particle-in-cell (PIC) simulations, we have recently demonstrated the self-consistent acceleration of nonthermal particles resulting from the development of the kink instability (KI) in relativistic magnetized pair plasma and pair-proton jets.7 These simulations revealed that the KI provides a viable means of producing radiating nonthermal particle distributions and of accelerating ultrahigh-energy cosmic rays in the jets of active galactic nuclei. These findings encourage further exploration of the KI's ability to efficiently accelerate particles in nonrelativistic electron-ion plasma jets of relevance to a range of space and laboratory conditions.

In this paper, we report on 3D fully kinetic simulations of the KI in nonrelativistic, electron-ion plasma jets and discuss the associated particle acceleration physics. We demonstrate that the nonlinear development of the KI in collisionless plasma results in the efficient dissipation of the jet's magnetic field energy, which is preferentially transferred to high-energy ions. Approximately 10% of the initial magnetic field energy is converted into nonthermal ions. The ion energy spectrum develops a power-law tail εp, with p approximately fixed at 2.6 over the parameter range explored in this work. We discuss the ion acceleration mechanism and how it depends on the size and magnetization of the jet and how the mass ratio between electrons and ions dictates the energy partition between species. In addition, we explore how the plasma collisionality impacts particle acceleration, which is relevant for a variety of laboratory experiments. Previous experiments using plasma guns,16,17 modified plasma Z-pinch platforms,11,18–20 and high-power lasers21 have successfully studied MHD instabilities in jets. In some cases, the observation of nonthermal ions has been reported,11 yet without a clear understanding of the underlying acceleration mechanism. Our work helps establish the conditions for which the present acceleration mechanism can be studied in laboratory experiments.

This paper is organized as follows. In Sec. II, we describe the initial physical configuration of the plasma jet and our simulation setup. We then discuss the evolution of the electric and magnetic fields produced by the KI and the associated particle acceleration dynamics in the collisionless regime in Sec. III. The effects of Coulomb collisions on the particle acceleration dynamics are discussed in Sec. IV, and conclusions are drawn in Sec. V.

We simulate the particle acceleration dynamics associated with the development of the KI in a nonrelativistic electron-ion plasma jet using the 3D PIC code OSIRIS 4.0.22,23 We consider a purely toroidal magnetic field profile of the form Bϕ(r)=B0rRe1r/R, where B0 is the peak amplitude of the magnetic field and R is the characteristic radius of the jet. The current density J=c/4π×B is supported by symmetrically streaming electrons and ions along the jet axis. The thermal pressure profile of the plasma (P) is chosen to achieve hydromagnetic equilibrium, P=J×B; we consider plasma with uniform number density ne=ni=n0 and a nonuniform temperature profile with Te=Ti=T(r) such that P(r)=2n0kBT(r). Note that we have tested different initial conditions with jet current being carried solely by the electrons and no net flow of ions and have found that the particle acceleration dynamics remains unaffected. This is to be expected since the particle thermal velocities are significantly higher than the flow velocities required to support the electric current.

The jet's initial conditions can thus be characterized by the following parameters. A dimensionless measure of the jet's magnetization is ωci/ωpi, where ωci=eB0/mic is the ion cyclotron frequency and ωpi=4πn0e2/mi is the ion plasma frequency; the dimensionless jet radius is R/di, with di=c/ωpi being the ion inertial length; and the ion to electron mass ratio is mi/me. Note that due to pressure balance n0kBT0=B02/4π, we have that the typical Larmor radius of thermal ions at the core of the jet is given by ρ¯i=mikBT0c/eB0=di.

We explore jet magnetizations in the range of 0.08ωci/ωpi0.25 [corresponding to nonrelativistic magnetic energy densities σi(ωci/ωpi)2 in the range of 0.006–0.06], and we simulate jet radii in the range of R/di = 5–10. Given the high computational cost of simulating realistic ions to electron mass ratios, we perform simulations with reduced mass ratios mi/me between 4 and 36. By progressively increasing the mass ratio, we uncover how the particle acceleration physics scales with mi/me, allowing us to infer the behavior for realistic mass-ratio conditions.

The dimensions of our simulation domain are 20R×20R×10R, with the jet placed at the center of the domain and oriented along ẑ. The grid resolution is chosen to resolve the gyroradius of thermal electrons at the core of the jet ρ¯e=ρ¯ime/mi with >2 points, and we use 12 particles per cell per species with quadratic splines for the particle shapes. Finer spatial resolutions and more particles per cell were tested to ensure numerical convergence. Periodic boundary conditions are imposed in all directions; the transverse (x̂ and ŷ) dimensions of the domain are sufficiently large to avoid artificial recirculation of particles during the time scale of interest. We follow the evolution of the system until the particle acceleration dynamics terminates, which occurs at 2530R/vA(vA=B0/4πn0mi is the characteristic Alfvén speed).

The general dynamics of the KI in collisionless electron-ion plasma is illustrated in Fig. 1 with the case of ωci/ωpi=0.125(σi=0.016),mi/me=16, and R/di=5(=R/ρ¯i); all other simulations explored in this work exhibited similar behavior. The onset of the KI is triggered by small transverse displacements of the jet about its equilibrium. Those displacements produce imbalances in the magnetic pressure across the jet, inducing motions that reinforce the initial distortion. This ultimately manifests as a growing helical modulation of the jet structure (left column of Fig. 1), with a wavelength of 2.5R and developing at a rate of vA/R, in agreement with linear MHD theory.1 

FIG. 1.

Development of the kink instability in a collisionless, nonrelativistic electron-ion jet (with ωci/ωpi=0.125,R/di=5, and mi/me=16). (a) Current density, (b) magnetic field lines, and (c) axial electric field, taken at times (1) vAt/R=8 and (2) vAt/R=12.5. These times correspond to the linear and nonlinear stages of the kink instability. Note that a quarter of the simulation box has been removed in (b1), (b2), and (c2) to reveal the inner field structure of the jet.

FIG. 1.

Development of the kink instability in a collisionless, nonrelativistic electron-ion jet (with ωci/ωpi=0.125,R/di=5, and mi/me=16). (a) Current density, (b) magnetic field lines, and (c) axial electric field, taken at times (1) vAt/R=8 and (2) vAt/R=12.5. These times correspond to the linear and nonlinear stages of the kink instability. Note that a quarter of the simulation box has been removed in (b1), (b2), and (c2) to reveal the inner field structure of the jet.

Close modal

The radial motions of the magnetized plasma jet give rise to an inductive electric field E=(v/c)×B, which at early times is also shown to have a helical and harmonic structure in Fig. 1(c1). Similar to the behavior found in the relativistic pair plasma regime,7,24 the nonlinear distortions of the jet (distortions comparable to the jet radius R) result in the formation of a coherent structure in the axial component of the electric field Ez [Fig. 1(c2)]; the mean amplitude of the electric field along the jet axis is found to be Ezz0.3vAB0/c. The nonlinear distortions of the jet current density also result in a highly tangled magnetic field structure [Fig. 1(b2)]. It is this configuration of electric and magnetic fields that mediates the efficient conversion of magnetic energy [≃65% of initial magnetic energy εB(0)] into plasma kinetic energy [Fig. 2(a)]. This efficient accelerating field structure persists during the transit time of the kink perturbations across the diameter of the jet, τKI2R/vKI=6R/vA. After this period, the fields decay and particle acceleration ceases.

FIG. 2.

(a) Temporal evolution of the change in magnetic (ΔεB), electric (ΔεE), and particle kinetic energies (ΔεK) integrated over the simulation domain. The separate evolution of the change in electron (ΔεKe) and ion kinetic energies (ΔεKi) is also shown. Temporal evolution of the ion (b1) and electron (b2) energy spectra, integrated over the simulation domain. Note that the double hump structure of both ion and electron spectra at t = 0 corresponds to the contributions of the warm jet plasma (that balances the hoop stress of the magnetic field) and the cold ambient plasma outside the jet (also providing confinement of the jet). The initial energy distribution of the warm jet plasma is represented by the black dashed-dotted curves in (b1) and (b2).

FIG. 2.

(a) Temporal evolution of the change in magnetic (ΔεB), electric (ΔεE), and particle kinetic energies (ΔεK) integrated over the simulation domain. The separate evolution of the change in electron (ΔεKe) and ion kinetic energies (ΔεKi) is also shown. Temporal evolution of the ion (b1) and electron (b2) energy spectra, integrated over the simulation domain. Note that the double hump structure of both ion and electron spectra at t = 0 corresponds to the contributions of the warm jet plasma (that balances the hoop stress of the magnetic field) and the cold ambient plasma outside the jet (also providing confinement of the jet). The initial energy distribution of the warm jet plasma is represented by the black dashed-dotted curves in (b1) and (b2).

Close modal

As shown in Fig. 2(a), the dissipated magnetic energy is preferentially transferred to the jet ions, gaining 2× as much energy as the electron population. More interestingly, this process results in the acceleration of nonthermal ions, forming a high-energy power-law tail in the ion energy spectrum εp with p = 2.6 [Fig. 2(b1)]. We find that 10% of the total initial magnetic field energy is transferred to nonthermal ions with ε>5KBT0. We further verify that ions are accelerated up to the confinement energy of the jet εconf(eB0R)2/2mic2=kBT0(R/di)2 (valid for subrelativistic ion energies), i.e., the energy beyond which the ion Larmor radius ρi exceeds the system size R. In the astrophysical context, this limiting energy is also known as the Hillas energy or the Hillas constraint.25 For the simulated parameters, we find that the cutoff energy of the ion spectrum occurs at ε40kBT0=1.6εconf. This is consistent with the maximum energy gain of an ion accelerating freely in the coherent axial electric field structure Ez during the time scale τKI, εmax(eEzτKI)2/2miεconf.

The jet electrons, however, are not accelerated to nonthermal energies by the KI [Fig. 2(b2)]. In fact, the electrons within the jet (within r < R) cool down as shown by the receding spectral tail in Fig. 2(b2). It is the electrons at the periphery of the jet (rR) that absorb a fraction of the dissipated magnetic energy but are not accelerated efficiently to many times their initial thermal energy. The difference between the acceleration dynamics of ions and electrons will be further discussed in Sec. III C.

We find that the mechanism by which nonthermal ions are accelerated is similar to that identified in the regime of relativistic pair plasma.7 By following the trajectories of a representative sample of nonthermal ions [gray curves in Fig. 3(a)], we have verified that their acceleration is due to the work of the motional electric field E=v×BE. This is shown for both the ion that attained the highest energy in the sample (red curve) and the mean energy gain of the sample (blue curve) in Fig. 3(a). This indicates that nonideal electric fields (E), commonly associated with reconnecting current layers, play a negligible role in the acceleration dynamics of ions.

FIG. 3.

(a) Evolution of the energy gain Δε of a representative sample of 2000 ions that reach the nonthermal power-law tail (gray curves), self-consistently accelerated via the development of the KI. The ions with the highest energy gain of the sample and the mean energy gain of the sample are represented by the solid red and blue curves, respectively. The evolution of the work performed by the motional electric field E is represented by the dashed red and blue curves for comparison. (b) The energy gain of a random subset of 3 ion trajectories (from the original sample of 2000 trajectories), represented by the solid curves, is compared with the work performed by E along the curvature drift trajectory of their guiding centers (dashed curves); the dashed curves are interrupted when the ion Larmor radii become a significant fraction of the jet radius, resulting in the breakdown of the guiding center drift description of the ion trajectories. (c) Evolution of the distribution of magnetic field curvature κB in the simulated domain at vAt/R=8, 12, and 17. The vertical dashed lines mark the points where the magnetic curvature equals the Larmor scale of thermal ions (κBρ¯i=1) and electrons (κBρ¯e=1).

FIG. 3.

(a) Evolution of the energy gain Δε of a representative sample of 2000 ions that reach the nonthermal power-law tail (gray curves), self-consistently accelerated via the development of the KI. The ions with the highest energy gain of the sample and the mean energy gain of the sample are represented by the solid red and blue curves, respectively. The evolution of the work performed by the motional electric field E is represented by the dashed red and blue curves for comparison. (b) The energy gain of a random subset of 3 ion trajectories (from the original sample of 2000 trajectories), represented by the solid curves, is compared with the work performed by E along the curvature drift trajectory of their guiding centers (dashed curves); the dashed curves are interrupted when the ion Larmor radii become a significant fraction of the jet radius, resulting in the breakdown of the guiding center drift description of the ion trajectories. (c) Evolution of the distribution of magnetic field curvature κB in the simulated domain at vAt/R=8, 12, and 17. The vertical dashed lines mark the points where the magnetic curvature equals the Larmor scale of thermal ions (κBρ¯i=1) and electrons (κBρ¯e=1).

Close modal

The efficient acceleration of ions by the motional electric field E implies that ions must move efficiently transverse to the local magnetic field. Indeed, we find that the highly inhomogeneous and highly tangled structure of the magnetic field that develops in the nonlinear phase of the KI [Fig. 1(b2)] facilitates the displacement of ions across magnetic field lines via guiding center drift motions. In particular, we find that curvature drift motions of the ions play a dominant role. This is verified in Fig. 3(b), which shows that the energy gain of a random subset of 3 particle tracks (from the same sample) is well described by the work performed by the motional electric field along their guiding center curvature drift trajectories, i.e., Δεevcurv·E, where vcurv=v2B×κB/ωci|B| is the curvature drift velocity and κB=B·B/|B|2 is the magnetic curvature vector field. The guiding center description of the ion trajectories breaks down when they achieve a large fraction of the confinement energy of the jet, 0.3εconf. Beyond these energies, ions become effectively unmagnetized, moving nearly freely along the motion electric field in meandering type orbits until they escape the acceleration region or until the accelerating fields decay.

Figure 3(c) shows the evolution of the distribution of magnetic curvature κB during the nonlinear development of the KI. Enhancement in the distribution of magnetic curvature is observed in between the scales of the jet radius κBR1 and Larmor radius of thermal electrons κBρ¯e1. However, this enhancement is significantly larger at the Larmor scale of thermal ions when compared to that of thermal electrons. This result indicates that thermal ions are more likely to encounter the magnetic field curvature that approaches the scale of their Larmor radii than electrons, allowing them to be more easily injected into a rapid accelerating phase via curvature drifts. Moreover, ions can become locally unmagnetized (when they experience κBρi1) with higher probability than electrons. It is likely that this distribution of magnetic curvature is key to explaining the difference in acceleration efficiency between the two species. A more in detailed analysis of the electron dynamics and acceleration efficiency for different initial conditions will be the subject of future work.

We have performed a set of simulations of varying magnetization (0.08ωci/ωpi0.25), system sizes (5R/di10), and ion to electron mass ratios (4mi/me36) to probe how these parameters influence the particle acceleration dynamics of the KI. For the range of parameters explored in this work, we have verified that all simulations exhibit identical growth and saturation of the KI upon normalizing space and time to R and vA/R, respectively. We systematically observe the conversion of 65% of the initial magnetic energy in the system into plasma kinetic energy after 25R/vA in all our simulations. The magnitude of the accelerating electric field structure that develops in the nonlinear phase of the KI is also similar across our simulations when normalized to the characteristic value of vA/cB0 (Ezz0.3vAB0/c).

We observe the persistent acceleration of nonthermal ions in all our simulations, while electrons are always found to remain thermal. The spectra of accelerated ions for varying mi/me and ωci/ωpi at a fixed system size R/di=5 are shown in Fig. 4(a). Interestingly, we find that the power-law index of the nonthermal tail remains approximately constant at −2.6 ± 0.1 for varying mi/me and ωci/ωpi (within the explored parameter range), and thus, these results may be extrapolated to realistic mass ratios. We further verify in all simulations that the nonthermal ion tail extends up to the confinement energy of the jet εconf=kBT0(R/di)2 [Fig. 4(a)]. For a fixed normalized jet radius R/di=5, we find that εconf/kBT0=(R/di)2=25, which is independent of the ion to electron mass ratio and jet magnetization, as seen in Fig. 4(a). In all the simulated cases, we find that the cutoff energy of the spectrum is εmax/kBT040=1.6εconf/kBT0.

FIG. 4.

(a) Energy spectrum of accelerated ions for varying ion to electron mass ratios (mi/me) and jet magnetization (ωci/ωpi) at the fixed dimensionless jet radius R/di=5. Note that the energy scale is normalized to the thermal energy kBT0, which varies with jet magnetization. Further note that the slight energy shift observed in the spectrum of the mi/me=36 case was due to the use of a slightly higher initial plasma temperature that raised the plasma thermal pressure uniformly throughout the simulation domain (without affecting the initial hydromagnetic equilibrium). This slight change in the initial conditions did not influence the development of the high-energy nonthermal component of the spectrum. (b) Dependence of the electron to ion energy gain (ΔεKe/ΔεKi) on the mass ratio. Matching colors between the spectra in (a) and the points in (b) correspond to the same simulated parameters.

FIG. 4.

(a) Energy spectrum of accelerated ions for varying ion to electron mass ratios (mi/me) and jet magnetization (ωci/ωpi) at the fixed dimensionless jet radius R/di=5. Note that the energy scale is normalized to the thermal energy kBT0, which varies with jet magnetization. Further note that the slight energy shift observed in the spectrum of the mi/me=36 case was due to the use of a slightly higher initial plasma temperature that raised the plasma thermal pressure uniformly throughout the simulation domain (without affecting the initial hydromagnetic equilibrium). This slight change in the initial conditions did not influence the development of the high-energy nonthermal component of the spectrum. (b) Dependence of the electron to ion energy gain (ΔεKe/ΔεKi) on the mass ratio. Matching colors between the spectra in (a) and the points in (b) correspond to the same simulated parameters.

Close modal

While the fraction of initial magnetic energy that is dissipated is approximately constant in all our simulations, the ratio of the electron to ion energy gain (ΔεKe/ΔεKi) is shown to depend on both the mass ratio and the magnetization, as illustrated in Fig. 4(b). We observe that an increasing fraction of the dissipated magnetic energy is transferred to the ions with both the increasing mass ratio and increasing magnetization. We note that a similar dependence of ΔεKe/ΔεKi on mi/me has been reported in 3D PIC simulations of relativistic magnetized plasma turbulence.26 

There is a significant experimental effort to study the MHD stability of plasma jets in the laboratory in conditions relevant to astrophysical environments.11,16–21 However, laboratory-produced plasma jets are generally far more collisional than those naturally occurring in space or astrophysical settings. Thus, in order to assess if the ion acceleration mechanism discussed in Sec. III can be studied in laboratory plasma experiments, it is important to consider the effects of Coulomb collisions. One naturally anticipates that if the jet plasma is too collisional, nonthermal particle distributions will not be allowed to develop as they will rapidly relax to thermal equilibrium.

We have investigated the impact of a finite Coulomb collision frequency on the particle acceleration dynamics of the KI, with the aim of determining the collisional threshold above which nonthermal particle acceleration is suppressed. We include the physics of binary Coulomb collisions (the Spitzer-Harm model for weakly collisional plasmas) in our PIC simulations using a Monte Carlo approach.27–30 This method randomly pairs particles locally within a cell and scatters their momenta such that energy and momentum are conserved on each collision (equal particle weights are used). We use this method to model electron-ion and ion-ion collisions in our simulations of the KI, and we vary the collision frequency by artificially varying the Coulomb logarithm. In this way, all other physical parameters of the system can be fixed (plasma density, temperature, and magnetic field strength) and the effect of Coulomb collisions can be varied independently.

We have examined the effect of Coulomb collisions on the illustrative case explored in Sec. III, i.e., a jet with ωci/ωpi=0.125,mi/me=16, and R/di=5 (note that R/di=5 is close to the jet radii produced in recent experiments11,19,21). We progressively increased the ion collision frequency νi relative to the dynamical time of the KI (which also corresponds to the acceleration time scale of the KI) from νiR/vA=0 to 0.4; note that νiR/vA can also be conveniently interpreted as the ratio of the jet radius to the ion mean-free-path, νiR/vAR/λi, since vAkBT0/mi (as imposed by the initial equilibrium configuration). The resulting spectra of accelerated ions are shown in Fig. 5(a) for varying νi values. We find that the spectrum of accelerated ions remains unaffected for νiR/vA0.01, relative to the collisionless case: the slope of the nonthermal tail, the number of nonthermal particles, and the maximum energy reached remain unchanged by Coulomb collisions below νiR/vA0.01. At νiR/vA0.1, the ion collision frequency is no longer negligible compared to the acceleration time scale of the KI. As ions gain energy through the KI-induced electric field, they also lose energy through collisions and heat up the background particles. This effect leads to a strong reduction in the acceleration efficiency of nonthermal ions as shown in Fig. 5(a): the number of nonthermal particles is strongly reduced and the slope of the tail of the distribution is hardened. At νiR/vA0.4, nonthermal ion acceleration is nearly fully suppressed.

FIG. 5.

Impact of Coulomb collisions on ion acceleration efficiency. (a) Energy spectrum of accelerated ions for varying ion-collision frequencies (νi); the jet parameters are R/di=5, ωci/ωpi=0.125, and mi/me = 16. The ion collision frequency is varied by artificially varying the Coulomb logarithm. The solid black curve represents the initial ion spectrum, before the development of the KI. Panels (b1), (b2), and (b3) show the cross-sections of the ẑ-component of the electric field at vAt/R12.5 for the different collisionalities νiR/vA=0, 0.1, and 0.4, respectively.

FIG. 5.

Impact of Coulomb collisions on ion acceleration efficiency. (a) Energy spectrum of accelerated ions for varying ion-collision frequencies (νi); the jet parameters are R/di=5, ωci/ωpi=0.125, and mi/me = 16. The ion collision frequency is varied by artificially varying the Coulomb logarithm. The solid black curve represents the initial ion spectrum, before the development of the KI. Panels (b1), (b2), and (b3) show the cross-sections of the ẑ-component of the electric field at vAt/R12.5 for the different collisionalities νiR/vA=0, 0.1, and 0.4, respectively.

Close modal

Note that as the collision frequency was varied, no significant changes were observed in the morphology or temporal evolution of the KI-induced electric and magnetic fields relative to the collisionless regime. This is illustrated Figs. 5(b1)–5(b3) by the snapshots of the cross section of Ez for varying collisionality. These snapshots were all taken at vAt/R12.5, revealing that the temporal development and spatial structure of the KI-induced electric field are nearly unaffected by collisions; only in the highly collisional case of νiR/vA0.4 does the amplitude of Ez become noticeably lower, mainly as a consequence of the increased magnetic diffusivity that lowers the peak amplitude of the magnetic field and hence of the motional electric field.

The laboratory plasma jets produced by the radial wire array Z-pinch19 are characterized by low-temperature and high-Z plasmas, achieving R/λi106. The acceleration of nonthermal ions by the mechanism reported here is thus expected to be significantly suppressed within the dense body of the jet. However, high-energy ions have indeed been reported, resulting from the development of MHD instabilities in the jets produced by this platform.11 It is likely these high-energy ions are accelerated outside of the dense jet in the low-density ambient plasma where the collisionality is reduced. Our simulations have not considered the role of a low-density plasma background surrounding the jet, which could allow ambient particles to interact with the fields produced by the KI of the jet while remaining in a weakly collisional environment. Indeed, we do see in our simulations that particles outside of the jet, within R<r<2R, do interact with the KI-induced fields and absorb a significant fraction of the dissipated magnetic energy. It is therefore possible that the mechanism reported here can participate in the acceleration of nonthermal particles in the low-density plasma surrounding the dense jet. A more detailed exploration of our work in the conditions of these experiments will be pursued in the near future.

Laser-driven high-energy-density plasmas are more likely to produce weakly collisional jets in the laboratory. Recent laser-driven plasma experiments produced near kilo-electron-volt-temperature plasma jets, with R/λi102, aimed at investigating the development of the KI in astrophysically relevant conditions.21 Future experiments using a more energetic laser drive may be able to produce jets with R/λi0.1, making it possible to directly probe the efficient particle acceleration physics of the KI within the body of the jet, as described in this work.

We have shown via 3D PIC simulations that the development of the KI in nonrelativistic, electron-ion plasma results in the efficient acceleration of nonthermal ions. Approximately 10% of initial magnetic energy is transferred into nonthermal ions over the course of a few 10s of dynamical times of the KI (R/vA), forming a power-law tail in the energy spectrum with index 2.6. We showed that the power-law index of the nonthermal tail remains nearly constant over the range of jet magnetizations explored in our work and that the maximum ion energy systematically reaches the confinement energy of the jet εconf/kBT0(R/di)2.

The ion acceleration mechanism is similar to the mechanism found to operate in relativistic pair plasma jets.7 The nonlinear development of the KI produces a coherent motional electric field along the axis of the jet that is embedded in a highly tangled magnetic field. Ions experience fast curvature drift motions in the highly tangled magnetic field, which permit their efficient displacement along the motional electric field and hence their efficient acceleration. Our results indicate that this is a viable mechanism to explain efficient acceleration of nonthermal protons in nonrelativistic jets in space and astrophysical environments.

The similarity between the particle acceleration physics in the relativistic and nonrelativistic regimes of the KI should also motivate the development of laboratory experimental platforms capable of investigating the particle acceleration mechanisms relevant to astrophysical jets. Our work indicates that in order to study this acceleration mechanism in the laboratory, the plasma collisionality needs to be significantly reduced, such that the ion mean free path becomes greater than 10× the jet radius. These conditions can likely be produced in laser-driven high-energy-density plasma experiments.31 

This work was supported by the U.S. Department of Energy SLAC Contract No. DE-AC02-76SF00515, by the U.S. DOE Office of Science, Fusion Energy Sciences under FWP 100237, and by the U.S. DOE Early Career Research Program under FWP 100331. The authors acknowledge the OSIRIS Consortium, consisting of UCLA and IST (Portugal) for the use of the OSIRIS 4.0 framework and the visXD framework. Simulations were run on Mira (ALCF) through an ALCC award.

1.
G.
Bateman
,
MHD Instabilities
(
MIT Press
,
Cambridge, MA
,
1978
), p.
270
.
2.
E.
Pariat
,
K.
Dalmasse
,
C. R.
DeVore
,
S. K.
Antiochos
, and
J. T.
Karpen
,
Astron. Astrophys.
573
,
A130
(
2015
).
3.
R.
Bučík
,
D. E.
Innes
,
G. M.
Mason
,
M. E.
Wiedenbeck
,
R.
Gómez-Herrero
, and
N. V.
Nitta
,
Astrophys. J.
852
,
76
(
2018
).
4.
R.
Bučík
,
M. E.
Wiedenbeck
,
G. M.
Mason
,
R.
Gómez-Herrero
,
N. V.
Nitta
, and
L.
Wang
,
Astrophys. J. Lett.
869
,
21
(
2018
).
5.
D.
Giannios
and
H. C.
Spruit
,
Astron. Astrophys.
450
,
887
(
2006
).
6.
O.
Bromberg
and
A.
Tchekhovskoy
,
Mon. Not. R. Astron. Soc.
456
,
1739
(
2016
).
7.
E. P.
Alves
,
J.
Zrake
, and
F.
Fiúza
,
Phys. Rev. Lett.
121
,
245101
(
2018
).
8.
L. E.
Zakharov
,
S. A.
Galkin
,
S. N.
Gerasimov
, and
J.-E. Contributors
,
Phys. Plasmas
19
,
055703
(
2012
).
9.
D. D.
Ryutov
,
M. S.
Derzon
, and
M. K.
Matzen
,
Rev. Mod. Phys.
72
,
167
(
2000
).
10.
M. G.
Haines
,
Plasma Phys. Controlled Fusion
53
,
093001
(
2011
).
11.
F.
Suzuki-Vidal
,
S.
Patankar
,
S. V.
Lebedev
,
S. N.
Bland
,
H.
Doyle
,
D.
Bigourd
,
G.
Burdiak
,
P.
de Grouchy
,
G. N.
Hall
,
A. J.
Harvey-Thompson
,
E.
Khoory
,
L.
Pickworth
,
J.
Skidmore
,
R. A.
Smith
, and
G. F.
Swadling
,
New J. Phys.
15
,
125008
(
2013
).
12.
A.
Tchekhovskoy
and
O.
Bromberg
,
Mon. Not. R. Astron. Soc.: Lett.
461
,
L46
(
2016
).
13.
A.
Ciardi
,
S. V.
Lebedev
,
A.
Frank
,
E. G.
Blackman
,
J. P.
Chittenden
,
C. J.
Jennings
,
D. J.
Ampleford
,
S. N.
Bland
,
S. C.
Bott
,
J.
Rapley
,
G. N.
Hall
,
F. A.
Suzuki-Vidal
,
A.
Marocchino
,
T.
Lery
, and
C.
Stehle
,
Phys. Plasmas
14
,
056501
(
2007
).
14.
Y.
Mizuno
,
Y.
Lyubarsky
,
K.-I.
Nishikawa
, and
P. E.
Hardee
,
Astrophys. J.
728
,
90
(
2011
).
15.
B.
Ripperda
,
O.
Porth
,
C.
Xia
, and
R.
Keppens
,
Mon. Not. R. Astron. Soc.
471
,
3465
(
2017
).
16.
S. C.
Hsu
and
P. M.
Bellan
,
Mon. Not. R. Astron. Soc.
334
,
257
(
2002
).
17.
S. C.
Hsu
and
P. M.
Bellan
,
Phys. Rev. Lett.
90
,
215002
(
2003
).
18.
S. V.
Lebedev
,
J. P.
Chittenden
,
F. N.
Beg
,
S. N.
Bland
,
A.
Ciardi
,
D.
Ampleford
,
S.
Hughes
,
M. G.
Haines
,
A.
Frank
,
E. G.
Blackman
, and
T.
Gardiner
,
Astrophys. J.
564
,
113
(
2002
).
19.
S. V.
Lebedev
,
A.
Ciardi
,
D. J.
Ampleford
,
S. N.
Bland
,
S. C.
Bott
,
J. P.
Chittenden
,
G. N.
Hall
,
J.
Rapley
,
C. A.
Jennings
,
A.
Frank
,
E. G.
Blackman
, and
T.
Lery
,
Mon. Not. R. Astron. Soc.
361
,
97
(
2005
).
20.
S. V.
Lebedev
,
A.
Frank
, and
D. D.
Ryutov
,
Rev. Mod. Phys.
91
,
025002
(
2019
).
21.
C. K.
Li
,
P.
Tzeferacos
,
D.
Lamb
,
G.
Gregori
,
P. A.
Norreys
,
M. J.
Rosenberg
,
R. K.
Follett
,
D. H.
Froula
,
M.
Koenig
,
F. H.
Séguin
,
J. A.
Frenje
,
H. G.
Rinderknecht
,
H.
Sio
,
A. B.
Zylstra
,
R. D.
Petrasso
,
P. A.
Amendt
,
H. S.
Park
,
B. A.
Remington
,
D. D.
Ryutov
,
S. C.
Wilks
,
R.
Betti
,
A.
Frank
,
S. X.
Hu
,
T. C.
Sangster
,
P.
Hartigan
,
R. P.
Drake
,
C. C.
Kuranz
,
S. V.
Lebedev
, and
N. C.
Woolsey
,
Nat. Commun.
7
,
13081
(
2016
).
22.
R. A.
Fonseca
,
L. O.
Silva
,
F. S.
Tsung
,
V. K.
Decyk
,
W.
Lu
,
C.
Ren
,
W. B.
Mori
,
S.
Deng
,
S.
Lee
,
T.
Katsouleas
, and
J. C.
Adam
, in
Proceedings of the Computational Science-ICCS 2002
(
2002
), Part III, Vol.
2331
, p.
342
.
23.
R. A.
Fonseca
,
S. F.
Martins
,
L. O.
Silva
,
J. W.
Tonge
,
F. S.
Tsung
, and
W. B.
Mori
,
Plasma Phys. Controlled Fusion
50
,
124034
(
2008
).
24.
S.
Zenitani
and
M.
Hoshino
,
Astrophys. J.
618
,
L111
(
2005
).
25.
A. M.
Hillas
,
Annu. Rev. Astron. Astrophys.
22
,
425
(
1984
).
26.
V.
Zhdankin
,
D. A.
Uzdensky
,
G. R.
Werner
, and
M. C.
Begelman
,
Phys. Rev. Lett.
122
,
055101
(
2019
).
27.
T.
Takizuka
and
H.
Abe
,
J. Comput. Phys.
25
,
205
(
1977
).
28.
K.
Nanbu
and
S.
Yonemura
,
J. Comput. Phys.
145
,
639
(
1998
).
29.
Y.
Sentoku
and
A. J.
Kemp
,
J. Comput. Phys.
227
,
6846
(
2008
).
30.
F.
Pérez
,
L.
Gremillet
,
A.
Decoster
,
M.
Drouin
, and
E.
Lefebvre
,
Phys. Plasmas
19
,
083104
(
2012
).
31.
C. M.
Huntington
,
F.
Fiúza
,
J. S.
Ross
,
A. B.
Zylstra
,
R. P.
Drake
,
D. H.
Froula
,
G.
Gregori
,
N. L.
Kugland
,
C. C.
Kuranz
,
M. C.
Levy
,
C. K.
Li
,
J.
Meinecke
,
T.
Morita
,
R.
Petrasso
,
C.
Plechaty
,
B. A.
Remington
,
D. D.
Ryutov
,
Y.
Sakawa
,
A.
Spitkovsky
,
H.
Takabe
, and
H. S.
Park
,
Nat. Phys.
11
,
173
(
2015
).