Electron Energy Distribution Function (EEDF)

• EEDF is the plasma electron energy distribution function that quantifies how electrons populate different kinetic energy intervals, influencing ionization, excitation, and transport.
• Techniques such as Langmuir probe analysis, inversion of x-ray spectra, and DEM-based reconstruction are employed to accurately measure and reconstruct the EEDF in various plasma settings.
• The shape of the EEDF, from Maxwellian to non-Maxwellian forms, critically affects plasma chemistry and ion flux-energy control, reflecting underlying heating and collisional mechanisms.

Searching arXiv for recent and foundational papers on EEDF to ground the article in the literature. The electron energy distribution function (EEDF) is the energy-space representation of the electron population in a plasma: it specifies how many electrons occupy each kinetic-energy interval, and therefore governs electron-impact ionization, excitation, dissociation, transport, and power absorption. In low-temperature plasma studies it is commonly written as $F(\varepsilon)$ or $f(\varepsilon)$, with $n_e = \int F(\varepsilon)\, d\varepsilon$, whereas in solar-flare analysis the corresponding observable is often the volume-averaged electron flux spectrum $\langle n V F \rangle(E)$ (Sharma et al., 2018, Battaglia et al., 2013). Across capacitively coupled plasmas, atmospheric-pressure microjets, ECR sources, positive columns, HiPIMS discharges, and flare plasmas, the EEDF is the central kinetic quantity linking microscopic collision physics to macroscopic plasma behavior (Vass et al., 2021, Rudolph et al., 2021).

1. Definition and mathematical representations

In low-temperature plasma kinetics, the EEDF $F(\varepsilon)$ is defined so that $F(\varepsilon)\, d\varepsilon$ is the number density of electrons with energy in $[\varepsilon,\varepsilon+\Delta\varepsilon]$, and the total density is the zeroth moment,

$$n_e = \int_0^{\varepsilon_{\max}} F(\varepsilon)\, d\varepsilon .$$

An effective electron temperature can be defined from the mean energy as

$$T_\mathrm{eff} = \frac{2}{3k_B}\; \frac{\int_0^{\varepsilon_{\max}} \varepsilon F(\varepsilon)\, d\varepsilon}{\int_0^{\varepsilon_{\max}} F(\varepsilon)\, d\varepsilon}.$$

These relations are used explicitly in particle-in-cell analyses of capacitively coupled argon discharges (Sharma et al., 2018).

For isotropic distributions, the energy variable is related to the velocity distribution by

$$\varepsilon = \frac{1}{2} m_e v^2, \qquad f_\varepsilon(\varepsilon)\, d\varepsilon = f_v(v)\, 4\pi v^2\, dv,$$

and, in the standard Maxwellian case,

$f(\varepsilon)$0

In semi-logarithmic representations, a Maxwellian therefore appears as a straight line, whereas deviations in slope identify non-Maxwellian structure such as bi-Maxwellian or Druyvesteyn-like behavior (Sharma et al., 2019, Sharma et al., 2018).

A related but distinct formulation appears in flare physics, where the inferred quantity is the mean electron flux spectrum

$f(\varepsilon)$1

and, for a uniform isothermal source,

$f(\varepsilon)$2

When the plasma is multi-thermal, the EEDF is obtained from the differential emission measure $f(\varepsilon)$3 through

$f(\varepsilon)$4

This representation extends EEDF inference in flares from $f(\varepsilon)$5 keV to a few tens of keV (Battaglia et al., 2013).

More general non-Maxwellian parameterizations are also used. In the Tsallis framework,

$f(\varepsilon)$6

with

$f(\varepsilon)$7

so that $f(\varepsilon)$8 produces a hard tail cut-off and $f(\varepsilon)$9 produces a $n_e = \int F(\varepsilon)\, d\varepsilon$0-like power-law tail (Boumali, 27 Apr 2026).

2. Measurement, reconstruction, and inversion

In laboratory low-temperature plasmas, the EEDF is classically obtained from Langmuir-probe I–V characteristics by the Druyvesteyn method. For an RF-compensated probe, the electron energy probability function follows from the second derivative of the electron current with respect to probe bias, and in one implementation the EEDF is written as

$n_e = \int F(\varepsilon)\, d\varepsilon$1

while the plasma density and effective temperature are obtained from

$n_e = \int F(\varepsilon)\, d\varepsilon$2

A second harmonic technique (SHT) has been used to obtain the direct measurement of EEDF in a magnetized cylindrical CCP, thereby avoiding numerical second differentiation of noisy probe traces (Khandelwal et al., 2024).

A broader review of electric-probe processing emphasizes that EEDF retrieval requires robust estimation of $n_e = \int F(\varepsilon)\, d\varepsilon$3, and compares analog differentiation, Savitzky–Golay filtering, B-spline fitting, Gaussian filtering, and Blackman windows with respect to dynamic range, energy resolution, and signal distortion (Caldarelli et al., 2022). In the RF argon case analyzed there, analog differentiation yielded the largest dynamic range for the EEPF, whereas Savitzky–Golay and B-spline filtering produced comparable $n_e = \int F(\varepsilon)\, d\varepsilon$4, $n_e = \int F(\varepsilon)\, d\varepsilon$5, and $n_e = \int F(\varepsilon)\, d\varepsilon$6 for parameter extraction (Caldarelli et al., 2022).

In x-ray diagnostics, the EEDF can be reconstructed from the x-ray energy distribution function by inverting a detector-response model that includes Bremsstrahlung emission, transmission, and finite energy resolution. In PFRC-II, a Poisson-regularized inversion of pulse-height spectra from an Amptek Silicon Drift Detector recovered EEDFs from below 200 eV to over 8 keV and spanning five orders-of-magnitude in intensity (Swanson et al., 2017). The forward model is written as

$n_e = \int F(\varepsilon)\, d\varepsilon$7

and the inversion minimizes the Poisson cost

$n_e = \int F(\varepsilon)\, d\varepsilon$8

together with a smoothness prior (Swanson et al., 2017).

In solar-flare diagnostics, the EEDF is reconstructed indirectly from EUV and hard-X-ray observables. AIA data yield the DEM $n_e = \int F(\varepsilon)\, d\varepsilon$9, RHESSI constrains the thermal and non-thermal high-energy parts, and the combined analysis allows the electron distribution function to be inferred over the broad energy range from $\langle n V F \rangle(E)$0 keV up to a few tens of keV (Battaglia et al., 2013). This demonstrates that EEDF reconstruction is not tied to a single diagnostic modality, but can be posed as a general inversion problem.

3. Canonical forms and departures from Maxwellian behavior

The Maxwellian EEDF remains the reference equilibrium form, but the literature summarized here shows that practical plasmas frequently exhibit Druyvesteyn, bi-Maxwellian, three-temperature, and other generalized distributions. In one-dimensional PIC/MCC simulations of argon CCPs at 100 mTorr and 13.56–100 MHz, the EEDF with metastables changes its shape from Druyvesteyn type, at low excitation frequency, to bi-Maxwellian, at high frequency plasma excitation; without metastable atoms, a three-temperature EEDF is observed (Sharma et al., 2018).

In symmetric VHF capacitive discharges at 1 Pa, several distinct transitions have been reported. When the electron plasma frequency is kept approximately constant at $\langle n V F \rangle(E)$1 MHz by adjusting the driving voltage, the EEDF at the discharge center evolves from bi-Maxwellian at 27.12 MHz to nearly Maxwellian at 60 and 100 MHz, with a strong enhancement in the population of electrons between 1 eV and 20 eV (Sharma et al., 2019). Under constant power density, the sequence is more structured: the EEDF changes from bi-Maxwellian to nearly Maxwellian up to a transition frequency, and then transforms into a nearly bi-Maxwellian at higher driving frequencies (Sharma et al., 2020).

Waveform tailoring at low pressure introduces further structure. In a current-driven argon CCP at 5 mTorr and 13.56 MHz, a square waveform produces an EEDF with 3 electron temperatures, with a bulk electron temperature around 6 eV, a mid-energy range from about 6 eV to 33 eV, and a high-energy tail above 33 eV (Sharma et al., 2021). In a magnetized cylindrical CCP, the EEDF undergoes a transition from bi-Maxwellian for unmagnetized to Maxwellian at intermediate $\langle n V F \rangle(E)$2 and finally becomes a weakly bi-Maxwellian at higher values of $\langle n V F \rangle(E)$3 (Dahiya et al., 2023).

Outside classical low-temperature discharges, the same pattern of composite distributions appears. In HiPIMS, the ionization region model assumes a bi-Maxwellian EEDF consisting of a cold bulk electron group and a hot secondary electron group, and the OBELIX Boltzmann solution shows very good agreement with that assumption (Rudolph et al., 2021). In solar flares, the inferred electron distribution functions generally show one or two nearly Maxwellian components at energies below $\langle n V F \rangle(E)$4 keV and a non-thermal tail above (Battaglia et al., 2013). In ECR plasmas, the reconstructed warm-electron population is better fitted by multicomponent Maxwellian/Druyvesteyn combinations than by a single Maxwellian (Mishra et al., 2021).

Taken together, these results indicate that “the EEDF” is often not a single-temperature object but a composite structure whose form depends on the balance between heating, collisions, confinement, and loss.

4. Mechanisms that shape the EEDF

The most direct controls on EEDF shape are the electron heating mechanism and the spatio-temporal electric-field structure. In low-pressure CCPs, sheath dynamics and higher harmonics are particularly important. In argon at 1 Pa and a 3.2 cm gap, increasing the driving frequency from 27.12 MHz to 100 MHz at constant electron plasma frequency increases electric-field non-linearity and higher-harmonic content; simultaneously, a positive $\langle n V F \rangle(E)$5 appears in the bulk plasma, and the EEDF changes from bi-Maxwellian to nearly Maxwellian through enhancement of mid-energy electrons (Sharma et al., 2019). Under constant power density, a transition frequency is observed: below it the density decreases and electron temperature increases; above it the trend is reversed, and the EEDF changes accordingly (Sharma et al., 2020).

At atmospheric pressure, the heating channel is different. In a He/N$\langle n V F \rangle(E)$6 micro atmospheric-pressure plasma jet driven by tailored “valleys” waveforms, the local nature of transport makes ohmic power absorption dominant. Increasing the number of harmonics shortens the sheath collapse at the grounded electrode, drives a high current through the discharge during that short collapse, and creates a high ohmic electric field where the electron density is low. This accelerates electrons to high energies, produces strong ionization, and forms a local density maximum that generates an ambipolar field reversal; these effects together modify the high-energy tail of the EEDF near the grounded electrode (Vass et al., 2021).

Metastable kinetics can also restructure the EEDF. In 100 mTorr argon, inclusion of Ar$\langle n V F \rangle(E)$7 and Ar$\langle n V F \rangle(E)$8 introduces strong inelastic channels at 11.6 eV and 13.1 eV, depletes the hot tail at low frequency, and increases the mid-energy population that is important for metastable production. As the RF frequency rises, the contribution of multistep ionization increases, and the EEDF changes from Druyvesteyn to bi-Maxwellian (Sharma et al., 2018).

In HiPIMS, the EEDF is shaped by the coexistence of Ohmic heating in the ionization region and sheath acceleration of secondary electrons to $\langle n V F \rangle(E)$9. The resulting distribution consists of a cold Maxwellian-like bulk plus a hot tail; the latter is fed by sheath-accelerated secondaries and is responsible for a substantial fraction of ionization from Ar ground state (Rudolph et al., 2021).

These studies jointly show that EEDF evolution is not reducible to a single mechanism. Depending on pressure and geometry, the dominant agents are sheath expansion, beam formation, higher-harmonic field transients, Ohmic heating, ambipolar fields, metastable-assisted energy loss and ionization, or direct injection of hot secondaries.

5. Spatial nonlocality, temporal modulation, anisotropy, and magnetization

A central result of recent kinetic work is that the EEDF often cannot be factorized into a purely energy-dependent shape multiplied by a local density profile. In hybrid kinetic-fluid simulations of an AC argon positive column, substantial spatial nonlocality and temporal nonlocality of the EEDF arise when the energy relaxation length is comparable to the tube radius and the drive period lies between the energy-relaxation and ambipolar-diffusion times. In the dynamic regime, $F(\varepsilon)$0, the EEDF changes strongly over the AC cycle, develops non-monotonic peaks associated with Penning ionization and superelastic collisions, and produces off-axis maxima in excitation rate, metastable density, and plasma density (Humphrey et al., 2023). This indicates that time-averaged or local-Maxwellian closures can miss essential phase-space transport.

Magnetized discharges provide a complementary example. In a geometrically asymmetric cylindrical CCP with an axisymmetric magnetic field, an intermediate $F(\varepsilon)$1 range produces high efficiency, lower electron temperature, and a transition of the center EEDF from bi-Maxwellian to Maxwellian; at larger $F(\varepsilon)$2, the density drops and the EEDF becomes weakly bi-Maxwellian again (Dahiya et al., 2023). In a coaxial cylindrical magnetized CCP, increasing the axial magnetic field from 0 G to 60 G raises the density by about a factor of three, lowers the plasma potential from about 56 V to about 26 V, and changes the EEDF from nearly Maxwellian to Druyvesteyn-like, because hot electrons are relatively depleted while cold and mid-energy electrons increase more strongly (Khandelwal et al., 2024).

In ECR plasmas the issue is explicitly anisotropic and non-homogeneous electron populations. A numerical tool based on ROI-dependent trial-and-error EEDF fitting was developed for space-resolved warm electrons in the energy range 2–20 keV, and the reconstructed continuous EEDFs depend on the region of configuration space and on the relative cold, warm, and hot populations (Mishra et al., 2021). In the tokamak context, a broader class of equilibrium distribution functions depending on $F(\varepsilon)$3, $F(\varepsilon)$4, and $F(\varepsilon)$5 was derived probabilistically by Bayes’ theorem for anisotropic equilibria such as NBI and ICRH scenarios (Troia, 2015). This suggests that in strongly magnetized systems an “EEDF” in the narrow energy-only sense may be insufficient, and constants-of-motion dependence becomes intrinsic to the equilibrium description.

6. EEDF as the control variable for rates, chemistry, and ion flux–energy decoupling

The practical importance of the EEDF follows from the fact that electron-impact processes are functionals of $F(\varepsilon)$6. In low-temperature plasma chemistry this is written as

$F(\varepsilon)$7

or, in local form,

$F(\varepsilon)$8

Accordingly, changing the EEDF changes ionization, excitation, metastable production, and ultimately plasma density and species composition (Sharma et al., 2018, Vass et al., 2021).

In applied CCP physics, EEDF control is directly tied to ion-flux and ion-energy control. In the constant-$F(\varepsilon)$9 VHF argon study, increasing frequency from 27.12 MHz to 100 MHz at constant ion flux causes the mean ion energy at the electrode to decrease from about 75 eV to about 35 eV, while the EEDF shifts toward a nearly Maxwellian form through enhancement of 1–20 eV electrons (Sharma et al., 2019). Under constant power density, increasing frequency above the transition frequency raises density and lowers ion energy, while the EEDF at high power becomes nearly bi-Maxwellian; this provides a route to high ion flux with reduced ion bombardment energy (Sharma et al., 2020). In the magnetized cylindrical CCP, the intermediate-$F(\varepsilon)\, d\varepsilon$0 regime likewise permits nearly independent control of ion energy and ion flux through the associated changes in self-bias, density, and EEDF shape (Dahiya et al., 2023).

In atmospheric-pressure jets, voltage waveform tailoring controls the EEDF through the spatio-temporal pattern of electron power absorption, which in turn allows selective enhancement of metastable and radical production (Vass et al., 2021). In ECR plasmas, continuous ROI-dependent EEDFs are explicitly intended to be inserted into rate integrals of the form

$F(\varepsilon)\, d\varepsilon$1

for K$F(\varepsilon)\, d\varepsilon$2 emission and other electron-driven reactions (Mishra et al., 2021).

A recent benchmark for neutral He, Li, and Be makes this dependence fully explicit. The single-ionization rate coefficient is

$F(\varepsilon)\, d\varepsilon$3

and varying the EEDF shape from a strict Maxwellian to a two-temperature Tsallis distribution changes the rates strongly, especially when $F(\varepsilon)\, d\varepsilon$4 is large. Sub-extensive distributions $F(\varepsilon)\, d\varepsilon$5 suppress ionization through a hard tail cut-off, while super-extensive distributions $F(\varepsilon)\, d\varepsilon$6 enhance low-temperature ionization through a $F(\varepsilon)\, d\varepsilon$7-like power-law tail; He responds most strongly and Li least (Boumali, 27 Apr 2026). This provides a clear quantitative demonstration that EEDF shape and cross-section model are independent uncertainty axes in collisional-radiative modeling.

7. Theoretical synthesis and unresolved structure

Recent theoretical work has attempted to unify the observed “metamorphosis” of the EEDF across kinetic and hydrodynamic regimes. A renormalization-group treatment of the Vlasov–Boltzmann equation attributes non-Maxwellian EEDF structure to dynamically broken scale invariance and derives non-perturbative solutions valid across a range of collisionality (Saucedo et al., 7 Jul 2025). In that framework, the kinetic limit yields a bimodal EEDF, the hydrodynamic limit yields a generalized exponential, and the two regimes are separated by a critical pressure. The same formalism predicts analytic scaling relations for electron heating and interprets the stable RG fixed point as a state of minimum entropy production (Saucedo et al., 7 Jul 2025).

This theoretical picture is consistent with the empirical transitions reported in low-pressure CCPs, positive columns, and magnetized discharges: bi-Maxwellian, Druyvesteyn-like, nearly Maxwellian, weakly bi-Maxwellian, and three-temperature forms appear not as isolated anomalies but as regime-dependent manifestations of collisionality, nonlocality, and energy-selective loss or heating. At the same time, the diagnostics literature shows that recovering the EEDF with sufficient dynamic range and resolution remains nontrivial, especially when line radiation, RF modulation, or poor signal-to-noise complicate inversion (Swanson et al., 2017, Caldarelli et al., 2022).

The EEDF therefore occupies a dual role. It is both a measurable object—through probes, x-ray inversion, and DEM-based reconstruction—and a reduced kinetic descriptor that encodes the physically relevant consequences of power absorption, confinement, collisions, and boundary conditions. The studies surveyed here indicate that accurate plasma modeling increasingly requires treating the EEDF as a structured, often nonlocal and non-Maxwellian quantity rather than as a single-temperature closure.