RIT-2.5 RF Ion Thruster

• RIT-2.5 Radiofrequency Ion Thruster is a compact, micro-Newton device that employs inductive RF excitation to generate xenon plasma and uses a two-grid system for electrostatic ion extraction.
• The device integrates advanced diagnostics including optical emission spectroscopy, Faraday cup measurements, and particle-in-cell simulations to characterize beam current, energy distribution, and thrust performance.
• Experimental and theoretical studies demonstrate that operating parameters such as RF power, xenon flow, and extraction voltage critically influence thrust, energy spread, and specific impulse for nanosatellite applications.

Searching arXiv for the specified RIT-2.5 papers and closely related work to ground the article. The RIT-2.5 Radiofrequency Ion Thruster is a compact gridded radiofrequency ion thruster in the micro-Newton-class, designed around xenon plasma generation by inductive RF excitation and electrostatic ion extraction through a two-grid optics system. In the literature, the designation refers both to a specific hardware configuration with an overall diameter of 2.5 cm and to a model system for studying ion extraction, sheath dynamics, and beam formation in radiofrequency ion propulsion. A microscopic theoretical treatment links its thrust generation to particle trajectories in electric fields and to space-charge-limited extraction (Kirmse, 2013), while a later experimental and simulation study characterizes its discharge, beam current, and ion energy distribution using optical emission spectroscopy, Faraday-cup measurements, retarding field energy analysis, and particle-in-cell simulation (Joshi et al., 12 Sep 2025).

1. Device class and physical configuration

The RIT-2.5 is described as a compact, micro-Newton-class radiofrequency ion thruster with an overall diameter of 2.5 cm (Joshi et al., 12 Sep 2025). Its discharge chamber consists of a hemispherical dome (radius $R = 12.5\ \mathrm{mm}$) seamlessly joined to a cylindrical section 9.9 mm in length. RF power is coupled into xenon plasma by eight RF-excitation coils (four around the dome, four around the cylinder), which provide rotationally symmetric coupling of the 2.2 MHz drive into the xenon plasma (Joshi et al., 12 Sep 2025).

The downstream ion optics comprise a two-grid ion optics assembly. The grounded screen grid (Grid 1) is 0.25 mm thick and contains 37 circular apertures (radius $r_1 = 0.95\ \mathrm{mm}$) arranged in a six-fold symmetry. The acceleration grid (Grid 2) is 1.5 mm thick, has hole radius $r_2 = 0.60\ \mathrm{mm}$, and is held at potentials up to $V_{\rm ext} = 1\,300\ \mathrm{V}$. The gap between the two grids is $d = 20.85\ \mathrm{mm}$. Xenon is supplied through a peripheral gas inlet near the extraction plane (Joshi et al., 12 Sep 2025).

This architecture places the RIT-2.5 within the family of gridded RF ion thrusters, where RF power primarily sustains and heats the discharge plasma, while thrust is generated by electrostatic extraction of positively charged ions. The theoretical treatment associated with the RIT class emphasizes that the extraction model can be generalized for all thruster types that use electrostatic fields to extract positively charged ions (Kirmse, 2013). This suggests that the RIT-2.5 is not only a propulsion device but also a tractable reference geometry for studying generic beam-extraction physics.

2. Operating regime and measured plasma state

During the experiments reported for the RIT-2.5, the vacuum facility operated at a base pressure $< 5\times10^{-6}\ \mathrm{mbar}$; with xenon flow, the working pressure rose to $(2–5)\times10^{-5}\ \mathrm{mbar}$ (Joshi et al., 12 Sep 2025). A calibrated mass-flow controller delivered $0.12–0.24\ \mathrm{sccm}$ of Xe, corresponding to neutral densities of order $10^{18}\ \mathrm{m}^{-3}$ in the chamber. The forward RF power was varied between 22 W and 46 W by adjusting the RF voltage between 14 V and 20 V at fixed frequency 2.2 MHz, and the primary ionic species extracted is Xe$^+$ (Joshi et al., 12 Sep 2025).

Electron temperature was diagnosed by optical emission spectroscopy (OES) using two Xe I lines at $r_1 = 0.95\ \mathrm{mm}$0 and $r_1 = 0.95\ \mathrm{mm}$1, interpreted under the coronal-model and assuming a Maxwellian electron energy distribution. The photon-flux ratio was written as

$r_1 = 0.95\ \mathrm{mm}$2

with $r_1 = 0.95\ \mathrm{mm}$3 the electron-impact excitation cross sections. The collisional-radiative model LPP0D was used to compute $r_1 = 0.95\ \mathrm{mm}$4, yielding $r_1 = 0.95\ \mathrm{mm}$5 in the range 3–7 eV over the experimental conditions (Joshi et al., 12 Sep 2025). The reported trend is that $r_1 = 0.95\ \mathrm{mm}$6 increases modestly with RF power and decreases slightly with higher mass-flow rates; a representative value is $r_1 = 0.95\ \mathrm{mm}$7 at $r_1 = 0.95\ \mathrm{mm}$8 and $r_1 = 0.95\ \mathrm{mm}$9 (Joshi et al., 12 Sep 2025).

A central result of the combined experimental-numerical analysis is an inferred plasma density of $r_2 = 0.60\ \mathrm{mm}$0, obtained by fitting simulated ion energy distributions to measured RFEA spectra (Joshi et al., 12 Sep 2025). This density could not be accessed directly because of the lack of intrusive diagnostics. The study therefore frames the RIT-2.5 as a case where non-intrusive diagnostics and simulation must be combined to reconstruct the discharge state.

3. Ion extraction and sheath physics

The microscopic description developed for the RIT class begins from a one-dimensional, planar plasma-grid separation with ions of charge $r_2 = 0.60\ \mathrm{mm}$1 and mass $r_2 = 0.60\ \mathrm{mm}$2 drawn across a gap $r_2 = 0.60\ \mathrm{mm}$3 by an applied potential $r_2 = 0.60\ \mathrm{mm}$4. In steady state, the ion current density is

$r_2 = 0.60\ \mathrm{mm}$5

and under the cold-ion approximation, energy conservation gives

$r_2 = 0.60\ \mathrm{mm}$6

Combining this with Poisson’s equation leads to the planar Child-Langmuir law for space-charge-limited ion current density (Kirmse, 2013):

$r_2 = 0.60\ \mathrm{mm}$7

The later RIT-2.5 study explicitly adopts the same law, writing

$r_2 = 0.60\ \mathrm{mm}$8

and noting that the coarse grid geometry and sheath curvature modify $r_2 = 0.60\ \mathrm{mm}$9 slightly, but the $V_{\rm ext} = 1\,300\ \mathrm{V}$0 scaling is observed experimentally in the Faraday-cup currents (Joshi et al., 12 Sep 2025). This establishes continuity between the idealized extraction model and measured device behavior.

At the plasma-grid interface, the microscopic treatment considers an RF voltage $V_{\rm ext} = 1\,300\ \mathrm{V}$1 superimposed on a dc bias, producing a time-varying sheath voltage

$V_{\rm ext} = 1\,300\ \mathrm{V}$2

A simplified collisionless oscillating-sheath model gives an asymptotic scaling for the instantaneous sheath thickness (Kirmse, 2013):

$V_{\rm ext} = 1\,300\ \mathrm{V}$3

To first order, the sheath potential profile is taken as

$V_{\rm ext} = 1\,300\ \mathrm{V}$4

and ions crossing the sheath acquire energy approximately equal to the sheath voltage at their entry time, yielding an ion energy distribution function with a spread of order

$V_{\rm ext} = 1\,300\ \mathrm{V}$5

The same theoretical source notes that, in most designs, $V_{\rm ext} = 1\,300\ \mathrm{V}$6, so the energy spread remains a few tens of eV around the mean $V_{\rm ext} = 1\,300\ \mathrm{V}$7 (Kirmse, 2013). In the RIT-2.5 experimental study, the measured beam energy spreads are indeed of order tens of eV, which is consistent with that description (Joshi et al., 12 Sep 2025).

The later study adds that ions enter the sheath at at least the Bohm velocity

$V_{\rm ext} = 1\,300\ \mathrm{V}$8

which is cited as the condition ensuring sheath stability (Joshi et al., 12 Sep 2025). This connects plasma thermodynamics, sheath formation, and stable extraction in a compact RF ion source.

4. Beam formation, diagnostics, and energy-distribution measurements

Beam current was measured using a stainless-steel Faraday cup located $V_{\rm ext} = 1\,300\ \mathrm{V}$9 downstream on axis, with a suppression grid at $d = 20.85\ \mathrm{mm}$0 to prevent secondary-electron escape. From the collected current $d = 20.85\ \mathrm{mm}$1 and the geometrical acceptance, the on-axis beam current density $d = 20.85\ \mathrm{mm}$2 was obtained (Joshi et al., 12 Sep 2025). The ion beam current scales approximately linearly with RF power, rising from $d = 20.85\ \mathrm{mm}$3 at 22 W to $d = 20.85\ \mathrm{mm}$4 at 46 W for $d = 20.85\ \mathrm{mm}$5 (Joshi et al., 12 Sep 2025).

Ion energy distribution functions were measured with a three-electrode retarding field energy analyzer (RFEA) consisting of a grounded entrance grid, a retarding grid biased from 0 to $d = 20.85\ \mathrm{mm}$6 in 1 V steps at $d = 20.85\ \mathrm{mm}$7, and a collector. After smoothing the $d = 20.85\ \mathrm{mm}$8-$d = 20.85\ \mathrm{mm}$9 trace with a Savitzky-Golay filter, the differential $< 5\times10^{-6}\ \mathrm{mbar}$0 yielded the energy distribution up to a resolution of $< 5\times10^{-6}\ \mathrm{mbar}$1 (Joshi et al., 12 Sep 2025).

The measured distributions show a relatively narrow dominant beam component. At $< 5\times10^{-6}\ \mathrm{mbar}$2, the EDF displays a pronounced peak at $< 5\times10^{-6}\ \mathrm{mbar}$3 and a full width at half maximum $< 5\times10^{-6}\ \mathrm{mbar}$4 (Joshi et al., 12 Sep 2025). As RF power or mass flow is varied, the peak shifts by up to $< 5\times10^{-6}\ \mathrm{mbar}$5 and $< 5\times10^{-6}\ \mathrm{mbar}$6 changes by 20–50 eV (Joshi et al., 12 Sep 2025). These values are technically significant because they show that the extraction system produces a beam with an energy close to the applied extraction potential, but with a measurable spread that encodes discharge conditions upstream of the grids.

The microscopic treatment of particle motion models the extracted ion through a combined electric field

$< 5\times10^{-6}\ \mathrm{mbar}$7

with equation of motion

$< 5\times10^{-6}\ \mathrm{mbar}$8

Under the approximations $< 5\times10^{-6}\ \mathrm{mbar}$9, negligible magnetic fields and collisions, and nearly uniform field across the sheath, integration gives

$(2–5)\times10^{-5}\ \mathrm{mbar}$0

and the transit time $(2–5)\times10^{-5}\ \mathrm{mbar}$1 is determined by

$(2–5)\times10^{-5}\ \mathrm{mbar}$2

For $(2–5)\times10^{-5}\ \mathrm{mbar}$3, the RF correction is small and

$(2–5)\times10^{-5}\ \mathrm{mbar}$4

This model provides the microphysical basis for the interpretation of the measured beam as predominantly electrostatically accelerated, with RF modulation acting primarily through the sheath and energy spread rather than by dominating the net acceleration (Kirmse, 2013).

5. Thrust, specific impulse, and reported performance

In the electrostatic beam picture, if the total beam current is $(2–5)\times10^{-5}\ \mathrm{mbar}$5, then the mass-flow rate is

$(2–5)\times10^{-5}\ \mathrm{mbar}$6

the exhaust velocity is

$(2–5)\times10^{-5}\ \mathrm{mbar}$7

and the thrust is

$(2–5)\times10^{-5}\ \mathrm{mbar}$8

The specific impulse follows as

$(2–5)\times10^{-5}\ \mathrm{mbar}$9

These relations are given explicitly in the RIT microscopic treatment (Kirmse, 2013). The later RIT-2.5 study states the same scaling in condensed form: $0.12–0.24\ \mathrm{sccm}$0 and $0.12–0.24\ \mathrm{sccm}$1 (Joshi et al., 12 Sep 2025).

The experimental characterization reports that the total thrust, calculated as $0.12–0.24\ \mathrm{sccm}$2, lies in the 5–15 $0.12–0.24\ \mathrm{sccm}$3 range (Joshi et al., 12 Sep 2025). It further states that, for micro-Newton thrust with $0.12–0.24\ \mathrm{sccm}$4 and $0.12–0.24\ \mathrm{sccm}$5, the optimal operating point is found near $0.12–0.24\ \mathrm{sccm}$6, $0.12–0.24\ \mathrm{sccm}$7, and $0.12–0.24\ \mathrm{sccm}$8 (Joshi et al., 12 Sep 2025).

The theoretical treatment includes a separate numerical application to a typical RIT-2.5 operating point using Xe$0.12–0.24\ \mathrm{sccm}$9, $10^{18}\ \mathrm{m}^{-3}$0, $10^{18}\ \mathrm{m}^{-3}$1, $10^{18}\ \mathrm{m}^{-3}$2, and discharge pressure $10^{18}\ \mathrm{m}^{-3}$3. Under the assumptions stated there, it yields Child-Langmuir limited $10^{18}\ \mathrm{m}^{-3}$4, beam current $10^{18}\ \mathrm{m}^{-3}$5, thrust $10^{18}\ \mathrm{m}^{-3}$6, exhaust velocity $10^{18}\ \mathrm{m}^{-3}$7, and $10^{18}\ \mathrm{m}^{-3}$8, with beam-power efficiency $10^{18}\ \mathrm{m}^{-3}$9 and overall $^+$0 in practice (Kirmse, 2013).

Because the later experimental paper reports a grid gap $^+$1 and RF frequency $^+$2 for the characterized device (Joshi et al., 12 Sep 2025), these numerical values should not be read as a direct measurement of the same hardware configuration. A plausible implication is that the earlier numerical evaluation functions as an idealized or class-level example for a thruster of the RIT-2.5 type, whereas the later paper provides experimentally resolved performance for a specific realized configuration.

6. Simulation framework, validation, and parameter dependencies

The RIT-2.5 has been analyzed with a two-dimensional, axisymmetric particle-in-cell (PIC) code, PlasmaPIC, used to model both the discharge plasma and ion extraction (Joshi et al., 12 Sep 2025). The simulation solves the Vlasov equation for each species $^+$3,

$^+$4

together with Poisson’s equation,

$^+$5

Ions and electrons evolve under Newton’s equations, and collisions with neutrals are included through a Monte Carlo collision (MCC) scheme using standard Xe cross sections (Joshi et al., 12 Sep 2025). The RF drive is imposed as a time-varying boundary potential on the coil region at 2.2 MHz. The computational domain includes a minimal discharge region (radius 15 mm, axial length 20 mm) and an extended extraction region up to 150 mm downstream; the mesh is typically $^+$6 cells, chosen to resolve the Debye length ($^+$7) and RF period ($^+$8) (Joshi et al., 12 Sep 2025).

For the nominal case $^+$9, $r_1 = 0.95\ \mathrm{mm}$00, the simulation shows that a sheathed plasma forms near the extraction apertures, accelerating ions into a collimated beam (Joshi et al., 12 Sep 2025). The simulated EDF just downstream of the grids exhibits a narrow peak at $r_1 = 0.95\ \mathrm{mm}$01 and $r_1 = 0.95\ \mathrm{mm}$02, matching the RFEA measurement. The reported quantitative agreement is $r_1 = 0.95\ \mathrm{mm}$03 in $r_1 = 0.95\ \mathrm{mm}$04 and $r_1 = 0.95\ \mathrm{mm}$05 in peak, which the authors interpret as validation of the combined experimental-numerical approach (Joshi et al., 12 Sep 2025).

A sweep over plasma density and electron temperature produced the following simulated EDF parameters at $r_1 = 0.95\ \mathrm{mm}$06 (Joshi et al., 12 Sep 2025):

Case	$r_1 = 0.95\ \mathrm{mm}$07 ($r_1 = 0.95\ \mathrm{mm}$08)	EDF Peak / $r_1 = 0.95\ \mathrm{mm}$09 (eV)
A	0.5	1192 / 18
B	0.8	1191 / 24
C	1.0	1188 / 36
D	1.2	1185 / 42
E	1.4	1182 / 53
F	1.6	1178 / 66

The best fit to experiment occurs for Case D, with $r_1 = 0.95\ \mathrm{mm}$10 and $r_1 = 0.95\ \mathrm{mm}$11 (Joshi et al., 12 Sep 2025). The study further states that $r_1 = 0.95\ \mathrm{mm}$12 is approximately linearly dependent on $r_1 = 0.95\ \mathrm{mm}$13 for fixed $r_1 = 0.95\ \mathrm{mm}$14, and proposes a simple linear fit

$r_1 = 0.95\ \mathrm{mm}$15

with $r_1 = 0.95\ \mathrm{mm}$16 and $r_1 = 0.95\ \mathrm{mm}$17 per eV of electron temperature (Joshi et al., 12 Sep 2025). The reported interpretation is that high $r_1 = 0.95\ \mathrm{mm}$18 increases beam current and thrust but broadens $r_1 = 0.95\ \mathrm{mm}$19, whereas low $r_1 = 0.95\ \mathrm{mm}$20 narrows $r_1 = 0.95\ \mathrm{mm}$21 but may limit ionization efficiency (Joshi et al., 12 Sep 2025).

7. Applications, interpretation, and technical caveats

The reported application domain is nanosatellite attitude and orbit control, especially where micro-N-class thrust is adequate and fine impulse resolution is required (Joshi et al., 12 Sep 2025). The study explicitly states that for CubeSat attitude control, the RIT-2.5’s 5–15 $r_1 = 0.95\ \mathrm{mm}$22 thrust range enables fine pointing and drag compensation (Joshi et al., 12 Sep 2025). The combination of OES, RFEA, and PIC is presented as a non-intrusive yet robust pathway for performance optimization, as well as a validation benchmark for plasma simulation codes (Joshi et al., 12 Sep 2025).

Two technical caveats are important. First, the literature associated with the RIT-2.5 contains both idealized analytic scaling arguments and device-specific measurements. The Child-Langmuir framework, Bohm criterion, and oscillating-sheath picture describe essential physics, but the experimental study explicitly notes that coarse grid geometry and sheath curvature modify $r_1 = 0.95\ \mathrm{mm}$23 slightly (Joshi et al., 12 Sep 2025). Second, the energy spread of the extracted beam should not be interpreted solely as an extraction-voltage artifact. The experimental-simulation campaign concludes that the energy spread is strongly dependent on plasma density and electron temperature (Joshi et al., 12 Sep 2025), and the theoretical microscopic analysis likewise ties the ion energy distribution to the dynamics of an oscillating sheath (Kirmse, 2013).

Future work identified in the experimental study includes mapping beam divergence, total efficiency, and lifetime, as well as investigation of alternate propellants (Joshi et al., 12 Sep 2025). This suggests that, although the RIT-2.5 is already sufficiently characterized to support model validation and operating-point optimization, its full engineering envelope remains an active subject of study. The available evidence positions it as both a practical compact ion source and a well-instrumented platform for connecting microscopic plasma-sheath physics to measurable propulsion performance [(Kirmse, 2013); (Joshi et al., 12 Sep 2025)].