Electric Propulsion: Why Ion Drives Are Changing Deep Space Exploration

The exhaust from a chemical rocket leaves at roughly 3–4.5 km/s. The exhaust from an ion thruster leaves at 20–80 km/s. That difference — a factor of 10 to 20 in exhaust velocity — is the entire argument for electric propulsion, expressed in a single comparison.

Specific impulse (Isp), the standard measure of propellant efficiency, is proportional to exhaust velocity. A high-Isp engine extracts far more velocity change from each kilogram of propellant it consumes. The consequence, through the Tsiolkovsky rocket equation, is that a spacecraft with an ion drive can achieve far greater total delta-v from a given propellant mass than any chemical rocket — or, equivalently, can reach the same destination carrying far less propellant and far more payload.

The catch is thrust. Ion drives produce millinewtons to a few newtons of force. Chemical rockets produce meganewtons. Ion drives cannot fight Earth’s gravity; they cannot launch a spacecraft or execute rapid orbital manoeuvres. But in the vacuum between planets, with months or years to run continuously, the cumulative velocity change they deliver is extraordinary.

Key parameters

Parameter	Gridded Ion (NSTAR)	Hall Thruster (SPT-100)
Specific impulse (Isp)	~3,100 s	~1,600 s
Thrust	92 mN	83 mN
Power	2.3 kW	1.35 kW
Propellant	Xenon	Xenon
Exhaust velocity	~30–40 km/s	~15–20 km/s

How Ion Drives Work

An ion thruster operates by ionising a propellant gas — typically xenon, chosen for its high atomic mass, chemical inertness, and convenient room-temperature storage — and then accelerating the resulting ions through an electric field.

In a gridded ion engine (the type used on NASA’s Dawn spacecraft), xenon is fed into an ionisation chamber where electron bombardment strips each atom of one electron, creating a Xe⁺ ion. The ions are then drawn toward a pair of electrically charged grids. The inner grid (the screen grid) is at high positive voltage, typically 1,000–1,500 volts. The outer grid (the accelerator grid) is at a few hundred volts negative. The electric field between the grids accelerates each ion to exhaust velocities of 30–40 km/s. A neutraliser cathode sprays electrons into the exhaust to prevent the spacecraft from accumulating a net charge that would decelerate the departing ions.

The result is a faint blue-white exhaust — the visible glow of xenon plasma — and a thrust force that could not lift a sheet of paper at Earth’s surface but can, over three years of continuous operation, change a spacecraft’s velocity by more than 11 km/s.

Hall Effect Thrusters: The Alternative Architecture

Hall thrusters, developed extensively by Soviet engineers from the 1960s onward and now the dominant electric propulsion technology in commercial satellite operations, use a different ion acceleration mechanism. Rather than electrostatic grids, they trap electrons in a magnetic field (the Hall effect), and use these trapped electrons to ionise propellant and accelerate ions through an electric potential maintained between an anode at the back of the thruster and the plasma boundary at the exit.

Hall thrusters operate at lower specific impulse than gridded ion engines (typically 1,500–2,000 s versus 3,000–10,000 s for gridded engines) but produce higher thrust density and are mechanically simpler — no erosion-prone grids. These properties make them the thruster of choice for stationkeeping on GEO satellites and increasingly for orbit-raising on large commercial satellites, reducing the amount of chemical propellant that must be launched.

The SPT-100, developed by Fakel in Russia in the 1970s, became the standard Hall thruster of the Soviet/Russian satellite industry and was widely licensed internationally. More than 250 SPT-100s had flown by 2020. SpaceX uses a custom Hall thruster variant in its Starlink satellites for orbit maintenance and disposal. ESA’s Alphasat carried a Snecma PPS-5000 Hall thruster capable of 250 mN thrust — at the time the most powerful Hall thruster flown in Europe.

Dawn: The Mission That Proved the Technology

NASA’s Dawn mission, launched in 2007 and managed by Marc Rayman at JPL, was the first spacecraft to orbit two extraterrestrial bodies in sequence — the asteroid Vesta and the dwarf planet Ceres — and it did so entirely because of ion propulsion.

Dawn’s three NSTAR gridded ion engines (a development of the technology tested on the Deep Space 1 technology demonstrator in 1998) operated on xenon propellant. Over the mission’s ten years, Dawn accumulated 5.9 km/s of delta-v just from its ion engines, on top of the delta-v provided by its launcher and a Mars gravity assist. The total delta-v from ion propulsion exceeded what any chemical rocket of equivalent mass could have provided by a factor of roughly 10.

The NSTAR thruster operates at 2.3 kW power consumption, producing 92 mN of thrust and a specific impulse of ~3,100 s. Running continuously for months is normal — the engines were throttled rather than shut down during deep space cruise, with interruptions only for navigation tracking and scheduled maintenance windows.

When Dawn arrived at Vesta in 2011, it spent 14 months in orbit, mapping the surface in detail, before ion-propelling itself out of Vesta’s gravity well and on to Ceres. No chemical propulsion system could have achieved this orbital flexibility at that spacecraft mass. Without ion drives, the mission architecture would have required separate spacecraft for each body.

Hayabusa and the Sample Return Revolution

JAXA’s Hayabusa missions demonstrated something equally significant: that electric propulsion enables sample return from small bodies at modest mission cost.

Hayabusa (2003–2010) used four Mitsubishi Electric μ10 ion engines, each producing 8 mN of thrust, to travel to near-Earth asteroid Itokawa, touch down twice (with difficulty — the surface sampling mechanism partially failed), and return a capsule containing approximately 1,500 microscopic grains of asteroid material to Earth. The ion engines accumulated 31,000 hours of operation. The trajectory required multiple Earth gravity assists and years of patient ion thrusting.

Hayabusa2 (2014–2020) repeated and improved the concept at asteroid Ryugu, successfully deploying impactors to expose subsurface material, collecting samples from both surface and subsurface, and returning 5.4 grams of pristine carbonaceous asteroid material — among the most scientifically valuable extraterrestrial material collected since the Apollo samples. The mission cost approximately ¥30 billion (~€200 million), modest by outer solar system mission standards.

NASA’s OSIRIS-REx (now OSIRIS-APEX), which returned 121 grams of Bennu material in September 2023, uses a different approach — monopropellant hydrazine thrusters rather than ion drives — but the Hayabusa lineage established the sample-return-from-small-body mission class as routine.

BepiColombo: Ion Propulsion to Mercury

The European-Japanese BepiColombo mission to Mercury, launched in 2018 and arriving at Mercury in 2026, provides the most demanding demonstration of ion propulsion in the inner solar system.

Mercury is thermodynamically difficult to reach. Unlike the outer planets, where gravity assists from Jupiter or Saturn can provide large velocity gains, reaching Mercury requires shedding angular momentum relative to the Sun — slowing down, not speeding up. A direct chemical trajectory to Mercury requires enormous delta-v. BepiColombo uses four QinetiQ T6 ion thrusters, each operating at 145 mN thrust and ~4,500 s specific impulse, combined with multiple gravity assists (one Earth, two Venus, six Mercury flybys) to thread a trajectory that would be impossible for chemical propulsion alone at the mission’s 4,000 kg launch mass.

The T6 thrusters run at 6.9 kW apiece, powered by BepiColombo’s solar panels — which at Mercury’s distance from the Sun receive roughly six times the solar irradiance of Earth, offsetting the efficiency losses from operating in higher-temperature conditions.

Power: The Fundamental Constraint

Ion propulsion systems are ultimately limited by available electrical power. Thrust is proportional to power input. In the inner solar system, solar panels are effective; at 1 AU, a 10 kW solar array is readily achievable at reasonable mass. At Jupiter’s distance, solar irradiance drops to 3.7% of Earth’s value — 10 kW requires a panel 27 times larger. At Saturn, it drops to 1.1%.

This is why deep outer solar system missions — Europa Clipper, Dragonfly (Titan), Cassini — use nuclear power sources (RTGs or fission reactors) rather than electric propulsion. NASA and DARPA are actively developing the DRACO program (Demonstration Rocket for Agile Cislunar Operations) to demonstrate a nuclear thermal engine, which combines nuclear reactor heat with hydrogen propellant to achieve Isp of ~800–900 s — far below ion drives, but with thrust levels approaching chemical rockets.

For the near term, ion propulsion and Hall thrusters will continue their expansion into commercial satellite operations, deep-space small missions, and sample return. The technology that produced a faint blue glow in a JPL test chamber in the 1960s has become standard equipment for the most capable robotic explorers humanity has ever built.

For the trajectory mathematics that determine where these engines need to point, see delta-v and orbital mechanics.