The Tyranny of Chemistry
Every rocket that has ever carried humans to space has burned chemical propellants—hydrogen and oxygen, kerosene and oxygen, hydrazine, or other combinations. Chemical propulsion works. It provides the thrust needed to escape Earth's gravity and the reliability demanded by crewed missions.
But chemical propulsion is approaching fundamental limits. The energy released by breaking and forming chemical bonds can only push exhaust so fast. The best chemical rockets achieve a specific impulse of roughly 450 seconds—a measure of propellant efficiency.¹ This is close to the theoretical maximum for hydrogen-oxygen combustion.
For getting off Earth, where high thrust matters more than efficiency, chemical rockets will remain dominant for the foreseeable future. But once in space, where thrust can be applied gradually over days or weeks, other options become attractive. An electric thruster using ionized xenon can achieve 3,000 seconds of specific impulse—seven times the efficiency of chemical rockets. It produces tiny thrust, but over time that efficiency compounds into enormous velocity changes.
For truly ambitious missions—rapid transit to Mars, outer planet exploration in reasonable timeframes, and eventually interstellar precursors—something beyond chemical propulsion becomes necessary. Nuclear thermal rockets, tested in the 1960s, could nearly double chemical performance. Nuclear electric systems could match ion drives while providing more power. And further out, concepts from solar sails to fusion drives to antimatter propulsion inhabit various stages of speculation and development.
This chapter examines propulsion options beyond chemical rockets: what works now, what's in development, and what physics allows even if engineering doesn't yet permit.
2026 Snapshot — Propulsion State of the Art
Chemical Propulsion (Still Dominant)
Liquid engines remain the workhorses:
- Staged combustion (SpaceX Raptor, RS-25): High efficiency, complex engineering
- Gas generator (Merlin, F-1): Simpler, slightly less efficient
- Pressure-fed: Simple but limited to smaller engines
- Methane (Raptor, BE-4): Emerging as preferred fuel for reusability and Mars ISRU
Solid rockets provide high thrust in boosters (SLS, Ariane) but cannot be throttled or shut down.
Performance plateau: Modern engines approach theoretical limits. Improvements come from materials, manufacturing, and operations rather than physics breakthroughs.
Electric Propulsion (Operational and Growing)
Hall thrusters and ion engines are standard for satellite station-keeping and increasingly for primary propulsion:
- Starlink satellites use Hall thrusters for orbit raising and maintenance
- Boeing 702SP satellites use all-electric propulsion
- NASA's Dawn mission used ion propulsion to orbit two asteroids
- ESA's BepiColombo is using ion engines for Mercury trajectory
Performance: Specific impulse 1,500-3,500 seconds; thrust measured in millinewtons; power limited by solar arrays (typically 1-30 kW).
Limitations: Very low thrust means very long maneuver times. A geostationary satellite might take six months to raise its orbit using electric propulsion versus hours for chemical.
Nuclear Propulsion (Research Stage)
No nuclear rockets currently fly, but interest has revived:
- NASA's DRACO program (Demonstration Rocket for Agile Cislunar Operations) aimed to demonstrate nuclear thermal propulsion by 2027 but was cancelled in May 2025
- DARPA partnership with General Atomics, Lockheed Martin, and BWX Technologies
- Estimated performance: ~900 seconds specific impulse—double chemical rockets
Challenges: Nuclear safety for launch, political acceptability, regulatory frameworks, and decades since the US last tested nuclear rockets (NERVA program ended 1972).
Advanced Concepts (Experimental/Theoretical)
Various technologies are in research or early development:
- Variable Specific Impulse Magnetoplasma Rocket (VASIMR): High-power electric thruster; ISS testing planned
- Solar sails: Demonstrated (JAXA's IKAROS, The Planetary Society's LightSail 2); provides continuous thrust from sunlight
- Pulsed plasma thrusters: Low-power electric propulsion for small satellites
- Electrospray thrusters: Very small thrusters using ionic liquids
Notable Players
Electric Propulsion Manufacturers
Busek: US company providing Hall thrusters for small and medium satellites.
Aerojet Rocketdyne: Ion engines (including those for Dawn mission) and Hall thrusters.
Safran (formerly Snecma): European electric propulsion systems including Hall thrusters.
SpaceX: Develops and manufactures Hall thrusters for Starlink constellation—one of the largest deployments of electric propulsion ever.
Exotrail, ThrustMe, Enpulsion, Phase Four: Startups developing electric propulsion for small satellites.
Nuclear Propulsion Development
BWXT (BWX Technologies): Nuclear fuel and reactor expertise; DRACO program partner.
General Atomics: Aerospace and nuclear experience; developing reactor concepts.
Lockheed Martin: Spacecraft integration for DRACO; broader nuclear propulsion interest.
NASA/DOE: Government programs providing funding and facilities for nuclear propulsion development.
Russia (Roscosmos): Has discussed nuclear propulsion concepts; claims development programs but limited public information.
Advanced Concepts
Ad Astra Rocket Company: Developing VASIMR; founded by former astronaut Franklin Chang-Díaz.
The Planetary Society: Demonstrated solar sail technology; advocates for space exploration.
Breakthrough Starshot: Funded research into laser-driven light sails for interstellar probes.
University and national lab research: Various programs exploring fusion propulsion, antimatter concepts, and physics-based advances.
Electric Propulsion: The Current Frontier
How It Works
Electric propulsion uses electrical energy to accelerate propellant, rather than chemical combustion:
Ion engines: Ionize a propellant (typically xenon) and accelerate ions through electric fields. Dawn's engine accelerated xenon ions to 90,000 mph. Very high efficiency; very low thrust.
Hall thrusters: Use magnetic fields to trap electrons, which ionize propellant and accelerate it electrostatically. Simpler than ion engines; slightly lower efficiency but higher thrust.
Pulsed plasma thrusters: Ablate solid propellant with electric arcs, creating plasma pulses. Low power; suitable for very small satellites.
Magnetoplasmadynamic (MPD) thrusters: Accelerate plasma using magnetic fields. Can achieve high power and thrust but have efficiency challenges.
Current Capabilities
Power levels: Most operational systems use 1-20 kW. Higher power requires larger solar arrays or nuclear power sources.
Propellants: Xenon dominates (inert, high atomic mass, easily ionized) but is expensive and limited in supply. Krypton (cheaper but less efficient) and alternatives are being developed.
Thrust: Millinewtons to tens of millinewtons—a fraction of a percent of chemical engine thrust. Useful for gradual orbit changes, not launch.
Specific impulse: 1,500-4,000 seconds depending on technology—3-10 times chemical rocket efficiency.
What's Advancing
Higher power systems: VASIMR targets 200 kW and beyond, suitable for fast interplanetary transit.
Alternative propellants: Iodine, water, air-breathing systems for very low orbit—expanding options beyond expensive xenon.
Miniaturization: Electric propulsion for CubeSats and small satellites, enabling deep space missions with tiny spacecraft.
Solar Electric Propulsion (SEP) missions: NASA's Gateway lunar station will use large SEP for orbit maintenance. Commercial interest in SEP tugs for cislunar transportation.
Limitations
Thrust-to-weight: Electric propulsion cannot escape Earth's gravity—chemical rockets remain necessary for launch.
Power requirements: High-performance electric propulsion needs significant power. Solar works in inner solar system; nuclear required for outer planets.
Trip time: Low thrust means long spirals out of gravity wells. A Mars transit using electric propulsion might take longer than chemical unless very high power is available.
Nuclear Propulsion: The Dormant Giant
Nuclear Thermal Propulsion (NTP)
How it works: A nuclear reactor heats hydrogen propellant to very high temperatures; the hot gas exits through a nozzle, producing thrust. No combustion—pure heat transfer.
Performance: Specific impulse ~900 seconds—double chemical rockets. Thrust comparable to chemical engines.
History: The US developed and tested NTP under Project NERVA (Nuclear Engine for Rocket Vehicle Application) from 1955-1972. Twenty reactors were built and tested. The technology worked—but the program was canceled when Apollo ended.
Why it matters: NTP could cut Mars transit time from 7-9 months to 3-4 months. Faster transit reduces radiation exposure, consumables, and mission risk.
Current status: NASA's DRACO program aimed to demonstrate NTP in orbit by 2027 but was cancelled in May 2025 due to budget constraints. Revived interest in nuclear propulsion continues but without an active flight demonstration program.
Nuclear Electric Propulsion (NEP)
How it works: A nuclear reactor generates electricity, which powers electric thrusters (ion, Hall, or MPD). Separates power generation from propulsion.
Advantages: Nuclear reactors can provide high power far from the sun; electric thrusters are highly efficient. Best of both worlds for outer solar system missions.
Performance: Specific impulse of 3,000-10,000 seconds depending on thruster type; continuous thrust over months or years.
Challenges: Reactor mass; heat rejection in vacuum (requires large radiators); technology development needed.
Status: No NEP system has flown, but concepts are studied for outer planet missions and eventually human Mars missions.
Bimodal Nuclear Systems
Concept: A single reactor provides both thermal propulsion (high thrust for orbit changes) and electric power (for payload and life support).
Advantages: More flexible mission profiles; shared reactor mass.
Status: Conceptual studies; no hardware development yet.
Challenges and Concerns
Launch safety: Flying nuclear material raises concerns about accidents. Modern reactor designs can remain subcritical until activated in space. International guidelines exist.
Regulatory framework: No current US regulatory pathway for space nuclear propulsion. New frameworks needed.
Political will: Public perception of nuclear power affects space nuclear development. Program continuity across administrations is uncertain.
Technical atrophy: Decades since NERVA means lost expertise and manufacturing capability. Rebuilding takes time.
Testing: Ground testing nuclear engines is challenging. The Nevada test site used for NERVA is not currently configured for such tests.
Advanced and Speculative Propulsion
Solar Sails
How they work: Large reflective surfaces capture momentum from photons (sunlight), providing continuous thrust without propellant.
Demonstrated: JAXA's IKAROS (2010) flew to Venus using a solar sail. The Planetary Society's LightSail 2 demonstrated sail deployment and orbit raising.
Performance: Very low thrust (micronewtons to millinewtons), but no propellant means unlimited operation time. Suitable for missions requiring continuous thrust over years.
Applications: Interplanetary missions, station-keeping, potential interstellar precursors (if launched close to the Sun for high initial velocity).
Limitations: Thrust decreases with distance from Sun squared. Not useful beyond Jupiter. Sail deployment and control at large scales remain challenging.
Laser-Driven Light Sails
Breakthrough Starshot concept: Use ground-based or orbital lasers to accelerate tiny probes to relativistic speeds (20% of light speed). Could reach Alpha Centauri in 20 years.
Status: Funded research into materials, photonics, and system architecture. No hardware demonstrated at scale.
Challenges: Requires gigawatts of laser power focused on a target receding at kilometers per second. Materials must survive extreme acceleration. Communication from tiny probes at interstellar distances is extremely difficult.
Timeline: Decades at minimum; fundamental feasibility still being assessed.
Fusion Propulsion
Concept: Harness nuclear fusion (the process powering the Sun) for propulsion. Potential for very high specific impulse (10,000-1,000,000 seconds) and significant thrust.
Variants:
- Magnetic confinement: Use tokamak-like systems to contain fusion plasma, directing exhaust for thrust
- Inertial confinement: Ignite fusion pellets, capturing thrust from explosions
- Magneto-inertial fusion: Hybrid approaches
Status: Fusion for ground power isn't yet practical; fusion for propulsion is further out. Some concepts being studied by NASA and DARPA. Private companies (Helion, General Fusion) focus on power, but propulsion applications could follow.
If fusion propulsion works: Mars transit in weeks; outer solar system missions in months; interstellar precursors become thinkable.
Antimatter Propulsion
Concept: Annihilate matter and antimatter, releasing energy at E=mc² efficiency—orders of magnitude beyond any other reaction.
Reality check: Antimatter production is extraordinarily expensive. Total worldwide production is micrograms per year. Storage is extremely difficult. No antimatter rocket has ever been designed beyond paper studies.
If antimatter becomes cheap: Theoretical potential for interstellar travel at significant fractions of light speed. But this is physics fiction, not engineering reality, for the foreseeable future.
Reactionless and Exotic Drives
EmDrive and similar: Claimed thrust without propellant, violating conservation of momentum. Subsequent testing has not confirmed claimed effects. No known physics supports such drives.
Warp drives: Theoretical concepts (Alcubierre drive) for faster-than-light travel require exotic matter with negative energy density—not known to exist in usable form. See Chapter 21.
Breakthrough Propulsion Physics: NASA funded research from 1996-2002 exploring physics-based alternatives. No practical propulsion emerged; program ended.
The Path Forward
Near-Term Likely (2026-2032)
Electric propulsion expands: Higher power systems (50-100+ kW) enable faster orbit transfers. SEP tugs become commercial. Starlink and successors deploy thousands more electric thrusters.
Nuclear thermal research continues: Despite DRACO's cancellation, continued research into NTP technology. Future demonstration programs may emerge with renewed funding.
Solar sails mature: Multiple missions demonstrate sail technology; near-Earth asteroid missions use sails for propellantless operation.
Alternative propellants: Water, iodine, and other non-xenon propellants reduce costs and expand electric propulsion access.
Plausible (2032-2040)
NTP operational: Nuclear thermal propulsion used for fast cargo missions to Mars; possibly crewed missions. Transit times reduced to 3-4 months.
High-power NEP: 100+ kW nuclear electric systems enable outer planet missions with reasonable timelines. Europa and Titan missions use NEP.
Fusion propulsion research: Ground demonstrations of fusion relevant to propulsion; first space tests possible by late 2030s.
Laser-driven demonstrators: Small probes accelerated to significant fractions of solar escape velocity, validating concepts.
Wild Trajectory (2040+)
Fusion-powered spacecraft: If ground-based fusion succeeds, propulsion applications follow within a decade. Mars in weeks; Jupiter in months.
Interstellar precursors: Laser-driven sails or fusion probes reach tens of percent of light speed, heading for Alpha Centauri or other nearby stars with arrival in decades.
Breakthrough physics: Some advance in physics—not currently predicted but not impossible—enables propulsion beyond current understanding. This is speculation, not projection.
The AI Acceleration Factor
AI could accelerate propulsion development in multiple ways:
Materials discovery: AI identifies materials that withstand nuclear reactor cores, fusion plasmas, or extreme temperatures. Novel high-temperature superconductors could transform magnetic confinement.
Plasma physics: Fusion and plasma propulsion involve complex instabilities. AI could help control plasma behavior in real-time.
Design optimization: Generative design creates lighter, more efficient thruster components. Optimization across the full propulsion system improves performance.
Testing and validation: AI accelerates testing cycles, predicts failure modes, and enables virtual prototyping before expensive physical tests.
Mission design: AI optimizes trajectories for complex propulsion systems, finding paths that human planners might miss.
The development cycle for propulsion systems is typically 10-20 years. AI could compress this—but fundamental physics limits and safety certification requirements will remain.
Implications for the Next Decade
The propulsion landscape in 2035 will likely look significantly different from today:
Electric propulsion will be ubiquitous for satellites and tugs, with higher power and alternative propellants expanding applications.
Nuclear propulsion research continues despite DRACO's cancellation. Future demonstration programs will depend on renewed funding and political will to revive active development.
Chemical rockets will continue to dominate Earth-to-orbit launch but will be complemented by in-space propulsion for deep space.
Fusion and interstellar concepts will remain research programs, with timelines measured in decades.
The constraints are real: the rocket equation doesn't yield to clever engineering. But incremental improvements and the revival of nuclear propulsion could substantially expand what's possible in space—making Mars transit faster, outer planet exploration practical, and the solar system more accessible to human expansion.
Endnotes — Chapter 18
- Specific impulse (Isp) is measured in seconds and represents propellant efficiency—how much thrust is produced per unit of propellant consumed. Higher is better. Chemical rockets achieve 300-450 seconds; electric propulsion can exceed 3,000 seconds.
- Dawn mission (NASA/JPL) used ion propulsion to orbit Vesta and Ceres, demonstrating electric propulsion for primary mission propulsion.
- NERVA program (1955-1972) developed and tested nuclear thermal rockets at the Nuclear Rocket Development Station in Nevada. The program demonstrated feasibility but was canceled with post-Apollo budget cuts.
- DRACO program (DARPA/NASA) aimed to demonstrate NTP in orbit by 2027, with contracts awarded to General Atomics, Lockheed Martin, and BWX Technologies, but was cancelled in May 2025.
- VASIMR (Variable Specific Impulse Magnetoplasma Rocket) developed by Ad Astra Rocket Company targets high power (100-200 kW) electric propulsion with variable performance characteristics.
- LightSail 2 (The Planetary Society) demonstrated controlled solar sailing in Earth orbit from 2019-2022, proving the concept for small spacecraft.
- Breakthrough Starshot is a research program funded by Yuri Milner exploring laser-propelled light sails for interstellar probes, targeting 20% of light speed.
- Antimatter production rates are extremely low—CERN's antimatter factory produces nanograms per year at enormous cost. Practical antimatter propulsion would require production rates many orders of magnitude higher.
- EmDrive claims have been extensively tested by multiple groups; no confirmed thrust exceeding measurement error has been demonstrated. The concept lacks theoretical support.
- Alcubierre metric describes a theoretical warp drive requiring exotic matter with negative energy density. While mathematically valid in general relativity, no known matter has the required properties.