How can we fuel visits to Mars?

They say slow and steady wins the race. But that advice falls apart when the race lasts months, and while the competitors are bombarded by solar radiation and their bones and muscles weaken in microgravity.

That is the biggest challenge in sending humans to Mars. Every extra week in transit makes the spacecraft heavier with supplies, the life-support burden greater, and the crew’s living conditions more constrained. The propulsion system that the spacecraft uses determines how long the journey lasts and affects the size of the spacecraft, the cost of the mission, the health risks crews must survive, and even the launch window.

The realistic propulsion debate today includes a range of systems, including conventional chemical rockets, nuclear thermal engines, and ion drives. Each system offers various advantages in speed, efficiency, mission flexibility, and technical complexity.

Pursuing the proven path with chemical propulsion

“We don’t need any advanced propulsion to get to Mars,” says Robert Zubrin, founder of the Mars Society and the astronautical engineering company Pioneer Astronautics. “We can do it with chemical rockets.” They already work. They have launched every astronaut who has ever left Earth, and they remain the backbone of every serious near-term crewed Mars mission.

The rocket’s concept is simple: burning fuel releases energy.

Because combustion needs oxygen and space has none, rockets carry both fuel and oxygen. The two components react in a combustion chamber, producing hot gases that expand through a nozzle. As exhaust shoots backward, the rocket moves forward, following Newton’s third law: every action has an equal and opposite reaction. That concept explains how almost any rocket—known or theorized—works.

Chemical rockets’ main advantage is high thrust, which determines how quickly a rocket can accelerate. They can leave Earth’s gravity and reach space faster than any other rocket conceived of today.

The issue with chemical rockets is energy density. They need enormous amounts of fuel to move a heavy spacecraft. The more a rocket carries, the more fuel it needs. But fuel adds weight, creating a hard ceiling on how much propellant a chemical rocket can realistically carry, limiting its ultimate speed. Hydrogen is the most efficient chemical fuel in use today, requiring the least fuel to travel the farthest. And burning hydrogen in oxygen is a very clean combustion process; it produces only water vapor as exhaust. But for rocket propulsion, hydrogen and oxygen must be held in heavily insulated tanks with active cooling to remain in the liquid phase. Unfortunately, on a 6-to-9-month trip to Mars, the problem of keeping hydrogen cold enough to avoid losing it to evaporation remains unsolved.

Using hydrogen also forces the astronauts to carry fuel for the trip to Mars and back. That extra weight may not be necessary for the leading alternative chemical fuel.

Space X’s Starship, the most serious crewed Mars architecture being built today, burns liquid methane and oxygen. Methane is less efficient than hydrogen, but it stays liquid at temperatures almost 100 °C warmer than hydrogen. Methane’s bigger advantage comes from the Sabatier reaction.

Water ice in the Martian subsurface can be split by electrolysis into hydrogen and oxygen, an oxidizer. “CO₂ is 95% of the Martian atmosphere,” Zubrin says. Using the Sabatier reaction, “you react the hydrogen with carbon dioxide over a catalyst to produce methane and water,” he explains. That water can be electrolyzed again, creating a loop that converts Martian resources into rocket fuel.

“You now have the ability to manufacture the propellants to come home,” Zubrin says.

But producing enough propellant to fuel a return mission is a chemical engineering challenge. “At Pioneer Astronautics, we did around 5 kg a day,” Zubrin says. “To fuel a Starship, you need about a ton [0.9 metric tons] of propellant a day.” That’s 200 times more.

Even if astronauts can make fuel on Mars, chemical propulsion still has a speed limit. “When we want to go faster with chemical, you run into the issue of energy,” says Ryan Gosse, an aerospace engineer at Florida A&M University–Florida State University College of Engineering. “But with nuclear power, you have, from a certain perspective, an infinite amount of energy.”

Nuclear propulsion needs next-generation materials

Nuclear fission is well understood on Earth. The process releases energy by splitting atomic nuclei. Because nuclei contain far more binding energy than that held in chemical bonds, for a given amount of fuel, nuclear fission produces much more energy than chemical combustion. “A gram of uranium has on the order of a million times as much energy as a chemical fuel,” Zubrin says. This decreases the mass of the fuel that the spacecraft needs to carry.

A nuclear thermal propulsion (NTP) rocket splits uranium atoms and uses the released energy to heat liquid hydrogen to about 2500 °C. The process turns hydrogen into a superheated gas that expands through a nozzle and produces thrust.

With this extra energy density, NTP engines are expected to halve the transit time to Mars. “Transit time matters because it reduces the amount of exposure that astronauts would have to cosmic rays and radiation,” says Samuel Cohen, a physicist at Princeton University. “It also reduces the money spent on [supplies]. There are financial benefits and health benefits.”

But uranium is tricky. It melts at 1135 °C, which is far below the temperature an NTP reactor needs to run. In early rocket designs, uranium was mixed into a graphite matrix to keep it stable at high temperatures. NASA’s NERVA program tested this engine in the 1960s. It demonstrated efficiency that no chemical engine has since matched, but the project was canceled for political and funding reasons.

Researchers later found that graphite reacts with liquid hydrogen flowing through the reactor, thus degrading the fuel. So they found a better alternative to keep uranium from melting: ceramic-metal composites, or cermets, typically uranium nitride particles locked in a tungsten matrix. Though tungsten cermets do resist hydrogen attack, Gosse says qualifying them for a Mars mission’s full temperature and radiation environment is difficult to do.

Hydrogen attack is only part of the problem. Because hydrogen atoms are small, they slip into gaps between metal atoms in engine structures. As hydrogen accumulates, the metal becomes brittle and prone to cracking under stress. This hydrogen embrittlement limits materials choices for structures such as propellant lines and fuel tanks.

Then there is the radiation problem. Nuclear propulsion reactors produce radiation that is harmful to humans, and building radiation shielding adds mass—the enemy of every Mars mission. Making the shield light enough to preserve NTP’s mass savings that were gained from the dense propellant is another design challenge.

The materials difficulties of NTP are significant, but they are not what has kept the technology grounded. DRACO (Demonstration Rocket for Agile Cislunar Operations), a NASA program to fly a nuclear thermal engine in space, was canceled in 2025 because of shifting administrative priorities and funding issues. It was the second time in 50 years that a functioning NTP program was shut down for nonengineering reasons. NASA has since shifted its near-term focus to nuclear electric propulsion.

Efficient electric engines

Hundreds of SpaceX’s Starlink satellites use ion thrusters called Hall thrusters to maintain their positions in orbit. Instead of burning propellant, these devices use electricity to ionize a propellant gas such as xenon. The thrusters then accelerate those ions to extremely high speeds using electric and magnetic fields, producing the characteristic blue exhaust glow of electric propulsion.

Xenon seems like an ideal fuel for interplanetary missions. “It is inert and heavy, which means it generates a lot of thrust,” says Kristina Lemmer, an aerospace engineer at Western Michigan University. “But krypton is what is used by SpaceX.” Krypton is a cheaper alternative to xenon, but because of its lighter weight it does not generate as much thrust as xenon does.

“Iodine is another emerging alternative,” Lemmer says. “It is solid at room temperature and sublimates directly into vapor.” That eliminates the need for a pressurized tank. “And it’s also easy to get, compared to xenon,” she adds. But because iodine is not inert, it forms molecules and might react with thruster components.

The propellant is one piece of the puzzle. Power is another. Today’s electric thrusters are powered by solar panels and fly on small, uncrewed spacecraft. “A crewed Mars ship would be a lot heavier and flying away from the sun, where sunlight is weaker,” Lemmer says. Solar panels can’t supply that much electric power. “For human missions you’re going to need a nuclear reactor,” she says. In that design, a compact fission reactor, which uses heat released from the nuclear process to generate electricity, would replace solar panels to power an ion drive.

“Nuclear electric technology, from a certain standpoint, is relatively mature,” Gosse says. “The new thing now is to scale those up to sizes that can provide propulsion power for crewed mission.” During the Cold War, both the US and Russia sent nuclear-powered ion thrusters to space, but the projects were brief, experimental, and never repeated. No nuclear electric spacecraft has operated in deep space (further than 2 million km from Earth) since then.

That may soon change. NASA’s SR-1 Freedom, an uncrewed spacecraft planned for a December 2028 Mars transit, would be the first test of nuclear electric propulsion (NEP) in more than 60 years—and the first nuclear-powered interplanetary spacecraft.

A crewed Mars NEP vehicle would need a reactor generating megawatts of heat, compared with SR-1 Freedom’s kilowatts. This is because only a portion of the heat can be used to generate electricity since no machine uses heat with perfect efficiency. “This is the real pain point,” Gosse says. “Getting rid of the excess heat needs a big radiator.”

One solution is to extend large, flat panels that radiate the heat into cold space. “These radiators become a big component of the mass,” Gosse says. So, the challenge is finding light materials that radiate heat effectively. NASA’s research target for next-generation panels is no more than 3 kg/m² while still rejecting enough heat to keep a megawatt reactor running safely.

Still, NEP reactors run cooler than NTP engines, which makes their materials challenges less severe. NEP rockets are more efficient than chemical or NTP rockets because they can push propellants out at much higher speeds. But their thrust is low because the rate at which they propel exhaust is very small. NEP thrust is measured in millinewtons—roughly the force of a piece of paper resting on your hand.

“It’s like a car that goes from 0 to 60 in 2 weeks but gets 1,000 miles to the gallon,” Zubrin says. “It’s perfect for a road trip on a highway around the world but terrible to get to nearby places like the mall or a grocery store.” For a crewed Mars mission, he argues, the time needed to reach useful speeds could exceed the trip itself.

SR-1 Freedom will use a separate vehicle with chemical thrusters to leave Earth’s orbit, then switches on its NEP ion thrusters for the cruise to Mars. Even so, Gosse says that SR-1 is not expected to reach Mars significantly faster than a chemical rocket.

“But the mission could be very interesting,” Lemmer says. “We have the opportunity to learn a lot about how NEP would work in space.”

Is fusion the future?

Whether to use chemical or nuclear power to get to Mars is hotly debated. Zubrin argues that chemical propulsion is not a compromise but the rational choice. He thinks direct cost is more important than a mission’s carrying capacity. NTP rockets may allow astronauts to carry twice the mass, but the cost savings of SpaceX’s chemical fuel is more significant. SpaceX has already cut the cost of reaching orbit by a factor of five and could cut it by another five with Starship.

“NTP is not an enabling technology for human Mars missions; it’s an enhancing technology,” Zubrin says. In his view, nuclear energy is better used not for propulsion but for a surface power plant that converts Martian resources to chemical fuel.

Gosse, on the other hand, is a proponent of nuclear. He wants a hybrid rocket engine that uses both NTP and NEP. “Think of it like the hybrid car,” Gosse says. “It has batteries and an engine in it, and it switches back and forth at the most opportune time, taking advantage of both.” His design uses NTP for the high thrust burns at the start and end of the journey, and NEP for the efficient cruise phase. This approach could cut the trip time to about 2 months.

“When we work with chemical propulsion, we’re limited in the sense of how much energy we can store in the tanks,” Gosse says. This limits how frequently spacecraft can be sent to Mars: the alignment between Earth and Mars that results in the shortest distance between the planets occurs once every 26 months.

“Chemical is stuck with that [timeline] due to its velocities being relatively low,” Gosse says. Higher efficiency nuclear propulsion could create more departure windows.

Whichever system we choose, flights to Mars remain several technological advances away from takeoff. To get there, Gosse says, all propulsion systems should be funded. “There isn’t one solution,” he says. “I think a lot of people push the technology they’re invested in.”

Beyond all of them sits fusion—the propulsion technology that would make the chemical versus nuclear fission debate look like an argument about horse-drawn carriages versus steam engines. “I think the Martians [people that move to Mars in the future] are going to be very interested in fusion energy,” Zubrin says. “Once you have airplanes, you don’t cross the Atlantic by boat.”

Scientists use deuterium as fuel in fusion reactors. And “Mars has five times as much deuterium as Earth,” Zubrin says. “The process of getting it out of water is much simpler than enriching uranium for NTP.”

The problem is that fusion has not yet been demonstrated as a useful energy source on Earth. Cohen, whose Princeton Field-Reversed Configuration reactor is among the most credible fusion propulsion concepts in development, gives a timeline measured in decades, not years.

So, until then, the race to Mars will be between the inefficient turtle and the unstable hare.