CIPA Home | Popular Articles

Prospects for an Interstellar Mission:
Hard Technology Limits but Surprising Physics Possibilites

How can we reach another star in a timely fashion?
Extrapolating our best current technology into the future is like trying to somehow
soup-up Magellan’s sailing ship to circumnavigate the globe in ninety minutes.

by Bernard Haisch and Alfonso Rueda

published in MERCURY, Vol. 29, No. 4, July/August 2000
copyright 2000, Astronomical Society of the Pacific (posted with permission)


The concept of interstellar missions of exploration, discovery, and, let’s be honest, sheer adventure, took hold of our cultural imagination with the advent of the Star Trek television series in 1966. Dressed up in clever, quasi-scientific language like "warp drive" and "impulse engine," the idea had a certain air of plausibility for the public.

Indeed, the notion of interstellar travel has probably come to seem inevitable to the public at large. Once upon a time it would have seemed like a miracle to cross the ocean…but Columbus did that by and by (even though long after the Vikings!). So why not assume we will one day sail the ocean of space?

Unfortunately, the problems are more fundamental. They have more to do with basic physics than with "mere" technology. But a sea change has occurred nonetheless. Although there are no known or plausible technologies that would make interstellar travel possible, the concept of an interstellar mission has become a legitimate topic for scientific discussion in NASA circles.

In the first part of this two-part discussion, we present some of the ideas and the limits to ideas based on current technology. In the second part, which will appear in the next issue of Mercury, we will move beyond technology to possible new physics based on our work on the electromagnetic zero-point field.

That Vision Thing

NASA has not just recovered from the dark days of the mid 1980s-when the Space Shuttle Challenger exploded and grounded NASA and its dreams for more than a year-it is again formulating an inspirational path of exploration for a new millennium. Planet after planet is being discovered around nearby stars using ground-based telescopes (see "Prowling for Planets," p. XX), and the NASA Administrator is determined to challenge and goad and coax astronomers into designing space-based observatories that will eventually photograph these worlds. This would take huge telescopes linked together in arrays that are hundreds of kilometers across, perhaps located in the outer parts of the Solar System. One of us (BH) personally heard him issuing the challenge to astronomers at a national meeting to come up with ways to photograph clouds and mountains on planets in other solar systems. Naturally, alien cities would be even more exciting!

These are grand visions. For now they are impossible; in a few more years they will still be nearly impossible; and in a few more years beyond that they will transition to merely difficult and challenging. There is no doubt in my mind that the NASA space program is the best thing our government does. This is one agency of the federal government that actually succeeds in being a source of inspiration.

Once upon a time it would have
seemed like a miracle to cross the ocean...

In 1996 NASA initiated a very modest but intellectually ambitious Breakthrough Propulsion Physics Program whose goals were to "seek the ultimate breakthroughs in space transportation." It serves as a way to formally bring together researchers working on such concepts as wormholes and other distorted space-time metrics (Consider, for example, the "warp drive" metric proposed in 1994 by Mexican general-relativity expert Miguel Alcubierre—then working on his doctorate at the University of Wales and now a researcher at the Max-Planck-Institut für Gravitationsphysik—which effectively suggests superluminal motion should be a possibility owing to theoretically allowed, space-time-metric distortions within general relativity. In this model, motion between two locations could take place at effectively hyperlight speed without violating special relativity because the motion is not through space at speeds greater than the speed of light, but rather within a space-time distortion: somewhat like the "stretching of space" itself implied by the Hubble expansion. Alcubierre’s concept would, indeed, be a "warp drive." Unfortunately, Michael Pfenning and Lawrence Ford at Tufts Institute of Cosmology in 1997 and, more recently, Chris van den Broeck at the private Starlab think-tank in Brussels demonstrated that, while the theory may be correct in principle, the necessary conditions are physically unattainable.), quantum tunneling, vacuum zero-point fluctuations, and the possible coupling between gravitation and electromagnetism. The Breakthrough Propulsion Physics Program (see Marc Millis’s BPPP website at is designed to address the most visionary end of the scale of concepts within NASA’s Advanced Space Transportation Program: the ideas and speculations that might be cultivated now to lead in the future to the possibility of interstellar travel.

The program objective is to produce near-term, credible, and measureable progress toward revolutionary space travel by organizing and cultivating a network to foster collaboration among established researchers. Incremental advances in existing technology, and most likely even existing physics, will not suffice. The requirement is that researchers need to be both visionary and credible, a difficult combination.

Naturally what we want is a starship, an Enterprise-class Federation vehicle complete with warp drive…cloaking device optional. There are three small problems: propulsion mass, energy, and the speed of light. The goal of the Breakthrough Propulsion Physics Program is to look for ways to conquer all of these. This may, of course, prove to be impossible, and that is acknowledged right up front. But the program cleverly turns impossibility into a thinking tool via a technique called "Horizon Mission Methodology." While this does begin to sound a bit like bureaucratic jargon, the intention is good. It simply means setting such impossible hypothetical mission goals to keep the mind from falling into the trap of assuming that better fill-in-the-blanks will solve the problem. A better anything will not do: you need something completely different.

Naturally what we want is a starship, an
Enterprise-class Federation vehicle complete with
warp drive—yet three small problems are
propulsion mass, energy, and the speed of light

Pushing Tin (and fuel)

Looking for a better rocket to get us to the stars would be like trying to upgrade Columbus’s Nina, Pinta, and Santa Maria with wings to speed up the Atlantic crossing time. A jet airliner is not a better sailing ship. It is a different thing entirely.

A typical car weighing perhaps 3000 pounds carries about 100 pounds of gasoline in its fuel tank. The fuel is a small percentage of the total mass, in this case under four percent. The ratio of fuel to vehicle is much less than one (1/30), and that works fine because you can stop and "fill ’er up" anytime. For a 747 jumbo jet leaving San Francisco for London, the ratio of fuel to dry weight of the airplane is much higher: the jet fuel may amount to as much as 1/3 of the mass of the unfueled aircraft because you want to make the trip without refueling. This ratio of fuel to vehicle gets larger and larger for a rocket. The propellant weighs more than the rocket itself, making the ratio greater than one. And that is the beginning of a major problem.

Once the ratio gets to be greater than one, you quickly enter a no-win situation. If you wanted to put the Space Shuttle into a higher orbit, you could-in principle anyway-use more propellant, but you have to then launch more propellant and a bigger rocket to carry the additional propellant. Pretty soon you are using your additional propellant almost exclusively to launch more propellant. Your gain in either more payload or longer range gets less and less the higher the ratio of propellant to rocket. And it just keeps getting worse.

The other killer is that it takes as much propellant to slow down as to speed up. Slowing down is just speeding up in reverse. Now, there may be mitigating circumstances. The Moon trips were possible because the Moon’s gravity is much less than Earth’s, so it took much less energy (per unit mass) to fight the Moon’s lower gravity and achieve a soft landing than it took to launch off Earth. You gain twice this way because leaving the Moon is also easy in the same proportion. Then, finally, the astronauts used Earth’s atmosphere to achieve braking. All of these mitigating circumstances enabled the Moon landing and return adventure to succeed, but we still had to launch a mighty, multi-stage Saturn rocket as tall as a 35-story building…just to get back one little capsule and three guys splashing down in the ocean.

How much conventional propellant would you need to launch the Space Shuttle to a speed that would carry it in 100 years to Alpha Centauri, the nearest star system and a mere 4.3 lightyears away? The answer is that a rocket the size of Earth filled with chemical fuel would be insufficient. Even a rocket the size of the Sun would not do. And that is for only a flyby. If you want to slow down, land on a planet—which one hopes is there to land on—and then launch back home again finally to land safely on Earth, you are totally out of luck. A rocket the size of the entire visible Universe would be too small. The problem of adding more and more propellant just to propel the propellant skyrockets to infinity. A better chemical rocket is simply not an interstellar option.

This exemplifies the propulsion-mass problem. Indeed, even within the Solar System, if you want to visit and return from other planets, the problem with rockets looms large. However, here it is at least possible to consider refueling by somehow processing material to be found at or near the destination. Carrying your propulsion fuel with you, like a turtle and his shell, becomes a show stopper, though, for interstellar exploration.

But moving beyond chemical propulsion, what if such an efficient propellant could be found that, like a car with its gas tank, the ratio of fuel to rocket mass were again well below one? It is the fact that the propellant to rocket ratio is over one that sends the problem soaring out of contral. Where can we look to find the real world analogs of Captain Kirk’s maneageable-sized dilithium crystals (which must be pretty compact since there always seems to be plenty of room left over for Commander Scotty to be tinkering and frantically running around in his engine room)?

A better chemical rocket
is simply not an interstellar option

Propellant Matters

A survey of ideas took place in July 1998. The Advanced Concepts Office at the Jet Propulsion Laboratory and the Office of Space Science at NASA headquarters jointly sponsored a workshop at Caltech entitled "Robotic Interstellar Exploration in the Next Century" (at which one of us, BH, presented an invited talk). Essentially all known, credible, interstellar, propulsion ideas were discussed.

In broad terms there are four possible types of onboard engines: chemical, fission, fusion, and antimatter. Rocket-type propulsion works in the vacuum of space because you do not need to push or pull on any medium the way a propeller does in air or water. You carry along your own matter (this becomes the problem of course) which you expel out the back of the rocket to push it forward. The vacuum not withstanding, you could push yourself forward by throwing bricks out the rear; naturally this is not very efficient.

There are two properties that characterize a propellant: for each kilogram of the stuff, how much force do you generate when it is expelled? This is measured by a quantity called specific impulse which tells you how much acceleration you can get over how long a period of time from a given propellant. And the second important fuel property: how much energy can you extract from each kilogram of propellant? Call this the energy content.

You need both a high specific impulse and a high energy content to have an efficient rocket. They do not necessarily go together: a fuel with a relatively low energy content may give a relatively high specific impulse and vice versa. It is instructive, though, to take the absolutely best possible case and see where that leaves the possibility of interstellar travel when you have to carry your fuel along as in a rocket.

No matter what your specific impulse, you cannot do better than to convert all of your available propellant energy into the kinetic energy of the rocket. That is, of course, wildly optimistic, but it simplifies things by letting us concentrate solely on energy. A benchmark for discussion at the NASA/Caltech meeting was a forty-year mission. To get to Alpha Centauri in this time would require a speed of about one-tenth the speed of light (0.1 c), since Alpha Centauri is just over four lightyears distant. At that modest speed you can still calculate the kinetic energy of your starship using ordinary Newtonian physics. It is simply mv2, where m is the mass and v the speed of the starship.

There is no more energy-efficient fuel than antimatter. Letting matter and antimatter combine gives you up to 100 percent energy efficiency, so the available energy from the E=mc2 relationship is just mpc2, where mp is the mass of the matter-antimatter propellant combination. When you equate the rocket kinetic energy and the mass energy of the antimatter, you find that an amount of matter-antimatter fuel that is only 0.5 percent of the mass of the starship is all you would need to get to 0.1c. Of course, if you want to slow down at the end of your forty-year mission, you would need just as much again. So the bottom line appears to be that a starship with one percent matter-antimatter fuel could reach a speed of 0.1c, get to Alpha Centauri in forty years, and brake at some planet—that one hopes is there—in that star system.

At first glance this seems remarkably encouraging. Unfortunately ugly details quickly loom large. First of all, just because the matter-energy conversion is 100 percent efficient does not mean that you can give that energy with 100 percent efficiency to the starship. This goes back to specific impulse considerations. There is at least a factor of ten loss in efficiency, so raise the matter-antimatter fuel to rocket percentage from one percent to ten percent. Being about the same ratio as that of full gas tank to car, ten percent still looks acceptable, but next you have to consider how much antimatter this really amounts to and its implications.

an amount of matter-antimatter fuel
that is only 0.5 percent of the mass of the
starship is all you would need to reach
a speed of ten percent the speed of light

Since the plan is to be as optimistic as possible to find out where there is a true hard limit, a show-stopper no matter what, let us also assume that we will know how to put the crew into suspended animation so that we can use a starship the size of the Space Shuttle rather than the size of an aircraft carrier (and who would even want to live on the latter for four decades?). Assume a 100-ton vehicle, and we now have a requirement of ten tons of matter-antimatter fuel, or five tons of pure antimatter as a requirement. Double that to ten tons of antimatter for a round-trip.

There are no free sources of antimatter. Since it annihilates on contact with matter, there is clearly no supply around to tap; you cannot just go mine for it like uranium. But antimatter can be manufactured in particle accelerators and there are techniques to store small amounts of it. With some upgrading to tailor the facility to do this, the United States’s Brookhaven National Laboratory could turn out one-tenth of a billionth of a gram of antiprotons per year. Similar capabilities exist at the Center for European Nuclear Research (CERN) near Geneva and the Institute for High Energy Physics in Russia. But the required ten tons of antiprotons would be equivalent to about ten million grams. The difference between what we can manufacture in a year of production and what we would need for a single roundtrip becomes an enormous factor of 1017 (i.e., a hundred million billion). This is more than a big gap.

Even if we could easily manufacture the required quantity of antimatter, however, we would have a tremendous problem on our hands. The huge amount of energy contained in five tons of antimatter-which is great for propulsion-is, of course, a disastrous storage problem here on Earth. It would be equivalent to about 200,000 Megatons of TNT, which is ten million Hiroshima-energy bombs. For comparison, the largest (then) Soviet hydrogen bomb tests in the 1960s involved explosions of a mere 100 Megatons or so.

In terms of designing a starship for a single, one-way trip based on rocket-style, onboard propulsion, this is as good as it gets! Rocket-style, carry-your-fuel propulsion is the wrong approach for a starship.

Leave the Fuel Behind

An alternative to carrying your engine and fuel along is to leave both behind. This way you do not get into the exponentially losing situation of having to accelerate your propellant. A tiny fraction of the same energy supply will get the same job done if all you have to accelerate is the vehicle itself. The problem now, of course, is to transmit the propulsion from the source to the vehicle. Two such schemes have been proposed and were discussed at the NASA/Caltech workshop: lasers and particle beams.

When you are sitting next to a window with sunlight shining on you, you feel the energy of the light in the form of heat, but you do not notice any force or pressure from the light. But light does carry both energy and momentum. When sunlights lands on a small particle in space and is absorbed, the momentum of the light is transferred to the particle and that creates a measurable force on the particle. This is the very process that forms a comet’s tail. As the snowy iceball that is the comet’s nucleus approaches the Sun and begins to evaporate, radiation pressure from the sunlight pushes the evaporating gas and dust outward, creating the tail that is millions of miles long.

A powerful laser can exert a great deal of force, and that is the basis of the laser sail concept. Instead of carrying vast amounts of propellant to power an onboard engine, a laser-beam starship would literally sail off into space with no need for an engine at all. It would ride on the force provided by a beam of light from a giant laser that never has to leave the Solar System. A perfectly reflecting sail is even more efficient—by a factor of two—than one that absorbs the light, and this could be made out of very flimsy material, something like super high-tech aluminum foil, if constructed in Earth orbit.

Rocket-style, carry-your-fuel
propulsion is the wrong
approach for a starship

A major problem with this concept for interstellar considerations is size. A laser can focus a beam of light to its maximum theoretical concentration. If we assume that a sail could be assembled and held in place that was as large as 10 km in diameter, we would want the laser to still concentrate virtually all of its light on the sail even when that sail is halfway to the nearest star, otherwise we lose more and more of the force. If the light beam spreads to only ten times the size of the sail, the loss of force is 99 percent since it is area that matters. Yet to keep this high degree of focus, the laws of optics dictate that the laser must effectively be a lens or mirror 1000 km in diameter to achieve this for visible light (5000 Å). You can trade off the sail diameter against the laser lens diameter. Make the sail ten times bigger and you can use a laser with a ten-times smaller lens. You then have a 100-km sail and 100-km laser lens. So we see that all possible combinations stretch credibility.

Apart from the problem of requiring vast sails or gigantic lasers, there is a good reason to get worried even if you could manage to concentrate all the laser power on the sail even at the halfway point to Alpha Centauri. And that is, what do you do for the next half of the trip to slow down? It is a one-way push. You can jettison your sail and coast, but you cannot stop.

Even if that were not problem enough, there is a time-delay feedback problem that gets worse and worse. Taking the laser-driven sail as an example, let us assume that a mission propelled by a 10-km diameter light sail is halfway (2 x 1013 km) to Alpha Centauri when a beam problem reaches the vehicle. A misalignment of the laser of no more than one part in a trillion, which occurred back on Earth two years ago, is now reaching the vehicle causing the beam to miss the light sail. Owing to the speed-of-light limitation, it will be another two years before any news of this problem transmitted by the spacecraft can reach Earth. It will be yet another two years before the correction from Earth will reach the spacecraft. But by then the vehicle may have drifted out of its trajectory sufficiently, owing to the effects of interaction with the interstellar medium (or other causes), that it is still out of the beam. Indeed, any drift of the vehicle from a line-of-sight trajectory will cause the same uncorrectible problem in the first place since there is no way to know where the vehicle is "now" (in the sense of where the beam is supposed to hit). This illustrates the inherent problem of speed-of-light caused time delay in any feedback loop. It would be all too easy—in fact, probably unavoidable—to have a mission using this propulsion mechanism "lost in space without a paddle" due to the slightest error.

Although a laser beam does exert a force, it is a very inefficient force in comparison to the energy required. The ratio of energy to momentum for a light beam is equal to the speed of light, c. The ratio of energy to momentum for an ordinary object is v/2, where v is the object’s speed. Force results when you transfer momentum. In other words, if one has a tough enough sail, one can throw tennis balls at it and transfer momentum from the tennis balls to the sail when the balls bounce off and thereby generate a force. Because of the energy to momentum ratio, a tennis ball going 80 km/hr is 27 million times more efficient than a beam of light of the same energy in generating a force on a sail. For this reason, beams of particles have also been considered in place of a laser beam. A particle moving one-tenth the speed of light bouncing off a sail is still twenty times more efficient than a photon of light of the same energy reflecting off a sail in terms of providing a propulsive push.

Particle beams have problems of their own. Interstellar space may be empty by Earth standards, but there are both particles and magnetic fields filling the void. The easiest way to make a particle beam is to let the particles carry an electric charge so that electrostatic or magnetic forces can be used to provide the boost. But a charged particle beam will suffer deflections from the interstellar magnetic field. You somehow have to neutralize the particle beam, so that you wind up with uncharged particles that will not interact with the interstellar magnetic field. This is difficult to do. Moreover, even if you can get rid of the charge, a neutral beam will still tend to be dispersed when collisions take place with the interstellar medium. And finally, the problem of trying to collimate a particle beam and keep it pointed with such precision that it hits the target sail and spreads out only to the size of a sail at distances of trillions of kilometers makes the laser focusing problem seem like child’s play.

One other scheme that is clever in principle was presented at the NASA/Caltech meeting: the use of giant tethers. Such a tether is simply a very long wire dangling in space. When a conducting wire is moved perpendicular to a magnetic field, a current is induced in the wire, which can be used as a source of electrical power. This has been the motivation for several space experiments in recent years. But a rather different application can be envisioned using the interstellar magnetic field to generate a propulsive force.

Instead of using the interstellar magnetic field to originate a current, you use an onboard source of power to create your own current in a long tether. There is a law in electrodynamics that tells you that if you have a current flowing in a wire perpendicular to a magnetic field, there will be a (Lorentz) force perpendicular to both. A current plus the interstellar magnetic field gives you a potentially propulsive force.

So what are the drawbacks to such a scheme? For one, the interstellar magnetic field is extremely weak. It would take a lot of very long tethers to create enough force to accelerate a tether-driven starship with even very small "g" acceleration. A 1000-km long by 1000-km wide venetian blind configuration with wires several centimeters apart was discussed at the meeting. Another problem is the direction of the force. Nature only provides it perpendicular to the interstellar magnetic field, and we are far from certain what the strength and direction of that field is.

Crossing the Interstellar Sea

Four spacecraft are already on their way out of the Solar System. Pioneer 10 and 11 were launched in March 1972 and April 1973, respectively, as missions to explore Jupiter. After Pioneer 10 successfully encountered Jupiter in December 1973, Pioneer 11, which was a backup mission, was redirected enroute to intercept Saturn via a Jupiter gravitational assist maneuver. Five years later, in August and September 1977, Voyager 1 and Voyager 2 were sent on their way to Jupiter and Saturn, with Voyager 2 continuing on to Uranus and Neptune. The distances of these spacecraft are now six to seven billion miles from Earth, beyond the orbit of Pluto, headed in quite different directions from each other and destined to cross the heliopause-the abrupt transition between the outflowing solar wind and the interstellar medium-and, thus, leave the Solar System in another decade or two.

While these distances sound impressive, they correspond to a mere ten or so lighthours, not the lightyears that separate stars from each other. At the rates at which these probes are now traveling (about 55,000 km/hr), it would them about 80,000 years to get to Alpha Centauri. But, in fact, none of the four spacecraft are heading toward any particular star.

If we could settle for 100,000-year, unmanned missions, we could direct more sophisticated versions of spacecraft like Pioneer and Voyager to our nearest stellar neighbors using technology already in hand. But communication and mismatch with human and societal timescales would be huge limitations.

If we were willing to risk accumulating twenty million Hiroshima-energy stockpiles of antimatter, we could perhaps accelerate and decelerate a single rocket-type vehicle to one tenth the speed of light or so. Confinement of the antimatter would be a huge environental issue, to put it mildly…and, of course, providing the energy to make the antimatter in the first place is daunting.

The problem of trying to collimate a particle
beam and keep it pointed with such precision
that it hits the starship’s sail and spreads
out only to the size of the sail at a distance
of trillions of kilometers makes the problematic
use of laser propulsion seem like child’s play.

If we could construct enormous lasers and sails in space hundreds of kilometers across, we might shoot probes to other stars. Lack of a brake would be a major problem: with luck such a mission might gather an hour’s worth of data on an extrasolar planetary system as it races through at 0.1c after a hundred-year flight.

None of these options, nor any of their cousins, look attractive or feasible. Studies of such possibilities make interesting exercises and are worth considering even if only to rule them out. In our view, none of these things will ever happen because it would be like trying to fly a sailing ship across the ocean; with enough brute force, jerryrigging, and engineering cleverness one could probably make a stunt like that work in some fashion. But that is definitely not the way to go.

The electromagnetic zero-point field, or more generally, the quantum vacuum, offers four possibilities that may someday yield the technology to travel to other stars: extraction of energy, generation of forces, and, most intriguingly, perhaps even manipulation of inertia and gravitation. At present these are highly speculative possibilities, but the exploration of these ideas seems at least as worthwhile as trying to work around the formidable obstacles and insuperable limitations we’ve already discussed here. In the next issue, we will introduce the nature of the zero-point field and explore how it might be manipulated to power starships.

Pricing the Production of Antimatter

Let us say that you have chosen antimatter propulsion for your interstellar spaceship, and now it is time to get the fuel. Where do you turn? The manufacture of antiprotons is extremely inefficient. Techniques for creating antiprotons at the Center for European Nuclear Research (CERN) require approximately two and one-half million protons each accelerated to an energy of 26 GeV to create a single antiproton. This amounts to an energy efficiency of 1.5 x 10-8.

There is no more
energy-efficient fuel than antimatter

This is further reduced by a factor of ten or so for the efficiency of the proton accelerator, leaving a net efficiency of perhaps 1.5 x 10-9 — i.e., about one part in a billion! At a cost of five cents per kilowatt-hour of electricity, the cost of ten tons of antiprotons would be 8 x 1021 U. S. dollars. The best way to express this amount of money might be to say that it represents the total current U. S. federal budget (approximately $1.2 trillion per year) spent every year for the next seven billion years.


BERNARD HAISCH is a high-energy astrophysicist. He is the director of the California Institute for Physics and Astrophysics ( in Palo Alto and a Scientific Editor for the Astrophysical Journal. In his "spare time" he and his wife, Marsha Sims, are semi-professional pop and country songwriters ( He can be reached by email at

ALFONSO RUEDA is a professor at California State University Long Beach who teaches primarily for the Electrical Engineering Department but also for the Physics and Astronomy and Mathematics Departments. He has done extensive research in several aspects of the electromagnetic zero-point field and also on its connection to various astrophysical phenomena. He is rather quiet, but his wild side may appear in the Spanish or Latin moods of a typical Mediterranean or Latin-American party. He is available via email at

CIPA Home | Popular Articles