Jocelyn Bell Burnell is possibly the person most famous for not winning the Nobel Prize. In the summer of 1967, she was tramping through 57 tennis courts’ worth of countryside to set up 2,048 radio antennae. The data from the resulting radio telescope was set to be the focus of the young student’s doctorate thesis at the University of Cambridge. What transpired was the discovery that would make Bell Burnell one of the most famous names in astrophysics.
Two months into examining the data from the telescope, Bell Burnell noticed a strange signal. It was a radio pulse that repeated exactly every 1.337 seconds. The accuracy was so precise that Bell Burnell and her academic supervisor, Anthony Hewish, contemplated whether it might not be the result of extraterrestrial life. The regularity with which this pulse appeared challenged even the accuracy of atomic clocks, which seemed to indicate an artificial creation by an advanced civilisation. Bell Burnell and Hewish referred to the object as LGM-1, for Little Green Men 1.
The little green men idea was put to rest when Bell Burnell found the same signal originating in a different part of the sky. The sources were too widely spread to be a single civilisation and it was highly unlikely that identical signatures would be used by entirely separate life forms. Bell Burnell declared that she was relieved by this conclusion, since the timing of finding aliens so close to the end of her doctoral course was particularly bad. But what was out there that could rival an atomic clock in accuracy? The answer turned out to be a dead star.
A star’s gravity is perpetually trying to crush it to bits, but is thwarted by the heat created from the star’s burning fuel. This energy increases the speed of the star’s atoms that zip around and push against the collapse. Burning for a star does not involve the chemical combustion of campfires, but the fusing of light atoms into heavier ones. It is a process known as nuclear fusion.
Due to the smaller repelling positive charge of their nuclei, it is easier to fuse lighter atoms than heavier ones. Stars therefore begin by fusing hydrogen atoms into helium. This still requires ridiculously high temperatures so that the atoms collide fast enough to overcome their electric repulsion, with that of the Sun’s core reaching 15 million degrees. Once helium forms, the atom’s larger mass causes it to sink to the star’s centre, leaving hydrogen to continue to fuse in the outer shells. When a star runs out of fuel to burn, gravity wins. What happens next depends on the mass of the star.
For a star like our own Sun, the weighty helium core is compressed by its stronger gravity. This raises the temperature and the star begins to swell. As the outer layers expand, they cool to emit a red hue that earns the star the name red giant. Eventually, the temperature in the core reaches 100 million degrees and helium begins to fuse into carbon. The heavier carbon sinks below the helium to form an even denser core. Our Sun will not have enough mass to compress the carbon core to the temperature where carbon will begin to fuse. Instead, the heat from the core will blow away the outer layers of the dying Sun and leave a dense remnant around one half of the mass of the Sun, but squished down to the size of the Earth. This is known as a white dwarf.
For a star of more than 8 Sun masses (or solar masses), the ending is rather more dramatic. The greater mass crushes the core to burn carbon and onwards to steadily heavier elements. When it reaches iron, the nuclear fusion comes to a halt. Fusing iron does not release energy, but absorbs it. This means that the star gets no boost in support from the burning. With no way to keep countering the collapse, gravity wins and the star implodes. Pretty much everything gets fused in the resulting shock wave, and the star explodes in an event referred to as a supernova.
If the core left behind after the supernova explosion has enough mass, gravity will become an unstoppable force. The remains of the star will collapse until not even light can escape its gravitational pull. This is a black hole. If the explosion leaves a core of between 1.4 and 3 solar masses, then there isn’t enough mass to really finish the process. Instead of a black hole, gravity compresses the core so strongly that electrons and protons within the atoms combine to produce neutrons. The result is a hot naked ember known as a neutron star: the densest star in the Universe.
These stellar corpses have diameters than have shrunk from millions of kilometres to around 20km (12mi), but with masses over 40 per cent more than that of the Sun. The composition from surface to core consists of steadily more neutron-rich atomic nuclei, before even the nuclei structure breaks apart to form a soup of neutrons. A sugar-cube volume of a neutron star would weigh more than 100 million tonnes on Earth, and contain the (immensely squashed) entire population of our planet.
As the radius of the star contracts to form the city-sized neutron star, its spin is retained. The result can be compared to pulling your arms to your chest while whirling around on a rotating office chair. In both cases, you and the star will spin faster. 1 For a neutron star that has undergone such a massive contraction, the resulting rotation time drops to a matter of seconds.
While a neutron star mainly consists of neutral neutrons, about 10 per cent of charged protons and electrons remain to ensure that the star hangs on to its magnetic field. 2 The collapse compresses the field, which is magnified to become a trillion times stronger than that found around the Earth. The magnetic field rips over the surface of the rotating star, and pulls the remaining protons and electrons from the crust to be funnelled along the field lines towards the magnetic poles. As they change direction to twist around the field lines, the charged particles emit radio waves along with the more energetic X-rays, Gamma rays and visible light. Where the magnetic field lines converge at the poles, the radiation narrows into beams that radiate into space, along with a wind packed with the charged particles.
Figure 12 A pulsar is a neutron star whose radiation beams sweep past the Earth as the neutron star rotates. We observe this as a regular flash like that of a lighthouse.
The north and south poles of the magnetic field do not necessarily align with the star’s axis of rotation. This is also true on Earth, where our planet’s magnetic field is tilted by 11 degrees with respect to our rotation axis. This offset makes the neutron star’s radiation beams swing around the star like a lighthouse beacon as it rotates. If the path of the beam passes across the Earth, our planet is flashed by a regular pulse of radiation for each rapid rotation of the neutron star. It was this pulse that Bell Burnell identified as her little green men.
When speaking to the science correspondent for the Daily Telegraph newspaper in 1968, Bell Burnell was asked what these strange flashing sources should be called. The science correspondent suggested ‘pulsar’ as a variant on the word ‘quasar’; the bright but non-pulsating radio sources Bell Burnell had anticipated studying with her telescope. The name stuck and a new type of astronomical object was found.
Upon determining the source to be the rapidly rotating pulsar, Bell Burnell and Hewish renamed their first mysterious object from LGM-1 to CP 1919, for Cambridge Pulsar, with the ‘1919’ designating its angular position eastwards in the sky. It later acquired its current official name ‘PSR B1919+21’, for Pulsating Source of Radio, with the extra ‘21’ indicating that it was also 20 degrees north, and ‘B’ to specify this coordinate-naming convention.
In 1974, Hewish was awarded the Nobel Prize in Physics for the discovery of pulsars. That Bell Burnell was not acknowledged in the prize for her part in the discovery has long remained a point of controversy, although Bell Burnell herself accepted this graciously, noting that, ‘I reckon I’ve done pretty well out of not getting the Nobel Prize!’ She went on to win a large number of other prestigious awards during her career, and has been president of both the Royal Astronomical Society and the Institute of Physics in the UK. Meanwhile, pulsar discoveries became curiouser and curiouser.
Late in the 1970s, a radio source was found just a few degrees away from Bell Burnell and Hewish’s discovery. Incredibly compact, it was initially suspected of being a new pulsar. However, when the sky was monitored for the signature flash, nothing was found. The source appeared to be a steady beam of radio waves rather than a pulsing lighthouse signal.
Suspecting that a pulsar’s blink could be missed if it was rotating incredibly fast, further searches were conducted in March 1982. These hunted for pulsars with rotation periods down to 4 milliseconds (or 250 times every second). The fastest known pulsar at the time lay in the Crab Nebula with a rotation speed of once every 33 milliseconds. This new search would therefore pick up anything that was a substantial 10 times faster. Yet there was still no sign of a pulsar’s signature flash until autumn of that year.
It was the Arecibo radio telescope in Puerto Rico that finally caught the pulsating signal. This observatory’s 305m (1,000ft) dish has a movie-star history, having searched for extraterrestrial life in the adaption of Carl Sagan’s novel Contact, and been dramatically destroyed in the climax of the James Bond film Goldeneye. When the huge dish imaged the sky at a blistering rate of every half a millisecond in 1982, it spotted the lighthouse flash of a record-breaking pulsar. The new pulsar had a spin period of 1.558 milliseconds, corresponding to a staggering 642 revolutions each second. This was 20 times faster than the Crab Pulsar and created a record the pulsar would hold for the next quarter of a century.
While the discovery of a millisecond pulsar resolved the quandary of the source of the radio emission, it cracked open another can of worms. Since pulsars are continuously pumping out energy in radio waves and other radiation, over time they gradually slow down. Young pulsars therefore spin faster than old pulsars. Since the millisecond pulsar was the fastest ever discovered, this should mean that it was incredibly young. But the evidence suggested otherwise.
If the pulsar had been caught near its birth, it should be surrounded by signs of the giant supernova explosion that expelled the star’s outer layers. The gas that is flung outwards from the dying star is known as a supernovae remnant, and is typically visible for more than 10,000 years. The Crab Nebula is the supernova remnant of the Crab Pulsar, with an estimated age of 960 years. The new millisecond pulsar should be far younger than this, but there was no sign of the gaseous remnant.
More strangely still, the pulsar appeared to be slowing down extremely slowly. Models for a pulsar’s changing speed suggest that young, fast pulsars slow down quickly, and such a millisecond whip-around should be decaying rapidly over a matter of years. Measurements for this pulsar’s slowdown were far lower than predicted and pointed to an age of around 230 million years. This was far, far older than any pulsar previously discovered. How could a pulsar, radiating energy into space, be both the fastest and the oldest? It would turn out that this pulsar had cannibalised its twin.
The story of millisecond pulsars begins with a pair of orbiting binary stars. While twinned by their motion in the sky, these stars are not identical and one is far more massive than the other. Size is not a healthy attribute for a star, since the extra mass causes it to burn more swiftly through its nuclear fuel supply. The bigger sibling therefore reaches the end of its normal starry life first and explodes in a supernova. Such a huge explosion at close quarters risks the smaller star being blown to bits, but if it survives it can find itself paired with a neutron star.
While the neutron star is tiny, it remains incredibly heavy. Its sibling therefore continues to feel the neutron star’s gravitational pull and the two remain orbiting one another. If the neutron star’s magnetic poles are aligned towards the Earth, its radio beams sweep across our planet to be detected as a pulsar. As time passes, the pulsar begins to slow. Over about 100,000 years, the pulsar’s radio signal gradually weakens until it is undetectable and the pulsar falls silent. The pulsar’s mass remains unchanged by this slowing, though, so its sibling star continues around its orbit. However, the sibling too is now finally reaching the end of its life.
The gravity of each star dominates a surrounding region of space called the star’s Roche lobe. The concept of the Roche lobe is similar to that of the Hill radius for when the objects involved are similar masses. Rather than being spherical regions around the stars, the Roche lobes resemble teardrops that meet at their tapered point. At this touch point, the gravitational pulls from both stars cancel one another like the lip between two mountain valleys. Step towards one star, and its gravity would pull you towards it. Move in the opposite direction, and it would be its sibling that dragged you inwards.
As the smaller star runs out of hydrogen to burn, it swells to a red giant. The star’s radius becomes so big that it overspills its Roche lobe and is drawn into the neutron star’s domain. This overflow is the same mechanism that meant chthonian super Earths could form from overflowing hot Jupiters, as described in Chapter 6.
As the red giant’s outer layers pour on to the neutron star, this receives a kick that causes its rotation to once again increase. As more of the red giant sibling is transferred, the neutron star’s spin increases to incredible millisecond speeds. The material that hits the neutron star’s surface is heated to extreme temperatures of up to 10 million degrees. Such fantastically hot material does not emit infrared, but the higher-energy X-rays. If these are detected on Earth, the paired stars earn the intermediate term low mass X-ray binary system.
Figure 13 The Roche lobes of an orbiting binary, consisting of a regular star and a pulsar. Matter inside the Roche lobe is gravitationally pulled towards the central star. When the regular star becomes a red giant, its outer layers can overflow its Roche lobe and be drawn on to the pulsar, which spins up to reach millisecond speeds.
Eventually, the outer layers of the red giant have been sucked on to the neutron star, leaving a white dwarf orbiting a millisecond pulsar. This rejuvenation of a pulsar’s spin has led to millisecond pulsars being referred to as recycled pulsars. Even more than the regular pulsars, the timing of millisecond pulsars is exceedingly exact – so exact, that the influence of even a tiny object is detectable.
The very first exoplanet
51 Pegasi b is frequently remembered as the first exoplanet to be discovered. Yet in truth, the hot Jupiter was only the first planet to be found around another Sun-like star. The title for the very first exoplanet is shared between two worlds orbiting a millisecond pulsar labelled PSR B1257+12.
The discovery of PSR B1257+12 is unusual since it begins not with a brand-new telescope, but with a broken one. In 1990, the Arecibo radio telescope – the same instrument that found the first millisecond pulsar – needed repairs. Cracks had been found in the telescope’s structure and no one wanted to take any risks, especially since the 90m (300ft) radio telescope at Green Bank in the US had collapsed suddenly a few years earlier due to structural failure. The Arecibo radio telescope was still usable during the repairs, but had to be held stationary in one position rather than moving continuously as it tracked an object over the night sky. This restriction greatly limited the projects Arecibo could do, resulting in a large drop in demand for the telescope’s time. It was a reduction of which Alex Wolszczan, a Polish astronomer based at Arecibo, took full advantage. His plan was to survey the sky for undiscovered millisecond pulsars; a proposal that required almost one-third of the then world’s largest radio telescope’s time for a month. Under normal circumstances, it was a proposal that would have been dismissed. However, with demand down and Wolszczan based at the telescope, the dedication of time to this project was accepted.
The result of Wolszczan’s survey was two new pulsars. The first was a pulsar in a binary, whose sibling was a second neutron star. This initially looked the more exciting system, but then Wolszczan noticed that there was trouble with the other pulsar’s rotation time.
From Wolszczan’s survey, PSR B1257+12 became the fifth millisecond pulsar detected. The pulsar had a rotation time of 6.2 milliseconds, making 161 rotations each second. Yet when Wolszczan tried to predict how frequently the radio beacon should be seen on Earth, he could not seem to get it right. This was particularly strange for a millisecond pulsar. Recycled by their companion, these old neutron stars suffer less from the quakes and shudders that can affect the spin of their younger and slower counterparts. One possibility was that the irregularity was due to the pulsar’s orbit with its sibling. As the two stars circled one another, the distance to the Earth would vary slightly and change the arrival times of the pulses. However, no companion could be seen (in itself a strange phenomenon for a recycled pulsar), and the variations in the pulsar’s timing seemed too small to be due to an orbit with a star-sized neighbour. A smaller companion also made no sense, since surely such an object would have been vaporised during the pulsar’s red giant phase, or blown free of the pulsar’s gravity as its mass shrank during the supernova explosion?
Wolszczan’s next thought was that this was a problem with pinpointing the pulsar’s position. If the pulsar’s location was wrong, then the distance to the Earth would be estimated incorrectly. Such an error would change the expected arrival times of the radio pulses and thereby knock Wolszczan’s predictions off base. To get a better marker, Wolszczan contacted Dale Frail at the US National Radio Astronomy Observatory’s imaginatively named Very Large Array, or VLA for short. Based in New Mexico, the VLA consists of 27 separate radio dishes arranged in a giant ‘Y’ shape. These can combine their data to achieve extremely high precision measurements.
While Frail was working on pinning down the pulsar’s location, the news headlines went wild: a planet had been discovered orbiting a different pulsar. The article splashed across the front cover of the journal Nature, on 26 July 1991, proclaimed that the ‘First planet outside our Solar System’ had been discovered.
The discoverers were British astronomers Andrew Lyne and Matthew Bailes, and graduate student Setnam Shemar. The planet-hosting pulsar was PSR B1829-10, a regular pulsar 30,000 light years away in the constellation of Scutum, the Shield. Variations in the pulsar’s signal indicated a planetary companion about 10 times the mass of the Earth, and an orbital time of about six months.
The news gave Wolszczan mixed emotions. With planets around pulsars an option now firmly on the table, he wondered whether he had just been beaten to making history. Were planets also the explanation for PSR B1257+12’s strange motion? It was a possibility that he had considered, but he did not yet have enough evidence to repeat the claim.
Frail had also read the news about the exoplanet discovery. He faxed the new updated position coordinates of their millisecond pulsar from the VLA observations to Wolszczan, joking as he did, ‘Don’t find any planets!’ Wolszczan updated his model for the new data, and was forced to email back his reply with the news that they had just found two.
The planets were both approximately 4 Earth masses, on slightly elliptical orbits lasting 65 and 98 days, circling the pulsar on either side of Mercury’s position from the Sun. By including the effects of these planets in his model, Wolszczan was able to fit the frequency of the pulsar’s beam perfectly.
Before Wolszczan and Frail could publish their discovery, news of the dual planets was leaked in the popular press. On 29 October 1991, the Independent, a UK newspaper, hinted at the detection of two new worlds around a pulsar based on comments from Lyne. The article was cautious about the discovery, noting that ‘Professor Wolszczan was not prepared to talk about his research because he feared it might prejudice the chances of it being published in a scientific journal. He also stressed that other astronomers have not yet had a chance to examine his results.’ The Independent article was followed up by a second piece in the science magazine New Scientist, on 14 December 1991. Despite sounding more confident about the discovery, the short article was a surprisingly lacklustre description of the first few planets ever found beyond our Solar System. The lack of fanfare might have resulted from scepticism that such odd planets had truly been found, or been due to awaiting the scientifically reviewed journal paper. Despite Wolszczan’s concerns, his and Frail’s publication appeared in the 9 January 1992 edition of Nature, and the two planets around PSR B1257+12 were formally announced.
The journal article appeared just before the American Astronomical Society winter meeting, one of the major events in the astronomy community’s calendar. It was held that year in Atlanta, in the US, and was scheduled to hear from both pulsar planet discovery groups in succession. Lyne would speak on the first exoplanet to be discovered, and Wolszczan would follow to describe their new planet pair. But Lyne’s talk was not the one originally planned. Standing before his audience, Lyne admitted that there had been an error in their calculations: there was no planet orbiting PSR B1829-10. The warning alarm had been the supposed planet’s six-month period. Such an exact fraction of the Earth’s cycle around the Sun indicated a possible problem with the accuracy of the pulsar’s known position, causing it to appear to wobble due to the Earth’s own motion. Despite the care the team had taken, this error had slipped through. Once corrected, the pulsar’s regular flash appeared on schedule, unaffected by the tug of a hidden companion. ‘Our embarrassment is unbounded,’ Lyne concluded. ‘And we are truly sorry.’
Lyne had discovered the error only days before the meeting, but had attended the international conference to announce the mistake. This confession elicited shock from his audience, which turned to respect at Lyne’s integrity and bravery at publicly admitting the error. The end of his talk was greeted with huge applause. This was what science was all about: trying, improving and continuously adjusting ideas to fit the new data.
For Wolszczan following Lyne’s talk, this was a difficult moment. The idea that a resurrected star like a pulsar could have orbiting planets was already very hard to believe. Now the first observation of such a system had been proved false. Yet, with Frail’s accurate measurements of the millisecond pulsar’s position with the VLA, the two scientists had avoided the same pitfall. The planets around millisecond pulsar PSR B1257+12 were real.
The discovery stood up to scrutiny. Six months later, PSR B1257+12 was observed independently using the 43m (140ft) radio telescope at Green Bank. The oscillation in the millisecond pulsar’s signal was confirmed, strengthening the claim that this planet pair was no experimental error.
Wolszczan continued to monitor the two pulsar planets over the next couple of years, searching for any extra details hidden in the arrival time of the pulsar’s beacon. In 1994, he found it. There was another object orbiting the pulsar, smaller and closer than the previously discovered planets. The quick, weak signal was difficult to spot, allowing it to escape detection until then.
The announcement did spur some scepticism. Like the false signal of the first pulsar planet, this third planet’s orbit matched an orbit within our own Solar System; that of the Sun’s rotation. Out near the edge of our Solar System, the US Pioneer 10 space probe had detected a fluctuation in the solar wind; a stream of charged particles that are ejected from the Sun’s surface. These variations matched the Sun’s rotation and the orbit of the supposed third planet. The concern was that the solar wind was dispersing the pulsar’s signal as it travelled to Earth, with the strength of this weakening varying to produce a signal fluctuation that looked like a planet.
The ability of the solar wind to disperse a pulsar signal depends on the frequency of the emitted radio waves. Wolszczan therefore observed the pulsar’s beacon at different radio frequencies to see if the strength of the signal changed. The whisper of the third planet remained in place: it was really there.
The new planet was just twice the size of the Moon, with an orbit taking 25.4 days. The fact that such a small planet could be detected more than 2,000 light years away was a testament to the incredible accuracy of the timing of a millisecond pulsar’s beacon, allowing even tiny variations to be picked up. In principal, the technique is so sensitive that a planet the mass of a large asteroid could be detected. More than 20 years on from Wolszczn’s detections, this moon-sized world remains the smallest planet ever found.
The confirmation of planets around a pulsar was great news, but it left a glaring question: how does a long-dead star have a planetary system?
Salamander planets
The most intuitive way in which a pulsar could be orbited by planets is if the system formed in the same way as our own did, at the beginning of the star’s normal life. The difficulty is for the planetary system to survive the star morphing into a pulsar.
Named the Salamander Scenario by California Institute of Technology scientists E. Sterl Phinney and Brad Hansen, after the fire-loving mythical lizard, closely orbiting planets have to first survive being enveloped by the outer layers of their star as its size rapidly expands to a red giant. Being inside a star is not a health spa for planets. The worlds risk vaporisation and being knocked by the expanding gas to skitter deeper into the hotter reaches of the star. How far out the red giant’s swollen surface extends depends on the star’s mass. Around our Sun at 1au, the Earth risks being enveloped when our Sun swells in its red giant phase. The three planets around the far more massive precursor to the pulsar PSR B1257+12 would certainly be in the star’s fiery belly.
An even bigger problem occurs when the star explodes as a supernova. The resultant loss of stellar material will see the star shrink to a fraction of its initial mass. The mass loss will cause a big drop in the star’s gravitational pull that will most probably untie a small object such as a planet, and see the world drift away from the star. Such a fate could be avoided if the supernova blast were asymmetric, kicking the remaining neutron star towards the planets to permit a lucky recapture. Frankly, this seems exceptionally improbable, especially for grabbing hold of three planets.
A final problem is that the planets around PSR B1257+12 orbit in the same plane, which suggests that they have been fairly undisturbed since their initial birth in a disc. A shared orbital plane also makes an additional option of planet capture unlikely. A close encounter between a pulsar and another star could potentially allow planets to transfer stellar parents to orbit the pulsar after its dangerous death had occurred. Yet if the three PSR B1257+12 planets had been dragged out of their original orbits around one star to circle another, their paths should skew randomly about the pulsar. This is not what is seen. What we need is a formation option with a less violent evolution than an explosion plus recapture.
Memnonides planets
If pulsar planets orbit in a tidy, disc-like plane, then perhaps a new protoplanetary disc was created after the star had become a pulsar. This new disc could then trigger a very late generation of planet formation. In keeping with the mythological analogy, this idea became the Memnonides Scenario, after the Memnonides birds in Greek mythology that rose from the funeral pyre of the fallen warrior, Memnon.
A new disc fits the observation of the PSR B1257+12 system, but opens the question of the origin of the disc material. The original protoplanetary disc would have long since been dispersed, so the pulsar would need a fresh supply of planet-building dust.
One source for a new disc might be the outer layers of the red giant star thrown off during the supernova explosion. If this material fails to escape the gravitational pull of the remaining stellar corpse, it will fall back to encircle the freshly made pulsar. Assuming that the incoming material can rotate fast enough to stop falling on to the pulsar, then a disc is formed. How much of the red giant’s outer layers could be recycled into a planet-forming disc is uncertain. But if the initial star were large enough, then sufficient material could fall back to create the small planets circling PSR B1257+12.
What is harder to explain is the fact that PSR B1257+12 is a millisecond pulsar, yet there is no sign of a companion star. If the outer layers from its stellar sibling were needed to spin the pulsar to sub-second speeds, the remnant of that star should remain as a white dwarf. Where is it and could this be linked with the planets?
The most promising idea for PSR B1257+12 is to blame its missing stellar sibling for the whole planet-making action. In this macabre scenario, the companion star is ripped to shreds by the pulsar and the new protoplanetary disc is formed from its broken remains.
One way to destroy the sibling is during the supernova. If the explosion is asymmetric, the newly forming pulsar could be slammed into its companion. This collision would rip apart the sibling star to form a protoplanetary disc around the pulsar. Collisions between stars are extremely rare, but pulsar planets are not yet considered common.
Alternatively, the pulsar could blowtorch its companion to pieces. Named black widow pulsars after the spiders that devour their mate, these cannibalistic dead stars orbit their sibling so closely that their radiation vaporises the other star to form a disc of its remains. One such stellar homicide is presently being committed by pulsar PSR J1311-3430. 3
In 2012, a faint star was discovered that seemed to be varying in colour from bright blue to dull red. Its location was also a strong source of high-energy Gamma rays, but there were only intermittent intervals of radio-wave emission.
The high levels of radiation led to a suspicion that a pulsar was at the heart of this mystery. The challenge was spotting the telltale lighthouse pulse in the Gamma ray emission. Due to the high energy of this radiation, pulsars emit far fewer Gamma rays than radio waves, and this makes it hard to spot a rapid flash. Careful analysis of four years’ worth of data taken by NASA’s Fermi Gamma-ray Space Telescope eventually yielded a successful result; the colour-changing star was indeed orbiting a pulsar, the first to be identified purely by its Gamma ray flash.
PSR J1311-3430 is a 2.5 millisecond pulsar, rotating 390 times per second. The pulsar and its sibling were incredibly close, orbiting at a separation just 40 per cent greater than the distance between the Earth and the Moon. This led to an orbital time of 93 minutes – less than the average round-trip commuter time in the UK. It was this proximity to the pulsar’s lighthouse beacons that was the root of the companion star’s colour change.
The side of the companion star facing its dead sibling felt the full assault of the pulsar’s radiation. This pounding drove its temperature up to 12,000°C (21,600°F) – more than twice that at the Sun’s surface – and turned it a bright blue. The far side of the star was a cooler red, corresponding to a temperature of just 2,700°C (4,900°F). As the star rotated about the compact pulsar, its red and blue faces were alternately visible from the Earth.
The pulsar also explained the star’s faintness. The star was tiny, at only about 12 times the mass of Jupiter. Since the pulsar had a millisecond spin, the companion star must have previously donated its outer layers to recycle the pulsar’s rotation. What remained was probably a helium core, possibly too light to have fully compressed into a white dwarf. Exposure to the pulsar’s relentless radiation had then whittled the star down to a near planet-sized object. Its disintegrating body trailed behind the star to form a barrier encircling the pulsar. The pulsar’s radio waves were scattered or absorbed by these shredded remains, allowing only the higher-energy Gamma rays to punch through and be seen on Earth. As the star continues to be vaporised away, its remains may condense into a disc around the pulsar. The result would be a lone millisecond pulsar and the starting disc for a new generation of planets.
If the pulsar is not close enough to blowtorch apart its sibling, this companion will eventual die and become a white dwarf. Tucked inside its own Roche lobe, the white dwarf can safely orbit the pulsar. Yet, this safety may not last forever. The reason is gravitational waves.
One hundred years ago, Albert Einstein predicted that the fabric of space should be rippling with waves. He pictured the Universe as a taut rubber sheet on which massive objects create curved indentations. Gravity is the result of these curves, forcing lighter objects to move towards the more deeply embedded heavier ones. As objects move, the sheet flexes to reflect their new positions, creating oscillations that travel outwards as a gravitational wave.
The first direct detection of gravitational waves was announced on 11 February 2016. It was possibly the worst-kept secret in scientific history, with news of a successful detection rumoured since the end of the previous year. The find was made by the US-based detector LIGO, which spotted the signature of two coalescing black holes. As the densest objects in the Universe, the ripples produced during a black hole merger are one of the strongest gravitational wave signals imaginable. Next in line are the vibrations from other interacting stellar remnants.
As a pulsar and white dwarf binary orbit one another, the continual flexing of space produces a steady source of gravitational waves. The energy to power these waves is pulled from the orbit motion, causing the pair to move closer together. 4 As the two approach, the pulsar’s stronger gravitational pull will shrink the Roche lobe of the less massive white dwarf until it overspills for a second time.
While a white dwarf has not quite crushed itself into neutrons as in a pulsar, its incredible density causes it to behave differently from normal material. When it loses mass as its layers are pulled towards the pulsar, a white dwarf expands rather than shrinks. This causes more and more of the dead star to overspill on to the pulsar until it is entirely disrupted. The overflowed remains can form the disc for a generation of planets.
A protoplanetary disc formed from the ashes of a dead star has a couple of interesting properties. First, such discs are short lived. The shredded stellar material is pounded by the pulsar’s radiation and rapidly heats, causing the disc to spread outwards. The disc density lowers until eventually it cannot form planets. Estimates for the lifetime of these recycled discs are around 100,000 years, compared with 10 million for a Sun-like protoplanetary disc. Such a short time period makes the formation of gas giants unlikely, but there is the possibility of close-in terrestrial planet formation, which could explain the trio seen around PSR B1257+12.
A disc made from the remains of a supernova explosion or the shredded body of a white dwarf is likely to have a strange composition. With nuclear fusion in the white dwarf’s star predecessor largely stopping at helium, such discs would probably be carbon rich. The resulting terrestrial worlds could therefore be examples of the diamond planets mentioned in Chapter 7.
While it would not apply to the trio of worlds around PSR B1257+12, it is worth noting that there is one other very strange way to build a single diamond world around a dead star. This involves converting a star directly into a planet.
The star that became a planet
In December 2009, a pulsar was discovered with a spin time of 5.7 milliseconds, or 175 rotations per second. As a second star was needed to spin the pulsar up to millisecond speeds, the skies were searched for its sibling. This search initially yielded nothing: pulsar PSR J1719-1438 seemed to be alone.
The dead star’s discovery had been made using the Parkes Radio Telescope in Australia. The 64m (210ft) dish is famous for receiving the majority of Neil Armstrong’s iconic Moon-landing broadcasts. But perhaps for astronomers the telescope’s greater claim to fame is as the record holder for the world’s most successful pulsar finder. This pulsar was discovered 4,000 light years away in the constellation of Serpens, the Snake. Nearly two years later, Parkes, along with the Lovell 76m (250ft) telescope at Jodrell Bank Observatory in the UK, was used to uncover its extremely strange sibling.
The slight variations in the timing of the pulsar’s flash revealed that it remained in a binary with an orbit lasting two hours and ten minutes. However, the pulsar’s sibling was a major lightweight, with a mass similar to that of Jupiter. So was this truly a star or a planet?
Due to the swift orbital time, the two companions were closely packed, separated by a distance of 600,000km (370,000mi); slightly less than the radius of the Sun. Since there was no X-ray emission, the pulsar’s companion could not currently be overspilling its layers on to the pulsar. This meant that the companion’s size had to sit within its Roche lobe, capping it at about 5 Earth radii for its current distance from the pulsar. This sibling was therefore a super Earth with the mass of Jupiter. A gas giant with a thick hydrogen atmosphere could never squeeze into a radius that small. This left the alternative of a very small white dwarf.
As the lighter versions of neutron stars, white dwarfs have a typical mass of two-thirds of the Sun packed down to the size of the Earth. To be the mass of Jupiter, PSR J1719-1438’s strange companion must have donated around 99.8 per cent of its mass to its pulsar. Due to the strange configuration of matter inside a white dwarf, this would have expanded the dead star to its larger radii. Yet despite this colossal mass loss, the star had avoided being completely disrupted.
As the white dwarf’s mass is transferred to the pulsar, the gravity of each changes. This alters the shape and extent of the pair’s Roche lobes. If the distance between the sibling stars is just right, then this adjustment can allow the white dwarf to duck back inside its own Roche lobe and stop the overflow on to the pulsar. It is a tricky game. If the binary pair are far apart, then the white dwarf never overspills its mass on to the pulsar. If the pair are too close, then the overspill can never be stopped and completely disrupts the white dwarf.
Made primarily of incredibly dense carbon, this white dwarf world has a density above 23g/cm3, far higher than the Earth’s 5.5g/cm3. At such values, carbon would have crystallised into a genuine solid diamond world.
This has to be one of the strangest objects in the Universe: a planet made from diamond, orbiting a sibling the size of a city, which once used to be a star.
Notes
1. I bet you just tried that. So did I.
2. As we saw in Chapter 6, charged particles can both create a magnetic field and feel a force in an existing field.
3. Pulsar names with a ‘J’ use a more up-to-date and precise position coordinate for their location in the sky.
4. The orbit’s energy also powers tidal heating: in Chapter 5, the puffy hot Jupiter WASP-17b was tugged on to a closer circular orbit as it was heated by the varying gravitational pull from the star.