There is a beautiful pulsar paper coming out in Nature tomorrow, 28th of October issue (Demorest et al 467, 1081, 2010)
Green Bank Telescope measurements of PSR J1614-2230 show it to be a 3ms binary pulsar with a white dwarf companion in an orbit aligned near perfectly edge on to our line of sight.
Measurements of the Shapiro Delay provide a measurement of the mass of the white dwarf, allowing the mass of the neutron star to be calculated from the known orbit of the white dwarf.
The resultant inferred mass is 1.97 +/- 0.04 solar masses.
This is a very nice result.
There has been a lot of speculation over the decades about what the maximum mass of neutron stars would be.
We know there has to be a maximum mass for neutron stats. The central density of neutron stars increases with the mass of the neutron star, and if you extrapolate the nuclear equation of states to higher densities, the sound speed is inferred to exceed the speed of light (ok c/SQRT(3)) which is not possible, so something has to give before you get to those densities.
Question is how high can you go.
Naive extrapolation suggests maximum masses of maybe a bit over 2 solar masses, but definitely no more than about 2.7 solar masses.
On the other hand, nuclear and particle physicists are fond of speculating that there could be phase transitions at higher pressures, leading to an equilibrium kaon abundance, or some other net strangeness quark puddle.
Such quark stars, or strange stars, generally have soft equations of state, the density increases steeply with increased pressure, and therefore they have low estimated maximum masses.
Strange stars, generally, have estimated maximum masses of more like 1.7 solar masses.
Therefore this result rules out a range of conceivable equations of state at high densities and pressures, and allows, necessarily, high neutron star masses.
Since neutron stars are thought to form with a narrow range of of initial masses, maybe 1.2-1.4 solar masses, they have to gain weight somehow - and a natural way to do so is by accreting material from a stellar companion, as it evolves onto the giant branch.
Some of the companion envelope is transferred to the neutron star, which becomes more massive and also gets spun up, leaving the observed massive neutron star - white dwarf binary system.
A nice secondary implication is that this implies some binaries will contain higher mass neutron stars. As a very rough approximation, the signal amplitude of gravitational radiation from coalescing neutron stars goes up with the mass of the neutron star, so this implies some such coalescences may have signal amplitudes about 50% larger than otherwise expected. Better still, at fixed sensitivity the volume that can be probed increases as the cube of the amplitude (for nearby sources), so the estimated rate can be up to three times higher.
Which is promising for the prospects of future detection of gravitational radiation signals by the advanced LIGO observatory.
- Log in to post comments
The mechanism they used to measure the mass is also quite interesting.
So what happens if a neutron star accretes sufficient mass to actually reach the mass limit? Are they predicted to undergo their own variant of type Ia supernovae, and if so are there any known candidate events?
@Birger - yup, the Shapiro effect is nifty and not a lot of systems which are aligned well enough to see it.
@andy - if the NS goes over the maximum ass it collapses to a black hole. If it is rapidly rotating collapse stalls while angular momentum is shed, and this is a likely origin scenario for short gamma-ray bursts - though there what you accrete is another neutron star, probably.
Energy release is neutrinos, and parts of the neutron star is opaque enough to neutrinos for heating to be significant; there is also opportunity some ferocious dynamo winding and magnetic fields reaching quantum electrodynamic limits "shorting space".
Definitely opportunity for strong jet outflows, possibly some anomalous blow out of ~ 1+% of solar mass - material would have weird isotopic composition, due to strong free neutron flux, and may be significant component of anomalous r-process element abundance in the universe.
About time this paper came out! Ransom gave us a preview in a colloquium back in May; I was worried that something horrible had happened to the result.
Not so sure I get that part about the constraint due to speed of sound limitations. Admittedly I'm not a physicist, but perhaps you can elucidate the argument. I would think the constraints are particles and information cannot exceed the speed of light. I suspect there is a relativistic correction for the sound speed. My uninformed hunch is the relativistic correction would prevent the sound speed from exceeding c.
Of course we have a dynamic going on here, softer EOS, allows the NS to compress more which leads to much higher pressure and density, which I presume implies the max mass goes down slowly with increasing stiffness.
Classically the sound speed is just the derivative of pressure with respect to density, square root thereof, with some dimensionless numbers thrown in.
There is a relativistic correction, Oppenheimer worked out a lot of the details. As a first approximation, relativistic fluids are worse - the self-energy shows up as extra energy density, so you approach speed of light faster.
Full equation of state can be solved exactly for ideal relativistic fluids, makes for a good homework problem.
There's also an idea (been around quite a while) that neutron stars may have a larger range of birth masses, but that the best-measured NS masses (those in double NS systems) are all low due to the particular binary evolution pathways that produce those systems. Several recycled NSs have "normal" mass measurements, suggesting it doesn't take much mass to spin them up, maybe ~0.1 Msun. Or it could be that the transferred mass varies among different systems, remains an interesting question.
This is indeed a very nice result; all previous claims of high-mass NSs had much larger statistical (and systematic) uncertainties, but this one nails it down.
Btw, one not so well known fact about Shapiro delay is that every photon we see from
any telescope undergoes Shapiro delay.
See http://adsabs.harvard.edu/abs/1988PhRvL..60..173L
What is the difference between anomalous R-process and normal R-process?
btw do the results on eqns of state still hold true in alternate gravity theories?