LUNA Results Suggest Really Old Stars Exist Despite Popular Models

[shrumeo]

There is a news column in Science that talks about soon-to-be-published results from the group at LUNA in Italy.
http://www.sciencemag.org/cgi/content/full/304/5675/1226b

Their results suggest that the current model for the CNO cycle (carbon-nitrogen-oxygen) was flawed and the calculated rates were about twice too high. This suggests that the oldest stars we have observed are older than 14 billion years.

Yet, the data from WMAP has allowed folks to calculate the age of the universe at 13.7 billion years, with an accuracy or +/- 1%.

Will someone please give us an explanation?

 

[Labguy]

There is a news column in Science that talks about soon-to-be-published results from the group at LUNA in Italy.
http://www.sciencemag.org/cgi/content/full/304/5675/1226b

Their results suggest that the current model for the CNO cycle (carbon-nitrogen-oxygen) was flawed and the calculated rates were about twice too high. This suggests that the oldest stars we have observed are older than 14 billion years.

Yet, the data from WMAP has allowed folks to calculate the age of the universe at 13.7 billion years, with an accuracy or +/- 1%.

Will someone please give us an explanation?

How do you get to the article without being a member?

 

[shrumeo]

huh?
oops, I didn't know you needed a subscription just to see it.
here I'll copy and paste the column:
-------------------------------------------
NUCLEAR ASTROPHYSICS:
New Measurement of Stellar Fusion Makes Old Stars Even Older
Kim Krieger

A key nuclear reaction inside stars takes significantly longer than standard models assume, European researchers have discovered. The result, which nuclear physicists at the Laboratory for Underground Nuclear Astrophysics (LUNA) in Gran Sasso, Italy, report in a pair of online papers, implies that the most ancient star clusters are at least 700 million years older than previously believed.

"The LUNA experiment is beautiful," says John Bahcall, an astrophysicist at the Institute for Advanced Study in Princeton, New Jersey, praising the group as "magically gifted experimentalists."

The LUNA team used an underground particle accelerator at Gran Sasso to measure the speed of the carbon-nitrogen-oxygen (CNO) cycle, one of the pathways by which stars fuse hydrogen into helium, releasing energy (see diagram). The cycle determines how long it takes a youthful hydrogen-burning star to turn into a giant helium burner. Astrophysicists can estimate the age of a star on the cusp of that transition by measuring its mass and then calculating how long it took to reach its current state.

The CNO cycle, however, is only as fast as its slowest step: a nuclear reaction in which the isotope nitrogen-14 absorbs a proton from hydrogen and turns into oxygen-15. Researchers had estimated the rate of the reaction by shooting protons at nitrogen-14 in particle accelerators. But the measurements were marred by noise from cosmic rays, and astrophysicists suspected they erred on the speedy side.

In papers scheduled to be published in Physics Letters B and Astronomy and Astrophysics, the LUNA researchers report that the limiting step is indeed only half as rapid as previously assumed. Working 1400 meters underground to shield their detectors from cosmic radiation, they smashed protons into a nitrogen-14 target and then measured the gamma rays the nitrogen released as it became oxygen-15. The results push the age of the oldest stars to almost 14 billion years. That's close to the figure of 13.7 billion years for the age of the universe that physicists derived from measurements by the Wilkinson Microwave Anisotropy Probe (Science, 14 February 2003, p. 991), although both still have significant uncertainties, Bahcall says.

The team plans to repeat the experiment at more realistic collision energies, says Carlo Broggini, spokesperson for the LUNA project. The first set of experiments was run at energies above 140 kilo-electron volts (KeV), Broggini says. A new gamma ray detector should allow researchers to study collisions at close to 25 KeV, the peak energy level at which the reaction occurs in stars.

 

[shrumeo]

I would imagine the answer lies in the "although both still have significant uncertainties" part. Or maybe they'll get different results at lower energies (?)

 

[Chronos]

I would be interested to see the 25K results. I am, however, a little gun-shy about this approach. There are other things going on in heart of the stellar furnace that would be hard to duplicate in the lab but might affect the process.

 

[shrumeo]

true, just because you get one set of results in a particle accelerator doesn't mean you can extrapolate them to represent a star (or does it?)

edit:
also, downstairs, i hooked up with a physicist so we could use his Van der Graaf acc. to shoot 2 MeV protons at a certain target. it's amazing that the nuclei in the center of a star have ~0.01 the energy. I must be misunderstanding it.

 

[Labguy]

The oldest stars, and any that would be left over from "first generation stars, do not fuse H to He by the CNO Cycle. Any first generation star (therefore old) would be small in mass and fuse by the Proton-Proton Chain process; that is, stars with a core temperature below about 16 million K. Look at the abundances and mostly life-spans of the G, K and M stars shown at:

http://anzwers.org/free/universe/startype.html

Notice that G's, like our sun, live ~10 billion years and the smaller K and M class stars live longer than the (estimated) current age of the universe. I would think that these stars, measured in globular clusters, would represent the oldest known stars. Any CNO burning star would be gone by now unless it is several generations from initial star formation.

 

[shrumeo]

ok, so that pushes the age of the first generation stars back to before 14 billion years, well before

the first gen. would have to make the C, N, and O for the next gen. to burn

??

edit: all the article says is that this pushes back the estimates of the first stars formed (i think)

 

[Labguy]

ok, so that pushes the age of the first generation stars back to before 14 billion years, well before

the first gen. would have to make the C, N, and O for the next gen. to burn

??

edit: all the article says is that this pushes back the estimates of the first stars formed (i think)

No, or at least I don't see how. The first generation stars can live as long as shown on the chart, but we (the universe) haven't reached the age yet of the maximum life of the smaller stars. There aren't many (much) of the heavier isotopes in an esentially H and He star, but some do exist and some others are created while stars burn by the Proton-Proton chain. At a mass over ~ 1.1 solar masses, the CNO Cycle will dominate and will, of course, in any second generation+ stars that exceed the 1.1 figure.

The P-P Chain and the CNO Cycle do predominate in stars of the given masses, but there is almost always at least a bit of each going on in any star with a core below 100 million K. Shells of heavier and heavier elements (isotopes) form at cores while H fusion still happens further from the core, even in the massive stars. The following sites aren't too great on detail of the development of the shells, but see:

http://csep10.phys.utk.edu/astr162/lect/energy/cno-pp.html
(Note last comments at bottom)

http://csep10.phys.utk.edu/astr162/lect/energy/cno.html

http://www.shef.ac.uk/physics/people/vdhillon/teaching/phy213/phy213_fusion3.html

http://aether.lbl.gov/www/tour/elements/stellar/CNO.html

 

[shrumeo]

all i was saying is that the column says that the authors say that it pushes back estimates of the age of the oldest stars.

that does not mean that the oldest stars ran or run on the CNO cycle
it just means they used the estimates of the ages of CNO stars to estimate the ages of the first generation, i guess

 

[Nereid]

The results push the age of the oldest stars to almost 14 billion years.

This seems to be the critical sentence in the article which shrumeo posted. However, without reading the Physics Letters B and Astronomy and Astrophysics papers (or preprints), we don't know how well this sentence reflects the LUNA's team's views. Perhaps they plugged the new cross sections into well-established stellar models, and checked the revised results against observational data on old stars? Perhaps there'll be a series of papers over the next few years showing how stellar models need further tweaking, in light of the LUNA results?

It certainly sounds like an interesting result, and goes to show that there are still many interesting areas of research in 'old' topics!

Anyone know if theoreticians feel they can/could calculate this (and other) cross sections sufficiently accurately that the experimental results will require some homework revision for nuclear physics?

 

[turbo]

Astronomy & Astrophysics article

http://www.edpsciences.org/articles/aa/abs/2004/23/aa0020-04/aa0020-04.html

The article is abstracted at the link above. I don't have a subscription (200 Euros for the on-line version), but the article is out now and perhaps someone with access to a subscribing library can review the article and give us an overview of the LUNA team methodology.

 

[shrumeo]

Thanks for the link.

The PDF is here:


www.geocities.com/shrumeo

Um, it's the little link above the ads. :)

 

[shrumeo]

In section 4, they seem to "nicely" question the lambdaCDM model that was used to calculate the 13.7 Gy. But they also say that any changes made to the ages of old globular clusters would still be within experimental error.

 

[turbo]

Thanks for the link, Shrumeo. As I understand it, the thrust of the project was to model the slowest part of the CNO reaction at lower energies than has been done in the past, since extrapolation from high-energy reactions seemed to result in dimmer and redder turnoffs from main sequence. Their results imply that globular clusters may contain stars significantly older than the age of the universe (13.7 billion years) commonly accepted. If the results of this study are accepted, will the Hubble constant have to be revised downward from its ~74 value, or will the age of the universe have to be bumped up? There may be more ways out of the conundrum, but those are the quick and dirty ones.

The globular cluster stars pose another problem for cosmologists, anyway. Even their current (pre LUNA) age estimates should make us ask how did these stars form so early in the universe. A parallel problem exists with quasars. Few of us would argue that a quasar z>6 with the luminosity of a thousand galaxies represents a VERY high level of order. How could that level of organization have occurred so soon after (or nearly concurrent with) hydrogen reionization? Every new record-setting redshift for these objects just compounds the problem.

 

[Nereid]

It's certainly clear that there are many questions and lots of things to research

However, given how difficult some of this stuff is to nail down ... look at how many decades of work were needed to get a half-way decent estimate of the (local) Hubble constant (and the 95% CL range is still quite large) ... I'd say it'll be decades more of painstaking work before a variety of the $$\Lambda CDM$$ is on the same firm experimental footing as, say, main sequence stellar evolution models

 

[turbo]

Hi, Neried!

How do you manage to add real formatted text to posts? I'm sorry if this is a goofy question with a trivial answer, but I really wanted to use the superscripts and symbols in my previous post but ended up using the really imprecise and lame expression:

"slowest part of the CNO reaction"

instead.

TIA

PS - Did you notice how casually I slipped in that quasar reference at the end? I really would like to address the excess-redshift associated with quasars, bLacs, Seyferts, etc at some point, because they are going to be a big problem in cosmology very soon - especially after the giant binocular telescope comes on line and some grad student starts publishing about z>8 quasars.

 

[Chronos]

Very doubtful the new observations will contradict the very solid math as we now know it. If GR and QT were terribly wrong, we would already have seen the contradictions. Mainstream physics leaves many questions. Quack physics leaves more. My biggest objection to crap like Arp puts out is it deflects bright minds like yours from thinking about real issues.

 

[shrumeo]

I think it's people like Arp that keeps everyone on their toes, so to speak. Not just in physics. We need people to question generally accepted theories so down the line folks don't become complacent in flawed theories just because they are the "standard." I think it's funny every time I read the line "standard cold dark matter model" and then all the problems associated with it are discussed. Why is it the "standard"? Is it just because there needs to be a standard?

Like this quote from the intro paragraph in the paper:

"Their age practically coincides with the time
elapsed since the epoch of the formation of the first stars in
the Universe and provides an independent check of the reliability
of standard (and non-standard) cosmological models."

Why didn't they just say "check the reliability of cosmological models" ?

 

[Nereid]

I think it's people like Arp that keeps everyone on their toes, so to speak. Not just in physics. We need people to question generally accepted theories so down the line folks don't become complacent in flawed theories just because they are the "standard." I think it's funny every time I read the line "standard cold dark matter model" and then all the problems associated with it are discussed. Why is it the "standard"? Is it just because there needs to be a standard?

No; rather because it's a model which contains most (or all) of the elements which are generally accepted within the relevant community, and none of those which are currently not so accepted. This makes it easier to do work, you don't have to spend dozens of pages presenting details of the standard model - and the observations which are consistent with it - and can concentrate on what's new, different, or disputed.

Why didn't they just say "check the reliability of cosmological models" ?

Because the primary purpose is to check the reliablity of the standard models; if they can also check non-standard ones, that's a bonus.

 

[Nereid]

The PDF is here:


www.geocities.com/shrumeo

Um, it's the little link above the ads. :)

Excellent, thanks shrumeo!

also, downstairs, i hooked up with a physicist so we could use his Van der Graaf acc. to shoot 2 MeV protons at a certain target. it's amazing that the nuclei in the center of a star have ~0.01 the energy. I must be misunderstanding it.

You understood perfectly well; it's pretty easy, in the lab, to accelerate particles to 'temperatures' (aka energies) considerably higher than those found in the cores of Main Sequence stars; in fact, as the LUNA team's paper make clear, it's quite difficult to simulate star core processes in the lab (as Chronos pointed out).

This seems to be the critical sentence in the article which shrumeo posted. However, without reading the Physics Letters B and Astronomy and Astrophysics papers (or preprints), we don't know how well this sentence reflects the LUNA's team's views. Perhaps they plugged the new cross sections into well-established stellar models, and checked the revised results against observational data on old stars?

Yep, that's just what they did, the 'observational data on old stars' being high quality work on nearby, low metallicity globular clusters.

Perhaps there'll be a series of papers over the next few years showing how stellar models need further tweaking, in light of the LUNA results?

... and here's a quote from the A&A paper: "New stellar models have been computed with the same code described in Straniero et al. (1997), but updating the rate of ¹⁴N(p, γ)¹⁵O." IOW, only one change, assumes secondary effects are either negligible or already accounted for in the models. We'll see.

Their results imply that globular clusters may contain stars significantly older than the age of the universe (13.7 billion years) commonly accepted. If the results of this study are accepted, will the Hubble constant have to be revised downward from its ~74 value, or will the age of the universe have to be bumped up? There may be more ways out of the conundrum, but those are the quick and dirty ones.

The globular cluster stars pose another problem for cosmologists, anyway. Even their current (pre LUNA) age estimates should make us ask how did these stars form so early in the universe.

It's actually a bit more subtle than that. First, there's no doubting that the old globulars formed pretty much just as soon as stars could, certainly within 1 Gy of the BB. Next, every parameter estimate comes with a range within which is it consistent with observational data (95% CL, 99% CL, it's a personal choice). Third, the LUNA authors are very clear that their results - on their own - aren't inconsistent with the standard $$\Lambda CDM$$; indeed, they explicitly suggest that they may be able to help refine it (as shrumeo's post, and my response, clearly show).

Finally, the best part - for those interested in experimental tests of cosmology theories - comes in the penultimate para: "An exhaustive comparison between stellar and cosmological ages requires a detailed statistical analysis taking into account all sources of errors (experimental and theoretical). This is beyond the purpose of the present paper and will be presented elsewhere."

IOW, it's but one more step in a very long journey.

 

[Chronos]

Speaking of old stars, some interesting stuff released by ESO a few days ago.

http://www.eso.org/outreach/press-rel/pr-2004/pr-17-04.html

 

[Nereid]

Speaking of old stars, some interesting stuff released by ESO a few days ago.

http://www.eso.org/outreach/press-rel/pr-2004/pr-17-04.html

Funny that; I recall another PF member posted a link to the same PR, in another thread ...