Dust and cosmology

Can a galaxy build enough dust in 600-800 million years so that it is
visible from earth at z = 6?

That is the question.

B.T. Draine is an astronomer at Princeton that has studied dust. Since a
long time: he has published since the eighties about this subject. I
came across his text:

Interstellar Dust Models and Evolutionary Implications

available at
ftp://ftp.astro.princeton.edu/draine...CS_414_453.pdf

I found it a very clear, introductory article into the realm of dust.

When I started reading that paper, I saw immediately why the arguments
about huge quantities of dust generated by supernovae are wrong.

quote
the average mass of dust per SN does not appear to be large: the
Type II SN 1987a produced less than 8 e" 10M-bM-^H'4 solar masses of dust.
end quote (Page 462 of the above article)

Problem is, super novae destroy most of the created dust in the huge
explosion...

Most of the dust in the milky way is generated by red-giant stars
(0.0025 solar masses per year), stars that are at the end of their life
of normal, sun-like stars. All this dust production can't happen in 800
My since a normal star like the sun, starts producing dust in its
red-giant phase, after 7-8 Gy.

Planetary nebulae produce 0.002 solar masses of dust per year. The same
argument: all those progenitor stars live much longer than 7-8 Gy!

Super novae can't help us very much since a single super nova can
destroy approx. 1000 solar masses of dust in their explosion. Since we
are just 800 My from the supposed "bang", the over-abundance of super
novae needed makes this removal of dust all the more probable!

A further problem in this context is the fact that most of the dust (at
least in our galaxy) is not star dust, but condensation of molecular
gases around existing dust in the inter stellar medium. And this is a
very slow process of chance collisions in molecular clouds.

At the end of his paper, Mr Draine comes explicitely to the cosmological
dust problem:

"Growth of Dust at High z: the example of J114816+525150" (page 469)

He writes:
quote (469)
The time available for stellar evolution prior to z = 6.42 is limited:
if star formation began at z = 10, the oldest stars are only 400 Myr old
at z = 6.42, and only massive stars will have been able to evolve--
insufficient time for low-mass stars to evolve to the asymptotic giant
branch which dominates production of stardust in the Milky Way
end quote

This is very clear: there is no time to evolve all that dust if we
assume that the conditions then were like in the milky way today, i.e.
that we are seeing a normal galaxy.

He continues:
quote
This has led a number of authors to propose that supernovae are
responsible for the dust in high-z galaxies (e.g. Maiolino et al.
2004; Sugerman et al. 2006; Bianchi & Schneider 2007). Dwek et al.
(2007) discuss the dust in J1148+5251 and conclude that supernovae would
have to produce approx 1M0 of dust per supernova to explain the
observations, but note that this is considerably in excess of what has
been observed in SN ejecta.
end quote

Note the careful wording. He wrote a few pages above that SN 1987a
generated less than 8 x 10-4 solar masses, i.e. 10 thousand times less
than what would be needed. :-)

He continues:
quote
J1148+5251 contains a large mass of molecular gas, detected in CO:
J = 7 -- 6, 6 -- 5, and 3 -- 2 (Bertoldi et al. 2003b; Walter et al.
2004) The CO that is observed, and the H2 that must accompany the CO, is
not supernova-produced: even if those molecules do form in the ejecta,
they are efficiently destroyed by the reverse shock when the
high-velocity ejecta are decelerated.
end quote

But if they are not super nova produced, they need a whole generation of
normal stars AT LEAST. So, this dust must be AT LEAST older than 7 Gy
only 800 million years after the bang.

That is the problem for big bang proponents... How can this OLD DUST
appear at z6?

He continues:
quote
Instead, the H2 must be formed by the mechanism that dominates H2
formation in the Milky Way: catalysis on grain surfaces. Given that the
massive stars present in these galaxies will destroy H2
molecules--primarily through photodissociation--each H nucleon in the gas
must, on average, have collided with grain surfaces many times in the
age of this galaxy. Metal atoms and ions move only a few times more
slowly than H, and will also collide with grain surfaces many times; if
they stick, they will form new grain material. This is the process that
dominates grain formation in the Milky Way, and there is no reason
not to expect it to dominate growth of grain material in J1148+5251.
end quote

This is also very clear: there is no time to produce all this old dust!

He finishes with this sentence
quote
Supernovae are of course required to produce the metals that compose the
grains, and to provide some supernova-condensed "stardust" to provide
some surface area on which to grow more material in the ISM, but the
bulk of the dust mass in high-z galaxies will be primarily the result of
grain growth competing successfully with grain destruction in the ISM.
end quote

Problem is, the ISM right after the supposed "bang" must have been a
hellish place with UV radiation from millions of new super-novae
building in no time huge galaxies.

Anyway, the galaxy I cited in my last contribution was twice the milky
way at around z=6. The milky way produces around 0.005 solar masses of
more dust per year. It has around 2.5 x 10E7 solar masses of dust, so
starting at zero we need 2.5 x 10E7 / 0.005 around 5 Gy to make a
similar quantity of dust. Ignoring all removal processes mentioned above
it still doesn't work.

Conclusion:
The dust present in all this far away galaxies is an unexplained anomaly
of big bang theory.

[[Mod. note --
While low-mass stars may dominate dust production in the present-day
Milky Way, it's not obvious that that was the case in the early universe.
In fact, I rather doubt we know the mass distribution of stars (known
as the "mass function") of galaxies formed at (say) z=10.

If dust then was produced by more massive stars, 400 Myr may have been
plenty of time for their lifetimes. I'm 3 time zones away from my
stellar-evolution textbooks right now, but my recollection is that the
entire lifetime of a 50 M_sun star is much less than 10 million years.
-- jt]]

Le 01/12/2015 07:42, jacobnavia a écrit :
[[Mod. note --
While low-mass stars may dominate dust production in the present-day
Milky Way, it's not obvious that that was the case in the early universe.
In fact, I rather doubt we know the mass distribution of stars (known
as the "mass function") of galaxies formed at (say) z=10.

If dust then was produced by more massive stars, 400 Myr may have been
plenty of time for their lifetimes. I'm 3 time zones away from my
stellar-evolution textbooks right now, but my recollection is that the
entire lifetime of a 50 M_sun star is much less than 10 million years.
-- jt]]

1) You say:
While low-mass stars may dominate dust production in the present-day
Milky Way, it's not obvious that that was the case in the early universe.

Who knows? It is up to YOU to propose an alternative mechanism for
creating all this dust!

In my article I pointed out that all known processes for creating dust
do not work in a time frame of less than 1000 My. And I pointed out that
super-novae do NOT create significant quantities of dust to create a
dusty galaxy in only the given time.

So, it was not normal stars, nor super-novae that created that dust.

What then?

2) You say:

If dust then was produced by more massive stars, 400 Myr may have been
plenty of time for their lifetimes.

Sure, there is plenty of time for several generations of super-novae.
But they do not create dust in the required quantities!

I do not understand your answers since you seem to ignore the arguments
advanced in my post. Can you please explain how do you think that the
dust was created out of just H and He in only 700-800 My?

Thanks

In article ,
jacobnavia writes:
B.T. Draine is an astronomer at Princeton that has studied dust.

He is one of the leading experts.

ftp://ftp.astro.princeton.edu/draine...CS_414_453.pdf
I found it a very clear, introductory article into the realm of dust.

Indeed so, though one area that has seen great progress since 2009 is
dust production in SNe. I'm not sure just where things stand at the
moment but expect some ideas to have changed.

quote
the average mass of dust per SN does not appear to be large: the
Type II SN 1987a produced less than 8 e" 10M-bM-^H'4 solar masses of dust.
end quote (Page 462 of the above article)

Several comments: SN 1987a was an unusual type of SN. SNe in the
early universe differ from local SNe by coming from generally more
massive stars with lower initial metallicity. So what's true locally
need not be true at high-z.

Problem is, super novae destroy most of the created dust in the huge
explosion...

Not entirely clear, though they may do so.

Most of the dust in the milky way is generated by red-giant stars
(0.0025 solar masses per year), stars that are at the end of their life
of normal, sun-like stars. All this dust production can't happen in 800
My since a normal star like the sun, starts producing dust in its
red-giant phase, after 7-8 Gy.

More massive stars evolve more quickly, and stars just below the SN
cutoff reach the asymptotic giant branch in a few hundred Myr. Maybe
still not fast enough for the highest redshifts but not Gyr.

Planetary nebulae produce 0.002 solar masses of dust per year. The same
argument: all those progenitor stars live much longer than 7-8 Gy!

Yep; PN are no good for the early universe because their progenitors
are low-mass stars.

Super novae can't help us very much since a single super nova can
destroy approx. 1000 solar masses of dust in their explosion.

I'm afraid we just don't know that, even locally, let alone in the
early universe.

A further problem in this context is the fact that most of the dust (at
least in our galaxy) is not star dust, but condensation of molecular
gases around existing dust in the inter stellar medium. And this is a
very slow process of chance collisions in molecular clouds.

Why do you think it's slow? The entire lifetime of molecular clouds
is of order 1 Myr, and we see "grain mantles" (the usual term for
ices surrounding dust "cores") everywhere there is high density and
high extinction. Anyway, grain mantles are probably irrelevant at
high z.

At the end of his paper, Mr Draine comes explicitely to the cosmological
dust problem:
....
But if they are not super nova produced, they need a whole generation of
normal stars AT LEAST.

You missed Bruce's conclusion even after quoting it:
Metal atoms ... will also collide with grain surfaces many times; if
they stick, they will form new grain material. This is the process that
dominates grain formation in the Milky Way, and there is no reason
not to expect it to dominate growth of grain material in J1148+5251.

There need to be a few "seed" grains from SNe, but most of the grain
material forms by direct accretion in the ISM. That surprised me,
but Bruce is the expert.

He finishes with this sentence
quote
Supernovae are of course required to produce the metals that compose the
grains, and to provide some supernova-condensed "stardust" to provide
some surface area on which to grow more material in the ISM, but the
bulk of the dust mass in high-z galaxies will be primarily the result of
grain growth competing successfully with grain destruction in the ISM.
end quote

Problem is, the ISM right after the supposed "bang" must have been a
hellish place with UV radiation from millions of new super-novae
building in no time huge galaxies.

Can you make that argument quantitative?

It's probably worth mentioning one other complication: dust _mass_ is
very poorly determined. What's measured is _reddening_, which comes
from small particles. In local dust, nearly all the _mass_ is in the
big particles. (This is like stellar mass but in the opposite
direction: nearly all stellar light comes from big stars, but the
mass is in the small stars.) The dust masses quoted assume the local
distribution of grain masses, but the extinction would be the same
whether big grains exist or not. If they don't exist, total dust
mass would be much lower than the values given.

--
Help keep our newsgroup healthy; please don't feed the trolls.
Steve Willner Phone 617-495-7123
Cambridge, MA 02138 USA

On Friday, December 4, 2015 at 12:37:11 PM UTC-5, Steve Willner wrote:


Here is a brand new paper on the topic of dust production in the "early" ev=
olution of the observable universe.

http://arxiv.org/abs/1512.00849

Title: Dust production 0.7-1.5 billion years after the Big Bang

Au: Micha=C5=82 J. Micha=C5=82owski (IfA, Edinburgh)

RLO
http://www3.amherst.edu/~rloldershaw
Help keep our newsgroup healthy; please question those who are often wrong =
but never in doubt.


[[Mod. note -- I think the first line of this message was a mistake by
the present author -- the text starting "Here is a brand new paper" was
written by the present author (Robert L Oldershaw), not by Steve Willner.
-- jt]]

In article ,
"Robert L. Oldershaw" writes:
http://arxiv.org/abs/1512.00849
Title: Dust production 0.7-1.5 billion years after the Big Bang

Seems a decent, if brief, summary of current knowledge. The article
doesn't mention the difficulty of determining dust mass from the
observed spectral energy distribution. Also, the statement in the
first sentence that "dust can be found in almost every galaxy" isn't
relevant to _young_ galaxies. (If you are talking about galaxies
actually observed, almost all of them are old.)

--
Help keep our newsgroup healthy; please don't feed the trolls.
Steve Willner Phone 617-495-7123
Cambridge, MA 02138 USA

Le 08/12/2015 05:16, Steve Willner a écrit :
In article ,
"Robert L. Oldershaw" writes:
http://arxiv.org/abs/1512.00849
Title: Dust production 0.7-1.5 billion years after the Big Bang

Seems a decent, if brief, summary of current knowledge. The article
doesn't mention the difficulty of determining dust mass from the
observed spectral energy distribution. Also, the statement in the
first sentence that "dust can be found in almost every galaxy" isn't
relevant to _young_ galaxies. (If you are talking about galaxies
actually observed, almost all of them are old.)


He says:

quote
for most of the galaxies with detected dust emission between z = 4 and z
= 7.5 (1.5–0.7 billion years after the Big Bang) AGB stars are not
numerous and efficient enough to be responsible for the measured dust
masses. Supernovae could account for most of the dust, but only if all
of them had efficiencies close to the maximal theoretically allowed
value. This suggests that a different mechanism is responsible for dust
production at high redshifts, and the most likely possibility is the
grain growth in the interstellar medium.
end quote

This grain growth among massive super novae explosions?

A whole galaxy needs to be built in an incredible short time scale. A
galaxy so powerful that we see it across 12 BILLION years!

In such a time scale, no stars can build any dust, not even AGBs as the
author says.

The last thing available to BB theory is the interstellar medium
condensing the gases and atoms produced by the supernovae... in less
than a billion years a galaxy twice the Milky Way is built...

No problem, that is maybe possible.

And the "dark ages"? Gone with the wind. 450 My was the center of the
"dark ages" as you can read in older BB proponents stories. Nobody seems
to take that seriously now. A galaxy at 450 My is estimated OK.

On 12/2/2015 3:09 AM, jacobnavia wrote:
Who knows? It is up to YOU to propose an alternative mechanism for
creating all this dust!

Quazars, The super massive black holes that may be the embryos of the
galaxies would have created at tremendous amount of metals, I suspect.

It's beginning to look like magnetic fields had a big role in the
formation of both the super massive black hole and the stars of the
visible galaxies we see today.

I see the metallicity profile of the bulge as tending to point to a
central source of metals in the galaxy .... dust pretty much from the
beginning

[[Mod. note -- I've never heard of, and cannot offhand think of any
physical mechanism for, quasars producing metals (= chemical elements
other than hydrogen/helium).
-- jt]]

On 12/10/2015 10:18 PM, David Staup wrote:
I've never heard of, and cannot offhand think of any
physical mechanism for, quasars producing metals (= chemical elements
other than hydrogen/helium).

Really?

The energy seen in a quasar is far more than fusion could provide, no?

you have a compact energy source and in falling gas, not all of which
goes into the hole.... surely fusion temps and pressures are achieved
well outside the hole, no?

[[Mod. note --
Actually I was the person who wrote the quoted remark. As David Staup
points out, I was (very) wrong. An ADS search on title words
"accretion disk" and "nucleosynthesis" turns up plenty of references,
generally in the context of gamma-ray bursts.
-- jt]]

On 12/12/2015 11:57 PM, David Staup wrote:
On 12/10/2015 10:18 PM, David Staup wrote:
I've never heard of, and cannot offhand think of any
physical mechanism for, quasars producing metals (= chemical elements
other than hydrogen/helium).

Really?

The energy seen in a quasar is far more than fusion could provide, no?

you have a compact energy source and in falling gas, not all of which
goes into the hole.... surely fusion temps and pressures are achieved
well outside the hole, no?

[[Mod. note --
Actually I was the person who wrote the quoted remark. As David Staup
points out, I was (very) wrong. An ADS search on title words
"accretion disk" and "nucleosynthesis" turns up plenty of references,
generally in the context of gamma-ray bursts.
-- jt]]


It occurs to me that a quasar is the equivalent of a continuous
supernova or rather N continuous supernova.

I suspect that a very large percent of the metals (dust) in galaxies is
created by the quasar. We are not so much made of star dust but quasar
dust.

[[Mod. note --
1. The details of progenitor composition and nuclear-reaction pressure,
temperature, entropy, and time scales will probably be quite different
between a supernova (further divided up at least into core-collapse
vs white-dwarf runaway accretion) and a quasar accretion disk. That's
going to have a major effect on the details of s-process, r-process,
etc nucleosynthesis.
2. Non-supernova stars also make lots of stuff-other-than-hydrogen/helium
which forms part of a galaxy's dust.
-- jt]]

On 12/17/2015 1:17 PM, David Staup wrote:
It occurs to me that a quasar is the equivalent of a continuous
supernova or rather N continuous supernova.

I suspect that a very large percent of the metals (dust) in galaxies is
created by the quasar. We are not so much made of star dust but quasar
dust.

[[Mod. note --
1. The details of progenitor composition and nuclear-reaction pressure,
temperature, entropy, and time scales will probably be quite different
between a supernova (further divided up at least into core-collapse
vs white-dwarf runaway accretion) and a quasar accretion disk. That's
going to have a major effect on the details of s-process, r-process,
etc nucleosynthesis.
2. Non-supernova stars also make lots of stuff-other-than-hydrogen/helium
which forms part of a galaxy's dust.
-- jt]]

The brightest known quasars devour 1000 solar masses of material every
year. The largest known is estimated to consume matter equivalent to 600
Earths per minute.

Estimates of quasar lifespan range from millions to up to 2 billion years.

The dust from red dwarf stars must come from non hydrogen/helium
components of the matter that make up the star, where did that come from?
Non- supernova stars do not create anything in the time frame under
discussion here ~ 600 million years

or am I wrong?

[[Mod. note -- Non-supernova stars would do a lot of nucleosynthesis
in 600 MYr. -- jt]]