Stuff About Stars

John Baez

Populations I and II

April 26, 1993

I'm learning a bit about astronomy so I thought I would post a few articles describing different aspects of the classification of stars. I'm mainly interested in learning about stellar evolution for its gut-level appeal. Those who know nothing about astronomy may think of them all as a bunch of specks or maybe suns like ours. Those who know a little astronomy (like me) have heard of the Hertzprung-Russell diagram and main sequence and the basic story about how stars of different masses have different life stories, with the small ones living long lives of quiet desperation, while some of the biggest ones go out in a bang - a supernova - and then leave a neutron star or black hole behind. But what's fascinating me about stellar evolution now is that the story is much more complex than this. It really reminds me of biology, actually. The earliest stars were made almost solely out of hydrogen and helium, since that's what was around, but when some of those early ones produced heavier elements and spewed them out (either via solar wind, flares, novae or supernovae), later stars formed containing this stuff, which seems to affect their characteristics substantially. Whatever the reasons, there are lots of weird kinds of stars, and I thought I'd talk about them a bit, emphasizing the sensationalistic aspects, since I am just studying this for fun. I like the idea that stars, far from being roughly identical balls of gas, can be highly idiosyncratic beasts, and that the galaxy is like a majestic sort of Petri dish.

I emphasize that I don't know much about astrophysics, so that one of my main reasons for posting is to goad experts into correcting me and telling us all more about what people are finding out these days.

I'd like to find nice friendly books on stellar evolution without many equations in them (I'm a mathematical physicist, and I'm reading this just for fun, so I could do without the equations even though they're clearly essential for anyone really wanting to understand stellar evolution well). But I haven't succeeded in finding books without equations that also describe the really weird stars in much detail. So right now some of my information is coming out of

and some is from other books I will cite later.

For starters let me talk about a basic classification of stars in our galaxy, population I vs population II stars. (This and a few more basic posts will set the stage for my real interest, which is all the pathological special cases.)

Our galaxy is a spiral galaxy and can be thought of roughly as a flat spinning disc - the galactic plane - together with a ball near the middle extending out of the galactic plane - the halo. Stars in the galactic plane tend to be different from those in the halo.

"Population II" stars are older stars that tend to live in the halo - older meaning about 5 to 10 billion years old. They are ``metal-poor,'' that is, only .03 to .3 percent of the mass is anything other than hydrogen or helium, presumably since they were formed before much other stuff had been made. They tend to be about 1000-2000 parsecs away from the galactic plane. They tend to be moving fast relative to the galactic center. There are a lot of globular clusters of population I stars - more about these later.

Notation in astronomy is often a bit quirky since people classify things before they understand their explanations, so keeping this in mind it should be easy to remember that "Population I" stars are younger than Population II stars. Population I stars tend to live near the galactic plane, and hence the spiral arms are mainly these stars. They are metal rich, meaning about 1.5 to 3 percent stuff other than hydrogen or helium, and less than a billion years old (some are being born now). They tend to be moving slowly relative to the galactic center. There are lots of open clusters of population II stars - more about these later too.

Not surprisingly, there turn out to be stars intermediate in character between these two kinds. Indeed, until I get to individual peculiar stars, everything I say will be a generalization about a fuzzy class of stars, to be taken with a grain of salt - in addition to the bigger grain of salt to be taken due to the fact that I don't know astrophysics! :-)

Now for some questions for the experts. I get the impression that as our galaxy formed, it flattened out into its present shape, and thus the early Pop II stars tend to lie outside the galactic plane in the roughly spherical halo, while the Pop I stars were formed after things had flattened out. Is this right? What happens in galaxies that are not spirals - are there these two populations?

Globular Clusters

April 26, 1993

I really want to go on to the quirks of the oddest sorts of stars, but I need to set up some terminology in order to make any sense, so I should mention these important ways stars form groups.

Very old, metal-poor Pop II stars are often found in globular clusters. These are stunningly spherically symmetrical formations - a picture is worth a thousand words, so check out this picture of a globular cluster in a neighboring galaxy. They contain 10 thousand to 10 million stars! They range from 5 to 30 parsecs in diameter. About 125 of them have been seen but people suspect there are about 500 altogether in our galaxy. The globular clusters live in the halo of the galaxy, which (recall) is roughly a sphere about the galactic center. They move in roughly Keplerian orbits around the galactic center and they move fast, up to 300 km/sec. I guess they must hit the galactic plane now and then - has anyone seen this?

If you plot stars in a globular cluster on a Hertzsprung-Russell diagram we get something vaguely like this:

BRIGHT                              .     
                     ..  3 ..... ....
             .    .          ...
                    .  .   ...
                        4 ...

       BLUE     WHITE   YELLOW      RED       

Note some features: 1 shows that some have become red giants - it's the "red giant branch". 2 is the "subgiant branch." 3 is a very distinctive feature of globular clusters, the "horizontal branch" which has a gap in it called the RR Lyrae gap, where the horizontal branch is sliced by a certain "zone of instability. 4 is the main sequence going from dim red stars to bright bluish ones. There are a lot more dim ones than bright ones. They haven't had time to run out of hydrogen and act up yet.

Open Clusters, Associations, Wolf-Rayet Stars

April 27, 1993

Thanks for all the corrections and suggestions for reading material so far!

In addition to the book I cited already, I have been using the book

It's quite elementary but packed with nice pictures and is useful for getting oriented.

Okay - the old Pop II stars are often found in globular clusters, typically outside of the galactic plane. How about the new Pop I stars?

These, recall, lie mostly in the plane of the galaxy. They occaisionally form "open clusters" of 100 to 1000 stars. One nice example are the Pleiades, "up and to the right of Orion," which even I have seen many a time. Others are the Hyades (which I haven't seen) and the Jewel Box down south. There are about 1000 open clusters known and a total of 2000 estimated to exist. They are a completely different sort of beast from the globular clusters - but they differ from each other a fair amount too. Unlike the globular clusters, which are much denser near the middle than the edge, these have no concentration near the center. Also, they're a lot smaller than globular clusters; not only do they have fewer stars, they are about 1-5 parsecs in diameter, as opposed to 5-30 for the globulars.

If you plot stars in an open cluster on a Hertzsprung-Russell diagram we get something vaguely like this (I'm trying to draw the Hyades):

               .             .
                ..         .

       BLUE     WHITE   YELLOW      RED       

Note the big difference from the globular clusters. There are a lot of stars on the main sequence, because open clusters are comparatively young; the Hyades being about a billion years old. Only a few stars have had a chance to become red giants and such (the speckling in the upper right). The Pleiades are even younger, about 60 million years old, and line pretty much right on the main sequence.

To wrap up my rap about open clusters, I should also mention another Population I phenomenon, the OB associations. Note that unlike in the globular clusters there are lots of big hot blue stars above. These are the sort that live fast and die young, the O and B types. O's are about 50 times as heavy as the sun and live only a million years, while B's are about 10 times as heavy and live around 10 to a 100 million years.


O blue
B blue-white
A white
F yellow-white
G yellow
K orange
M orange-red
R orange-red (mild carbon stars)
S orange-red (carbon stars)
N red (mild carbon stars)

It turns out that O and B stars almost always come in gangs about 50 members, the so-called "OB associations". An example is Orion itself! These are not gravitationally bound like clusters and are quite dilute. They are big and sometimes hard to spot, about 100 parsecs in diameter. They tend to live really near the galactic plane, in a disk only 120 parsecs thick. All this points to the fact that they are a pretty new thing - only 30 million years old or so. Question: I guess then there may be lots of OB associations that have come and gone already? The idea seems to be (correct me if I'm wrong!) that O and B stars form in clumps and then drift apart. I'm curious about their relationship to the T Tauri stars. The T Tauri stars are infant stars that have recently been formed in "stellar nurseries" and again come in groups, the "T associations". Every OB association has T Tauri stars in it but not every T association has OB stars in it - yet? I went to a nice talk on protostars yesterday and they seem to have found some really interesting candidates for big masses of gas and dust that haven't quite collapsed to form stars. Unfortunately the dust makes it hard to see what's going on with visible light, but you can use infrared and submillimeter to see some of the stuff that goes on. Unlike the simple spherically symmetrical models astronomers used to like when pondering protostars, some of these candidates seem to emit intense jets of hot gas.

Okay, that's it for setting the stage... this isn't a course, after all. Now, on to some weird kinds of stars...

The Wolf-Rayet stars are rather shocking in that their spectra show NO TRACE OF HYDROGEN, although they do show lots of helium. They come in two very different kinds, the carbon-rich WCs and the nitrogen-rich WNs. The WN stars have a reasonable amount of carbon and oxygen, but the WCs have no nitrogen at all. I have a picture of one here, and it's surrounded by a dense shell of blue and red gas (called NGC 2359 for the experts out there). Wolf-Rayets do that - they spew out huge amounts of stuff and shroud themselves in gaseous nebulae. The predominant spectral lines are of singly ionized helium, triply ionized carbon and triply ionized nitrogen. I believe these are absorption lines due to the nebulae, but the fact that these lines are due to highly ionized matter indicates that the Wolf-Rayet inside is really hot, something like a class O star (a blue-hot star). The lines are broadened by Doppler shifts so we can tell that stuff is shooting out fast, meaning about 4000 km/sec as opposed to 200 km/sec or less for stars more like our sun. This is a serious stellar wind.

Their masses are very high, but it seems that they have lost a lot original matter by spewing stuff out. And this is the proposed explanation of the no-hydrogen puzzle. Apparently they have blown off all their hydrogen envelope!

From "Structure and Evolution of Single and Binary Stars" by C W H de Loore and C Doom I get the following story, which I may be misunderstanding. If we start with a star of 60 times the mass of the sun - a really big one - there is so much stellar wind that during the "core hydrogen burning phase" and early "shell hydrogen burning phase" all the outer layers of the star are blown off.

What are these phases? Well, at first hydrogen is burned to helium in the center of the star, the "core," which is surrounded by a convective envelope that transmits energy to the surface but doesn't actually do fusion itself. In the case at hand this core is big, almost 50 solar masses. After a while helium "ashes" build up in the center so one has an inert helium core, and all the burning goes on in a "shell" - the layer of hydrogen right next to the core. Eventually the envelope is all blown off and we're left with an almost pure helium core of about 40 solar masses and a hydrogen shell. Since the helium's not burning, the core contracts until the temperature hits 40 million degrees and the helium catches fire. Helium burns into carbon-12, but some hydrogen gets into the core and makes carbon-13 and nitrogen-14, and later (when the helium is almost all burnt) oxygen-16. All the while the stellar wind is increasing so that by the time the helium is all burnt our star is down to 15 solar masses. Sometime before then is when we have a classic Wolf-Rayet star: lots of stellar wind, almost no hydrogen, and lots of carbon and nitrogen!


I don't know how common these guys are - probably pretty rare, since smaller stars are more common than big ones, and these are up there near the biggest. But about how many have been seen? Are there enough to really affect the interstellar medium a lot? They certainly pump out gas. Do they pump out much dust? I recall an expert on interstellar dust grains saying something about dust from Wolf-Rayets... but he seemed to try to account for most of the dust in our vicinity (found in meteorites) by "asymptotic giant branch" stars. I wonder if anyone has puzzled much over the future of the galaxy, and how the present stars will affect the stars to come... certainly people ponder how shock waves from supernovae may trigger star formation, but I would like to peer into the future and find out what sort of stars might form in the vicinity of, for example, a Wolf-Rayet star that has put out lots of carbon and nitrogen. One can speculate idly (he idly speculated) that stars in which the CNO cycle was much stronger than typical for Pop I might be as different from Pop I as Pop I is from Pop II. Too bad we can't just simulate the whole galaxy and see what it'll be like in a while....

Tomorrow I'm going to hear Lee Smolin talk at UC Santa Barbara about "The Fate of Black Hole Singularities and the Parameters of the Standard Model" - his evolutionary cosmology scenario.

alpha2 Canum Venaticorum Stars

May 1, 1993

Wolf-Rayet stars, despite being odd, are at least fairly understandable and a rather natural stage in the development of a certain mass of star. Now I will talk about a kind of star that is apparently completely mysterious. (If anyone has figured these ones out, please let me know!) My information is from:

alpha2 Canum Venaticorum stars are a class of variable stars that was distinguished around 1950, named (as often the case) for a particular star in this class. They are type A stars (white stars) that show a completely abnormal (Petit's emphasis!) abundance of certain heavy metals but a deficiency of elements such as oxygen and light metals. There seem to be 3 kinds of a2CVn stars: ones with lots of silicon, ones with lots of manganese, and ones with lots of chromium, strontium and rare earth elements! The star alpha2 Canum Venaticorum itself is the silicon type.

These stars vary with the same period in their light intensity, their spectrum, and their magnetic fields. They have strong magnetic fields - several hundred gauss - that vary strongly. For example, the magnetic field of alpha2 Canum Venaticorum itself varies from +4000 to -4000 gauss. I think the way they detect this sort of thing is through the Zeeman effect - spectral lines of atoms change when they are put in a magnetic field. (Details and corrections are welcome!) It's great how much information one can get out of light. The amplitude of the luminosity variations depends STRONGLY on the wavelength. For one example, it goes from 2% in the "visible" (5500 angstroms) to a factor of two in the "ultraviolet" (3600 angstroms)! (Astronomers have wavelengths they call U, B, V, and R for "ultraviolet," "blue," "violet" and "red", but they are precise wavelengths.) Their periods range quite widely from star to star - most are about 2-3 days, but they go from half a day to several hundred days! Alpha_2 Canum Venaticorum itself varies has a period of about 5 days.

And now for the kicker: the period of these variations is just the observed period of rotation of the star itself!

How to explain all this?

Here's what Petit says: "It is thought that the over-abundant metals in the atmosphere are not uniformly distributed but, trapped by the magnetic field, are localized at certain points. This leads to an irregularity in brightness made perceptible by the rotation. Work carried out in recent years has shown that the localization takes the form of rings or segments and not just of irregular patches."

Wow! I wonder what people think, or know, now (Petit's book appeared in French in 1982, and he says there is much research on these stars, so maybe more is known.) Are we supposed to imagine a star with fragmentary rings of vaporized metal around it, held there by magnetic fields? Pretty neat.

R Coronae Borealis Stars

May 1, 1993

As of 1982 there were only 32 R Coronae Borealis stars listed, and only 20 that were definitely of this type according to Petit. So this is a very small class of stars.

R Coronae Borealis itself was discovered in 1795. It is normally of magnitude 6 with small fluctuations, but it occaisionally drops to magnitude 14. Recall that the bigger the magnitude, the dimmer the star, and that magnitude is a logarithmic scale, with 5 magnitudes corresponding to a factor of 100. So R Coronae is occaisionally getting dimmer by a factor of 1600 or so! I have a plot of the brightness of this star from 1842 to 1951. It looks roughly like this:

... . .   ..... . ... .... . ....  .        .   ..... .................  .
   . . ..      . .   .    .      .. .     .   .      .                 ..
         .                           .   .     .

. . .

In other words, it "flickers" with brief irregular periods of dimming.

R Coronae Borealis stars are type G (yellow) supergiants with absolute magnitude -6 or so. The sun's absolute magnitude is about 5 so these are about 25,000 times as bright as the sun. One of the observed ones is in the Large Magellanic Cloud (outside our galaxy). The specturm show lots of carbon and relatively little hydrogen. When they are dim, there are emission lines of neutral sodim, ionized calcium, titanium, and iron. Also, the radial velocity varies little (as can be checked by Doppler shifts) so it seems that matter is not being ejected.

How to explain these rascals? (I can imagine a nice quiz series - Explain These Stars.)

Petit says that "the interpretation of all these observations runs into difficulties. The stars are unusually rich in carbon and it is thought that this element plays a similar role in the star's atmosphere to that of water vapor in the atmosphere of the Earth, condensing to form clouds that obscure the sky unless they disappear in the form of rain. The carbon circulating in the star's atmosphere condenses in the upper regions in the form of fine grains of graphite. During their fall towards the surface under the influence of gravity, these `clouds of soot' become denser, and end up forming a thick envelope that observures the star and cuts out a large part of its radiation. However, when the clouds get nearer still to the surface, their temperature rises and the carbon again becomes gaseous by sublimation. The atmosphere is clear once more and everything is in order... until the next cycle starts."

This is a very charming image but I am puzzled about why it is supposed to explain the observed data. Why the metallic spectral lines? Why doesn't the soot build up slowly and cause a slow dimming, rather than an occaisional drastic dimming?

© 1993 John Baez