The Elephant in the Universe, page 9
In 1946, two years before Alpher and Gamow’s αβγ paper, Hoyle had already penned forty pages about stellar nucleosynthesis in Monthly Notices of the Royal Astronomical Society. Eight years later, he provided more detail in a twenty-five-page article in Astrophysical Journal Supplement. Eventually he teamed up with American physicist Willy Fowler and British-American astrophysicists Margaret and Geoffrey Burbidge—a collaboration that resulted in a monumental, landmark paper in Reviews of Modern Physics.4
Published in October 1957, “Synthesis of the Elements in Stars” became generally known as the B2FH paper, after the initials of the four authors. Another milestone in the history of astrophysics, and another quirky acronym-like nickname. But while the αβγ paper was just six paragraphs long, B2FH went on for a whopping 108 pages, its thirteen chapters dense with graphs, equations, tables, and diagrams of nuclear reactions.
We now know that Alpher and Gamow were right in claiming that helium was produced during the big bang; subsequent stellar nucleosynthesis added only small amounts of helium. But the Burbidges, Fowler, and Hoyle hit the nail on the head with their assertion that elements like carbon, nitrogen, oxygen, sodium, aluminum, silicon, chlorine, calcium, and even iron are cooked up in the nuclear cauldrons in the interiors of stars—Eddington’s great furnaces.
The B2FH paper opened with an apt quote from Shakespeare’s King Lear: “It is the stars, The stars above us, govern our conditions.” They do indeed, not in the astrological way, but in the most literal sense: our very substance—the carbon atoms in our muscles, the calcium in our bones, and the iron in our blood—has a stellar origin. “We are stardust, billion-year-old carbon,” as folk singer Joni Mitchell wrote in her 1969 ballad “Woodstock.”
OK, so the theory of stellar nucleosynthesis, as described in the comprehensive B2FH paper, tells us about the origin of the known, familiar world around us: the atomic building blocks of mice, motorcycles, and mountains. But in order to learn more about the universe’s mysterious dark matter, we need to focus again on the brief αβγ publication, as Peebles realized right after the discovery of the cosmic background radiation.
This 1964 discovery by Arno Penzias and Robert Wilson (briefly described in chapter 1) was the third and final nail in the coffin of Hoyle and Gold’s steady-state model. The first had been the gradual realization that the universe contains a large amount of helium—some 24 percent of the total atomic mass of the cosmos is in the form of this second-lightest element. (About 75 percent of the atomic mass of the universe is hydrogen; all other elements together comprise less than two percent.) Yes, helium is also produced in the hot interiors of stars, but not in such huge quantities—only big bang nucleosynthesis fits the bill.
The second line of support for the big bang theory came in the early 1960s. Radio astronomers discovered that galaxies in the very distant universe have different properties from galaxies in our cosmic neighborhood. Since a remote galaxy’s light may take billions of years to reach us, this observation means that galaxies looked different in the past than they do now. In other words: the universe is not in a steady state but is indeed evolving, in agreement with the big bang theory.
The discovery of the cosmic background radiation—also known as the cosmic microwave background, because of its peak wavelength of approximately one millimeter—finally closed the case in favor of the big bang: the radiation is rightfully known as the “afterglow of creation.” (I’ll get back to the peculiarities of the cosmic microwave background in chapter 17.) When Bob Dicke asked his Canadian postdoc, “So Jim, why don’t you delve into the theory behind all this?” Peebles knew that sinking his teeth into the early universe would shed light on the chemical makeup of the cosmos. Apart from working on the propagation of over- and underdensities in the hot, viscous soup of particles and radiation, he also studied the way in which the outcome of nuclear reactions in the first few minutes after the birth of the universe, and in particular the amount of deuterium, depends on the ever-decreasing cosmic matter density.
This is not a book about the big bang, so don’t expect me to go into too much detail, but what matters most here is that elementary particles known as quarks combined into the first nucleons (protons and neutrons, the constituents of all atomic nuclei) when the universe was about one second old. Because of their mass, protons and neutrons are also known as baryons, from the Greek word meaning “heavy.”
At first, the number of protons was almost equal to the number of neutrons, but before long, that started to change dramatically. Because of the extreme temperatures of the newborn universe, solitary baryons could not yet combine into nuclei. And while the neutrons in an atomic nucleus are stable, free neutrons slowly but surely decay into protons. Within a time span of just a couple of minutes, the number of neutrons decreased significantly, while protons became ever more abundant.
By the time temperatures had dropped to a billion degrees or so (low enough for atomic nuclei to start forming), only one-eighth of the total baryonic mass in the universe was in the form of neutrons; the rest was protons. Through various nuclear reactions, most of the neutrons eventually ended up in helium nuclei, which consist of two protons and two neutrons. The remaining protons were left behind as nuclei of hydrogen atoms. Almost no nuclei heavier than helium could form, and after a brief but extremely energetic period of nucleosynthesis, temperatures and densities in the early universe dropped even further, and fusion reactions came to a halt.
It’s now quite straightforward to calculate the outcome of this primordial nuclear mayhem. If you do the math properly, you’ll find that about three quarters of the total baryonic mass of the universe is in the form of hydrogen, while something like one quarter of the mass is in the form of helium, in satisfying agreement with measurements of the cosmic abundances of these two elements. In other words: big bang nucleosynthesis provides a neat explanation for the observed chemical composition of the universe—something that the steady-state model (or divine creation, for that matter) fails to accomplish.
So what about deuterium and cosmic density? Where does Peebles’s work come in? Well, helium nuclei don’t form in one fell swoop. It’s not like two solitary protons and two solitary neutrons happen to bounce into each other at exactly the same time. Instead, there are a number of possible intermediate steps, deuterium being the most important and the one we’re focusing on here.
While a hydrogen nucleus is just one proton, a nucleus of deuterium (sometimes called a deuteron) consists of one proton and one neutron, bound together by the strong nuclear force. It’s still hydrogen—chemical elements are defined by the number of protons in their nuclei—but it’s almost twice as massive, hence the common name “heavy hydrogen.” Pretty soon after their formation, deuterium nuclei take part in yet another range of nuclear reactions, eventually leading to the production of helium. But there’s not much time for these reactions to occur: as a result of the expansion of the universe, temperatures are dropping rapidly, and big bang nucleosynthesis stalls, with the result that some deuterium is left unused.
In a November 1966 paper in The Astrophysical Journal, Peebles showed that the relative amount of deuterium in the universe critically depends on the density of nuclear matter (baryonic matter) during the brief epoch of nucleosynthesis.5 The higher the density, the more efficiently the nuclear reactions proceed, and the less deuterium is left over as a residue. In contrast, a lower density during this critical epoch results in a higher deuterium abundance. Similar arguments hold for some other rare atomic nuclei, including helium-3, which contains two protons but just one neutron, but the dependency is most pronounced for deuterium.
Measuring the current abundance of deuterium in the universe would thus provide information about the cosmic density during big bang nucleosynthesis. And extrapolating from that, it’s not hard to calculate the current density of baryonic matter in the universe, after billions of years of cosmic expansion. In other words: precise deuterium measurements can tell you the average density of “normal” matter, mainly consisting of atomic nuclei (baryons).
Five months after Peebles published his calculations, they were confirmed in yet another (and much more detailed) Astrophysical Journal paper by Robert Wagoner and Willy Fowler of Caltech, joined by Hoyle, who was still critical of the big bang but nevertheless made important theoretical contributions to the idea.6 In January 1973 Wagoner, then at Cornell, published an updated account, “Big-Bang Nucleosynthesis Revisited.”7 It appeared that scientists had really solved the thorny issues of the origin of light elements in the primeval fireball. Now they only needed to gauge the relative deuterium abundance, and they would know the current baryonic mass density of the universe.
Directly measuring the cosmic abundance of deuterium is hard, but space science came to the rescue. On August 21, 1972, just a few months after I had my first telescopic look at Saturn, NASA launched its third Orbiting Astronomical Observatory (OAO-3), nicknamed Copernicus after the Polish astronomer whose 500th birthday was approaching in early 1973. One of the Copernicus observatory’s instruments was an 80-centimeter ultraviolet telescope-plus-spectrometer, developed at Princeton. Detailed ultraviolet spectra of bright stars (which cannot be acquired from the ground) would reveal absorption lines of both interstellar hydrogen and deuterium: these filter out certain wavelengths of ultraviolet light. From the relative amounts of absorption, the abundance ratio of deuterium could be calculated.
In December 1973, Peebles’s Princeton colleagues John Rogerson and Donald York published the first results in The Astrophysical Journal, based on Copernicus observations of the bright southern-hemisphere star Agena (Beta Centauri).8 Their conclusion: interstellar space contains just one deuterium nucleus for every 70,000 hydrogen nuclei. A tiny portion left over from the nuclear reactions in the newborn universe. From this value, the amount of baryonic matter in the universe—atomic nuclei, basically—could be derived.
Technicians testing the 81-centimeter mirror of the telescope aboard NASA’s Copernicus satellite. When launched in 1972, it was the largest space telescope ever flown.
Using the latest calculations by Wagoner, Rogerson and York arrived at an average cosmic baryon density of 1.5 × 10–31 grams per cubic centimeter. A 1976 paper based on ultraviolet observations of four more stars (including Spica, the brightest star in the zodiacal constellation Virgo) yielded more or less the same result.9 Finally astronomers had a reliable estimate for the total amount of “normal” matter in the universe. And the result had dire implications for our ideas about dark matter.
Peebles, Ostriker, and Yahil were well aware of the Copernicus results when they wrote their 1974 paper “The Size and Mass of Galaxies, and the Mass of the Universe.” As you may recall from chapter 4, they used a large variety of dynamical observations and arguments to determine the average mass-to-light ratio of galaxies. Then, by estimating the number of visible galaxies in a certain volume of space, they calculated the average mass density of the universe, arriving at a value of 2 × 10–30 grams per cubic centimeter—some thirteen times denser than what Rogerson and York found.
If the current density of the universe was really so much higher than the value inferred by Rogerson and York, then the density also must have been higher during the epoch of big bang nucleosynthesis. That would have resulted in a much smaller cosmic abundance of deuterium than the Copernicus measurements had indicated.
Unless.
There you go again. Unless most of the mass in the universe does not consist of baryons. The deuterium abundance only tells us about the density of atomic nuclei, the building blocks of “normal” matter that take part in nuclear reactions. What if the universe contains lots of abnormal matter? Matter that does not consist of atomic nuclei? Nonbaryonic matter? That could explain the dynamical observations without conflicting with the deuterium measurements.
It seems like a large leap of faith, and Ostriker, Peebles, and Yahil certainly didn’t jump to conclusions. Maybe some other process could have produced additional amounts of deuterium in the billions of years since the big bang, although no one has ever come up with a viable mechanism. Or maybe their estimates for the number of galaxies in the local universe were too high. Indeed, an independent analysis by Richard Gott, James Gunn, David Schramm, and Beatrice Tinsley, also published in 1974, arrived at a somewhat less disturbing result.10 But over the years, it slowly started to dawn upon cosmologists that nature really was telling us something important. If, for whatever reason, the total mass density of the universe is significantly higher than the value derived from big bang nucleosynthesis calculations, then we have to accept that there’s a lot of nonbaryonic matter out there. Dark and weird.
Kapteyn, Oort, and Zwicky; Peebles and Ostriker; and Rubin and Ford—they all found circumstantial evidence for the existence of substantial amounts of dark matter. But their results did not, in and of themselves, imply that “dark” means “mysterious.” Their findings could be explained by large amounts of dim dwarf stars, huge clouds of cold interstellar gas, or even a population of invisible black holes. That changed as the theory of the big bang became more convincing. Across billions of years of cosmic history, the big bang itself brought forward a new courtroom exhibit in the dark matter trial: not some kind of astrophysical object, but a strange, unknown particle. Suddenly, the old ideas—the easy explanations—appeared to be off the table. Nonbaryonic dark matter—we’d better get used to it.
Gott, Gunn, Schramm, and Tinsley started their comprehensive paper on the mass density of the universe with a quote from the Roman poet and philosopher Lucretius: “Desist from thrusting out reasoning from your mind because of its disconcerting novelty. Weigh it, rather, with a discerning judgment. Then, if it seems to you true, give in.”
Right—what else could one do but give in?
They might as well have added an unsettling Star Trek quote: “Resistance is futile.”
8
Radio Recollections
Frustrated?
Albert Bosma only has to think for a split second. Then he decisively replies, “I’m not easy to frustrate. But I keep notes of everything. Who knows, I might write a book someday.”
Most of his radio astronomer colleagues, as well as dark matter historians, would agree that Bosma’s discoveries in the mid-1970s were the first to really clinch the case for the existence of dark matter in galaxies. But beyond professional circles, he is hardly known. Instead, Vera Rubin’s name is all over the place. How could Bosma not be frustrated?
I am meeting the short, long-haired, and bearded astronomer in the support building of the Westerbork radio observatory, in the sparsely populated Dutch province of Drenthe.1 The observatory is built next to a former Nazi transit camp. Between 1942 and 1944, a staggering 97,776 Jews, Roma, and Sinti—including young Anne Frank and her family—were deported by train from here to Auschwitz and Sobibór, where almost all of them were killed. Today the area is used to peacefully study the mysteries of the universe. It’s a cold and drizzly November morning, but every now and then the Sun breaks through the clouds, spotlighting the kilometers-long row of fourteen 25-meter radio dishes outside.
Well past retirement age, Bosma is still an active radio astronomer at the Laboratoire d’Astrophysique in Marseille, France. He just returned from a trip to China and is now visiting his family in the Netherlands. Drenthe is where young Albert grew up, in the small village of Smilde. It’s where his math teacher Dr. Knol sparked his interest in astronomy. And the Westerbork observatory is where his groundbreaking observations were carried out, as part of his PhD research at the University of Groningen. The observations many people have never heard about.
Yes, you could choose to fight the battle, he says, but you would end up doing nothing else.
One thing’s for sure: the 1970 Rubin and Ford paper on the Andromeda galaxy, described in chapter 5, didn’t prove the existence of dark matter. It couldn’t. As Australian astronomer Ken Freeman and science teacher Geoff McNamara wrote in their 2006 book In Search of Dark Matter, “Flat optical rotation curves can rarely provide conclusive evidence for dark matter, because they do not reach out far enough from the center of the galaxy.”2
Neither were Rubin and Ford the first astronomers to note that something peculiar was going on with the rotation of disk galaxies. More than thirty years earlier, Horace Babcock—who would become director of Palomar Observatory in 1964—already found that the edge of the visible disk of Andromeda rotates faster than astronomers expected.3 Walter Baade and Nicholas Mayall obtained similar results in 1951.4
Moreover, in the same year Rubin and Ford published their Andromeda results, Freeman, too, discovered that some galaxies appeared to contain more matter than you would guess on the basis of visual observations. Freeman analyzed the distribution of starlight in the disks of thirty-six galaxies and derived the rotation curves expected of those systems, assuming they contained only stars. Back then, rotational data were available for just a handful of galaxies, and in at least two cases, M33 and NGC 300, the measured rotation curves deviated from the computed ones.5 “If [the data] are correct,” Freeman wrote in his Astrophysical Journal paper, “then there must be in these galaxies additional matter which is undetected.… Its mass must be at least as large as the mass of the detected galaxy, and its distribution must be quite different from the … distribution which holds for the optical galaxy.”
So yes, something surprising was going on, but nothing weird enough to convincingly show that galaxies are embedded in large dark matter halos. After all, optical observations could only reveal the mass distribution in the visible part of a galaxy. As Rubin and Ford had noted in their 1970 paper, “Extrapolation beyond that distance is clearly a matter of taste.”
