91av

IN the last few years, astronomers have found that all the ordinary matter we are familiar with is only a trace component of a cosmos dominated by “dark matter” of an unknown form, and even more abundant and mysterious “dark energy”. Hence, it appears we are not even made of the main ingredients of this cosmic soup.

Nevertheless, ordinary matter is proving surprisingly useful in the study of the stuff that makes up the rest of the universe. In the last year, NASA’s Wilkinson Microwave Anisotropy Probe (WMAP) has provided detailed knowledge of the composition and density of ordinary matter in the universe. The unprecedented accuracy of this information opens up new possibilities, not only for testing our theory of how this type of matter formed, but also to challenge some views of cosmology. It could even provide answers to questions about more mysterious kinds of matter and energy.

In the first seconds and minutes after the big bang, the universe acted as a huge nuclear reactor, producing the lightest elements: hydrogen, helium and lithium. This process, known as “big bang nucleosynthesis” or BBN, (see “The making of matter”) took place 14 billion years ago, 1 second after the big bang, when the universe was at a temperature of about 10 billion degrees and was much denser than it is today. Atomic nuclei simply could not exist in these extreme conditions, so baryons, particles of ordinary matter, existed as free protons and neutrons.

As the universe expanded and cooled, vigorous nuclear reactions combined neutrons and protons into atomic nuclei. The resulting elements were hydrogen atoms with single protons for nuclei, and helium-4 atoms with nuclei of two protons and two neutrons. Tiny traces of lithium were also made, as well as the rare, but stable, isotope deuterium, with a nucleus of one proton plus one neutron, and helium-3, which is made of two protons and one neutron. As the universe continued to expand, temperatures and pressures fell further and the cosmic nuclear reactor eventually shut down. Any heavier elements formed much later, in the nuclear furnaces of stars.

My colleagues and I have worked for years to understand the exact composition of matter in the early universe and work out the details of the nuclear reactions that created it. Our early calculations led to definite predictions for the relative amounts of the different light elements in the early universe. We estimated that hydrogen made up 76 per cent of the baryons, and helium-4 about 24 per cent. In fact, measurements of the proportion of helium-4 in hot gas clouds in nearby galaxies confirm that their make-up is about 24 per cent helium-4; the rest is hydrogen with traces of other elements.

Cosmologists study these galaxies because they contain only tiny amounts of the heavy elements that are produced by stars. In the billions of years since the big bang, nuclear reactions in stars would have altered the proportion of light elements, so we know these galaxies must be relatively uncontaminated samples of primordial material.

The match between the predicted and measured proportions of helium-4 shows that our theory of big bang nucleosynthesis is on the right track. But to test the theory more precisely, we have also been looking at the proportions of the other light elements – deuterium, helium-3 and lithium – which had not burnt all of the way to helium-4 by the time cosmic fusion ceased. The density of baryons in the early universe dictated how much of each of these elements were created.

For a quarter of a century astronomers estimated the baryon density by running BBN backwards from the proportions of deuterium, helium-3 or lithium in the universe today. In fact, plug any of these elements into the equations and they point to the same baryon density, at least within the margin of error for the measurements. This concordance marked a major success of BBN and cosmology in general.

However, a totally new and more precise way of measuring the baryon density has now arisen, at last providing a way to test the BBN result and to take things forward. The idea is to look at the cosmic microwave background radiation, or CMB. Just as BBN marked the “nuclear age” of the cosmos, when nuclear processes condensed protons and neutrons into nuclei, the CMB is a probe for the “atomic age”, when electrons and nuclei condensed into atoms.

The CMB gives us a snapshot of the universe 400,000 years after the big bang, when neutral atoms first formed. The spectrum of the microwaves at different positions in the sky measures the cosmic temperatures, and this is a probe of the densities of matter.

Small temperature fluctuations in the CMB were first detected a decade ago by NASA’s COBE satellite, and in February 2003 were measured in spectacular detail by WMAP (91av, 15 February 2003, p 21). We can extrapolate backwards from this information to the minutes when BBN reactions were taking place. In the last year, colleagues of mine have done this calculation and found that the WMAP measurements show that the baryon density is 4.7 ± 0.6 per cent of the critical density, that is, around 4.7 per cent of that needed to prevent the universe collapsing back on itself under its own gravity. This result is notable not only for its high precision, but also because it is a completely independent measurement of the cosmic baryon density, and yet matches the observations of light elements. This was by no means inevitable, and marks a great success of big bang cosmology.

Rather than relying on observations to give us the cosmic baryon density by running BBN backwards, we can pluck the baryon density from the CMB and put it straight into BBN calculations. The result is to make BBN a much sharper probe of the early universe.

For example, some theorists have suggested that exotic particles may make up the mysterious dark matter of our universe. But as David Schramm of the University of Chicago in Illinois and colleagues pointed out in the late 1970s, if these particles do exist they would inevitably alter the expansion and cooling of the universe. This in turn would have affected how much of each light element was made before nuclear reactions ceased. By combining the proportions of light elements with the CMB results, many of these theories on the nature of dark matter have been ruled out.

BBN is not only useful for understanding dark matter, but also for investigating the nature of dark energy. In 1998, observations of distant supernovae showed that the universe’s expansion was actually accelerating. Cosmologists reasoned that the culprit could be a new fifth force, to add to the four fundamental forces: strong, weak, electromagnetic and gravitational. This “fifth essence” or quintessence would explain the accelerating universe today in terms of the action of a new cosmic energy field. But while quintessence models are designed to explain the acceleration of the universe today, they also have consequences for the early universe in general and BBN in particular.

By computing the effect of quintessence models on the abundance of light elements, my colleagues and I have now found that the CMB places strict constraints on quintessence theories. Dominant as it is today, we argue, in a paper submitted to the journal Physical Review D, that dark energy could only have formed a tiny fraction of the total cosmic energy during the first seconds and minutes after the big bang. Any more, and nuclear reactions would have been so inhibited that the universe would be seeded with many more light elements. Crucially, this rules out the vast majority of quintessence theories, unless they can find a way to “shut off” dark energy during BBN. Theorists may choose to look to ideas other than quintessence to explain the nature of dark energy.

Another way to use the WMAP results is to compare the proportions of elements created in BBN to those we can actually measure, to tell us about later nuclear processing of light elements by stars in galaxies. For example, astronomers using NASA’s FUSE satellite recently found that deuterium within a few hundred parsecs of the sun is at about 55 per cent of its primordial abundance. Since deuterium is completely destroyed in stars, this shows that 55 per cent of matter near our galaxy is in the form of “survivor” atoms that have never been processed in a star, and date all the way back to the beginning of the universe. This kind of information is useful to astrophysicists who model galaxy formation and evolution, and study the rates of star birth and death.

We can also compare the amount of primordial lithium to that found in stars today, as a way of finding out more about star activity. This has thrown up a serious puzzle. The oldest and most primitive stars in the Milky Way apparently contain only ½ to ⅓ of the amount of lithium created soon after the big bang. Most of our lithium is missing.

There are several possible reasons for this discrepancy. It could point back to difficulties in measuring the properties of the stellar environments in which lithium is measured. For example, inferring the amount of lithium requires a very accurate measurement of the star’s temperature, and astronomers may be getting this stellar “thermometry” wrong. Or the discrepancy might suggest that these stars have destroyed some of their lithium, even though these stars were selected precisely because they should only be burning hydrogen and helium. Or perhaps our picture of BBN is incomplete. Could the action of mysterious exotic particles have altered the expansion of the universe and damped down the production of lithium? But then we are left to wonder how this did not affect the deuterium.

Each of these alternatives is now the subject of intense research. In my view, the first two alternatives are likely to play at least some part in resolving this problem, but the third possibility is the most intriguing since it might uncover new fundamental physics.

Combining our theory of BBN with the microwave background has given cosmologists a detailed and accurate understanding of ordinary matter. Getting to grips with the full implications of this will require unprecedented quantitative precision in BBN theory as well as more accurate observations, and, if the lithium problem persists, new ideas as well. But we can be confident that our efforts will eventually be rewarded.