1

Inte

nsity

400200 600

Temperature= 2.73 K

THE BIG BANG marks the birth of the universe, when space, time, and matter came into existence. ALL ILLUSTRATIONS BY ASTRONOMY: ROEN KELLY

▲ THE SPECTRUM of the cosmic microwave background (CMB) matches that of a black body with a temperature of 2.73 kelvins. This shows nearly all of the universe’s radiant energy was released within a year of the Big Bang.

▼ THE CMB formed approximately 380,000 years after the Big Bang. The di�erent colors denote tiny density enhancements that later condensed into the �rst structures. WMAP

0 200,000 300,000 380,000Age of universe (years)

Redshift (z)∞ 1,000

Wavelength (centimeters)0.5 0.1 0.05

Frequency (gigahertz)

The first billion years After the release of

the cosmic background radiation, darkness fell over the cosmos. ⁄ ⁄ ⁄ BY AdAm FrAnk

© 2013 Kalmbach Publishing Co. This material may not be reproduced in any form without permission from the publisher. www.Astronomy.com

2

Inte

nsity

400200 600

Temperature= 2.73 K

THE BIG BANG marks the birth of the universe, when space, time, and matter came into existence. ALL ILLUSTRATIONS BY ASTRONOMY: ROEN KELLY

▲ THE SPECTRUM of the cosmic microwave background (CMB) matches that of a black body with a temperature of 2.73 kelvins. This shows nearly all of the universe’s radiant energy was released within a year of the Big Bang.

▼ THE CMB formed approximately 380,000 years after the Big Bang. The di�erent colors denote tiny density enhancements that later condensed into the �rst structures. WMAP

0 200,000 300,000 380,000Age of universe (years)

Redshift (z)∞ 1,000

Wavelength (centimeters)0.5 0.1 0.05

Frequency (gigahertz)

B ig Bang theory tells us the entire universe — all space, time, matter, and dimension — emerged from a single titanic explosion that set the cosmos

in motion. Light brilliant beyond descrip­tion flooded the infant universe.

There’s a second part to the scientific story, however, that many people have not heard: Darkness soon returned with a ven­geance. The cosmic Dark Ages began less than 1 million years after the Big Bang and lasted for a billion or so years.

For astronomers, the story is still com­ing into focus as more powerful telescopes and faster computers illuminate the full scope of cosmic history. Using these tools, astronomers are taking their first steps toward understanding the strange uni­verse of the Dark Ages, a cosmos devoid of galaxies that started forming the ear­liest structures, with starlight shining for the first time.

Beginning at the beginningTo understand the Dark Ages, we must briefly touch on the first epoch of light: the Big Bang. After this violent birth, the universe was a smooth, hot, dense soup of exotic high­energy particles. Matter and radiation were mixed so closely that they shared a common temperature. In the parlance of physics, matter and radi­ation were strongly coupled. A photon could cross only a tiny fraction of the universe before being absorbed and then re­emitted by some particle of matter.

Then, as the universe expanded, it cooled, taking the particle­radiation mix through a series of dramatic transitions. They included the creation of protons and neutrons, building blocks of all atomic nuclei, about 1 second after the Big Bang. A scant 3 minutes later, the universe created the lightest nuclei: helium (2 neutrons, 2 protons) and a little lithium (3 protons, 3 neutrons).

During all these changes, matter and radiation remained strongly linked. After about 380,000 years, the universe had expanded and cooled to the point where

electrons and protons could catch each other and bind into atomic hydrogen. This was the great parting of ways between matter and light. The birth of atomic hydrogen, the most abundant element in the universe, meant the end of one cosmic era and the beginning of the Dark Ages.

Astronomers often call the era when neutral hydrogen formed “recombination” because electrons and protons combined to form atomic hydrogen. It’s a misnomer, however — this is the first time these atoms formed. In most stories of the Big Bang, the emphasis on this era lies with the sudden transparency of the universe to photons and the birth of the cosmic micro­wave background (CMB). When recombi­nation occurred, photons previously linked to naked electrons and protons found themselves with nothing to interact with. They were free to expand unimpeded with the universe, with their wavelength stretching along with cosmic expansion. We see them today as relics of the Big Bang. If the still­smooth universe became transparent to photons and thus light, why does recombination signal the beginning

Adam Frank is an astrophysicist at the University of Rochester in New York and a member of Astron­omy’s Editorial Advisory Board.

THE GOOdS SUrVEY uses advanced tele-scopes, including Hubble, Spitzer, and Chandra, to target a fairly empty region of space . The goal is to get a census of the universe and to probe back toward the beginning of cosmic structure. NASA/ESA

3

▼ THE CMB FORMED when the universe had cooled enough for electrons to join with protons, allowing CMB photons to travel unimpeded through the cosmos.

▲ THE DARK AGES reigned for about a billion years. During this era, neutral hydrogen absorbed most visible light, so the universe appeared dark. However, gravity was starting to pull matter into the structures we see today. NASA/ESA/STSCI

▼ STARS with hundreds of times the Sun’s mass began to form in the later stages of the Dark Ages. However, darkness still reigned because most of the universe still consisted of neutral hydrogen.

▼ GALAXIES also started to make their appearance before the end of the Dark Ages. This one existed just 900 million years after the Big Bang. NASA/ESA/STSCI

▼ THE DARK AGES came to an end a billion years after the Big Bang, once early stars and galaxies like this one reionized most of the universe. NASA/ESA/STSCI

380,000 200 million 400 million 600 million 800 million 1 billionAge of universe (years)

Redshift (z)1,000 5.7

Electron PhotonProton T H E D A R K A G E S

of the Dark Ages? The answer lies with visual light, the kind our eyes respond to.

While the neutral hydrogen gas could not absorb cosmic background photons, it efficiently absorbed visual and ultraviolet (UV) light. As soon as neutral hydrogen dominated the universe, visual and UV photons became trapped close to any source producing them. Most of the hydrogen gas in today’s universe is “ionized,” meaning it consists primarily of bare hydrogen nuclei and free electrons. Stellar UV light main-tains this ionization. These photons pack enough punch to tear electrons off any neutral hydrogen atoms that form by recombination. The current dominance of ionized hydrogen is one reason we can see so far with optical telescopes. The Dark Ages were the epoch of cosmic history between the initial formation of neutral hydrogen and its eventual destruction.

What was it, however, that illuminated the universe with the glow we now recog-nize as starlight? How did the cosmos go from a smooth, dark sea of neutral gas emerging from the recombination era to the multitudes of stars and galaxies that dominate today? The Dark Ages mark an era of transitions, not only from blackness to light, but also from formlessness to form.

The universe at recombination was extraordinarily smooth. From detailed studies of the CMB, astronomers know that any ripples, bumps, and lumps in the den-sity of cosmic plasma were 1⁄10,000 as small as matter’s average density. The universe we live in now bears little resemblance. Today’s lumps and blobs — stars and galaxies — are far denser than an average volume of space, which is pretty close to a vacuum. The

journey through the Dark Ages takes astronomers through the epoch when tiny, initial bumps, or perturbations, in the cosmic stew first began exerting their gravitational influence, feeding on sur-rounding gas in an effort to grow into the structures we see today.

“There is tension in the early universe between expansion and collapse,” explains Greg Bryan of Columbia University. Work-

ing with his former thesis advisor, Mike Norman of the University of California at San Diego, and Tom Abel of Stanford Uni-versity, Bryan has pioneered using advanced computer simulations to study the forma-tion of the universe’s first structures.

In the absence of cosmic expansion, a lump of matter denser than its surroundings will draw in material at an ever-increasing rate. The denser the growing lump gets, the stronger its gravity becomes, and the faster it draws new material inward — a runaway collapse. The expansion of the universe changes this process. Just as gravity tries to draw some over-dense lump together, the universe’s expansion pulls it apart.

“During the earlier years of cosmolog-ical study, this struggle between gravity growing structures and expansion diluting them caused astronomers a lot of conster-nation,” says Bryan. Originally, astronomers had hoped that any perturbation, even one as small as an atom, could grow into a clus-ter of galaxies. The push-pull of expansion and gravity dashed that hope.

“There needs to be a range of perturba-tions that already exists at recombination,” explains Bryan, “and they need to be large enough to allow gravity to do its work so the structures we see now can emerge.”

THE mILLEnnIUm rUn simulation tracked 10 billion dark-matter particles to see how cosmic structure formed. This sequence shows the universe at an age of 210 million years (z = 18.3), well within the Dark Ages. Each image left to right zooms in by a factor of 4.

Volk

Er S

priN

gEl

(MA

x pl

AN

ck iN

Stit

utE

fo

r A

Stro

phyS

icS)

, Et

Al.

DiD StructurE forMfrom the top down, or from the bottom up?

210 million years

4

▼ THE CMB FORMED when the universe had cooled enough for electrons to join with protons, allowing CMB photons to travel unimpeded through the cosmos.

▲ THE DARK AGES reigned for about a billion years. During this era, neutral hydrogen absorbed most visible light, so the universe appeared dark. However, gravity was starting to pull matter into the structures we see today. NASA/ESA/STSCI

▼ STARS with hundreds of times the Sun’s mass began to form in the later stages of the Dark Ages. However, darkness still reigned because most of the universe still consisted of neutral hydrogen.

▼ GALAXIES also started to make their appearance before the end of the Dark Ages. This one existed just 900 million years after the Big Bang. NASA/ESA/STSCI

▼ THE DARK AGES came to an end a billion years after the Big Bang, once early stars and galaxies like this one reionized most of the universe. NASA/ESA/STSCI

380,000 200 million 400 million 600 million 800 million 1 billionAge of universe (years)

Redshift (z)1,000 5.7

Electron PhotonProton T H E D A R K A G E S

Astronomers now understand that while the perturbations seen in the microwave background may be tiny, those bumps — imprinted at the Big Bang’s earliest stages — are large enough to give gravity the head start it needed once the Dark Ages began.

Perturbations grew slowly as the Dark Ages progressed. Eventually, the density contrast between a growing lump and its surroundings became obvious, and gravity halted expansion in that region.

Winners and losersWhen the cosmic plasma entered the Dark Ages, perturbations imprinted on it from the beginning ranged in size. Some were planet-sized wiggles, and some were vast undulations that stretched across galactic distances. For many years, astronomers did not know which of these wiggles grew fastest. Did structure form from the top down, or from the bottom up?

“You must remember that most of the matter in the universe is dark matter,” explains Bryan. Given dark matter’s pre-ponderance — it makes up more than 80 percent of the universe’s mass — its proper-ties determined how structure formed.

“For many years, people fought between hot-dark-matter and cold-dark-matter

models,” says Bryan. “Hot” and “cold” in this context mean fast and slow moving. The difference is important because fast-moving dark matter tends to stream out of any small clumps trying to grow. It’s a bit like flicking marbles across a flat surface and trying to catch them in shallow depres-sions. If the marbles move quickly, they barely know the depressions exist. If they roll slowly, they have a better chance of getting trapped.

Hot-dark-matter models need big col-lections of stuff to form before gravity can make the structure collapse. With hot dark matter, the largest wiggles grew fastest. Smaller clumps formed once the big ones were done — top-down structure formation.

Cold dark matter does the opposite. Lit-tle wiggles grew fastest, so structure formed from the bottom up. Once the clumps collapsed, they merged with other small clumps to form ever-larger structures. By the early 1990s, astronomers had ample evi-dence that only cold-dark-matter models could produce the kinds of structures we see in the current universe. “The cold-dark-matter models won because they compared better with observations,” says Bryan.

According to Bryan, the first objects to become gravitationally bound may have

been Earth-sized clumps of dark matter. But while the clump might have been about Earth’s size, it would have been tenuous — at best a dark-matter ghost of a planet.

Dark clump to bright starIn today’s universe, a giant halo of dark matter surrounds the visible part of every galaxy. Astronomers also use the term halos for the first dark-matter structures to form. At a redshift around z = 60, the dark matter halos had grown to contain almost 1,000 times the Sun’s mass. Ordinary matter had not yet joined the party. The 1,000 solar masses were too little to cause ordinary matter, mostly hydrogen gas, to clump. While dark matter began to form signifi-cant structures, ordinary gas was still too hot and moving too fast to be contained in the shallow gravity wells of nascent halos.

Only when the dark matter halos grew to 10,000 times the Sun’s mass could the hydrogen clump, too. The first clumping of ordinary matter — our kind of stuff — marked a critical moment in cosmic history. Unlike dark matter, normal (or “baryonic”) matter can dissipate its energy. In other words, it can cool and slow down, which allowed hydrogen and helium to condense at the center of a growing dark matter halo.

THE dArk AGES COnCLUdEd about 1 billion years after the Big Bang (z = 5.7). At this time, the first stars — massive beasts weighing more than 100 solar masses drawn together by cold dark matter — already had formed. Galaxies were just starting to form.

Volk

Er S

priN

gEl

(MA

x pl

AN

ck iN

Stit

utE

fo

r A

Stro

phyS

icS)

, Et

Al.

1 billion years

5

▲ BEAUTIFUL SPIRAL STRUCTURE de�nes UDF 423, a galaxy that existed 6 billion years after the Big Bang (43 percent of the current cosmic age), at a redshift of 1. NASA/ESA/STSCI

▼ INTERACTING GALAXIES NGC 6872 and IC 4970 lie “just” 300 million light-years distant (z = 0.02), close enough that Earth-based telescopes can show great detail. ESO

▼ ELLIPTICAL and spiral galaxies intermingle in the cluster CL 0053–37, located 2 billion light-years from Earth, so we see it as it was 11.7 billion years after the Big Bang. ESO

▼ LIGHT still dominates the universe today (epitomized by our neighboring galaxy, the Large Magellanic Cloud) because hot stars ionize most of the cosmos. NOAO/AURA/NSF

▲ THE GALAXY UDF 5225 existed about 1.2 billion years after the Big Bang. Its reddish color is typical of such distant galaxies, whose light has been shifted far to the red. NASA/ESA/STSCI

▲ GALAXY UDF 2881 lies at a redshift of 4.6, so we see it about 1.4 billion years after the Big Bang. Most such distant galaxies look ragged because they’re still forming. NASA/ESA/STSCI

1 billion 3 billion 5 billion 7 billion 9 billion 11 billion 13 billion 13.7 billion

Redshift (z)5.7 5.0 02.5 1.0 0.3 0.02

Age of universe (years)

This was the moment when the universe’s first star was poised to form.

To make a star, matter must collapse into the bottom of a gravitational well. The more gas that falls into the well, the higher the density. Eventually, nuclear fusion occurs, and the star turns on.

Astronomers understand how stars form in the current cosmos and know enough to paint the outline for forming the first gener-ation of stars. Life in the Dark Ages was not that simple, however. In astronomers’ quest to comprehend the births of the first stars, they have been thrown many curveballs.

“The absence of heavy elements is the most important difference between modern star formation and the creation of the first stars,” explains Bryan. Astronomers call every element more massive than helium a metal. And metals are, for the most part, created inside stars. Metals in an astrophys-ical gas can shed heat far more efficiently than hydrogen or helium atoms. The pho-tons emitted by metals stream out of the gas, taking energy with them. In this way, even traces of metals can act as highly effective refrigerants.

The first stars, however, formed from gas lacking metals. This primordial gas had almost no way to cool. “The cooling is very

slow until hydrogen atoms combine to form hydrogen molecules,” says Bryan. He also points out that molecular hydrogen in the modern universe forms on dust grains, which are made of metals. Without dust, molecular hydrogen formed so slowly that dark matter halos grew to a million solar masses before ordinary matter could really begin to cool. “These dark matter halos are literally microgalaxies, one-millionth the

mass of a present-day galaxy, by the time the first star can begin to form,” says Bryan.

The wait leads to disappointment on a cosmic scale. According to Bryan’s calcula-tions, the type of star that finally contracts from the primordial gas is a kind of mon-ster rarely seen in the current universe.

In today’s cosmos, any star larger than 8 solar masses is considered massive and has a harder time forming than its smaller

cousins. Stars more than 100 times the Sun’s mass are so rare that we know of only a handful in the entire Milky Way.

A giant’s birth and deathAccording to Bryan, the universe’s first stars may all have been this big. Over the past several years, Bryan, Norman, and their collaborators have used advanced computer simulations to track the evolu-tion of cosmic structure from the era of recombination to the formation of the first star. Unlike traditional simulation methods, Bryan’s code simultaneously tracks a huge range of scales, everything from vast arcs of still-forming dark matter halos millions of light-years across down to the outlines of the first star with a radius just a few times that of Earth’s orbit.

After running the codes continuously for years, the researchers found that each microgalaxy halo produces exactly one massive star. Each star weighs a few hun-dred solar masses. But the story doesn’t end there. Massive stars burn hot and die fast. In their brief lives, they dramatically affect their environment.

The massive stars in Bryan’s calculations produce a torrent of UV photons that ion-ize every hydrogen atom out to well beyond

GALAXIES And CLUSTErS appear prominent by the time the cosmos is 4.7 billion years old (z = 1.4). On the biggest scales (left), the universe still looks fairly smooth, but closer up, individual galaxies and clusters start to dominate.

Volk

Er S

priN

gEl

(MA

x pl

AN

ck iN

Stit

utE

fo

r A

Stro

phyS

icS)

, Et

Al.

thE uNiVErSE’S firStstars may all have been

hundreds of solar masses.

4.7 billion years

6

▲ BEAUTIFUL SPIRAL STRUCTURE de�nes UDF 423, a galaxy that existed 6 billion years after the Big Bang (43 percent of the current cosmic age), at a redshift of 1. NASA/ESA/STSCI

▼ INTERACTING GALAXIES NGC 6872 and IC 4970 lie “just” 300 million light-years distant (z = 0.02), close enough that Earth-based telescopes can show great detail. ESO

▼ ELLIPTICAL and spiral galaxies intermingle in the cluster CL 0053–37, located 2 billion light-years from Earth, so we see it as it was 11.7 billion years after the Big Bang. ESO

▼ LIGHT still dominates the universe today (epitomized by our neighboring galaxy, the Large Magellanic Cloud) because hot stars ionize most of the cosmos. NOAO/AURA/NSF

▲ THE GALAXY UDF 5225 existed about 1.2 billion years after the Big Bang. Its reddish color is typical of such distant galaxies, whose light has been shifted far to the red. NASA/ESA/STSCI

▲ GALAXY UDF 2881 lies at a redshift of 4.6, so we see it about 1.4 billion years after the Big Bang. Most such distant galaxies look ragged because they’re still forming. NASA/ESA/STSCI

1 billion 3 billion 5 billion 7 billion 9 billion 11 billion 13 billion 13.7 billion

Redshift (z)5.7 5.0 02.5 1.0 0.3 0.02

Age of universe (years)

1,000 light-years. “The ionization produced by the big stars heats the surrounding gas and, in the process, can completely unbind all the baryons from the still-forming dark matter halo,” explains Bryan. Even worse, when a star this size runs out of nuclear fuel — after only a million years — it ends its life in a titanic explosion that blows away the halo gas.

Only a star bigger than 300 times the Sun’s mass can escape this fate. Such hyper-giant stars may implode as a black hole without ever exploding.

from darkness to darknessAlthough the massive stars that form in Bryan’s simulations ionize lots of gas, they don’t pack enough punch to alter the bal-ance of neutral-to-ionized hydrogen atoms completely. The universe remains dark. “We really don’t know what happens next,” declares Bryan. “Does the next generation of stars form in blast waves from first- generation supernovae?” Bryan and other astronomers now are focusing their efforts on the era when the Dark Ages ended —sometime between redshift 17 and approx-imately redshift 6.

“We know that the growing dark matter halos create objects that probably look like

today’s dwarf galaxies,” says Bryan. On the baryon side of the equation, successive stel-lar generations eventually created enough metals to bring the star-formation process in line with what we see today. At that point, stars of all masses can form.

Still, astronomers have yet to fill in the details, and many of our current ideas may be wrong. For example, no one knows what role black holes played in re-ionizing the universe. Any black hole that formed dur-ing this epoch would have drawn in gas at a prodigious rate. As the gas fell toward the black hole, it emitted ionizing UV photons.

“People used to think that re-ionization and ‘first light’ occurred quickly, at a red-shift of around 6 or 7,” says Bryan. “These days, that is not so clear.” Did black hole accretion effectively contribute to filling the universe with visible light? How did the galaxy-formation process continue to the point of building the great spirals and ellipticals we see today? Answers to these questions remain unknown.

Seeing first lightWhile questions concerning the first billion years abound, hope for answers does as well. An array of new instruments, sched-uled to come online during the next decade

or two, holds the promise of directly prob-ing the Dark Ages and its end in the era of re-ionization.

The highest profile instrument will be Hubble’s successor: the James Webb Space Telescope (JWST). Ultraviolet and visible light emitted during the era of re-ionization now has been redshifted to the infrared. With a 6.5-meter mirror and high sensitiv-ity in the infrared, JWST will explore the nature of the objects that lit up the universe at the end of the first billion years.

In addition, astronomers hope the Square Kilometer Array will directly probe radio waves emitted by clumps of atomic hydrogen gas during the Dark Ages. The ability to map ripples in gas density at such early times and far-flung distances could prove to be a watershed. Astronomers would get their first clear view of the early universe’s initial structures.

From light to darkness to light again — this describes the unveiling of our uni-verse’s early years. By illuminating the Dark Ages, the age of first structures and massive primeval stars, astronomers are exploring a critical gap in cosmic history. With only the outlines filled and much still to be written, there can be no doubt we live in our own privileged epoch.

In TOdAY’S UnIVErSE, rich galaxy clusters follow filaments of dark matter, with large voids separating the denser regions. Ordinary matter now largely follows the lead established by dark matter, and dark matter halos surround individual galaxies.

Volk

Er S

priN

gEl

(MA

x pl

AN

ck iN

Stit

utE

fo

r A

Stro

phyS

icS)

, Et

Al.

13.7 billion years

7

The discovery of the cosmic microwave background (CMB) confirmed — in the eyes of science — the Big Bang theory. The CMB’s clumpiness gives astronomers evidence for theories ranging from what our universe’s contents are to how mod-ern structure formed. Astronomy: Roen Kelly

© 2013 Kalmbach Publishing Co. This material may not be reproduced in any form without permission from the publisher. www.Astronomy.com

8

While adjusting an antenna, two astronomers uncovered cosmic static, leading to one of the greatest

discoveries of all time. ⁄⁄⁄ By JAMeS TReFIl

The accident that saved theBig BangS ometimes the most profound insights come from the most mundane experiences.

So it is with the discovery of the cosmic microwave background (CMB), which manifested itself as an engineering nuisance.

You can learn a lot about the history of the universe and the CMB sitting around a campfire. Early in the evening, when the fire is hot, the coals are bright yellow, maybe even white. Later in the evening, before you roll out your sleeping bag, they glow a dull red. The next morning, they don’t glow at all, yet they feel warm when you hold your hand over them. The CMB has undergone a similar process, although spread over bil-lions of years instead of hours.

This illustrates one of the basic laws of physics: Every object at a temperature above absolute zero radiates into its environment, with the type of radiation depending on the object’s temperature. Your roaring campfire, for example, gives off visible light, with a lot of yellow — and a small amount of green and blue — wavelengths.

As the coals cool, the radiation slides to lower-energy (and longer-wavelength) red light. By morning, the cooling takes the radiation down below the visible range and into the infrared. This is the heat you still feel.

During the past 14 billion years or so, the universe itself has gone through a pro-cess similar to those coals in your fire. It started out bathed in high-energy gamma rays. As it expanded and cooled, the type of radiation slid to longer wavelengths — through visible light, infrared, and finally, down to microwave. The discovery of this

S p e c I A l c o S M o l o G y I S S u e

9

ULTRAVIOLET

X RAYS

GAMMA RAYS10–10 m10–14 m 10–12 m 10–16 m 10– 8 m 10– 6 m 10– 4 m 10–2 m 10–1 m 1 m

INFRARED

MICROWAVES

RADIO WAVES

VISIBLE

so-called cosmic microwave background and a full understanding of its significance is one of the most fascinating tales in the history of cosmology.

cosmic modelsIt may be hard to believe with today’s sci-entific evidence, but not so long ago, sci-entists had serious doubts about the Big Bang theory. Everyone knew the universe is expanding, of course — Edwin Hubble had settled that in the 1930s. But a group of British astrophysicists suggested a pic-ture of the universe quite different from the Big Bang.

Called the steady state model, the theory agreed with the observation of galaxies moving away from each other, but it said

new matter (and eventually new galaxies) is created in the spaces left behind.

In the 1960s, this model was proposed as a serious alternative to the Big Bang, although few believed in it. A problem in deciding which theory was correct was that both the Big Bang and steady state models explained the one piece of observational evidence we had about the universe — gal-axies are moving away from each other. Astronomers needed a new piece of data — a new observation — that would establish which theory was correct.

Signal in staticThe missing puzzle piece came from an unexpected source. The early 1960s was when trans-Atlantic television transmis-sions started. The system — unbelievably primitive by modern standards — bounced TV signals off an orbiting mylar balloon, about 100 feet (30 meters) across.

Getting these signals through was such a technological feat at the time that a leg-end “Live from Europe” would appear at

the bottom of the TV screen when they were shown.

Because scientists were pointing receivers at the sky to get a signal, they needed to know exactly what else was coming in from other sources — sources that interfered with the weak beam from the satellite. Two scientists at Bell Labs in New Jersey — Arno Penzias and Robert Wilson — worked to answer this question. Because TV signals are in the microwave part of the spectrum, the two took an old microwave antenna, pointed it at the sky, and began to survey what was out there.

They quickly ran into a problem. No matter which way they pointed their antenna, they found a faint whisper of microwaves raining down.

In the field of electronics, detecting a signal coming from every direction usu-ally means there is a problem with the equipment circuitry. To Penzias and Wil-son, the faint hiss was a red flag, an indi-cation that they needed to check their apparatus for flaws.

They evaluated one circuit after another. They even noted that pigeons had roosted in their antenna, coating the interior with a “white dielectric substance,” which they removed. But the hiss persisted.

Then, a colleague mentioned a group of astrophysicists at Princeton University who were examining the consequences of the Big Bang. In fact, those astrophysicists had determined the cooling universe should be bathed in microwave radiation, which could be thought of as an echo of the Big Bang.

observation becomes proofThe best way to understand this prediction is to refer back to the campfire example. This time, imagine you are watching it cool

The elecTromagneTic specTrum shows the difference between high-energy gamma rays and low-energy microwaves. The universe’s expansion stretched gamma-ray wavelengths from 10–12 meters to long, 0.2-centimeter micro-waves in the CMB. Astronomy: Roen Kelly

While surFing The radio, Robert Wilson (left) and Arno Penzias unexpectedly discov-ered the CMB with this horn-type antenna. This picture was taken in 1978, after they received their Nobel Prize. ASTRonoMIcAl SocIeTy oF The pAcIFIc

10

Inte

nsit

y (m

egaj

ansk

y pe

r ste

radi

an)

0.50

0.2 0.1 0.07Wavelength (centimeters)

Frequency (gigahertz)

0.05

60 300 400 500100

400

300

200 600

Temperature = 2.725 K

200

100

ULTRAVIOLET

X RAYS

GAMMA RAYS10–10 m10–14 m 10–12 m 10–16 m 10– 8 m 10– 6 m 10– 4 m 10–2 m 10–1 m 1 m

INFRARED

MICROWAVES

RADIO WAVES

VISIBLE

from inside the coals. At the beginning, you would see visible light coming at you from all directions. As the campfire cools, the radiation slides to red, then to infrared, but it is always coming from every direction.

In the same way, the theorists argued, observers on Earth should see radiation raining down from all directions. As the universe cooled and expanded, the radiation dropped to long-wavelength microwaves. The faint hiss the observers couldn’t get rid of was precisely the signal that should have been there. What had begun as a routine survey became a dazzling revelation about the cosmos. Penzias and Wilson shared the 1978 Nobel Prize in physics for their work.

The CMB discovery removed all doubts about the Big Bang and sent the steady state model to the dusty back room. During the next couple of decades, whatever contro-versy existed about the CMB centered on whether it is the kind of radiation we should expect from an object that has been cooling for 14 billion years.

A perfect fitThe laws of physics require that every object above absolute zero gives off radia-tion, and the lower the temperature, the longer the radiation wavelength. The laws say precisely how much of each type of radiation a body at a given temperature should emit. They also say an object radi-ates a range of wavelengths: a small amount of long- and short-wavelength radiation, with most of its energy radiated at a par-ticular wavelength that depends on the object’s temperature.

The Sun, for example, has an outer surface at about 5,800 kelvins, so it emits most of its radiation in the form of visible light. It also gives off small amounts of both radio and ultraviolet — the latter is obvious every time you get a sunburn. The curve that shows the amount of radiation at each frequency for a given object is called a blackbody curve (see “The universe’s temperature,” to right).

⁄⁄⁄ B l A c K B o d y R A d I A T I o n

Any object with a temperature above absolute zero radiates a range of energy. The energy radiated follows a spectral distribution curve — known as a blackbody curve. The curve is typically shown as intensity versus wavelength, and it peaks at a characteristic wavelength, which corresponds to a particular temperature.

The spectral distribution curve eventually led to the origins of quantum theory. When deriving the calculation, physicist Max planck determined the energy emitted or absorbed

must be in discrete packets — quanta, or photons. The blackbody curve is also known as the planck radiation curve. — Liz Kruesi

nasa’s coBe mea-sured the CMB’s spec-tral distribution and found it has a perfect blackbody curve — the best example found in nature. Its maximum-intensity peak, at roughly 0.2 centimeter, corresponds to a tem-perature of 2.725 K. Intensity is given as energy flux, jansky, per solid angle, steradian.

a roaring campFire gives off visible white light. As it cools, the emitted radiation slides to red, and even-tually infrared, which is the heat you feel the fol-lowing day. This is an example of a blackbody: an object that radiates energy, with the most promi-nent wavelength corre-sponding to a certain temperature.

The universe’s TemperaTure

Astr

on

om

y: R

oen

Kel

ly

Astr

on

om

y: R

oen

Kel

ly, A

FTeR

nA

SA/c

oBe

11

Earth’s atmosphere absorbs some micro-wave wavelengths, so parts of the expected spectral curve were not visible to ground-based antennae. The search to fill in the gaps of the blackbody curve was postponed until the technology was ready, in 1989, when NASA’s Cosmic Background Explorer (COBE) satellite was launched.

Sitting above Earth’s obscuring atmo-sphere, COBE measured the radiation with unparalleled accuracy. When University of California at Berkeley astronomer George Smoot first presented the COBE data at a scientific meeting in 1992, the audience of sober, tweed-clad scientists burst into wild

applause when the perfect blackbody curve was flashed on the screen.

Answers raise more questionsThe microwave radiation was exactly what it should have been for a universe that had cooled to an average temperature of 2.725 K above absolute zero. But there was another problem.

Although the microwaves fall on Earth from all directions, as expected, the radia-tion is the same no matter which direction we look. This is true to an accuracy of about one part in 100,000, which means that all parts of the universe, whether we look up, down, or sideways, are at the same temperature.

To understand why this poses a problem, think about another everyday experience. If

you’re in your bathtub and you feel the water cool off, you turn on the hot water. At first, the water is warmer near the tap than at the far end of the tub, but after a while, all the water comes to the same temperature. In the language of physics, the water in the tub eventually reaches thermal equilibrium.

The CMB tells us that the universe, like the water in the bathtub, is in thermal equilibrium. It has the same average tem-perature everywhere.

The problem is introduced when you trace the universe’s expansion backward. There wasn’t enough time for the universe to establish thermal equilibrium before the expansion moved the universe’s contents away from each other. It would be as if you turned on the hot water in your bathtub and the temperature of the water rose instantly, even in the most distant parts.

This was called the horizon problem, because the universe’s expansion would have moved matter from one part of the universe so much that, as seen by other parts, that matter would appear to have moved over the horizon. Different sec-tions could not communicate with each other to establish thermal equilibrium. This problem — along with others —

was resolved in the 1980s with the devel-opment of inflation theory (see “Seeing the dawn of time,” Astronomy, August 2005, for more information).

According to inflation, the universe in its earliest stages was much smaller than you would expect if you do a backward extension of the expansion; the universe was smaller than a sub-atomic particle. In this small state, thermal equilibrium was easy to establish. Then there was a period

The angular resoluTion difference between COBE and WMAP is staggering. COBE was launched in 1989; its data was released in 1992. The scientific community was in awe at the sky maps COBE took, limited as they were by the spacecraft’s low angular resolution (7°). WMAP followed in 2001, with its first data release in 2003. Its sky maps show far greater detail than COBE because the probe has an angular resolution of 0.3°, or 18'. nASA/WMAp ScIence TeAM

James Trefil, a member of Astronomy’s edito-rial board, is a professor of physics at George mason University in Fairfax, Virginia.

The cMB has given us new insights into both the beginning and the

end of the universe.

Wilkinsin Microwave Anisotropy probe (WMAp)

cosmic Background explorer (coBe)

12

of very rapid expansion — called inflation — driven by forces that can be calculated from the theory of elementary particles. In this expansion, the universe grew to the size of a softball, and eventually to the uni-verse we see today.

What those spots tell usSo the regularity of the background taught us one lesson about the early evolution of the universe, and the small differences — the ones that amount to one part in 100,000 — taught us another.

For the first 400,000 years or so, the uni-verse was so hot that no atoms could form. If an electron attached to a nucleus, the next collision would knock it off.

Matter consisted of what physicists call a plasma — gas that is so hot it is ionized and has loose electrons and nuclei knocking around. (The Sun is mostly plasma, and there is a partial plasma in every operating fluorescent lightbulb.) Plasmas absorb radi-ation, which means light and other forms of radiation get trapped in them and are con-stantly absorbed and re-emitted. Astro-physicists say the universe was “opaque” during this time.

Once the universe was about 400,000 years old, it had cooled to the point that neu-tral atoms could form — electrons and nuclei combined. This moment is called “recombination” (even though there wasn’t a previous “combination”). Atoms are largely transparent to radiation — which is why light can travel such long distances through Earth’s atmosphere. After recombination, the radiation that was initially trapped in the universe’s plasma was free to escape.

The microwave background is the high-energy photons that were released when atoms formed, and the photons have spent the last roughly 14 billion minus 400,000 years being stretched out into microwaves as the universe expanded.

This is why the seemingly insignificant differences (scientists call them “anisotro-pies”) in the CMB are so important. Places where matter was starting to clump together would have been at a slightly

higher temperature than the surroundings, and we should be able to see that difference by looking at temperature differences in the CMB’s spots.

The early concentrations of matter served as the points where matter — and eventually galaxies — condensed. Because of this, they often are called “seeds,” although I prefer the more poetic “ripples at the beginning of time.”

Ripples were first seen in data from the COBE satellite. So important is this infor-mation that in 2001 another satellite — the Wilkinson Microwave Anisotropy Probe (WMAP) — was launched with the specific mission of measuring the CMB’s spots to unparalleled accuracy. The comparison between COBE and WMAP is equivalent to putting on glasses to read the fine print in a document.

Forty years laterThe detailed measurements of the CMB provided by many sources, including

WMAP, have given astronomers precise information about our universe. We can now say with 99-percent certainty that the universe is 13.7 billion years old, and that the release of radiation from the plasma — the moment of last scattering — took place 379,000 years after the Big Bang.

WMAP also provides input into our current understanding that the universe is made of about 4-percent ordinary matter, 23-percent dark matter, and 73-percent dark energy. It was, in fact, corroborating data from WMAP that made scientists so willing to accept the notion that almost three-quarters of the universe is in the form of something called dark energy — some-thing whose composition we still have to discover, but which will ultimately govern the fate of the universe.

The CMB has given us new insights into both the beginning and the end of the uni-verse — not bad for an observation that started out as a calibration measurement designed to improve TV reception.

⁄⁄⁄ c o n T I n u I n G T h e Q u e S T

In February 2007, the european Space Agency (eSA) will launch the third-generation cMB probe: planck Surveyor. compared to WMAp, planck has roughly twice the angular resolu-tion — 10' versus WMAp’s 18'.

planck Surveyor involves crucial science because it will observe a wider frequency range than coBe or WMAp. Whereas WMAp covered the frequencies from 23 to 94 Ghz, planck will reach far past that range, to cover 30 to 857 Ghz. This will allow planck to improve on WMAp’s mea-surements by separating foreground sig-nals from cMB signals and by seeing 10 times the frequency range WMAp saw.

With the increased sensitivity, planck will provide information on the universe’s fundamental parameters — including spa-tial curvature and expansion — in addition to measuring structure size within the cMB.

The first planck data release may be in late 2010. — L. K.

The Third-generaTion cmB proBe, Planck Surveyor, will provide astrono-mers with even more detail about the composition and fate of the universe.

nexT up For The challenge

eSA

/Med

IAlA

B

13

Deuteron

Three quarks constitute neutrons and protons

Neutron

Energy release

Proton

Nuclear reactions in the universe’s

first minutes made the lightest

elements. How it happened laid the groundwork

for everything that followed.

⁄ ⁄ ⁄ BY AdAm FrAnk

How theBig Bangforged the first elements

The BirTh oF deuTerium. The most fragile of the light elements, deuterium (H2), formed in the universe’s first minutes when protons and neutrons stuck together. All astrophysical processes destroy this nucleus, so its abundance has been declining since the Big Bang. Deuterium’s absorption feature in the spectra of quasars helps astronomers pin down its original abundance. The fusion reactions illustrated here involve photon emission; other, faster reactions also were present.All illustrAtioNs: Astronomy: roeN Kelly

© 2013 Kalmbach Publishing Co. This material may not be reproduced in any form without permission from the publisher. www.Astronomy.com

14

Deuteron

Three quarks constitute neutrons and protons

Neutron

Energy release

Proton M oments after the Big Bang, as the universe quickly expanded from an unimaginably dense, impossibly hot state, something won-derful happened. Over the course of the first 3 minutes, the first elements were born.

Every instant of every day, evidence that the universe began in a cosmic fireball stares us in the face. Proof that the universe was once hot and dense resides in the very atoms from which the stars, planets, and we ourselves are built.

Big Bang nucleosynthesis — BBN, for short — is the field of astrophysics linking the observed abundances of the chemi-cal elements to theoretical predictions based on the Big Bang. Along with the universal redshift of galaxies and the cosmic microwave background, BBN is one of the great pillars on which modern cosmology stands.

BBN is a remarkable mix of precise astronomical observa-tion and exacting physical theory. Using only the abundances of the lightest elements, hydrogen and helium, BBN spins out a detailed picture of our cosmic beginnings. It is a remarkable tale and a grand triumph of science’s power and precision. Most amazing of all, the events that drive this story, with con-sequences stretching across space and time, unfolded in little more than the span of a typical TV commercial break.

Elemental originsWe recognize more than 116 distinct chemical elements today. Each appears different to us — copper is metallic and shiny,

while sulfur is yellow and powdery — because the atoms making each element differ. It’s hard to believe scientists were still vigorously debating the reality of atoms even 100 years ago. But once researchers confirmed the reality of the atom in the early decades of the previous century, they began probing its internal structure.

Every atom, they found, contains a central nucleus com-posed of one or more protons, which carry a positive electric charge. Hydrogen, the simplest and most abundant element, has a single proton in its nucleus. It’s the number of protons in a nucleus that distinguishes one element from another.

The nucleus also may contain another particle, called a neutron. It’s slightly heavier than the proton and lacks an electrical charge. The number of neutrons in a nucleus is what distinguishes one variation of a single element — called an isotope — from another.

A third kind of particle, the negatively charged electron, orbits each nucleus at a great distance. Compared to protons and neutrons, electrons weigh next to nothing. The discovery of atomic, and then nuclear, structure answered questions about the nature of matter that had haunted philosophers and scientists for 2,000 years.

Until the 1930s, physicists could not explain elemental abundances. Why is it so much easier to find hydrogen atoms than, say, iron atoms? And good luck finding a lutetium atom. Hydrogen is vastly more abundant than iron, which is vastly more abundant than lutetium. Why?

In 1937, German-American physicist Hans Bethe (1906–2005) was returning by train to his Ithaca, New York, home following a nuclear physics conference in Washington. It just

How theBig Bangforged the first elements

Adam Frank is an astrophysicist at the University of rochester in new york and a member of Astronomy’s Editorial Advisory Board.

15

Deuteron

Neutron

Energy release

Energy release

Proton

Triton (H3)

Normal helium (He4)

HiDing in astronomers’ journals

was a paper that could solve the puzzle, but its solution meant opening

a door to the dawn of time.

helium genesis, pArT 1. Primordial fusion also created tritium (H3), an unstable, radioactive form of hydrogen. Some tritium nuclei captured a pro-ton to make normal helium (He4). Stars also make helium, so this element is ever more common in the universe. ionized hydrogen gas clouds in other galaxies clue astronomers into helium’s abun-dance before stars shone.

might have been the most productive train ride in history: By using the time to explore equations for the newly developing science of nuclear physics, Bethe discovered the secrets of stellar fusion. Tak-ing into account the high temperatures and densities inferred by astronomers to exist at the centers of stars, Bethe showed how simple elements can be squeezed together to form more complex ones, a process that releases energy.

In a single stroke, Bethe showed how the fusion of elements fuels the stars, that stellar cores are alchemical furnaces transmuting one kind of matter into another. Bethe’s success con-vinced physicists and astronomers that the handiwork of stars could explain all the elements and their abundances.

They were both right and wrong.In 1957, British astronomers Geoffrey

and Margaret Burbidge, American astronomer Willy Fowler, and British astrophysicist Fred Hoyle published a monumental work that put the theory of stellar nucleosynthesis on firm ground. Often known as B2FH, the paper refined earlier studies into a single coherent picture that accounted for the observed abundances of elements — almost.

While the astronomers could nail down elements like carbon, oxygen, and iron, their model couldn’t get the simplest elements

right. The theory predicted hydrogen and helium

proportions that were completely different

from what astron-omers observe.

Stellar nucleosynthesis predicted a cosmos with too little helium. Observations show that helium makes up about 24 percent of the universe’s normal matter. Everything heavier accounts for less than 2 percent of the total, and all the rest is hydrogen. For years, astronomers were left scratching their heads at this glaring failure in the midst of a spectacular success.

In fact, the answer had already been found and forgot-ten. Hiding in their journals was a paper that could

solve the light-element puzzle. But accepting the solution it offered meant opening a door to

the dawn of time.

Beyond steady stateIn 1948, Ralph Alpher (1921–2007), a wiry, young graduate from George Washington University, wrote a doctoral thesis that began, for the first time, at the beginning. Under the tutelage of George

Gamow (1904–1968), a Russian-refugee physicist known as much for his heavy

drinking as for his genius, Alpher set out to describe nuclear physics in the realm of an

infant expanding universe.It’s difficult to imagine now how bold, how radical

this endeavor was. In 1948, few scientists were thinking about cosmology, and those who were had locked themselves into the so-called steady-state model. Steady-state cosmology held that, even with expansion, the universe never changed its appearance or its condition. The cosmos had always looked — and always would look — just as it does now.

Gamow and Alpher were beyond the leading edge. The origin of the elements had always been a cherished problem to Gamow.

He asked Alpher to imagine what might happen in a universe that started out small, hot, and dense

and expanded to its present enlarged, cold, tenuous state. Specifically, Gamow asked

Alpher to work out the nuclear reac-tions that might occur during the hot, dense period. Within the space of a year, Alpher had worked out many

16

Deuteron

Neutron

Energy release

Energy release

Proton

Triton (H3)

Normal helium (He4)

of the crucial aspects with meticulous attention to mathematical detail. It was a triumphant physics tour de force — but one the scientific establishment promptly forgot. While Alpher’s first cal-culations contained some missteps, they got the fundamentals of Big Bang nucleosynthesis correct.

Alpher, along with collaborator Robert Herman, spent the next few years refining his models and examining the implications of a cooling and expanding universe. The team even predicted the pres-ence and temperature of a cosmic microwave background from the redshifted light released when the universe had cooled enough that electrons could combine with nuclei to form atoms.

Alpher said he and Herman expended “a hell of a lot of energy” giving talks to convince astronomers that the results deserved a serious look. But their work received little attention, and, in a trag-edy of cosmic proportions, the two physicists ultimately gave up in frustration. Alpher left academia to work for General Electric while Herman moved on to General Motors Research Laboratories.

In the mid-1960s, the weight of new data finally forced the acceptance of Big Bang cosmology. But, even then, Alpher’s immense contribution was largely ignored, with credit given almost solely to Gamow.

In the years since, others have refined the picture Alpher and Gamow first glimpsed. Its predictions of simple-element abun-

dances prove that we understand something about cosmic origins. The secret of BBN, the secret Alpher, Gamow, and

Herman knew first, occurs just after the cosmos began.

Fusion and the Big BangAstronomers see galaxies rushing away from one another in today’s expanding universe. But if we could run cosmic evolution backward, everything would draw together. The cosmos would become denser and hotter. As the clock runs backward

toward the Big Bang, structures like galaxies melt into a thickening soup of primordial gas. Run the

clock back further, and the gas also breaks down into a smooth, ultrahot sea of protons, neutrons, and other

subatomic particles. At this point, the universe has a temperature of about 100 billion kelvins. A teaspoon of cosmic matter weighs more than 100,000 tons.

This is where BBN begins. By going back only to about 0.01 second after the beginning, physicists limit themselves to a tem-perature and density domain they can work with comfortably. More than 60 years of particle accelerator experiments validate their understanding. Running the clock forward from 0.01 second, BBN describes the universe’s next 3 minutes in astonishing detail.

From the chaos of those first moments, fusion physics leaves an unalterable imprint on the universe. To choreograph this dance, BBN requires two critical components — an understanding of fusion processes and the physical conditions in the young cosmos.

A hydrogen nucleus (denoted H) is a single proton. Helium nuclei (denoted He4) have two protons and two neutrons. Fusing hydrogen into helium is a battle between electromagnetism and the strong nuclear force, two of the four forces that govern the cosmos.

While it’s easy to push neutral neutrons together, every proton carries a positive electric charge. Like charges repel via the elec-tromagnetic force, which gets stronger as the particles get closer. (It’s like trying to force the same poles of two magnets together.) To fuse into more complex nuclei, protons must overcome this electromagnetic barrier.

The strong nuclear force is more powerful than electromagne-tism. But it has the odd property of kicking in only when protons and neutrons get really close to each other.

At a high enough density and temperature, protons whiz around fast enough that some collisions have the energy to push them past the electromagnetic barrier and trigger fusion. But because the universe is expanding and cooling, Big Bang nucleo-synthesis becomes a race against time.

Beat the clockThe universe’s rapid expansion and cooling leaves only a brief win-dow for nuclear fusion to occur. Einstein’s theory of relativity speci-fies the expansion rate; nuclear physics specifies the temperature and density at each moment in cosmic history. But as the young

Big BAngThe event that spawned space, time, and the expanding universe.

deuTeronA hydrogen nucleus (pro-ton) bound to a neutron; a nucleus of deuterium.

FusionThe merger of protons and neutrons to form atomic nuclei, accompanied by a characteristic energy release. The fusion of hydrogen into helium powers the Sun.

nucleonA proton or neutron.

nucleosYnThesisProcesses in stars and the early universe that create new atomic nuclei from existing protons and neutrons.

TriTonA hydrogen nucleus (proton) bound to two neutrons; a nucleus of tritium.rAlph Alpher and gamow

found that observed abun-dances of light elements, like hydrogen and helium, are a consequence of a hot, expanding early universe.

AiP

em

ilio

seg

rÈ V

isu

Al

Arc

HiV

es

george gAmoW, a Russian-American scientist and a pioneer in nuclear physics, suggested the universe originated from a hot sea of radiation and particles.

AiP

em

ilio

seg

rÈ V

isu

Al

Arc

HiV

es

nucleAr speAk

17

Triton

Lithium nucleus (Li7)

Energy release

Normal helium (He4)

Deuteron

Proton

“Light” helium (He3)

Neutron

Normal helium (He4)

Energy release

Energy release

helium genesis, pArT 2. “Light” helium (He3) also formed in the Big Bang’s opening act. Stars convert deuterium to He3 , but, beyond this, little is known. Some argue the actions of stellar furnaces result in little net He3 production or destruction. if this is true, the total amount of deuterium and He3 remains approximately constant.universe ages, each temperature

and density regime allows only cer-tain particles to exist and certain kinds of reactions between those particles.

Fusion can’t start until protons and neutrons — collectively, nucleons — form. A millionth of a second after the Big Bang, when the temperature is a mere 2 trillion K, the universe has cooled enough that quarks can coalesce into protons and neutrons.

About 1 second after the Big Bang, the ratio of neutrons to protons becomes fixed, and fusion reactions can begin. But this window of opportunity lasts only 3 minutes. After this time, the cosmos will have expanded and cooled so much, it won’t support fusion reactions at all.

As BBN begins, protons outnumber neutrons 7 to 1. The difference emerges because neutrons are slightly heavier than protons, and this mass difference allows a neutron to decay spontaneously into a proton, electron, and a ghostly particle called a neutrino. Left to its own devices, a lone neutron will, on average, decay into a pro-ton and an electron in just 15 minutes.

save the neutronFusion saved the neutrons. They collided with the abundant protons and fused together as a deuteron — the simplest compound nucleus. A deuteron, a nucleus of deuterium (denoted H2), is a second stable isotope of hydrogen.

Deuteron formation can’t start up in earnest until about 100 seconds after the Big Bang. Once it does, it triggers a cascade of reactions that leads to nuclei with 2 protons and 2 neutrons — helium. For example, a deuteron may collide with a neutron to make tritium (H3), which then collides with a proton to make normal helium (He4). Or the deuteron could collide with a proton to make a nucleus of light helium (He3), which then collides with a neutron to make He4.

Other reactions create a small amount of lithium and beryllium. But that’s as far down the periodic table as we can go before fusion

grinds to a halt. More complex elements must await the first stars — several hundred million years in the future.

Scientists must follow all possible reactions, their pace, and all their products. Most importantly, physicists must perform these calculations in a cosmic back-ground of continually changing temperature and density. It is a tremendous task. But when the smoke clears, BBN predicts exactly how much

hydrogen, helium, deuterium, and other light elements exist in the cosmos.

From H to usWhile stellar nucleosynthe-

sis could not match the observation that helium makes up one-quarter of the cosmos’ mass, Big Bang nucleosynthesis nails it right out of the gate. BBN’s main prediction is the copious early production of He4. This

result ends up being remarkably insensi-tive to details in the calculation. Barring

major changes to the basic scenario, BBN always leads to helium production close to the

observed amount. Fundamentally, all that really matters is that a Big Bang occurred.

Helium abundance isn’t all that sensitive to conditions in the early universe, but deuterium is another story. The denser the early universe was at the beginning of the fusion era, the more likely it is that all the deuterium would end up in helium nuclei. That some deuterium remains — even 0.01 percent relative to hydrogen — tells physicists something about the young universe.

This and other trace elements let physicists determine the universe’s baryonic density — a measure of matter like protons

FuSion reactions begin about 1 second

after the Big Bang and last only

3 minutes.

18

Triton

Lithium nucleus (Li7)

Energy release

Normal helium (He4)

Deuteron

Three quarks constitute neutrons and protons

Neutron

Energy release

Proton

Tritium (H3)

Temperature (kelvins)100 million K100 billion K 10 billion K 1 billion K

Time after Big Bang1 second 1 minute 5 minutes 1 hour

Neutrons

Protons (H)

Frac

tion

of t

otal

mas

s

Greater

Lesser

Deuterium (H2)

Heliu

m (H

e4)

Bery

llium

(Be7

)

Helium (He3)

Lith

ium

(Li7

)

Lithium (Li6)

and neutrons — with high accuracy. Using precise measurements of light-element abundances in regions as diverse as stars and intergalactic clouds, astrophysicists now can claim that the den-sity of normal matter in the cosmos is only around 2 percent of the value needed to halt the universe’s expansion in the future. Most astronomers and cosmologists believe the universe’s total density (the sum of all kinds of matter and energy) exactly equals this critical density.

Because BBN predicts such a tiny fraction for stuff like us, the rest of the universe must be composed of dark matter and dark energy — “dark” in the sense that astronomers don’t yet under-

stand what they are. In this way, BBN not only has provided proof that a Big Bang must have occurred, but it also gives us strong evi-dence that we have much to learn about our universe.

Much has changed from Alpher and Gamow’s first calculations 60 years ago to the current era of precision cosmology. Now, a wealth of high-quality data lets scientists test competing cos-mological models. But while astronomers have firm reasons for believing in the reality of the Big Bang, they don’t need to rely on cutting-edge physics to do so. Big Bang nucleosynthesis shows us that a brief period of well-understood physics has consequences that trickle down 13.7 billion years to the universe we observe.

The AmounT of deute-rium peaks about 100 seconds after the Big Bang, but much of it becomes swept into helium nuclei. Fusion with these helium nuclei then builds lithium and beryllium. But Be7 isn’t stable, and the nucleus decays to Li7 with a half-life of 53 days. Tritium also decays, with a 12-year half-life, to He3. none of the beryllium or tritium formed during BBn sur-vives today.

The early universe’s chemical content

consTrucTing liThium. The heftiest survivor of Big Bang nucleosynthesis is lithium (Li7). Some stars pro-duce the element, others destroy it. Astronomers study its abundance in the atmospheres of stars in our gal-axy’s halo to infer its original value. observations and stellar models suggest these stars contain about half of the lithium available before stars began to shine.

19

1 seco

nd

10–10 se

cond

10–15 se

cond

10–20 se

cond

10–25 se

cond

10–30 se

cond

10–35 se

cond

10–40 se

cond

10–45 se

cond

10–5 se

cond

IN ITS FIRST SECOND, the uni-verse witnessed an explosive rate of expan-sion known as inflation, the birth of the four fundamental forces, and the creation of a sea of relic neutrinos that is still with us. Astronomy: Roen Kelly

10–43 second — Gravity splits off

10–36 second — Strong nuclear force splits off

10–36 to 10–32 second — Inflation occurs

10–32 to 10–5 second — Sea of quarks and antiquarks

10–5 second — Protons and neutrons form

© 2013 Kalmbach Publishing Co. This material may not be reproduced in any form without permission from the publisher. www.Astronomy.com

20

1 seco

nd

10–10 se

cond

10–15 se

cond

10–20 se

cond

10–25 se

cond

10–30 se

cond

10–35 se

cond

10–40 se

cond

10–45 se

cond

10–5 se

cond

stronomers who study the early universe divide into two camps: theorists and observers. Theorists routinely

work on ideas like inflation, which began some 10 trillion-trillion-trillionths of a second after the Big Bang. Observers, on the other hand, hit a road-

block if they try to look back further than 380,000 years after the Big Bang — when photons of light were first set free and created the cosmic microwave background (CMB).

Nonetheless, observers may have one path to take them back further. The Big Bang created a flood of neutrinos — subatomic particles that rarely interact with matter — that should fill the universe with an estimated 300 neutrinos per cubic centimeter, producing a background like the CMB. And those primordial, or “relic,” neutrinos date back to a mere second after the Big Bang.

No one has detected a relic neutrino, which by now would have cooled to a frosty 1.95 kelvins, although scientists have pondered the problem for decades. The chief difficulty is that neutrinos are

The birth of the universe released a torrent of neutrinos. Where are they?

⁄ ⁄ ⁄ BY STEvE NaDIS

Contributing Editor Steve Nadis is a science writer living in Cambridge, massachusetts. He enjoys writing about cosmology and the early universe.

1 second

A

The Big Bang

10–12 second — Electromagnetic and weak nuclear forces split

1 second — Cosmic neutrino background forms

21

High-energycosmic-rayneutrino

Low-energy relic neutrino

Milky Way

Z-particle

Z-burst beamed

toward Earth

Z-particle decays in 3x10–25 second into an average of about:

1 baryon-antibaryon pair

10 neutral pions r Decay into 20 high-energy photons

17 charged pions r Decay into electron-positron and neutrino-antineutrino pairs

elusive by nature, even under the best circumstances, and their interactivity decreases with the square of their energy. After cooling for nearly 14 billion years, Big Bang neutrinos would pass through virtually everything without a trace. In the unlikely event that a cold neutrino did react with one of the traps astronomers have laid, the resulting signal would be “frustratingly minuscule,” according to University of Hawaii physicist John Learned.

Thomas Weiler of Vanderbilt University realized in 1982 that astronomers needed a different approach — something that went beyond conventional neutrino searches. Suppose some neutrinos are flying around at ridiculously high energies, roughly a billion times greater than our best particle accelerators can achieve. There is a special energy at which the probability of this hyperkinetic neutrino interacting with a cold relic neutrino (or antineutrino) goes way up. This occurs at the resonant energy of the Z-particle, which is the product of this neutrino-antineutrino collision.

THE MOON is being tar-geted by GLUE — the Gold-stone Lunar Ultra-high energy neutrino Experiment. Scientists hope to see flashes as neutri-

nos explode in the lunar soil. T. A. RecToR/I. P. Dell’AnTonIo/noAo/AURA/nSF

ULTRaHIGH-ENERGY NEUTRINOS in cosmic rays rarely should interact with the Big Bang’s relic neutrinos, creating a Z-particle and a pronounced dip in the cosmic-ray energy spectrum.

NEUTRINO PRIMERaNISOTROPY

A lack of uniformity in, say, the cos-mic background, seen when looking in different directions.

COBEThe Cosmic Background Explorer, a NASA satellite launched in October 1989 that made detailed measure-ments of the background radiation.

COSMIC MICROWavE BaCKGROUNDThe cooling afterglow of cosmic gen-esis, released about 380,000 years after the Big Bang at the time when matter and radiation parted ways.

COSMIC NEUTRINO BaCKGROUNDThe flood of neutrinos cosmologists suspect was released 1 to 2 seconds after the Big Bang.

COSMIC STRINGSOne-dimensional defects in the structure of space-time.

DOMaIN WaLLSTwo-dimensional defects in the structure of space-time.

GRaND UNIFICaTION THEORY (GUT)A theory that combines electromag-netism, the strong nuclear force, and the weak nuclear force.

INFLaTIONA period of rapid expansion that took place 10–36 second after the Big Bang.

MaGNETIC MONOPOLESHypothetical particles that contain only one magnetic pole.

Z-burst creation

THE 70-METER GOLDSTONE radio telescope looks for neutrinos crashing into the Moon. It hasn’t found any. nASA/JPl

Astr

on

om

y: R

oen

Kel

ly

22

To make a Z-particle, the fast-moving neutrino would need some 1022 to 1023 elec-tron volts (eV) of energy before slamming into its relatively inert counterpart. Anyone lucky enough to witness lots of these events would notice a dip at the Z resonance level — the energy at which the Z-particle is produced. Observing such a “Z-dip,” as Weiler calls it, would confirm a cosmic neutrino background that parallels the CMB — and predates it by 380,000 years.

But the technique would yield another important benefit: Physicists could calcu-late the neutrino mass because it is dictated by the Z resonance energy (already firmly established) and by the measured absorp-tion energy. With this new data, physicists could, in principle, calculate the masses of the three known neutrino types — elec-tron, muon, and tau — by identifying three separate dips.

In this way, Weiler says, we could com-pute the neutrino spectrum, just as investi-gators determine atomic spectra by shining a light on atoms and seeing dark lines at wavelengths where electrons absorb energy. Z-particles decay in a fraction of a second to make new forms of cosmic rays — pro-tons, neutrons, and pions — that decay, in turn, to make gamma rays and neutrinos. The hope is that Earth-based observatories might pick up some of these “Z-bursts” if they originate relatively nearby.

Weiler knew from the onset that two incredible things had to happen for his scheme to work. The first requirement, that neutrinos have mass, has been borne out in experiments since his original paper was published. The second prerequisite, that neutrinos are somehow accelerated to tremendous energies, is contingent on as-yet-unobserved physics. Researchers

have proposed two sources for ultrahigh-energy neutrinos. The first is the decay of so-called topological defects. These hypo-thetical entities — such as cosmic strings, domain walls, and magnetic monopoles — are like defects seen in ice and other crys-tals and pack lots of energy into a small space. The second possibility is the decay of heavy (and still theoretical) particles created when the universe was ultrahot.

The requirement for exotic sources, whose existence has not been verified, makes the proposition a long shot, Weiler concedes. No cosmic rays have ever been detected with energies much above 1020 eV,

yet he’s speculating about particles acceler-ated to energies 100 to 1,000 times greater. Scientists will need a new generation of neutrino and cosmic-ray detectors to con-firm this conjecture, along with lots of luck.

Yet the payoff would be great. The CMB offers a glimpse of the universe when it was 380,000 years old. That’s when the Big Bang photons decoupled from what had been an ionized plasma and started streaming

freely. Neutrinos, which interact much more weakly than photons, decoupled from matter and radiation when the universe was just a second or two old. “The neutrino background would enable us to look much farther back than the CMB, offering a direct probe of the universe 1 second after the Big Bang,” says Andreas Ringwald, a physicist at DESY in Hamburg, Germany.

Fluctuations in the neutrino back-ground, formed when the universe was much smaller and denser, would be on correspondingly smaller scales than the fluctuations observed in the CMB. This would afford scientists new insights on inflation and the initial conditions that led to structure formation. Relic neutrinos are also thought to have played vital roles in nucleosynthesis and the universe’s general evolution. “Learning about the physics of the universe at 1 second,” says Princeton University’s David Spergel, “could tell us about the strength of gravity, for example, while placing interesting constraints on the number of particles created.”

The contribution to fundamental phys-ics could be immense, claims MIT cosmol-ogist Max Tegmark. “Information about the number of neutrinos and their masses is the final frontier of the standard model of physics. It’s the big unknown — the one thing that hasn’t been well-measured.”

“The cosmic neutrino background is a profound prediction of Big Bang theory and a very frustrating one,” adds University of Hawaii physicist Peter Gorham. “It’s like nature’s joke on us: Here’s something that must exist, but you may never get to see it.” On the other hand, Gorham admits, “There’s no physical reason why there shouldn’t be ultrahigh-energy particles, so it makes sense to look.”

THE aNITa-LITE EXPERIMENT flew over Antarctica (with Mount Erebus in the background) aboard a scientific balloon in 2003. The experi-ment looked for high-energy neutrino interactions with the ice. It came up dry. nASA

THE FORTE (Fast On-orbit Recording of Transient Events) satellite observed Greenland in the late 1990s. It looked for lightning bolts in the ice cap created by the impact of a high-energy neu-trino. It had no success. loS AlAmoS/SAnDIA/Doe

23

He has, in fact, looked for these particles in what might seem to be the unlikeliest places: the Moon, Greenland, and Antarc-tica. In 2003, Gorham and his colleagues reported on the Goldstone Lunar Ultra-high energy neutrino Experiment (GLUE). The team trained the Goldstone radio tele-scope in California on the Moon to look for explosions resulting from super-energetic neutrinos (about 1021 eV) slamming into it. Such a collision would release an enormous amount of energy in the form of relativistic electrons that would, in turn, create a flash of light. “The burst would be so intense that for 1 nanosecond, it would be brighter than the brightest quasar,” explains UCLA phys-icist David Saltzberg, the project’s lead researcher. But nothing like this was seen.

Gorham and other researchers looked for the effects of an ultrahigh-energy neu-trino shower on the Greenland icecap, drawing on data collected by the FORTE (Fast On-orbit Recording of Transient Events) satellite from 1997 to 1999. An impact of this sort would have created a 10-meter-long lightning bolt inside the ice and an attendant radio pulse. But again, the team failed to identify such an event.

A 2004 balloon experiment over Antarc-tica called ANITA-LITE — a prototype of the more ambitious ANITA experiment that lifted off in December 2006 — also came up dry.

The three experiments, Gorham con-tends, rule out supercharged neutrinos as the main source of the most energetic cos-mic rays. In the 1990s, Weiler had proposed that collisions between neutrinos and anti-neutrinos in our galaxy’s halo would cause Z-bursts that could unleash the highest-energy cosmic rays ever detected on Earth. The origin of these cosmic rays remains

one of the great mysteries of science because ordinary cosmic rays can’t travel far without losing energy through inter-actions with CMB photons.

Weiler proposed that neutrinos could travel unfettered across the universe, creat-ing these cosmic rays relatively nearby through the Z-burst mechanism. “But our results show there would not be a high enough flux of ultrahigh-energy neutrinos to make those cosmic rays,” Gorham says.

“Of course, we have not ruled out the Z-burst process itself, which might still be the best way of directly observing the cosmic neutrino background.”

“While Tom Weiler’s Z-burst idea is very cool, we don’t know whether neutrinos get accelerated to high enough energies to make it happen,” comments FermiLab astrophysicist Scott Dodelson. Fortunately, he says, there are indirect ways of studying

relic neutrinos — an approach he compares to “seeing how the needle disturbs the hay-stack, without seeing the needle itself.”

The idea is to look for the imprint of the neutrino background on the CMB, on gal-axy distribution (as seen, for instance, by the Sloan Digital Sky Survey), and on the more general distribution of mass revealed by gravitational-lensing experiments. “The CMB is like money in the bank,” Dodelson says. “We’ve done these experiments for more than a decade, and they’ve always delivered more than expected. Those studies, combined with galaxy surveys and lensing experiments, offer a powerful suite of techniques and the surest way of getting at the neutrino background.”

Drawing on WMAP findings released in March 2006 and on large-scale structure and supernova data, Steen Hannestad, a physicist at the University of Aarhus in Denmark, claims to have seen the imprint of relic neutrinos on the CMB with better than 99.99 percent confidence. The effect is subtle, Hannestad explains, “But if the [relic] neutrinos weren’t there, you’d see a completely different temperature anisot-ropy pattern in the CMB.” Because neutri-nos constituted about 40 percent of the energy density in the early universe — putting them almost on a par with photons — they contributed to the expansion of the universe. However, they did not make a comparable contribution to structure for-mation, which has a pronounced bearing on the temperature pattern observed in the microwave background.

In a 2005 paper published in Physical Review Letters, Roberta Trotta of Imperial College London and Alessandro Melchiorri of the University of Rome carried the analy-sis a step further. “Not only can we detect

PLaNCK, a spacecraft launched in May 2009, will examine the cosmic micro-wave background in unprecedented detail. It seeks further evidence of the neutrino background. eSA

THE COSMIC MICROWavE BaCKGROUND (CMB) appears to show the subtle imprint of relic neutrinos. A simulation of the CMB with neutrinos (left) matches what WMAP observed. The net effect of the neutrinos (right) is small. oxFoRD UnIveRSITy

Searching for a neutrino signature

24

the sea of [Big Bang] neutrinos, we can also detect the fluctuations, or wiggles, on top of that sea,” Trotta says. These neutrino anisotropies have a gravitational effect on the CMB, smoothing out the clumpiness of the early universe on small scales by stream-ing out of dense regions at essentially the speed of light. “The cosmic neutrino back-ground affects the onset of gravitational collapse,” he adds. “If there were no neutri-nos, or fewer neutrinos, star formation and galaxy formation would start earlier.”

Trotta and Melchiorri ran computer simulations in the absence of neutrino anisotropies and compared the results to actual data from WMAP and the Sloan Survey. Their analysis affirmed the pres-ence of the neutrino anisotropies at the 95 percent confidence level. “We hope to learn more about the neutrino wiggles — not just confirming that the wiggles exist but also determining their structure and distribution,” Trotta says. “We’re really just at the beginning.”

The latest WMAP findings (released in March 2008) nearly cement the case for the cosmic neutrino background. The data confirm its existence to a confidence level of better than 99.5 percent. The neutrinos have such a big effect on the CMB that

scientists claim they made up 10 percent of the universe 380,000 years after the Big Bang. This compares with the 12 percent contribution from atoms, 15 percent from photons, and 63 percent from dark matter.

Tegmark expects to see rapid progress on this front. By way of comparison, he says, people started looking for fluctuations in the CMB soon after it was discovered in 1963. “But zero progress was made toward a deeper understanding of theory until 1992, when the COBE results were announced,” Tegmark says. “It’s the same with the neu-trino background: Now that the sensitivity has reached the point where we can learn interesting things about it, we’ll learn rapidly.”

Weiler is glad to see advances in indirect measurements, but he still believes it’s important to get direct observations “to make sure what we’re seeing really is the neutrino background and not the decay products of relic neutrinos.” FermiLab’s Dodelson, for instance, has proposed a “neutrinoless universe” scenario in which relic neutrinos annihilated themselves some 10,000 years after the Big Bang, leav-ing behind a background of other particles. The search for ultrahigh-energy neutrinos now awaits experiments that rely on bigger detector volumes like the full-scale ANITA.

Cosmic neutrino physics may eventually reach the required level of precision, says Ringwald. “It took decades to establish reli-able sources of [1012 eV] gamma rays, but now there are dozens of known sources. The same could happen with cosmic neutrinos.”

Confirmation of Weiler’s Z-dip and Z-burst hypothesis would have several profound ramifications. First would be the clear-cut detection of the cosmic neutrino background. Second would be a determination of the neutrino mass. Third would be the observation of physics at the grand unification theory (GUT) scale, pointing to exotic GUT particles or even to more exotic cosmic strings. Each of these would be important, warranting — in the opinion of Gorham and Dodelson — an automatic Nobel Prize.

Rather than getting carried away by dreams, Weiler is mindful of the harsh reality: Although the physics of Z-dips and Z-bursts looks robust, nature still has to cooperate by providing ultrahigh-energy neutrinos from sources that have not yet been identified — and might not even exist. “My 1982 idea may still be the best idea we have for directly detecting relic neutrinos,” Weiler says. “But 25 years have passed, and now we need an even better idea.”

THE PERSEUS GaLaXY CLUSTER is a rich collection of galaxies imaged by the Sloan Digital Sky Survey. Astronomers who map the dis-tribution of galaxies in the universe expect to see the imprint of the cosmic neutrino background. RobeRT lUPTon AnD The SDSS conSoRTIUm

25

Beforethere was

Astronomers are poised to explore the mysterious cosmic

Dark Ages. /// BY STEVE NADIS

The TransiTion from the cosmic Dark Ages to the first galaxies took hundreds of millions of years. Neutral hydrogen may be the key to what happened during this mys-terious epoch. ASTRONOMY: CHUCK BRAASCH AND ROEN KELLY

12 COSMOS ⁄ ⁄ ⁄ 2006

light© 2013 Kalmbach Publishing Co. This material may not be reproduced in any form without permission from the publisher. www.Astronomy.com

26

C old dark matter models of Big Bang cosmology predict the emergence of structure dur-ing the “Dark Ages” — the

period after the Big Bang’s heat was unleashed and before the first luminous sources turned on — but direct evidence is hard to come by.

Dark matter is invisible, by defini-tion, and has eluded detection efforts. However, the most abundant element of “ordinary matter,” hydrogen, has a con-venient observational signature. To astronomers hoping to probe this cos-mic dark era, neutral hydrogen’s 21- centimeter line may be all they have.

A neutral hydrogen atom consists of one proton and one electron. The elec-tron’s direction of “spin” — spin is an intrinsic property of particles — can be either aligned with or opposed to that of the proton. The atom has slightly greater energy if the spins are aligned. Because of this, when the electron flips its spin from aligned to opposed — from a higher to a lower energy state — it emits energy in the form of a photon with a wavelength of 21 centimeters [see “At the atomic level,” page 14].

“That’s why everyone is so excited about hydrogen’s 21-centimeter line,” says Ger de Bruyn of ASTRON, the Netherlands Foundation for Research in Astronomy.

So where’s the signal?Radio telescopes will be used to track primordial hydrogen because any 21cm radiation from early epochs has been stretched to meters-long radio waves by the universe’s expansion. Because of this, the new telescopes will look at wavelengths between 1.5 and 5.3 meters. Snaring these long-wavelength signals, says Harvard theorist Lars Hernquist, “is the next big frontier in observational cosmology” — a sentiment shared by the people behind the telescopes.

Indeed, radio-astronomy activity abounds these days, as researchers pre-pare to hunt for the “cosmic hydrogen background.” Prototypes have already been built for several radio surveys: the LOw-Frequency ARray (LOFAR) in the Netherlands; the PrimevAl Structure Telescope (PAST) in China; and the

Mileura Widefield Array (MWA) in Australia. “This is not just a few experi-ments,” notes Massachusetts Institute of Technology (MIT) astrophysicist Miguel Morales. “It’s the birth of a field.”

Jeff Peterson, a Carnegie Mellon physicist coleading PAST with Xiang-Ping Wu of China’s National Astronom-ical Observatory, is looking forward to practicing astronomy at redshift 17, some 200 million years after the Big Bang. “We have quasar astronomy at redshift 6 — a billion years after the Big Bang — and cosmic microwave back-ground [CMB] astronomy at redshift 1,000 — a few hundred thousand years after the Big Bang — but nothing in

between,” he says. “Here’s a chance to see 5 times farther than Hubble or Sloan [Digital Sky Survey] — a chance to advance astronomy fivefold.”

Astronomers hope to figure out where neutral hydrogen has accumu-lated over time and when it reverted to its ionized form. Space now contains vast regions of ionized hydrogen, desig-nated HII. After the Big Bang, all gas was hot and ionized. The universe cooled as it expanded, and hydrogen became neutral, with its constituent pro-tons and electrons joining together as neutral hydrogen at the moment of recombination. The CMB followed shortly afterward.

Neutral gas in the intergalactic medium reionized hundreds of millions of years later — the exact time is still to be determined — when it was heated by radiation from early stars, galaxies, and quasars. The universe went from ionized to neutral and back to ionized.

The last phase, called reionization, may be discernible as a bump in the sky’s radio background. “Our best

chance of detecting the reionization sig-nal is if the neutral gas disappeared quickly,” explains Steve Furlanetto of the California Institute of Technology. If the change was gradual, it will be harder to pick out amid the background noise of the Milky Way and other galaxies.

If astronomers can determine the frequency of the “bump” in the radio spectrum when the neutral gas abruptly vanished, they’ll immediately know the redshift, which will tell them when reionization happened. Next, they’ll look for how it happened: In other words, what sources lit up the universe?

Zooming in on reionizationThe origin of the first stars has fasci-nated humans for centuries, but interest in the reionization epoch is more recent. The research gained momentum in 2003, when investigators on the Wilkin-son Microwave Anisotropy Probe (WMAP) team announced that reion-ization of the universe’s intergalactic gas likely occurred roughly 200 million years after the Big Bang, rather than 1 billion years after, as previously thought.

WMAP narrowed down the onset of reionization to between 100 million years (redshift 30) and 400 million years (redshift 11) after the Big Bang, but it could not follow the process over an extended period of time. If reionization occurred early, at redshift 30, it’s doubt-ful any neutral-hydrogen instruments under development will have the sensi-tivity to detect it. Background emissions from radio galaxies are noisier at low frequencies because galaxies — includ-ing the Milky Way — give off more low-energy, and hence low-frequency, photons than high-energy photons.

An analysis of two distant quasars’ absorption spectra observed by the Sloan Survey suggests more than 10 per-cent of the universe’s hydrogen was still neutral 900 million years after the Big Bang, according to Harvard’s Abraham Loeb and University of Melbourne’s Stu-art Wyithe. The findings were welcome news to MIT radio astronomer Jacque-line Hewitt, who heads MWA’s reioniza-tion group. “That means it’s worth doing this experiment because reionization is still going on,” Hewitt says. Even with

“This is not just afew experiments. It’s the birth of a field.”— Miguel Moraleslight

27

Ly β

Ly γ

Ly α

Ly δ

H α

H β

H γ

H δ

3 α

3 β

3 γ

n=1

n=2

n=3

n= 4

n=5

n=6

Lyman series

(Ultr

aviolet)

Balmer series (Visible)

Paschen series

(Infrared)

Electron

Proton Proton

Electron

Spins aligned Spins opposed

flips

21cm photon

present capabilities, she adds, “There should be something to see.”

Experimental varietySome technology used in these efforts is not new. “A telescope like PAST could have been built years ago, but no one made the investment,” says Peterson. He and his

fellow astronomers are deploying standard TV and radio antennae — items that can be purchased at a local electronics store.

The ingenuity lies in linking large numbers of these devices in sophisticated ways to make them perform like one large dish. “It takes a lot of computing power to bring the information together, but com-puting costs keep falling, unlike the cost of steel,” notes Morales. “It’s an inexpensive way of increasing your sensitivity.”

The main projects now underway — LOFAR, PAST, and MWA — all rely on multiple antennae with no moving parts, but their search strategies differ. PAST will point in just one direction — the North Celestial Pole — gaining deep observa-tions of a small field of view.

Of these projects, LOFAR has the lon-gest baseline, a 62-mile-diameter (100 kilometers) array, which offers the greatest sensitivity. MWA is just a mile (1.5 km) across but has the largest field of view. By surveying the sky for longer times, MWA astronomers hope to match LOFAR’s sen-sitivity. Before long, says Morales, “We’ll see who guessed right.”

Having different instruments with dif-ferent designs is critical, says Peterson, “because this is difficult astronomy, pos-sibly more difficult than the CMB.” While most challenges for CMB astronomy are technical in nature, the main problems for 21cm astronomy lie in the sky.

The radio background’s temperature is 200 Kelvin — 75 percent of which is from the Milky Way, 25 percent from other radio galaxies and quasars. Whereas the neutral-hydrogen signal is a mere 20 mil-likelvin — 10,000 times dimmer. “We can’t identify individual radio galaxies and remove their signals,” Peterson says.

Meanwhile, the ionosphere — a layer of ionized particles between 50 and 200 miles (80–320 km) above Earth’s surface — intermittently reflects radio waves, creat-ing havoc for radio astronomers. At higher redshifts, the problem gets worse.

Radio-frequency interference from tele-vision and radio transmitters and air-planes further confounds matters. “It’s definitely going to be hard,” admits Chris Carilli of the National Radio Astronomy Observatory. “If it were easy, we’d have done it already.”

Resolving a historyAstronomers have been spurred on by the potential payoff. “21cm measurements offer the richest data set on the universe’s initial conditions that we can see on the sky,” says Loeb. Calculations he did with Harvard colleague Matias Zaldarriaga show neutral-hydrogen anisotropies “con-tain an amount of information orders of magnitude larger than any other cosmo-logical probe,” says Loeb.

Whereas CMB measurements are taken at a single redshift — one moment in time

hydrogen’s specTrum is the simplest of all elements. As an electron falls from a higher to lower energy level, the energy difference is emitted as a photon. The most studied lines in astronomy are hydrogen alpha (Hα), the smallest fall of the visible Balmer series; and Lyman alpha (Lyα), the smallest drop of the ultraviolet Lyman series. The Balmer series — Hα, Hβ, Hγ, and Hδ — gives rise to the lines com-monly seen in a spectroscopic view of hydrogen gas.

Zooming in on The lowesT energy level — neutral hydrogen’s ground state, n=1 in the top diagram — the electron’s spin can be either aligned to the proton’s spin (left) or opposed (right). As the electron’s spin flips from the higher energy state to the lower energy state, it will emit a 21cm wavelength photon. astronoMy: RoEn kElly

steve nadis, an Astronomy contributing edi-tor, lives in Cambridge, Massachusetts.

At the atomic level

28

21cm emission shell

21cm emission

Ionized gas shell

Quasar X rays

UV rays

antenna on each of VLA’s 27 dishes to intercept low-frequency radio signals. The initial plan is to examine the three most distant quasars, all lying beyond redshift 6.2 in the reionization epoch.

When other far-flung quasars are dis-covered, they’ll be added to the list. By taking advantage of VLA’s existing infra-structure, says Greenhill, “We can do this fast and cheap.” With the Smithsonian Astrophysical Observatory covering the $100,000 equipment costs, they just need to secure telescope time.

Greenhill and his colleagues will focus on the thin “emission shell” lying at the edge of the bubble — the most likely source of 21cm radiation. X rays from the quasar penetrate and heat the neutral gas, giving rise to 21cm emissions.

The quasar’s ultraviolet photons, which the neutral medium absorbs strongly, ion-ize the gas and thereby cease emissions [see “Making bubbles in space,” below]. Greenhill hopes to determine the size and growth rate of the bubbles, while gathering details about the surrounding medium.

After first tackling the most extreme cases, radio astronomers will attempt to observe ever-smaller fluctuations until the neutral hydrogen background is fully mapped. The ultimate goal, says Carilli, is “3-D tomography” — the equivalent of a CT scan of the universe. “Instruments like LOFAR won’t have the sensitivity for this experiment, but all hope is not lost,” he says. “Unlike WMAP, the COsmic Back-ground Explorer, did not map the CMB. Instead, it made a statistical measurement,

— neutral-hydrogen measurements can look at numerous redshifts, and hence dif-ferent times. The result should be the uni-verse’s definitive history of hydrogen.

To search the Dark Ages, astronomers have designed instruments to explore both available spectra: absorption and emission. In early epochs, when the universe’s gas was colder than the microwave back-ground, neutral hydrogen absorbed CMB flux, leaving a telltale imprint on the low-frequency (radio) end of the spectrum.

Astronomers can trace neutral hydro-gen’s distribution in this era by finding regions with fewer CMB photons. The technique, according to Loeb and Zaldar-riaga, could detect hydrogen clumps in the nascent universe 300 light-years across — a resolution roughly 1,000 times better than WMAP’s.

Once neutral hydrogen is heated above the CMB temperature by the first stars (but before reionization), the gas will emit 21cm radiation. Says Carilli, “The 21cm line is the only direct probe we have of the neutral intergalactic medium.”

The absorption signal has a lower fre-quency, and therefore, longer wavelengths than the emission signal, which occurred in a different epoch.

Bubbly universeThe distribution of this neutral gas, in turn, can reveal whether the ionizing sources are stars or quasars. Each primor-dial light source carves out a bubble of ion-ized gas within an otherwise neutral sea. Over time, the bubbles expand and overlap until reionization is complete.

Two scenarios exist. Either stars and galaxies caused ionization or quasars were to blame. Each leads to a geometrically distinct picture. “If quasars dominated the process, there should be a small number of large bubbles,” explains Loeb. “If normal galaxies drove reionization, there’d be a large number of small bubbles.”

From an observational standpoint, bubbles around quasars are the easiest places to start probing reionization because they’re dramatic features whose positions are well known. Lincoln Green-hill of the Harvard-Smithsonian Center for Astrophysics intends to use the Very Large Array (VLA) in New Mexico to study these bubbles.

In collaboration with Carilli and oth-ers, Greenhill wants to install a dipole

Quasars emit both X-ray and ultraviolet (UV) photons. Although both types of pho-tons travel at the speed of light, higher-energy X rays interact less with the surrounding media, and thus get out faster. UV photons ionize most of the neutral hydrogen because they are more abundant and the gas absorbs these lower-energy photons more easily. Energy from the UV photons ionizes the hydrogen by stripping away electrons from the hydrogen nuclei. As ionization continues — primarily by UV photons — a cavity of ionized gas forms and grows around the quasar, just like a steadily inflating bubble.

The X rays propagate through the gas with less energy loss — and penetrate far-ther — because they are less likely to be

absorbed. They heat the gas, therefore, without ionizing it, giving rise to 21cm radiation. As X rays propagate outward from the quasar, a thin emission shell — where 21cm emission takes place — forms just beyond the ionized bubble. When the UV photons reach this shell, the gas ion-izes, and emission ends. At that point, the X rays have penetrated farther into the neutral medium to heat the outlying neu-tral hydrogen, causing additional 21cm radiation. The emission shell moves out-ward radially, always one step ahead of the UV-driven ionization front. — s. n.

astr

on

oM

y: R

oEn

kEl

ly

a 21cm emission shell is always radiating away just ahead of the shell of ionized gas. By looking for these shells, or bubbles, astronomers will be able to map reionization.

making bubbles in space

29

Dark Ages

First stars

Present day

Cosmicmicrowavebackground

Protogalaxiesform

Galaxiesform

1

10

100

1,000

10,000

100,000

1 million

10 million

100 million

1 billion

10 billion

Years after the Big Bang (logarithmic scale)

Ionization phase

Reionization phase

and near-term instruments should be able to do something comparable.”

De Bruyn agrees, saying researchers will try a statistical approach first, looking for fluctuations in 21cm emissions com-pared to the average background level. To represent their data, astronomers will pro-duce a “power spectrum” that shows how fluctuations vary at different scales.

Furlanetto believes a power-spectrum analysis of 21cm emissions can help estab-

lish the characteristic size of hydrogen fluctuations throughout the universe’s his-tory because “these variations will peak on the scale of the bubble size.” Also, a differ-ent peak will occur at each redshift, and each of these peaks should grow over time. Therefore, Furlanetto says, “At each epoch in the reionization history, there should be a preferred scale for the size of bubbles.”

A computational model he devised with Hernquist and Zaldarriaga suggests a pre-

The era of lighT

ferred scale of about 10' (1/6 of a degree — equivalent to several millions of light-years in length) for bubbles in the mid-dle of reionization. Today’s instruments are aiming for that same scale.

Good “seeing”Initially, LOFAR will look for the charac-teristic sizes of ionized hydrogen bubbles; its search begins around 2007. The array will comprise 15,000 antennae deployed in a five-arm spiral spread out over a 62-mile-diameter (100 km) area in its initial phase. Morales is skeptical about LOFAR’s chance for success in such a densely populated area. With competing radio and TV signals, he says “It’s like doing astronomy in an outdoor disco.”

De Bruyn is confident his team can deal with those challenges, but he acknowledges the site’s shortcomings. For starters, they’ll have to curtail observa-tions at redshift 11.5 to avoid interference from FM radio broadcasts. But optical astronomers also have to adapt to clouds and inclement weather, he says. “Not everyone can observe in Hawaii or Chile.

A timeline of the universeeach cosmic epoch blends into the next because direct cutoffs are difficult to observe. Astronomers are working to understand parts of each era gradually, to see the full pic-ture eventually. Although many cosmological riddles are unanswered, and probably will remain so for many years, next-generation telescopes and detectors will allow astron-omers to look back hundreds of millions of years to probe the main question: How did structures form?

The cosmic Dark Ages ended once the universe’s neutral hydrogen was com-pletely reionized by the first stars and gal-axies. The luminous period that followed holds almost as many questions as the Dark Ages. Because astronomers know important steps in structure formation occurred shortly after the Dark Ages ended and the stars ignited, they want to explore this era soon. one of the instruments astronomers are pinning their hopes on is the James Webb Space Telescope. nASA plans to launch that instrument in 2013.

With the formation of the first galaxies, dust absorbed starlight and re-emitted infrared radiation. Astronomers think light from this process created a “cosmic infra-red background.” By studying this light,

astronomers can observe the “structural clumpiness” of the early universe.

The infrared radiation traces early pro-togalaxies, helping us understand how the universe’s structure evolved. — Liz Kruesi

scheduled for a 2013 launch, the James Webb Space Telescope — Hubble’s infrared successor — will look back to early galaxy formation. noRThRop GRUmmAn SpAcE TEchnoloGy

30

Dark Ages

First stars

Present day

Cosmicmicrowavebackground

Protogalaxiesform

Galaxiesform

1

10

100

1,000

10,000

100,000

1 million

10 million

100 million

1 billion

10 billion

Years after the Big Bang (logarithmic scale)

Ionization phase

Reionization phase

Just because one site is best doesn’t mean other sites can’t support good science.”

In contrast to LOFAR’s crowded set-ting in the Netherlands, PAST sits in the remote Ulastai Valley of western China, shielded from interference by high moun-tains. “We can choose our frequencies on the basis of cosmology,” says Peterson, rather than the whims of FM broadcast-ing. With a goal of 10,000 antennae installed upon completion of the project, PAST will survey the regions that existed 200 million years after the Big Bang, before they grew into the large-scale structures we see today.

In Australia, MWA has a similar tar-get, aiming for great big lumps of neutral hydrogen that will eventually become gal-axy clusters, rather than individual galax-ies. Although PAST will have a larger

collecting area — 80,000 square meters versus MWA’s

8,000 — MWA

can point in any direction. “We’ll look for holes in the galactic radio emissions,” says Morales, “seeking out the quietest, darkest places in the sky.” Test observations began in March 2005. Current plans call for an early 2007 start date, assuming the U.S. team can contribute its share of the cost.

Groundwork for the futureThe neutral-hydrogen signal will have to be confirmed by more than one instru-ment because of its faintness on the radio sky. Verification has proved crucial in CMB findings, notes Peterson, and it underscores the value of having multiple instruments and multiple approaches.

These early “pathfinder experiments” can help astronomers confront challenges posed by the ionosphere, radio galaxies, and interference so the next-generation radio telescope, the Square Kilometer Array (SKA), can be optimized before it’s built (in a location still to be determined) says Carilli, who chairs SKA’s science advisory team. Scheduled for operation in

2020, SKA will have roughly 10 times LOFAR’s collect-

ing area.

With the added sensitivity, says Carilli, “We can go from imaging the most extreme objects in the universe to imaging normal objects.” The goal is to examine small angular scales to chart the early clumpiness of neutral hydrogen, which should tell astronomers about structure formation, inflation, dark matter, and other cosmic riddles.

If things go well, and SKA lives up to its billing, there’s the potential to extract a wealth of information from the neutral-hydrogen signal, says Zaldarriaga. But he still considers the proposition dicey. “Whether or not this can be done in prac-tice remains to be seen.”

Ron Ekers, a radio astronomer based at the Commonwealth Scientific and Industrial Research Organization (CSIRO) — Australia’s national scientific research organization — is trying to pin-point the timing of reionization. What caught Ekers’ attention is the challenge. “I’m attracted to difficult experiments, which makes 21 centimeters a good area to work in,” he says. When asked at a recent conference about the highest red-shift astronomers can measure, Ekers replied, “There’s no cutoff. It’s difficult now, and it just gets harder and harder.”

Quick Terms

cosmology: The study of the universe.

phoTon: A “bullet” of energy; no mass; e.g., visible light, X rays.

dark maTTer: 22 percent of the universe; invisible; as of now, only detected from its gravitational effects.

redshifT: A measure of distance; as the universe expands, light’s wavelength is stretched and its color shifts toward red.

cosmic microwave background (cmb): The imprint from when light separated from matter; shows beginnings of structure.

astronoMy: RoEn kElly/WmAp ScIEncE TEAm

31

L ooking up at the star-studded sky on a clear night, it’s difficult to imagine there was a time when the universe contained no stars, no gal-axies, no shining celestial bodies of

any kind. Yet, for a significant stretch of the first roughly 100 million years after the Big Bang, total darkness permeated the cosmos.

Then came the first generation of stars, the brilliant and short-lived ancestors of our Sun and its kind, which burnt through the dense fog of primordial hydrogen and helium atoms. Investigating how these first stars ushered in the cosmic dawn and made possible all else that followed — including human life on Earth some 14 billion years later — is one of the hot-test research frontiers in cosmology today.

Hiding in the first billion yearsIn the beginning, there were no atoms, only a dense, hot soup of electrons, protons, and other elementary particles moving around and

scattering light at all wavelengths. As the infant universe expanded and cooled below a few thousand Kelvin, protons could hold on to electrons to make neutral atoms of hydrogen and helium for the first time. The cosmic microwave background — the relic glow of the hot, early universe — carries the imprint of cosmic structures at the time of this transition, 380,000 years after the Big Bang. That etching is remarkably smooth, with density variations of only 1 part in 100,000.

Cosmologists now use sophisticated com-puter simulations and deep observations with giant telescopes to unravel the tale of how the first stars and galaxies grew out of those tiny ripples. It’s a story shrouded in darkness, liter-ally. Between the time neutral atoms formed and the moment the first stars lit up, the uni-verse was in its so-called Dark Ages. Thus, there’s little chance for us to see directly the buildup of the first objects. Instead, we have to rely on theory to guide us.

Using sufficiently powerful telescopes, we can expect to see the effects of those first stars on their environment and their immediate descendants — and, in turn, test the theoretical

Detecting the earliest stars will help astronomers unlock secrets of the infant universe. /// BY RAY JAYAWARDHANA

In searchof the first

Ray Jayawardhana, a professor of astronomy and astrophysics at the University of Toronto, studies the formation of planets, brown dwarfs, and stars.

18 COSMOS ⁄ ⁄ ⁄ 2006© 2013 Kalmbach Publishing Co. This material may not be reproduced in any form without permission from the publisher. www.Astronomy.com

32

stars

THE FIRST STARS FORMED from the universe’s primordial material: almost all hydrogen and helium gas, with trace amounts of lithium gas. Because these stars were massive, they lived fast and died young, just 2 million years after their births. LYNette Cook foR AsTronomy

33

Proton

Neutrino

Deuteriumnucleus

Deuteriumnucleus

Helium-3nucleus

Helium-4nucleus

Helium-3nucleus

Gamma-ray photon

Neutron

Positron

Fusion

Fusion

Fusion

Fusion

Fusion

predictions. We have observed some galax-ies and quasars shining brightly a billion years after the Big Bang, so the first stars must have formed sometime earlier.

Making the first starsIt may seem futile to try to figure out the birth of the first stars in the distant, unob-servable past, when our understanding of present-day star formation is far from com-plete. But, in some ways, the recipe was simpler back then. We don’t need to include the complicating influences of magnetic fields threading gas clouds or shocks from nearby supernovae in the prescription for the very first stars.

The only ingredients in the mix were hydrogen, helium, and traces of deuterium and lithium — plus “dark matter.” Dark matter accounts for over 80 percent of mat-ter in the universe, and its influence is felt primarily through gravity. And one doesn’t need to worry about the cooling effects of heavier elements or dust produced by pre-vious generations of stars.

In recent years, astronomers have set up elaborate computer codes to model the for-mation of the first stars. Perhaps the most detailed simulations are those done by Tom Abel of Stanford University, Greg Bryan of Columbia University, and Michael Norman

of the University of California, San Diego (see “The universe’s first light” on page 22). Another group, including Volker Bromm of the University of Texas, Austin, and Yale University’s Richard Larson and Paolo Coppi, has done simulations using a sim-pler set of assumptions. However, this team explored a wider range of possibilities.

All of these simulations show that the minuscule density fluctuations — the lumps in the primordial soup — in the early universe acted as seeds for growing gas clouds. These fluctuations became the nodes of a network of filaments along which more gas continued to flow in. These clouds contracted under their own gravity and heated up to over 1,000 Kelvin.

A small number of hydrogen atoms paired up to become hydrogen molecules, which helped cool the gas by emitting infrared radiation. Once temperatures in the densest regions dropped to a few hun-dred Kelvin, the cloud clumps could con-tract further. Cooling was crucial: It allowed ordinary matter, which cooled by emitting radiation and, therefore, contin-ued contracting, to separate from dark mat-ter. The dark matter didn’t cool and, thus, remained scattered throughout the cloud.

Some of the densest gas clumps col-lapsed until they lit up. Because cooling by

STARS SHINE as a result of energy released during nuclear reactions. The simplest reaction — nuclear fusion — combines, or fuses, hydrogen nuclei together to produce helium and neutrinos. As a result, energy is released. Extremely high temperatures are required to combine nuclei. The universe’s primordial material was mostly hydrogen and helium nuclei. Once the gas collapsed enough because of gravity, nuclei were both hot enough and close enough to fuse. And then there was light: The first star lit up. AsTronomy: RoeN keLLY

small numbers of hydrogen molecules isn’t very efficient, these star-forming clumps were much warmer than their counterparts today. (Dust grains and other molecules cool present-day clouds more efficiently.) The gas clumps also had to be more mas-sive in order for gravity to overcome the outward pressure from the hotter gas. This means the first star-forming clumps were probably several hundred times more mas-sive than the Sun.

Did all of that mass go into a single star? Even the most-detailed simulations do not show any tendency for these clumps to fragment as they contract. Thus, theorists think the first stars were rather massive and luminous: Estimates for the upper mass limit range from about 300 to 1,000 solar masses, and the luminosities could have been millions of times that of the Sun.

the cosmic dawnThe birth of these first-generation stars marked the cosmic dawn. These behemoths were also extremely hot, with surface tem-peratures approaching 20 times the Sun’s. They emitted primarily ultraviolet light. Their energetic radiation would have started to heat and ionize the neutral hydrogen atoms in their vicinity, carving out a growing bubble of ionized gas around each one. As more and more stars formed over hundreds of millions of years, these

bubbles would have overlapped, until nearly all the universe’s gas became ionized.

Most of the first stars probably ended their brief, but brilliant, lives as exploding supernovae within a few million years. Theory predicts stars with masses between 140 and 260 times that of the Sun blow up completely, expelling the heavier elements they produce through nuclear fusion. The first stars’ ejecta seeded surrounding gas with ingredients for dust, planets, and even life, and were eventually incorporated into future generations of stars.

tHe fiRst stARs’ eJeCtA polluted surrounding gas with ingredients for dust,

planets, and even life.

Atomic billiards

34

Sun today(1 solar mass)

Red giant

Planetary nebula

White dwarf

Sun

Black hole

Brilliantexplosion

Massive blue star(100 solar masses)

Blue giant

First star

Protoplanetary nebula

www.aStrOnOMy.COM 21

FOR 10 BILLION YEARS, our Sun steadily burns, convert-ing hydrogen to helium in its core. After that point, the Sun will expand 100 times, and its outer layer will cool. It becomes a red giant. The Sun will fuse helium into heavier elements (carbon and oxygen), burning brighter and brighter for tens of millions of years. Finally, the Sun will shed its outer layers — initially as a protoplanetary nebula — releasing elements crucial to life. What was once the Sun’s core will contract into a white dwarf. Radi-ation from the white dwarf will cause the element-rich material to glow as a planetary nebula. AsTronomy: RoeN keLLY

ONE OF THE FIRST STARS would have been extremely massive — 100 solar masses in this example — formed mostly from hydrogen, helium, and a tiny amount of lith-ium gas. After just a few million years, the star burned its fuel and ended in fantastic style: as a huge explosion. The star’s material — including heavy elements — was ejected. Either its core collapsed as the first black hole, or the explosion was powerful enough to blow up com-pletely and scatter the star’s material throughout space.

the first stars and their descendants

Not to scale35

Both above and below this mass range, dying massive stars are expected to collapse into black holes without ejecting much of their mass. These stars didn’t enrich their progeny by much. But their remnant black holes may have lit up as “mini-quasars” as they accreted surrounding material, provid-ing additional sources of light and ionizing radiation that ended the Dark Ages.

Some of the stellar corpses may have merged with one another to build up the cores that seeded protogalaxies. This could explain why supermassive black holes appear to lurk in the centers of quasars and many galaxies. Thus, the demise of the first stars was perhaps even more important than their birth to all that came afterwards.

There is little chance of catching the first stars in their spectacular death throes, even with today’s largest telescopes. But some of them may have given rise to gamma-ray bursts, which are much more luminous than supernovae, at the edge of the observ-able universe. Detecting such a distant gamma-ray burst is our best hope for prob-ing the end of the Dark Ages directly.

from an ancient eraIn the meantime, other ways to investigate those early days exist: by looking for the effects of those first stars on their sur-roundings. In 2001, a group of astronomers led by Robert Becker of the University of California, Davis, detected possible signs of the final stages of cosmic reionization.

In the spectrum of one of the most dis-tant quasars known, dating to about 900

million years after the Big Bang (at a red-shift of 6.4), they found a telltale signature of neutral gas: Essentially, all of the quasar’s ultraviolet light had been absorbed by hydrogen atoms in the line of sight.

Slightly closer quasars do not show such complete absorption. These findings sug-gest the last patches of neutral hydrogen gas were ionized around that time.

There has been a new twist. Preliminary measurements of the degree of polarization in the cosmic background by NASA’s Wilkinson Microwave Anisotropy Probe (WMAP) satellite indicate the universe was reionized much earlier — sometime between 200 and 500 million years after the Big Bang (at a redshift of 10–20), rather than at 900 million years, as implied by the quasar observations. These findings may show the two ends of the reionization period: one the beginning, and one the end.

The apparent discrepancy between the WMAP and quasar results remains “a genu-ine big puzzle, and no one knows the physi-cal mechanism that can explain such an extended reionization history,” says Zoltan Haiman of Columbia University. The answer may come from the European Space Agency’s Planck satellite, to be launched in 2007, which will make more reliable mea-surements of polarization in the cosmic microwave background.

finding different methodsAstronomers would like to find other, per-haps even more distant, quasars and galax-ies that could contain the second generation

The universe’sfirst light

THE FIRST STAR to form was mas-sive — between 30 and 300 solar masses — and sat inside the first

forming protogalaxy, as shown by Tom Abel, Greg Bryan, and

Michael Norman in their computer simulations. Hydrogen and helium

gas — the universe’s primordial material — contracted to form the earliest generation of stars. These

huge stars burned bright, lived fast, and died young — after just

a 2-million-year life cycle. VisuALizAtioN: RALf kÄHLeR AND toM ABeL

siMuLAtioN: toM ABeL, GReG BRYAN, MiCHAeL NoRMAN

ASTROSPEAKDARK AGES

The roughly 200,000 years after the cosmic microwave back-ground radiation was released, but before the first luminous sources (stars) turned on.

DEuTERIuMAn isotope of hydrogen whose nucleus is composed of one neutron and one proton (com-pared to hydrogen, which has no neutrons), nicknamed “heavy hydrogen”; has twice the mass of hydrogen.

QuASAR(Abbreviated from “quasi- stellar object”) the most lumi-nous objects in the universe; thought to be powered by supermassive black holes; lie extremely far away.

REDSHIFTAs the universe expands, light’s wavelength is stretched and its apparent color becomes redder; used as a measure of cosmolog-ical distance with the symbol z; today is z = 0.

SuPERNOvAExplosive death of a star at least 8 times the mass of our Sun; expels elements into space that eventually seed planets and life.

At 65 light-years (20 parsecs) across, this shows cold gas in the center of one of the first protogalaxies.

Follow the contours to zoom inside clumps of different gas densities. One star will form in the center, where the density is highest.

36

of stars. Simulations predict these protogal-axies would be small and relatively faint.

Several surveys have been undertaken recently, using the Japanese-built Subaru Telescope in Hawaii and the European Southern Observatory’s Very Large Tele-scope in Chile, to look for such objects. A network of radio dishes now being built in northern Chile — the Atacama Large Mil-limeter Array — will search for distant star-forming galaxies at millimeter wavelengths.

The observers have come up empty so far. Unless they get lucky, success may have to await the launch of the James Webb Space Telescope, the planned successor to Hubble, or the construction of a 20- or 30-meter optical telescope on the ground.

Directly detecting neutral hydrogen is another way to probe the reionization era. Hydrogen atoms emit photons at a charac-teristic wavelength of 21 centimeters, in the radio. This radiation from the early universe would be redshifted to longer wavelengths — a few meters — by the time it reaches us.

Several new radio observatories are planned for the near future to look for hydrogen’s signature. The LOw-Frequency ARray (LOFAR), with 15,000 antennae spread over 62 miles (100 km), is under construction in the Netherlands, and its prototype is taking data. Ue-Li Pen of the Canadian Institute for Theoretical Astro-physics and his collaborators in the United States and China want to begin searching sooner and at low cost: They are placing thousands of commercially available TV antennae on a remote plateau in China and

expect to begin searching for hydrogen emission within the next year.

in our own neighborhoodOther astronomers are searching for clues closer to home. While the first stars lived fast and died young, low-mass stars that formed soon after their deaths may still lurk in the Milky Way’s halo. Element abundances in the atmospheres of these old and iron-poor second-generation stars could tell us how their massive progenitors lived and died.

Within the past few years, astronomers have identified two such “hyper-metal-poor” (HMP) stars. HE 0107–5240, located in the southern constellation Phoenix, and HE 1327–2326, which lies in Coma Bereni-ces, contain 1/200,000 and 1/300,000 of the Sun’s iron abundance, respectively. But these iron-deficient stars are enhanced in other (lighter) elements, such as carbon.

This was a surprise. Previous theoretical calculations suggested much iron, but only small amounts of carbon, would be ejected

during the death throes of the first stars and incorporated into the next generation. Timothy Beers of Michigan State University describes HMP stars as “scribes” of the early universe. “Their atmospheres retain the memory of the composition of the gas from which they formed,” he explains.

To account for the abundance pattern in these two stars, their progenitors must have masses as low as 25 times that of the Sun, according to a new theoretical model by Nobuyuki Iwamoto and his colleagues in Japan. To test this emerging scenario, astronomers now are looking for other examples of ancient, iron-poor stars. It isn’t easy to find them because they are dim and usually live in the outskirts of the galaxy.

The Sloan Digital Sky Survey’s recent extension includes a search called the Sloan Extension for Galactic Understanding and Evolution (SEGUE). This study will collect spectra of 250,000 stars in the Milky Way through summer 2008, and it may reveal other candidates for HMP stars.

“The SEGUE project will allow us, for the first time to get a ‘big picture’ of the structure of our Milky Way,” says Heidi Newberg of Rensselaer Polytechnic Insti-tute in New York.

Clearly, we haven’t unraveled the full story of the cosmic dawn yet. Some missing chapters may be deciphered in the spectra of old stars in the Milky Way’s halo. Other parts would be revealed in snapshots from billions of light-years away. And plot twists could emerge as astronomers put the pieces together over the next decade.

The gas density rises rapidly as the pri-mordial gas radiates and collapses faster and faster.

At the center of the protogalaxy, where the gas density is greatest, material con-tinues to collapse into a star.

The white-hot first star — roughly 10 times our solar system’s width — burns quickly because of its massive size.

tHe DeMise of the first stars

was perhaps even more important than

their birth.

37

SubScribe now!

1 month SubScription$3.99 - now available

1 year SubScription$42.99

P18482

choose the option that is best for you:

38