74 Science and technology
The Economist
September 22nd 2018
A
SMISMATCHES go, it’s a big one. When
physicists bring the StandardModel of
particle physics and Einstein’s general the-
ory of relativity together they get a clear
prediction. In the very early universe,
equal amounts of matter and antimatter
should have come into being. Since the
one famously annihilates the other, the re-
sult should be a universe full of radiation,
but without the stars, planets and nebulae
that make up galaxies. Yet stars, planets
and nebulae do exist. The inference is that
matter and antimatter are not quite as
equal and opposite as themodels predict.
This problem has troubled physics for
the past half-century, but itmaynowbe ap-
proaching resolution. At
CERN
, a particle-
physics laboratory near Geneva, three
teams of researchers are applying different
methods to answer the same question:
does antimatter fall down, or up? Relativ-
ity predicts “down”, just like matter. If it
falls up, that could hint at a difference be-
tween the two that allowed a matter-
dominated universe to form.
Testing this idea is tricky. Antimatter
needed for experiments is usuallymade in
accelerators, in the form of particles travel-
lingat nearly the speedof light. Touse it ina
gravitational test, it must be slowed down,
isolated from all other forces and kept in-
tact for long enough to track its fall (or rise).
All without allowing any matter to touch
it, and thus annihilate it.
This is possible only at
CERN
, which
has dedicated one of its smaller accelera-
tors to creatinga streamof low-energyanti-
protons. It does this by smashinga high-en-
ergy beam of protons into a block of
iridium, and then feeding the resulting
antiprotons into a ring 180 metres in cir-
cumference. This ring acts as a particle ac-
celerator in reverse, slowing the antipro-
tons to amere tenth of the speed of light.
Since antiprotons are electrically
charged (theyare negative; normal protons
are positive), gravity’s effect on them is
swamped by any electromagnetic forces
around, which will be much larger. All
three experiments deal with this by com-
bining their antiprotons with positrons
(antielectrons) to create antihydrogen at-
oms. These are electrically neutral, and
thus immune to such disturbances.
One team,
AEGIS
(Antihydrogen Ex-
periment: Gravity, Interferometry, Spec-
troscopy), guides the resulting beam of an-
tihydrogen into something called a Moiré
deflectometer. This splits the beam up and
then recombines it to create an interference
pattern. Gravitational effects should show
up as small shifts in the pattern.
A second group,
GBAR
(Gravitational
Behaviour of Antihydrogen at Rest), runs
twomore stages of deceleration, then adds
two positrons to each antiproton, to create
a positively charged antihydrogen ion.
Such ions, precisely because they are
charged, can be slowed down even further
using an electric field. This brings them to
something approaching human walking
pace. The field then guides them towards a
target point where they are pinned in place
by a second field and the extra positron is
shot off by pulses of laser light. At this
point, the ions become electrically neutral,
so gravity is the only remaining force act-
ing on them. The direction of their fall can
thus be observed directly.
The third team,
ALPHA
, maybe the clos-
est to success. This group has spent the past
18 years measuring every possible proper-
ty of antihydrogen, and has become adept
at trapping it and chilling it towithin half a
degree of absolute zero. (A cold atom is a
slow-moving atom.)
ALPHA
’s best attempt
so far gathered antihydrogen atoms in a
trap and turned off the magnetic field to
watch which way they drifted. But even at
this temperature, there was still plenty of
noise in the system. The team could con-
clude only that antimattermight fall down-
wards anywhere up to110 times faster than
normal matter does, or fall upwards any-
where up to 65 times faster than normal
matter falls down.
Now,
ALPHA
’s researchers may be able
to do better. As a consequence of work
they reported inAugust on a phenomenon
called the Lyman-alpha transition, they
think they can bring a technique called la-
ser cooling to bear on antihydrogen. This
should bring the antiatoms towithin a few
hundredths of a degree of absolute zero.
With antihydrogen that cold, a similar drift
test would give a much more precise mea-
surement of gravitational behaviour.
All three teams are now in a race to per-
fect their apparatus in time to make obser-
vations before
CERN
’s accelerators come to
the end of their current run and take a two-
yearmaintenance break. That is scheduled
for November. There appears to be a good
chance of at least a “sign test” by then (ie,
one that provides an answer to the ques-
tion of whether antimatter falls up or
down), even if it does not show exactly
howfast the atoms are falling or rising.
Few really believe that antimatter falls
up, because so much existing theory pre-
dicts it will not. But in 1887 existing theory
predicted that the apparent speed of light
would vary with the speed at which the
observer was moving. When Albert Mi-
chelson and EdwardMorley showed that it
does not, they tore up the rule book of
physics. It could happen again.
7
Antimatter
I go up, or down...
Three experiments probewhymatter
and antimatter are notmirror images
YUSAKU MAEZAWA is a Japanese fashion billionaire. If he gets his way, in 2023 he will
become the first person to visit the Moon since Eugene Cernan, an American astronaut,
left in 1972. Unlike Mr Cernan, Mr Maezawa will not land. Instead, he will zoom round the
Moon aboard a spaceship built by SpaceX, an American rocketry firm. That, at least, was
the plan announced on September 17th by Elon Musk, SpaceX’s boss. Whether it will
happen is another question. SpaceX is already late with plans to fly astronauts to the
International Space Station, which is far easier. A previous lunar trip scheduled for this
year has been shelved. And the rocket that will do the job does not yet exist. Even Mr
Musk was careful to describe the timeline as aspirational.
Destination: Moon. Maybe