of 5
VOLUME
72,
NUMBER
5
PHYSICAL
REVIE%
LETTERS
3l
JAXUWRV
1994
Search
for
Slowly
Moving
Magnetic
Monopoles
with
the
MACRO
Detector
S.
Ahlen,
(
)
M.
Ambrosio
(
z)
R.
Antolini,
(
)
(7')
G.
Auriemrna,
(
4)
(
R.
Baker,
(
)
A.
Baldini,
(
)
B.
B.
Bam
)
G.
C.
Barbarino(
)
B.
C.
Barish,
)
G.
Battistoni,
(6)
(
)
R.
Bellotti,
(
)
C.
Bemporad,
(
P.
Bernardini
(
)
H.
Bilokon
(
)
V.
Bisi
(~
)
C.
Bloise,
(
)
C.
Bower,
)
S.
Bussino,
(
)
F.
Cafagna,
(~)
M.
Calicchio
D.
Campana
(
P.
Campana
M.
Carboni
S.
Cecchini,
(
)
(')
F.
Cei,
'
)
V.
Chiarella,
(
R.
Cormack,
(
)
A.
Corona,
&
)
S.
Coutu
(4)
(&&
)
G.
De
Cataldo
(&)
H.
Dekhissj.
,
(
)
(d)
C.
De
Marzo,
(
)
M
De
Vincenzi,
(
4)
(
)
A.
Di
Credico,
(
)
E.
Diehl,
(
)
O.
Erriquez,
(
)
C.
Favuzzi,
(t)
D.
Ficenec,
(s)
(r)
C.
Forti
(
)
P.
Fusco,
(
)
G.
Giacomelli
(
)
G
Giannini,
(
)
(~)
N.
Giglietto
(
)
P.
Giubellino,
(
)
M.
Grassi,
(
P.
Green,
(~
8')
A.
Grillo
(
)
F.
Guarjno
(»)
p
Guarnaccja
(~)
C
Gustavino
(7)
A
Habjg,
(8)
R.
Heinz,
(8)
J.
T.
Hong,
&
)
(
')
E.
Iarocci,
(
)
(")
E.
Katsavounidis,
(
)
E.
Kearns,
(
)
S.
Klein,
(
)
(')
S.
Kyriazopoulou,
(4)
E.
Lamanna,
(
)
C.
Lane,
(
)
C.
Lee,
(i
)
D.
Levin,
(
)
(
')
P.
Lipari,
(i
)
G.
Liu,
(4)
R.
Liu,
(4)
M.
J.
Longo,
('i)
Y.
Lu(15)
G.
Ludl~,
(3)
G.
Mancarella,
(10)
G.
Mandrioll(2)
A.
Marglotta-Nerl(2)
A
Marin,
(3}
A.
Marini,
(
)
D.
Martello,
(
)
A.
Marzari
Chiesa,
(
)
M
Masera(
)
M
N.
Mazziotta,
(
)
D.
G.
Michael,
(
)
S.
Mikheyev,
(
)
(')
L.
Miller,
(
)
M.
Mittelbrunn,
(
)
P.
Monacelli,
(s)
M.
Monteno,
(
)
S.
Mufson,
(
)
J.
Musser
(s)
D.
Nicolo
(
)
R.
Nolty,
(
S.
Nutter
(s),
(it')
C.
Okada,
(s)
G.
Qsteria,
(t
)
O.
Palarnara
(to)
S
Parlati,
(
)
(7')
V.
Patera,
(6)
L.
Patrizii
(
)
B.
Pavesi,
(
)
R
Pazzi(
)
C
W.
Peck,
J.
Petrakis
(
)
(
7*)
S.
Petrera,
(
)
N
D.
Pignatano,
(
)
P.
Pistilli
(
)
J.
Reynoldson,
(7)
F.
Ronga,
(
)
G.
Sanzani,
(~)
A.
Sanzgiri
(
)
C.
Satriano
(
)
(
)
L.
Satta
(
)
(")
E.
Scapparone,
(
)
K.
Scholberg,
(
)
A.
Sciubba,
(is)
(")
P.
Serra
Lugaresi,
(
)
M.
Severi
(
)
M.
Sitta,
(~
)
P.
Spinelli
(
)
M.
Spinetti,
(6)
M.
Spurio,
(
)
J.
Steele,
(
)'(
*)
R.
Steinberg,
&
)
J.
L.
Stone,
(
)
L.
R.
Sulak,
(
)
A.
Surdo,
('
)
G.
Tarle,
(
)
V.
Valente,
(s)
C.
W.
Walter,
(
)
R.
Webb
(
)
and
W.
Worstell(
)
(MACRO
Collaboration)
'
Dipartimento
di
Fisica
dell'Universita
di
Bari,
Bari,
70186,
Italy
and
Istituto
Nazionale
di
Fisica
Nucleary
(INFN),
Bari,
70186,
Italy
Dipartimento
di
Fisica
dell'Universita
di
Bologna
and
INFN,
Bologna,
/01M,
Italy
Physics
Department,
Boston
University,
Boston,
Massachusetts
OM15
(
)
California
Institute
of
Technology,
Pasadena,
California
911g5
(
)Department
of
Physics,
Drexel
University,
Philadelphia,
Pennsylvania
1910$
Laboratori
Nazionali
di
Prascati
dell'INFN,
Prascati
(Roma),
000//,
Italy
Laboratori
Nazionali
del
Gran
Sasso
dell'INFN,
Assergi
(L'Aquila),
67010,
Italy
(
)Departments
of
Physics
and
of
Astronomy,
Indiana
University,
Bloonw
ngton,
Indi'ana
$7/05
Dipartimento
di
Fisica
dell'Universita
dell'Aquila
and
INFN,
I
'Aquila,
67100,
Italy
Dipartimento
di
Fisica
dell'Universita
di
Lecce
and
INFN,
Lecce,
78100,
Italy
)Department
of
Physics,
University
of
Michigan,
Ann
Arbor,
Michigan
$8109
Dipartimento
di
Fisica
dell'Universita
di
Napoli
and
INFN,
Napoli,
80125,
Italy
Dipartimento
di
Fisica
dell'Universita
di
Pisa
and
INFN,
Pisa,
56010,
Italy
Dipartimento
di
Fisica
dell'Universita
di
Roma
and
INFN,
Roma,
00185,
Italy
Physics
Department,
Texas
ASM
University,
College
Station,
Texas
776)8
Dipartimento
di
Fisica
dell'Universita
di
Torino
and
INFN,
Torino,
10125,
Italy
Bartol
Research
Institute,
University
of
Delaware,
Newark,
Delaware
19716
Sandia
National
Laboratory,
Albuquerque,
New
Mexico
87185
(Received
21
September
1993)
A
search
for
slowly
moving
magnetic
monopoles
in
the
cosmic
radiation
was
conducted
from
October
1989
to
November
1991
using
the
large
liquid
scintillator
detector
subsystem
of
the
first
supermodule
of
the
MACRO
detector
at
the
Gran
Sasso
underground
laboratory.
The
absence
of
candidates
established
an
upper
limit
on
the
monopole
Aux
of
5.
6
x
10
cm
sr
s
at
90%
confidence
level
in
the
velocity
range
of
10
~
P
(
4
x
10
.
This
result
places
a
new
constraint
on
the
abundance
of
monopoles
trapped
in
our
solar
system.
PACS
numbers:
14.
80.
Hv,
96.
40.
De,
98.
70.
Sa
Magnetic
monopoles
are
predicted
in
grand
unified
theories
(GUTs)
[1].
With
an
expected
mass
approx-
imately
2
orders
of
magnitude
greater
than
the
unifica-
tion
scale,
GUT
monopoles
could
not
have
been
produced
at
accelerators.
Produced
in
the
early
universe,
relic
monopoles
in
the
galaxy
are
expected
to
be
slowly
mov-
ing
with
velocities
comparable
to
typical
galactic
veloci-
ties
of
10
sc
[2].
The
survival
of
the
galactic
magnetic
field
requires
that
this
monopole
Aux
should
not
exceed
the
Parker
bound
of
10
tscm
2sr
is
[2].
Recently
608
0031-9007/94/72
(5)/608
(5)
$06.
00
1994
The
American
Physical
Society
VOLUME
72,
NUMBER
5
PHYSICAL
REVIEW
LETTERS
31
JANUARY
1994
an
extended
bound
of
1.
2
x
10
cm
sr
s
has
been
established
for
monopoles
of
mass
10i7GeVicz
by
con-
sidering
the
survival
of
a
small
galactic
seed
field
[3].
It
has
been
argued
[4,
5]
that
monopoles
may
be
trapped
in
the
solar
system
and
thus
their
local
flux
may
be
en-
hanced
above
the
Parker
bound.
We
concentrate
in
this
paper
on
a
search
for
these
trapped
monopoles
that
are
expected
to
travel
at
velocities
as
low
as
10
4c
(the
orbital
velocity
about
the
Sun
at
1
astronomical
unit).
The
large
kinetic
energy
of
GUT
monopoles
associ-
ated
with
their
large
mass
implies
that
they
are
highly
penetrating.
Therefore,
searches
for
GUT
monopoles
can
be
performed
using
underground
detectors
to
reduce
backgrounds
from
the
cosmic
radiation.
The
MACRO
(Monopole,
Astrophysics,
and
Cosmic
Ray
Observatory)
detector
[6],
a
large
underground
detector,
is
nearing
completion
at
the
Gran
Sasso
Laboratory
in
Italy,
with
the
primary
goal
of
searching
for
monopoles
at
flux
lev-
els
below
the
Parker
bound.
The
rock
overburden
has
a
minimum
thickness
of
3200
meters
of
water
equivalent
and
attenuates
the
fiux
of
cosmic
ray
muons
by
a
fac-
tor
of
about
10s.
When
completed,
MACRO
will
have
an
acceptance
of
10000
m~sr
for
an
isotropic
flux
of
penetrating
particles,
corresponding
to
about
three
monopoles
per
year
for
a
flux
level
at
the
Parker
bound.
The
experimental
data
reported
in
this
paper
uses
only
the
first
operational
supermodule
of
the
MACRO
detec-
tor,
which
is
documented
in
detail
elsewhere
[6,
7].
Its
dimensions
are
12.
6m
x
12m
x
4.
8m
and
its
acceptance
is
870
mzsr.
It
is
surrounded
on
five
sides
by
planes
of
liquid
scintillator
counters:
32
horizontal
counters
in
two
horizontal
planes
and
21
vertical
counters
in
three
vertical
planes.
Each
counter
is
an
11m
long
tank
of
liq-
uid
scintillator
viewed
by
20cm
diam
hemispherical
pho-
tomultiplier
tubes.
A
horizontal
counter
has
two
pho-
totubes
at
each
end,
while
a
vertical
counter
has
only
one
phototube
at
each
end.
In
addition,
there
are
ten
horizontal
planes
and
six
vertical
planes
(covering
three
vertical
sides)
of
limited
streamer
tubes,
one
horizontal
plane
of
plastic
track-etch
detectors
and
seven
horizontal
layers
of
passive
rock
absorber.
In
the
liquid
scintillator
subsystem,
we
have
employed
two
types
of
specialized
monopole
triggers,
which
cover
different
velocity
regions
[7].
In
this
Letter,
we
report
the
monopole
search
results
obtained
from
data
collected
using
the
trigger
which
covers
the
lower
velocity
region
[7]
and
has
also
been
applied
to
a
search
for
nuclearites
(strange
quark
matter)
[8].
Based
on
the
time
of
pas-
sage
of
slowly
moving
particles
through
the
19cm
thick
scintillator,
this
trigger
selects
wide
phototube
pulses
or
long
trains
of
single
photoelectron
pulses
generated
by
slowly
moving
particles,
and
rejects
with
high
eKciency
any
single
large
but
narrow
pulse
from
the
residual
pene-
trating
cosmic
ray
muons
or
from
radioactive
decay
prod-
ucts
from
walls
of
the
hall
or
detector
materials.
In
or-
der
to
pick
up
single
photoelectron
pulses,
whose
average
102
K
10
v
OC
o
1
O
I
I
I
I
IIIII
I
I
I
I
I
IIII
I
I
I
I
I
II4
0
HORIZONTAL
COUNTER
4
VERTICAL
COUNTER
10
10
I
I
I/I
I
IIII/
I
I
I
I
I
IIII
I
I
I
I
lllL
10
p=v/c
10
10
FIG.
1.
The
measured
slow
monopole
trigger
sensitivity
compared
with
the
expected
light
yields
of
monopoles
and
dyons.
The
probability
for
a
particle
with
light
yield
above
either
measured
curve
to
fire
the
corresponding
counter
is
greater
than
90%.
pulse
height
is
3mV,
the
front-end
discriminator
thresh-
old
is
set
at
only
2.
5
mV,
making
it
sensitive
to
electrical
noise
as
well
as
to
long
pulse
trains
of
low
light
level.
To
help
discriminate
between
them,
wave
forms
of
the
pho-
totube
signal
are
recorded
by
two
complementary
sets
of
wave
form
digitizers.
The
sensitivity
of
this
trigger
to
slow
monopoles
was
measured
by
simulating
the
expected
signals
using
light-
emitting
diodes
(LEDs)
in
representative
counters
[7].
Figure
1
shows
the
measured
amount
of
light
(normal-
ized
to
the
light
yield
of
minimum
ionizing
particles)
re-
quired
to
achieve
90'%%uo
trigger
efficiency
as
a
function
of
the
monopole
velocity.
Also
shown
are
the
light
yields
from
bare
monopoles
and
dyons
(monopoles
carrying
a
unit
electric
charge
or
monopole-proton
composites)
es-
timated
by
Ficenec
et
al.
based
on
the
best
fit
of
their
measured
scintillation
from
protons
with
velocities
as
low
as
2.
5
x
10
4c
[9].
As
Ficenec
et
aL
have
noted,
the
esti-
mate
for
dyons
is
more
certain
than
that
for
monopoles,
due
to
the
electric
charge
carried
by
dyons.
We
note
that
Kleber
[10]
has
shown
that
monopoles
will
induce
molec-
ular
ring
currents
when
passing
through
benzenelike
ring
molecules
in
organic
scintillator,
which
may
decay
opti-
cally
and
give
more
light
than
Ficenec
et
al.
's
estimate,
but
a
detailed
calculation
is
difficult.
The
data
were
collected
from
October
1989
to
Novem-
ber
1991
with
an
accumulated
live
time
of
542
days,
during
which
there
were
583999
events
with
the
slow
monopole
trigger
present
in
at
least
one
scintillator
plane.
After
vetoing
events
which
also
fired
a
two-plane
muon
trigger
requiring
time
of
fiight
(
1
ps
(corresponding
to
p
)
1.
5
x
10
z),
541918
events
remained.
The
majority
of
these
single
plane
monopole
triggers
were
due
to
ra-
dioactivity
pileups
(i.
e.
,
many
background
radioactivity-
induced
pulses
accidentally
occurring
within
a
short
time
609
VOLUME
72,
NUMBER
5
PH
YSICAL
REVI
E%
LETTERS
31
JANUARY
1994
interval).
Requiring
the
trigger
to
be
present
in
two
separate
planes
within
600ps
(the
time
of
flight
for
a
P
=
10
particle
to
cross
the
apparatus
with
the
longest
possible
path
length),
as
expected
for
slow
monopoles,
yielded
723
events,
some
of
which
were
caused
by
power
glitches
which
eventually
ended
the
run.
To
eliminate
these,
candidate
events
were
then
required
to
occur
at
least
0.
015
h
before
the
end
of
run.
This
end-of-run
cut
reduced
the
total
live
time
to
541
days
and
573
candidates
survived,
each
of
which
was
examined
and
classified
using
a
wave
form
analysis.
The
majority
of
these
candidates
(565
events)
were
easily
identified
as
due
to
electrical
noise
by
the
follow-
ing
characteristics:
the
presence
of
bipolar
oscillations
in
their
recorded
wave
forms
(388);
having
no
feature
other
than
occasional
isolated
radioactivity-induced
pulses
in
the
wave
form
(169),
interpreted
as
being
caused
by
elec-
trical
noise
on
the
trigger
input;
or
having
long
pulse
trains
()
4
ps)
simultaneously
present
in
every
channel
(8),
inconsistent
with
passage
of
particles.
The
remaining
eight
candidates
are
of
nonelectrical
ori-
gins.
Among
them,
two
candidates
were
identified
as
muons
because
of
their
time-of-flight
and
pulse
shapes;
they
escaped
the
fast
muon
veto
because
they
occurred
during
a
period
when
the
fast
muon
trigger
malfunc-
tioned.
Three
other
candidates
had
muon
signals
in
one
plane
and
radioactivity
pileups
in
the
other
plane.
The
remaining
three
candidates
had
wave
forms
con-
sisting
of
4
8
narrow
pulses
in
sequence,
where
each
pulse
typically
had
a
pulse
height
at
least
several
times
larger
than
the
average
single
photoelectron
pulse
height,
and
therefore
inconsistent
with
the
expectation
for
monopoles.
Instead,
these
events
are
consistent
with
being
due
to
accidental
coincidences
between
radioactiv-
ity
pileups
in
difFerent
planes,
for
which
the
expected
number
is
calculated
to
be
2.
6,
compared
to
the
three
that
were
observed.
We
note
that
for
the
passage
of
slow
particles,
the
photoelectrons
should
be
randomly
but
uniformly
distributed
to
produce
much
smoother
pulse
trains
than
the
observed
"spiky"
pulse
trains.
To
quan-
tify
this
analysis,
the
"spikiness"
of
a
pulse
train
has
been
represented
by
the
Campbell
quantity
o2/p,
,
where
y,
is
the
average
pulse
height
of
the
pulse
train
and
o
is
the
standard
deviation
of
the
pulse
height
about
its
mean.
Campbell's
theorem
[ll]
indicates
that,
if
the
time
dis-
tribution
of
the
pulses
making
up
the
train
is
Poissonian,
this
quantity
is
a
constant
independent
of
the
average
pulse
height,
and
thus
is
independent
of
the
light
yield
and
velocity
of
the
traversing
particle.
We
computed
this
quantity
for
the
three
candidate
events
as
well
as
for
monopolelike
pulse
trains
generated
by
LEDs
(Fig.
2).
For
selection
of
monopoles,
we
required
at
least
one
of
the
two
triggered
counters
to
have
at
least
one
end
satis-
fying
o
jp
(
2mV.
None
of
the
remaining
three
candi-
date
events
satisfies
this
criterion.
Under
this
criterion,
the
probability
of
rejecting
a
real
monopole
event
was
30
25
20
E
1
5
~
r
o
10
4J
5
I
1
I
I
l
I
I
I
I
l
I
I
l
t
l
l
l
I
t
l
I
l
I
I
I
~
MONOPOLE-LIKE
LEO
PULSES
I
0
GNQN)ATE
3
b,
CANOIATE
3
I
I
b.
I
0
~
I
0
I'
CI
~
I
~
~
'4
~
I
~
1TIE
l
I
I
Ill
l
I
i
i
)1f'i
i
»
l
0
5
10
15
20
25
END
0:
o'/p,
(mV)
I
0
I
I
30
FIG.
2.
The
Campbell
quantity
o
jp,
for
wave
forms
at
both
ends
of
a
scintillator
counter.
Each
of
the
three
candi-
date
events
has
two
entries
because
each
triggered
two
coun-
ters.
Most
of
the
881
monopolelike
LED
pulse
trains
are
clus-
tered
around
o
/p
=
1mV,
and
the
outiiers
are
caused
by
the
radioactivity-induced
pulses
superimposed
on
top
of
the
LED
pulse
trains
by
accidental
coincidences.
determined
to
be
5
x
10
from
Fig,
2,
while
the
proba-
bility
for
a
radioactivity
pileup
event
to
be
mistaken
for
a
monopole
was
determined
to
be
2.
2%
by
studying
single
plane
pileup
events.
Therefore,
the
expected
number
of
background
events
which
would
satisfy
this
cut
for
the
entire
data
set
was
0.
06.
As
a
final
cross-cheek,
we
have
looked
at
the
streamer
tubes
for
slow
particle
triggers
or
tracks
for
the
sample
of
the
573
candidates,
and
found
agreement
with
the
above
classifications
based
on
the
scintillator
wave
forms.
For
future
more
sensitive
searches,
the
streamer
tube
subsys-
tem
will
give
an
additional
strong
handle.
Furthermore,
we
note
that
if
any
candidate
events
survive,
we
can
per-
form
a
rigorous
inspection
of
the
tracks
in
the
track-etch
subsystem.
Rubakov
[12]
and
Callan
[13]
have
speculated
that
for
grand
unified
theories
which
do
not
conserve
baryon
number,
GUT
monopoles
may
strongly
catalyze
nu-
cleon
decay.
With
the
current
electronics
configura-
tion,
the
pulses
from
relativistic
decay
products
may
disrupt
the
proper
recording
of
wave
form
signals
of
slow
monopoles,
and
thus
this
search
may
be
insensi-
tive
to
those
monopoles
with
a
catalysis
cross
section
10
cm
.
However,
no
evidence
for
any
baryon
number-violating
process
has
ever
been
observed.
Fur-
thermore,
it
has
been
argued
that
the
catalysis
cross
sec-
tion
may
be
suppressed
as
compared
to
the
Rubakov-
Callan
prediction
[14],
that
the
Rubakov-Callan
eifeet
may
vanish
in
the
SU(5)
GUT
[15]
or
in
some
other
GUTs
[16],
and
that
proton
decay
does
not
occur
in
some
GUTs.
In
conclusion,
we
have
found
no
evidence
for
the
passage
of
a
slowly
moving
ionizing
particle
through
MACRO
and
have
established
an
upper
limit
on
the
isotropic
Aux
of
GUT
monopoles
at
5.
6
x
610