hep-ph/0305062

revised, June 2003

arXiv:hep-ph/0305062v2 29 Jun 2003

Destruction of Nuclear Bombs Using

Ultra-High Energy Neutrino Beam

— dedicated to Professor Masatoshi Koshiba —

Hirotaka Sugawara∗ Hiroyuki Hagura† Toshiya Sanami‡

Abstract

We discuss the possibility of utilizing the ultra-high energy neutrino beam (≃

1000 TeV) to detect and destroy the nuclear bombs wherever they are and whoever
possess them.

∗
e-mail:[email protected], Department of Physics and Astronomy, University of Hawaii at
Manoa, HI, USA
†
e-mail:[email protected], Radiation Science Center of KEK, Tsukuba-shi, Japan
‡
e-mail:[email protected], Radiation Science Center of KEK, Tsukuba-shi, Japan

1
1 Introduction
Twentieth century physicists produced one of the most powerful weapons on earth [1]
and they were used twice as an actual weapon with “Results Excellent.”1 The number of
countries which possess or will possess nuclear weapons could increase in spite of the existence
of Treaty on the Non-Proliferation of Nuclear Weapons (NPT). There is no guarantee that
these countries which already possess nuclear weapons always behave humanistically. Arms
control negotiations may stabilize the world temporarily but, again, there is no guarantee
that the long lasting peace on earth will come true in the future. We discuss in this article a
rather futuristic but not necessarily impossible technology which will expose the possessors
of nuclear weapons in an extreme danger in some cases.
Our basic idea is to use an extremely high energy neutrino beam which penetrates the
earth and interacts just a few meters away from a potentially concealed nuclear weapon. The
appropriate energy turns out to be about 1000 TeV. This is the energy where the neutrino
mean free path becomes approximately equal to the diameter of the earth. The neutrino
beam produces a hadron shower and the shower hits the plutonium or the uranium in the
bomb and causes fission reactions. These reactions will heat up the bomb and either melt it
down or ignite the nuclear reactions if the explosives already surround the plutonium. We
will calculate the intensity of the neutrino beam required and the duration of time which
the whole process will take place for a given intensity.
We emphasize that the whole technology is futuristic and the reason should be clear to
all the accelerator experts. Actually, even the simplest prototype of our proposal, i.e. the
neutrino factory of GeV range needs substantial R & D work. We also note that a 1000 TeV
machine requires the accelerator circumference of the order of 1000 km with the magnets
of ≃ 10 Tesla which is totally ridiculous. Only if we can invent a magnet which can reach
almost one order of magnitude higher field than the currently available magnet, the proposal
can approach the reality. Even if it becomes the reality, the cost of the construction is of
the order of or more than 100 billion US$. Also we note that the power required for the
operation of the machine may exceed 50 GW taking the efficiency into account. This is above
the total power of Great Britain. This implies that no single country will be able to afford
the construction of this machine and also the operation time must be strictly restricted. We
believe the only way this machine may be built is when all the countries on earth agree to
do it by creating an organization which may be called the “World Government” for which
this device becomes the means of enforcement.
Section 2 gives a rough estimation and section 3 deals with a computer simulation using
1
A coded message sent to President Harry S. Truman from Richard Nelson, the youngest crewman of
Enola Gay, who died February 7, 2003 of complication of emphysema (New York Times, February 7, 2003).

2
a Monte-Carlo generator MCNPX [2]. Section 4 gives a conclusion with various comments.
In addition, we give the calculation of the mean free path of neutrino inside the earth in
appendix A. In appendix B, we also describe a possible accelerator scheme.

2 Rough estimation
The neutrino beam is already hazardous at the energy of several TeV as have been
analyzed by B. J. King [3] and N. V. Mokhov and A. Van Ginneken [4] in connection with
the study of the muon colliders. If one constructs a race track shaped muon storage ring
as shown in fig. 1, most of the muons decay into the two opposite directions of the straight
sections. These two directions are the most hazardous directions (“hot spots”) but the
circular parts also emit the neutrinos which may also be hazardous in the vicinity of the
storage ring.

µ
neutrino radiation “hot spot”

muon storage ring

neutrino radiation “hot spot”

µ

straight section
Figure 1: Neutrino radiation from a race track shaped muon storage ring. The decay of
muons will produce the neutrino radiation emanating out tangentially everywhere from the
ring. In particular, the straight sections in the ring will cause radiation “hot spots” where
all of the decays line up into a pencil beam.

Here we would like to consider a situation where one of the straight lines is directed
toward the nuclear bomb which is located somewhere on the opposite side of the earth
(fig. 2). We must choose the energy of the neutrino beam in such a way that the mean
free path of the neutrino is compatible to the diameter of the earth. Fig. 3 shows the mean
free path of (anti-)neutrino vs. its energy calculated assuming that the deep inelastic cross
section dominates in the relevant energy region2 . From fig. 3 we conclude that the energy of
2
See appendix A for the detailed calculations of the mean free paths shown in fig. 3.

3
 nuclear bomb
muon storage ring

ν
neutrino beam

inside of the earth

Figure 2: Neutrino beam is aimed at the nuclear bomb that is placed on the opposite side
of the earth. The beam is emitted downstream from one of the straight sections of the muon
storage ring (see fig. 1), and reaches the bomb after passing through the inside of the earth.

the neutrino beam must be about 1000 TeV to have approximately single interaction before
the neutrino beam hits the bomb. The size of the beam at the point of the bomb is given by

mµ c2 0.1 (GeV) × 107 (m)

r= d ≃ = 1 (m) ,
Eν 106 (GeV)

where mµ and c stand for the muon mass and the speed of light, and d is the distance from
the muon storage ring to the position of the bomb which we take to be the diameter of
the earth (≃ 107 m). The beam spread due to the transverse momentum of the beam is
negligible at this energy if the current value of the ionization cooling of Pt = 1 (MeV) is
adopted. The range of the neutrino is 107 meters and the effective neutrino interaction is
restricted within a few meters away from the bomb because of the interaction range of the
hadrons. Therefore, the probability of getting an effective reaction from the beam is 1/107 .
As a result, the energy deposit from the beam for the unit area (m2 ) is given by

Edep = 1015 × 10−7 × I = 108 I (eV/sec · m2 ) , (1)

where I stands for the neutrino beam intensity. For example, we get

Edep ≃ 1000 (Joule/sec · m2 ) for the intensity of I = 1014 (1/sec) . (2)

This is equivalent to about 1 SV /sec. We note that this value of the radiation dose is very
large, compared with the U.S. Federal off-site limit of 1 mSV /year.3
The above estimation can be summarized in the following formula:
!−2
Rh mµ c2
 
Edep = Eν · ·I · d (eV/sec · m2 ) , (3)
Rν Eν
3
The unit SV corresponds, in alternative units, to 100 rem.

4
 8
10

mean free path of ν̄µ

10
7
mean free path of νµ

Mean free path [km] 6

10

5
10

4
10

3
10 −1 0 1 2 3
10 10 10 10 10
Energy of (anti−)neutrino [TeV]
Figure 3: Mean free path of (anti-)neutrino vs. its energy. This is calculated under the
assumption that in this energy region the deep inelastic cross sections dominate. For the
detail of the calculation, see Appendix A.

where Rh and Rν stand for the average hadron and the neutrino mean free paths. Rν is
proportional to Eν−1 below 1000 TeV corresponding to σνtot (cm2 ) ∼ 10−38 Eν (GeV) which
leads to

Edep ∼ Eν4 .

Edep drops sharply to 0.1 Joule/s for 100 TeV neutrino energy. It is, therefore, rather crucial
to keep the energy as high as 1000 TeV.
To proceed further we need to know a little about the structure of the nuclear bomb.
Since no official information is available to us, we rely on popular books [5, 6] and unclassified
papers [7] on the subject. As a possible model for the bomb we consider a 10 kg ball of
239 238
Pu which has the critical mass of 15 kg, surrounded by the U tamper, the reflector
and the explosive material (fig. 4). We also consider a system without explosive material
surrounding the plutonium ball since we have no way to know how these bombs are stored. A
crucial parameter in the former case is the number of fissions in the system which provides
the temperature rise enough to ignite the surrounding explosives. If we assume that the
explosive has the ignition temperature of 300 ◦ C (TNT (TriNitroToluene) has its ignition
temperature 210 ◦ C), the number of fissions, Nfission , required turns out to be about 1016 per
10 kg of plutonium [7]. One order of magnitude larger value 1017 should be enough to melt
down the system in the latter case.

5
 explosive A reflector

uranium tamper

initiator
plutonium core

trigger device

explosive B

Figure 4: A model for the plutonium bomb of implosion type [6]. The whole profile of the
bomb is in the shape of a spherical body.

Let us estimate how much time it takes for this process to happen when the energy
238
deposit from the neutrino beam is given by eq. (1). Since the tamper U can also be
regarded as the source of the fission when the neutron energy is as high as 10 MeV, which
is the typical energy, ǫsp , of the spallation neutrons, we take the area exposed to the hadron
shower to be 0.1 (m2 ). In this case the total energy deposit in the bomb fission system is
T
Edep = 0.1 × 108 I = 1021 (eV/sec) for I = 1014 (1/sec) .

The number of spallation neutrons, therefore, is equal to

T
Edep 1021
nsp = = = 1014 (1/sec) .
ǫsp 107
If we assume a spallation neutron causes a single fission and none are lost, the time required
to ignite the explosive is
Nfission 1016
= 14 = 100 (sec) ,
nsp 10
and the melt down time is about 1000 sec. The next question is whether we are going to
have a full fledged explosion or a kind of “fizzle explosion” in the case when the explosive sets
ignited. The problem was studied in the appendix of ref. [7] by F. von Hippel and E. Lyman.
This analysis shows that the neutron from the spontaneous fission which contaminates the
239
Pu system gives the probability of the occurrence of “fizzle explosion” given by

P = 1 − e−cnspon , (4)

6
where nspon is the number of neutrons from the spontaneous emission and c is a certain
constant. We can replace this number nspon by the number of spallation neutrons caused
by the neutrino beam, which is 9 order of magnitude larger than the number of neutrons
240
from the spontaneous emission of Pu. Therefore, the probability of the “fizzle explosion”
is practically equal to 1 in this case. This results in an energy yield of the explosion by the
neutrino beam to be about 3% of the full explosion.

3 Simulation
Having discussed the rough estimation, let us now turn to numerical simulations to study
the system in a more precise way. We divide our simulation procedure into two parts: The
first part deals with a neutrino beam and its development into a hadron shower (see fig. 5).
The second one follows the first one to calculate the nuclear reactions in the target bomb (see
fig. 6). We give detailed discussions of the two parts separately in the subsequent sections 3.1
and 3.2.

neutrino beam

interaction point

soil ≤ 100 m

hadron shower

surface of the earth

Figure 5: Hadron shower arising near the target bomb. The neutrino beam passing through
the soil interacts with nuclei near the surface of the earth, resulting in a hadron shower in a
place a few meters close to the bomb.

3.1 Incident neutrino beam and hadron shower

The first part is to start from a given neutrino beam of certain energy and intensity. We
simulate the process of the neutrino beam hitting a target nucleus in the soil and follow the
development of a hadron shower initiated by the neutrino interaction, as shown in fig. 5.

7
The former process can be simulated by a generator HERWIG [8], in which we can include
processes with the incident neutrino beam, such as ν + p (or n) → hadrons + leptons. In the
latter process, subsequently, we simulate the interactions of the hadron shower with nuclei
of the soil by using another generatorS GEANT4 [9] and MARS [10]. The purpose of this
part is to obtain the multiplicity of the shower when the shower is going out of the earth.
The neutrino interaction which occurs near the surface of the earth is relevant. We consider,
therefore, a system which is shown in fig. 5.
The result of this part will appear in a separate publication [12].

3.2 Nuclear reactions inside the target

The second part of our simulation is to calculate the temperature increase of the plu-
tonium system caused by the hadron shower. We consider a system shown in fig. 6. Our
calculation of this part is carried out using the MCNPX code4 .

hadron shower

239
Pu

238
tamper U

explosive

Figure 6: A hadron shower going into the plutonium bomb. It will induce the fission
reactions inside the plutonium system and cause the temperature increase as a result.

In order to simplify the system and save the computation time, we replace the parallel
hadron shower of fig. 6 by a neutron source which is situated at the center of the 239 Pu core,
as shown in fig. 7. We also assume that the incident neutron of its energy 1 GeV starts from
a point inside the core.
To make clear the algorithm of the MCNPX code, we illustrate a simple example of
nuclear reactions in fig. 7, where the first collision occurs at event 1 in the Pu core. The
4
MCNPX has been used extensively in nuclear reactor physics and its applications, and developed for
a long time since 1940’s [2]. Furthermore, many physicists and programmers are still developing it for
improvement even at present.

8
 6
5
2 3

4
incident
1 neutron

239
Pu
track of neutron
7 track of photon
238
U

Figure 7: A history of a neutron incident on the 239 Pu core that can undergo nuclear fission.
1. Neutron scattering and photon production in the core. 2. Fission and photon production
in the 238 U tamper. 3. Neutron capture in the tamper. 4. Neutron leakage out of the tamper.
5. Photon scattering in the tamper. 6. Photon leakage out of the tamper. 7. Photon capture
in the tamper.

neutron is scattered in the direction shown, which is selected randomly from the physical
scattering distribution. A photon is also produced and is temporarily stored, or banked, for
later analysis. At event 2, fission occurs, resulting in the termination of the incoming neutron
and the birth of two outgoing neutrons and one photon. One neutron and the photon are
banked for later analysis. The first fission is captured at event 3 and terminated. The banked
neutron is now retrieved and, by random sampling, leaks out of the core at event 4. The
fission-produced photon has a collision at event 5 and leaks out at event 6. The remaining
photon generated at event 1 is now followed with a capture at event 7. Note that MCNPX
retrieved banked particles such that the last particle stored in the bank is the first particle
taken out.
This neutron history in the core and the tamper is now complete. As more and more
such histories are followed, the neutron and photon distributions become better known. The
quantities of interest, such as the total energy arising in the reactions, are tallied along
with estimates of the statistical precision of the results. Hence, after repeating the similar
calculations, we could obtain the average value, ǫfission , of the fission energy deposition:

ǫfission = 0.6260 ± 0.0032 (MeV/g) .

This is the contribution on the average from one incident neutron. Thus, if we prepare Nin

9
 239
neutrons incident on the Pu core, the increase in temperature can be estimated as
Nin ǫfission
∆T = = 0.9547 × 10−12 Nin (K) , (5)
CPu
239
where CPu is the specific heat of Pu, whose numerical value is given by

4.186 (J/cal) × (6.0/239) (cal/g · K)

CPu = = 6.557 × 1011 (MeV/g · K) .
1.602 × 10 −13
(J/MeV)

Therefore, in order to obtain a temperature increase ∆T = 250 (K), which corresponds to

the ignition temperature of TNT, the total number of the incident neutrons should be
250
Nin = = 2.619 × 1014 . (6)
0.9547 × 10−12
This value of Nin ∼ O(1014) is close to the estimated value in section 2, and it is not
unrealistic in the future technology of the muon colliders.
To verify our results of simulation, the experiment can be devised in which a hadron beam
of GeV energy hits a ball of plutonium (not necessarily of weapon grade) and increases its
temperature. The experiment is similar to the one which is performed to study the target
material for the neutron spallation sources.

4 Conclusion
We have shown that it is possible to eliminate the nuclear bombs from the surface of the
earth utilizing the extremely high energy neutrino beam. When the neutrino beam hits a
bomb, it will cause the fizzle explosion with 3% of the full strength. It seems that it is not
possible to decrease the magnitude of the explosion smaller than this number at this stage.
It is important to decrease this number to destroy bombs safely. We are not sure what this
means when the plutonium or uranium is used to ignite the hydrogen bomb. We may just
break the bomb or may lead to a full explosion. The whole process takes a matter of a few
minutes in the case considered in this paper although, of course, it depends on the intensity
of the neutrino beam. When the bombs are stored in the form of plutonium ball separated
from the explosives, what we can do is to melt them down or vapor them away. It takes
substantially longer time for this process to occur.
To justify the above statements we performed a detailed simulation calculation and the
part of its results is explained in this paper although the full content will be published later.
After the high energy neutrino beam passes through the soil, it causes a hadron shower
239
near the surface of the earth, and subsequently, neutrons in the shower will strike the Pu
core. In order to estimate the number of the incident neutrons which is large enough to

10
make the temperature of the TNT surrounding the core increase to its ignition one, we have
carried out the numerical simulations using MCNPX under the simplified conditions. As
a consequence, we obtained the value of Nin ∼ O(1014 ). This value is consistent with the
estimation obtained roughly, and it is expected to be realistic in the future technology.
We utilize 1000 TeV5 neutrino beam for our purposes and we do not have the technology
yet to produce such a high energy neutrino. We may start an R & D now and proceed step
by step. Yet it may take even an order of a century to achieve the goal.

We describe below a possible scenario for the whole project:

(1) First, we should construct a neutrino factory which could have substantially lower
energy than even 1 TeV. The purpose is to fully understand the properties of the
neutrino including mass, mixing angles, CP violating phase, Majorana property and
the interactions with other particles [15].
(2) The next step is to construct a muon collider in the multi-TeV energy range. The
energy should be beyond 10 TeV. These two steps still require a fair amount of R & D
but we believe that they are on the straightforward extension of the currently available
technology.
(3) The third step is to construct a muon collider of more than 100 TeV energy and to
reach even 1000 TeV. We do not have the technology yet for this kind of machine. We
may need a magnet with one to two orders of magnitude higher field than the currently
available one to construct a machine of a reasonable scale. A completely new approach
may be necessary to make this possible.
(4) If the third step becomes real, then the fourth step is to actually build the 1000 TeV
muon collider with the movable straight sections.

We want to emphasize the importance of realizing the first two steps since the technology
is within the reach and its contribution to the basic science is enormous. The study may
even show that the neutrino interaction increases more rapidly with the energy owing to the
extra-dimension as in some models [16]. In that case we need not go as far as 1000 TeV to
achieve our goal. The last two steps require tremendous amount of effort in developing the
5
Actually, the neutrino mean free path which leads us to consider 1000 TeV is not quite accurate. As is
mentioned in Appendix A, we did not include the contribution of the heavier quarks. In addition, we did
not take into consideration the “neutrino transport theory” [11] at all here. The inclusion of there effects
on the deep inelastic cross-sections will lead to the mean free path which is almost 1/3 of the value we have
used in this paper. This change will lead to the energy of 300 TeV and, therefore, we need 27 times higher
intensity than considered in section 2. Of course, targeting the bomb becomes much easier.

11
necessary technology. It is also true that fair amount of financial and human resources will
have to be introduced to accomplish the last two steps.
The neutrino beam could also be used to detect the nuclear bombs with much less energy
and with much less intensity. The necessary technology is the detection of the fission products
from a reasonable distance. It could be rather difficult if the bombs are stored in a deep
underground location.
Another useful application of high energy neutrino beam is to the study of the inner
structure of the earth [17]. We may not need the neutrino energy to be as high as 1000 TeV
in this case. The detailed study is being performed on this subject and we will describe it
in a forthcoming publication.
We are certainly aware of the fact that this kind of device can not only target the nuclear
bombs but other kinds of weapons of mass destruction and also, unfortunately, any kind
of living object including human. But we should emphasize that the device itself is not
a weapon of mass destruction. The reason is as follows: The calculation in section 2 and
section 3 shows that it takes 1 second for this device to cover a 1 m2 with the radiation dose
of 1 SV . It takes more than a year to cover the area of 10 km2 with this value of dose per
unit area. It is extremely unlikely that no measure is taken after a few minutes of exposure
of this kind. Moreover, as is emphasized in the introduction and also in the appendix B,
the construction cost and the power required for the operation make it almost impossible
for even the richest country to build and operate it all by itself. We strongly object to the
ungrounded worry that this kind of device, even its downgraded version could be used by
certain irresponsible organization as a weapon of mass destruction. On the contrary, we
sincerely hope that our proposal will motivate and stimulate the revival of the old idea of
“World Government” which has so far been discarded as unrealistic.
Lastly we would like to point out that at least the first two steps described in this
section have nothing to do with the weapon research. They belong to the most fundamental
scientific research activities. The suitable organizational structure to perform such a research,
therefore, is through the world-wide collaboration. Another worry could be expressed on the
neutrino hazard around the machine. It depends crucially where one builds the machine.
The concrete proposal explained in appendix B has two hazardous planes and two dangerous
(P3 P4 and Q3 Q4 ) directions. ”No fly zones” should be set to avoid these hazardous regions.
The duration time of an operation should be minimized for the security reason and also for
the reason of power consumption.

12
Appendices

A Mean free path of (anti-)neutrino

This section is devoted to the derivation of the mean free path of (anti-)neutrino passing
through the inside of the earth. T. Abe, who is on the staff of Accelerator Laboratory of
KEK, has mainly contributed to the following calculations and actually made fig. 3 under
some assumptions. Here we discuss the assumptions in some detail and further derive the
mean free paths in a numerical way.
First, we assume that the deep inelastic scattering, including the contributions of both
the charged and neutral currents, dominates in the energy region multi TeV ∼ 103 TeV; the
relevant cross sections are given at the tree level [13] as
!2 (
d2 σcc
ν
G2 mN Eν x 2
)
MW 
¯

= F Zd(x) + Nu(x) + (1 − y) 2
Z ū(x) + N d(x) , (7)
dxdy π −q 2 + MW
2
!2 (
d2 σcc
ν̄
G2 mN Eν x 2
)
MW ¯ + N ū(x) + (1 − y)2 Zu(x) + Nd(x)
 
= F Z d(x) , (8)
dxdy π −q 2 + MW
2
!2
d2 σnc
ν
G2 mN Eν x MZ2
= F
dxdy π −q 2 + MZ2
" ( 2  2 
2 2  1 
× (1 − y)2 sin θW Zu(x) + Nd(x) + sin2 θW Zd(x) + Nu(x)
3 3
2  2  )
1 2 2 1 1
 
¯ ¯ + N ū(x)
 
+ − + sin θW Z ū(x) + N d(x) + − sin2 θW Z d(x)
2 3 2 3
( 2  2 
2 2 ¯
 1
¯ + N ū(x)

+ sin θW Z ū(x) + N d(x) + sin2 θW Z d(x)
3 3
2  2  )#
1 2 2   1 1 2
 
+ − sin θW Zu(x) + Nd(x) + − + sin θW Zd(x) + Nu(x) , (9)
2 3 2 3
d2 σnc
ν̄  
= exchange the factor (1 − y)2 with the factor 1 in the above expression (9) , (10)
dxdy
where GF , θW , mN and Eν stand for the Fermi coupling constant, the Weinberg angle, nu-
cleon mass and incident neutrino energy. u(x) and d(x) are the parton distribution densities
for u and d quarks. Z and N stand for the proton number and the neutron number inside
a given nuclei, and q denotes the momentum transfer for the processes considered. Here the
variables x and y are defined as

−q 2 (Eν − Eµ )
x= , y= ,
2mN (Eν − Eµ ) Eν

where Eµ stands for the energy of the muon in the final state. In eqs. (7) and (8) we take

13
the following scattering processes into consideration at the tree level:

νµ + d → µ− + u , νµ + ū → µ− + d¯,
ν̄µ + u → µ+ + d , ν̄µ + d¯ → µ+ + ū ,

Here we ignored the contribution of strange quark for simplicity. Similarly, in eqs. (9) and
(10) we include the following neutral current scatterings:

νµ + q → νµ + q , νµ + q̄ → νµ + q̄ ,
ν̄µ + q → ν̄µ + q , ν̄µ + q̄ → ν̄µ + q̄ ,

where q = u or d, and again we do not include the s quark contribution at all.

Second, we adopt a set of parton distributions called CTEQ5L [14] as the structure func-
tot tot tot
tions of proton and neutron. Using the set, we calculate the total cross sections σνp , σνn , σν̄p
tot
and σν̄n for the νµ p, νµ n, ν̄µ p and ν̄µ n scattering processes, respectively. Actually, we have
carried out these calculations by the Monte-Carlo method with a high precision of 0.1 %.
Finally, we obtain the mean free path of neutrino Rν and that of anti-neutrino Rν̄ in a
usual manner. For the purpose we first must know the number density of protons, Np , inside
the earth and that of neutrons, Nn . Indeed, we can easily obtain these densities, because
we know that the number of protons and that of neutrons are on the average in the ratio
of 49.5 to 50.5. In addition, the average mass density of the earth, ρearth , is measured to be
5.52 × 103 (kg/m3 ). Hence, one can obtain the number densities:
ρearth 49.5
Np = × (1/m3 ) ,
mp 49.5 + 50.5
ρearth 50.5
Nn = × (1/m3 ) ,
mn 49.5 + 50.5
where mp and mn stand for the proton and the neutron masses. Thus, we obtain the mean
free path of (anti-) neutrino
1
Rν = tot tot (m) , (11)
Np σνp + Nn σνn
1
Rν̄ = tot tot (m) . (12)
Np σν̄p + Nn σν̄n
Throughout the process of the above calculations, we did not include higher order cor-
rections at all. However, it is plausible that if we include s and other heavier quarks into the
parton distributions, the total cross sections will probably become a few times larger than
those obtained here. Furthermore, QCD corrections and the “regeneration processes” [11]
will also have other effects on the cross sections. Our next step is to take these effects into
consideration on the basis of a detailed Monte-Carlo study, which remains to be solved in
the near future.

14
B Possible accelerator scheme
We first look for a mountain like in fig. 8 whose surface does not touch many of the
straight lines depicted as P1 P2 , P3 P4 , Q1 Q2 or Q3 Q4 . We construct two synchrotron A and
B which are both revolvable. A should be larger than B. Muon beam is injected into the
synchrotron A first and accelerated to a sufficient energy. Injection system could be installed
in a tunnel in the mountain. Then it is stored either in the path P2 P3 P4 P1 P2 or Q2 Q3 Q4 Q1 Q2
depending on the direction of the beam in the synchrotron A. The beam is either µ+ or
µ− . The straight sections P1 P2 , P3 P4 , Q1 Q2 and Q3 Q4 are made of chambers separated by
many bellow structures so that they can have a flexible length. We probably have to prepare
several chambers to cover from the minimum to the maximum length continuously. When
we rotate A or B the chambers must follow until we steer the straight section to a given
target.The next question is how precisely we can steer it. From the discussion given in the
text the required accuracy is 10−7 . This is 1/10 micron per meter. We believe this is not an
outrageous number. The current effort toward the construction of a linear collider is aiming
at approximately 1 micron per meter. Future technology certainly will reach our required
number sooner or later.

P4 Q4
synchrotron B hazardous plane
P1 Q1

Q3 P3
synchrotron A hazardous plane
Q2 P2
injection system

Figure 8: Accelerator scheme.

Another issue is the power consumption and the radiation hazard. Power required is
1014 × 10−19 × 1015 W ≃ 10 GW. Actually, we may need something like 50 GW (considering
the efficiency) which is exactly the whole capacity of Japanese nuclear power. But the energy
consumption could be as small as 102 /108 = 10−6 times the whole consumption of 50 GW
power. This should be quite tolerable. For the radiation hazard we have two planes in fig. 8

15
which should not be crossed by anyone during the operation and one direction toward the
sky where no one is allowed to touch. The other direction is, of course, toward the target.
Almost all the energy is lost in the earth and only 10−7 times the whole energy hits the
target. People working near the target should be warned unless they are working to conceal
the weapons.
We can perform µ+ µ+ , µ− µ− and µ+ µ− colliding experiment in this scheme by injecting
two beams simultaneously although the detector should be placed on a very steep slope
between A and B synchrotrons. We believe it is not unreasonable to build this kind of
accelerator complex first with much lower energy beam to study the inside of the earth and
simultaneously performing the muon collider experiments and also the neutrino experiments.

Acknowledgment
We are grateful to Tetsuo Abe for giving us fig. 3 of the mean free paths. We thank
Masayoshi Kawai for his useful comments on MCNPX. We appreciate Yoshinobu Takaiwa
for his helpful suggestion of some Monte-Carlo generators. We would like to acknowledge
valuable discussions with Prof. Tokushi Shibata and members of Theory Division of KEK.
The questions and comments by Yoshitaka Kimura, Sakuei Yamada, Frank von Hippel,
Sydney Drell, Burton Richter and colleagues of H.S. at the University of Hawaii are deeply
appreciated.

References
[1] Peter Goodchild, J. Robert Oppenheimer: ‘Shatterer of Worlds’, New York: Fromm
International Publishing Corp. (1985).
[2] Laurie S. Waters, editor, MCNPXTM User’s Manual Version 2.3.0, April 2002,
LA-UR-02-2607.
Judith F. Briesmeister, editor, MCNPTM – A General Monte Carlo N-Particle Transport
Code Version 4B, March 1997, LA-12625-M, UC 705 and UC 700.
[3] Bruce J. King, “Potential Hazardous from Neutrino Radiation at Muon Colliders”, avail-
able form LANL preprint archive as physics/9908017; “Neutrino Radiation Challenges
and Proposed Solutions for Many-TeV Muon Colliders”, in Proc. HECM’99 Workshop
– Studies on Colliders and Collider Physics at the Highest Energies: Muon Colliders at
10 TeV to 100 TeV, Montauk, NY, September 27-October. 1, 1999.

16
 [4] N. V. Mokhov and A. Van Ginneken, “Neutrino-Induced Radiation at Muon Colliders”,
FERMILAB-Conf-99/067, at the 1999 Particle Accelerator Conference, New York, NY,
March 9-April 2, 1999.
[5] R. Server, The Los Alamos Primer, University of California Press (1992).
[6] Katsuya Yamada, Genshi Bakudan – Sono Riron to Rekishi (in Japanese), Bluebacks B-
1128, Kodansha, Tokyo (1996).
Richard Rhodes, The Making of The Atomic Bomb, Simon & Scuster, New York (1988).
Paul P. Craig and John A. Jungerman, Nuclear Arms Race, Technology and Society,
McGraw-Hill Book Company, New York (1986).
Hoddeson, Henriksen, Meade and Westfall, Critical Assembly, A technical History of Los
Alamos during the Oppenheimer years 1943-1945, Cambridge University Press (1993).
[7] J. Carson Mark, “Explosive Properties of Reactor-Grade Plutonium”, Science & Global
Security Vol. 4 (1993) 111-128.
[8] G. Corcella, I. G. Knowles, G. Marchesini, S. Moretti, K. Odagiri, P. Richardson,
M. H. Seymour and B. R. Webber, “HERWIG 6: an event generator for Hadron
Emission Reactions With Interfering Gluons (including supersymmetric processes)”,
Cavendish-HEP-99/03, CERN-TH/2000-284, RAL-TR-2000-048.
[9] Source codes of GEANT4 are available from the following site:
http://wwwinfo.cern.ch/asd/geant4/source/source.html
and also see references therein.
[10] Information on the generator MARS is available from the following site:
http://www-ap.fnal.gov/MARS/
and also see references therein.
[11] Vadim A. Naumov and Lorenzo Perrone, “Neutrino Propagation Through Matter”,
Astropart. Phys. 10 (1999) 239, hep-ph/9804301.
[12] Our calculation is now in progress on the basis of the algorithm adopted in the generator
HERWIG or MARS. In addition, we must take the neutrino transport theory into
consideration. The result will be reported in the near future.
[13] Ian J. R. Aichison and Anthony J. G. Hey, Gauge Theories in Particle Physics – A
Practical Introduction, the University of Sussex Press (1982), and references therein.
[14] H. L. Lai et al. [CTEQ Collaboration], “Global QCD analysis of parton structure of the
nucleon: CTEQ5 parton distributions”, Eur. Phys. J. C12 (2000) 375.
[15] For a detailed review of neutrino physics, see the following reference,
M. Fukugita and A. Suzuki, Physics and Astrophysics of Neutrinos, Springer-Verlag,
Tokyo (1994).

17
[16] J. Kubo, H. Terao and G. Zoupanos, “Running couplings in extra dimensions”, talk
given at 30th International Conference on High-Energy Physics (ICHEP 2000), Osaka,
Japan, 27 July-2 August 2000, hep-ph/0010069.
[17] A. De Rújula, S. L. Glashow, R. R. Wilson and G. Charpak, “Neutrino Exploration of
the Earth”, Phys. Rep. 99, No. 6 (1983) 341-396.
[18] This is a modified version of “Flying Linac” which was discussed many years ago. See:
Edward A. Heighway (LANL), ”An Accelerator Technology Legacy”, Proceedings of the
1994 International Linac Conference, pp. 664 and the references therein.
[19] There are studies on the shock wave inside the mercury target. The study must be
extended to the solid target for our proposal. See the following references:
L. J. Briggs, J. Appl. Phys. 24-4 (1953) 488.
S. Ishikura, H. Kogawa, M. Futakawa, K. Kikuchi, R. Hiro and C. Arakawa, “Bubble
dynamics in the thermal shock problem of the liquid metal target”, in Proceedings of
the Fifth International Workshop on Spallation Materials Technology, Special volume
of the Journal of Nuclear Materials (2003).

18