U.S. patent number 3,925,676 [Application Number 05/493,415] was granted by the patent office on 1975-12-09 for superconducting cyclotron neutron source for therapy.
This patent grant is currently assigned to Atomic Energy of Canada Limited. Invention is credited to Clifford B. Bigham, Harvey R. Schneider.
United States Patent |
3,925,676 |
Bigham , et al. |
December 9, 1975 |
Superconducting cyclotron neutron source for therapy
Abstract
A neutron source for medical therapy purposes comprising a
cyclotron comprising an iron metal housing acting as a magnetic
yoke, magnetic shield, radiation shield, and vacuum vessel, a pair
of superconducting coils mounted in a cavity in the housing said
coils being cooled to superconducting temperatures by passage of a
refrigerant fluid therethrough and connected to an electrical
energy source such that a high current flows in the coils producing
an intense magnetic field inwardly between the coils, an ion
orbiting region defined by pairs of sectoral-shaped RF electrode
structures energized at an RF frequency and focussing flutter poles
mounted in the intense magnetic field between coils, a source of
ions positioned centrally of the ion orbiting region to provide a
stream of ions that will be orbited in the orbit region; an ion
target positioned internally of said iron housing at an outer
position in the orbit region such that orbiting ions strike the
target and produce neutrons; a channel formed in the iron housing
from the target to the exterior for passage of the beam of neutrons
formed at the target, said channel acting as a beam collimator; and
a mounting structure for movably mounting the cyclotron and target
such that the neutron beam produced can be employed at more than
one position.
Inventors: |
Bigham; Clifford B. (Deep
River, CA), Schneider; Harvey R. (Deep River,
CA) |
Assignee: |
Atomic Energy of Canada Limited
(Ottawa, CA)
|
Family
ID: |
23960140 |
Appl.
No.: |
05/493,415 |
Filed: |
July 31, 1974 |
Current U.S.
Class: |
376/112; 313/62;
315/502 |
Current CPC
Class: |
A61N
5/10 (20130101); H05H 3/06 (20130101); H05H
13/00 (20130101); A61N 2005/109 (20130101) |
Current International
Class: |
A61N
5/10 (20060101); H05H 3/00 (20060101); H05H
13/00 (20060101); H05H 3/06 (20060101); G21G
004/02 () |
Field of
Search: |
;250/499-502
;313/61R,61S,62 |
References Cited
[Referenced By]
U.S. Patent Documents
Primary Examiner: Willis; Davis L.
Attorney, Agent or Firm: Hughes; James R.
Claims
We claim:
1. A neutron source for medical therapy purposes comprising:
a. a cyclotron comprising an iron metal housing acting as a
magnetic yoke, magnetic shield, radiation shield, and vacuum
vessel, a pair of superconducting coils mounted in a cavity in the
housing said coils being cooled to superconducting temperatures by
passage of a refrigerant fluid therethrough and connected to an
electrical energy source such that a high current flows in the
coils producing an intense magnetic field inwardly between the
coils, an ion orbiting region defined by pairs of sectoral-shaped
RF electrode structures energized at an Rf frequency and focussing
flutter poles mounted in the intense magnetic field between coils,
a source of ions positioned centrally of the ion orbiting region to
provide a stream of ions that will be orbited in the orbit
region;
b. an ion target positioned internally of said iron housing at an
outer position in the orbit region such that orbiting ions strike
the target and produce neutrons;
c. a channel formed in the iron housing from the target to the
exterior for passage of the beam of neutrons formed at the target,
said channel acting as a beam collimator; and
d. a mounting structure for movably mounting the cyclotron and
target such that the neutron beam produced can be employed at more
than one position.
2. A neutron source as in claim 1 wherein the ions orbited in the
cyclotron are deuterons and the target is a thick beryllium metal
layer.
3. A neutron source as in claim 1 wherein the ions orbited in the
cyclotron are deuterons and the target is a thick lithium metal
layer.
4. A neutron source as in claim 1 wherein the mounting structure is
a gantry allowing movement of the cyclotron and thus the neutron
beam in an isocentric path.
5. A neutron source as in claim 1 wherein the flutter poles are
four in number and are formed as spiral-edged sectoral raised
portions of the iron housing structure defining valleys between
with the said RF electrode structures being positioned in the
valley regions.
6. A neutron source as in claim 1 wherein the refrigerant fluid is
supercritical helium at a temperature of 4.2K.
Description
This invention relates to a superconducting cyclotron neutron
source for medical therapy purposes.
Recent experiments in the irradiation of tumors by neutrons have
given encouraging results. One series of experiments were done with
a 16 MeV deuteron beam on a beryllium target approximately 100
mg/cm.sup.2 thick. This provided a neutron spectrum with a mean
energy about 8 MeV and a "depth-dose" of 43% at 10 cm. The does
rate at a distance of 120 cm from the target for a 100 .mu.A beam
was about 50 rad per minute requiring 2.4 minutes for the usual
required fraction of 120 rads. Although these results are quite
good it is considered that higher mean neutron energy would improve
the treatments. A depth dose 50% at 10 cm would appear to be
desirable. This can be achieved with 14 MeV neutrons from a d-T
(deuterium beam-tritium target) source at a source distance of 75
cm. A d-T source producing 5 x 10.sup.12 n/sec. gives a dose rate
of 30 rad/min. at 75 cm. The difficulty with d-T sources is in the
target technology required to obtain higher dose rates, say 50
rad/min., with a reasonable target life. The present target
lifetimes are unacceptable.
A cyclotron source can provide the desired dose rate, say 120
rad/minute for one minute irradiations to obtain the desired
depth-dose from a thin beryllium target requires a deuteron beam
with an energy of about 25 MeV. Conventional designs of cyclotrons
capable of producing 25 MeV deuterons are very heavy and bulky
(about 115 tonnes in weight). It is highly desirable that a neutron
source for this purpose have an isocentric mounting for multiple
beam directions in fractionated treatment. This would suggest the
mounting of the neutron source on a gantry similar to that used for
cobalt irradiators. it appears from this that a conventional
cyclotron producing the required beam energy would have to have its
beam extracted and transported to an isocentrically mounted target
and this is out of the question for this application and would be
uncompetitive with the d-T type source which is relatively small
and which can be readily mounted in an isocentric manner.
It is therefore an object of the present invention to provide a
neutron source for medical therapy that has high intensity is
relatively light in weight, and which can be readily mounted
isocentrically.
This and other objects of the invention are achieved by a neutron
source comprising a cyclotron using an air core superconducting
magnet to provide high intensity magnetic fields for producing a
relatively high energy beam of ions, a target made of a material
chosen from the group lighium, beryllium positioned in the ion beam
path such that a high energy neutron beam is produced, and an
isocentric mounting for said cyclotron and target.
A superconducting cyclotron of a type suitable for the present
purpose is described in applicants' pending application Ser. No.
419,034.
In drawings which illustrate an embodiment of the invention,
FIG. 1 is a cross section of a superconducting cyclotron on an
isocentric mounting,
FIG. 2 is a cross section of the cyclotron showing the neutron beam
from the internal target,
FIG. 3 is a graph of the mid phase magnetic field of the cyclotron
with and without flutter poles,
FIG. 4 is a cross section of the cyclotron and the target,
FIG. 5A and 5B are midplane and vertical cross sections of the RF
accelerating structure, and
FIG. 6A and 6B are vertical and midplane cross-sections of the ion
source region.
A first type of neutron reaction considered for this apparatus is
the .sup.9 Be (d,n).sup.10 B with a Q of 4.35 MeV. At these energy
levels, the stripping reaction predominates so the output neutrons
have about half of the deuteron energy. For a thick target the mean
energy is about 0.36 E.sub.d. It has been found that a mean energy
of about 14 MeV can be obtained with 25 MeV deuterons if the
beryllium target is made about 50 mg/cm.sup.2 in thickness. The
neutron yield is then approximately one-third that for a thick
target requiring beam intensities about three times larger. Other
types of neutron reaction is .sup.7 Li(d,n).sup.8 Be with Q of 15.0
MeV and .sup.9 Be(P,n).sup.9 B with Q = 1.9 MeV. The neutron yield
from this reaction is a little smaller (80% of the beryllium target
yield at 16 MeV) but the neutron spectrum has a higher mean energy
and a component going to much higher energies. From known data, it
is considered that a 25 MeV, a thick lithium target would have
about twice the yield of a 50 mg/cm.sup.2 beryllium target, a mean
energy of approximately 16 MeV and about 1/3 of the neutrons in a
distribution extending to 26 MeV. These high energy neutrons would
be very effective in the C.sup.12 (n,n.sup.1) 3.sup.4 He and
O.sup.16 (n,n.sup.1) 4.sup.4 He reactions which are believed to be
important in therapy.
Referring to FIG. 1 a cyclotron 10 is mounted isocentrically by
supports 11a and 11b on a strong, vertical, rotating pillar 11 such
that the neutron beam 12 of the device can be moved over a fairly
long arc. The cyclotron has two superconducting coils 13 and 14
mounted in an iron housing 15 that has the combined function of a
magnetic yoke, magnetic shield, radiation shield, and vacuum
vessel. The coils are mounted in a cavity in the iron housing in a
radiation shield 16. Adjustments to the positioning of the coils
can be made by means of a supporting and adjusting elements 17a,
17b, 17c, and 17d. Helium cooling fluid and electrical leads to the
coils pass through port 18 into the interior of shield 16. The
coils are Nb.sub.3 Sn superconductors providing a 7 tesla average
magnetic field in orbit region 19 having a BR product of 1 for a
deuteron output energy of 25 MeV (B = field strength and R =
radius). An ion source 20 provides a stream of ions (deuterons or
protons) that are introduced into the orbit region via passage 21.
These ions orbit outwardly in the orbital region between plates 22
and 23 forming an RF resonator 36 mounted on RF input structures 24
and 25 and energized from an appropriate RF source via line 27. A
vacuum pump 28 connects to the internal cyclotron cavity via line
29 to provide the necessary vacuum level required for cyclotron
operation. A target 30, either beam path inside the cyclotron as
shown through port 31. Neutrons produced at the internal target
pass through window 33 are collimated by a channel 32 in the magnet
support spool and the iron yoke 15.
Fig. 2 shows the described features and more detail of the four
sector flutter pole geometry (sectors 34a, 34b, 34c, 34d) and the
RF accelerating structure (resonator) 36. The flutter poles provide
azimuthal focussing and although four sectors are illustrated,
other numbers e.g., three might be used. These are formed as raised
portions of the iron housing structure leaving in effect four
valley sectors 35a, 35b, 35c, 35d). The superconducting coils in
conjunction with the flutter poles provide an isochronous field
profile to an outer radius R.sub. 0. Since the energy level is
fixed by design superconducting trim coils or normal trim coils are
not required. RF accelerating electrodes 36 are positioned in
valleys 35a and 35c and are mounted on "leg" 24 and 25 extending
axially to form a .mu./2 resonator at 212.8 MHz, four times the
cyclotron frequency of 53.2 MHz. A large energy gain per turn is
not required since the beam is not extracted and orbit separation
is not necessary. The ions from the central ion source are
extracted by the RF field and this determines the RF voltage
required. Typical beam power for a 200 .mu.A beam is 5 KW. An RF
power of 10 KW will drive the electrodes to about 20 KV peak with
full beam.
The total weight of the cyclotron of beam power levels described
above is about 4 tonnes. This compares favorably with the weight of
shielding required for an isocentrically mounted cobalt-60 therapy
head. The design provides a minimum of 30 cm. of iron
shielding.
FIG. 3 shows the required radial field profile for 25 MeV deuterons
at a final orbit radious (R.sub.o) of 15 cm. The midplane magnetic
field for isochronism is curve A. Because the flutter poles enhance
the midplane field and also cause some radial variation in the
azimuthally averaged field, the field profile required from the
magnet coils is as shown by curve B which is the isochronous field
with flutter pole field substracted.
FIG. 4 is a cross-section through the cyclotron showing the RF
resonator 24 on one side and a flutter pole 34a on the other. The
latter is formed as part of the iron shield and vacuum vessel
structure 15.
FIG. 5A and 5B shown the reasonator cavity 36 more clearly. Two
interconnected sectors are mounted in opposite valleys and joined
at the centre to form a single .mu./2 resonator. Two sectors
instead of four give a simpler structure at the ion source. The
resonator is energized at RF by a probe loop 39 attached to leg 24
which may be a tube or rod. Other methods of energizaton may be
used. A large energy gain is not necessary but at least two sectors
are necessary to provide symmetry. An rf power of .about. 10 KW
should be sufficient to run the resonator at about 25 KV peak and
supply a beam power up to about 5 KW (200 .mu.A at 25 MeV). A fixed
frequency system comprisng a stable oscillator, intermediate
amplifier and power amplifier with 0.2 stability in power level and
frequency would be adequate.
FIG. 6A and 6B show vertical and midplane sections of the ion
source 21. A tungsten filament arc source 37 in conjunction with a
D.sub.2 gas stream provides a supply of deuterons into the orbit
region 19. If protons are used as the ions, then an H+ gas stream
would be used. A repeller electrode 38 causes the ion beam to
travel in the correct paths to begin orbiting.
The azimuthal focussing is provided by a flutter field produced by
iron poles extending to the end sheild which then forms a yoke.
(See FIGS. 2 and 4). The estimated field increase between the poles
.DELTA.B = 1.6 T giving for an average field, B, a flutter factor
##EQU1## for h .apprxeq. 0.5 . At the extraction orbit .gamma.=
1.0133 for 25 MeV deuterons. Using ##EQU2## gives a maximum spiral
angle .xi. = 45.degree.. The flutter field cannot extend inward to
the ion source so a magnetic "hill" is used to produce "weak"
focussing out to where the flutter focussing becomes effective.
This can be provided by a suitable rion "shim" in the central
region. The radial betatron frequencey .nu..sub.r = .gamma. in the
isochronous region reaching a maximum value of 1.0133. In the
central hill region however .nu..sub.r <1 and must pass through
unity at a small radius where imperfections could drive the .nu.= 1
resonance. This is not expected to cause difficulty since it would
only spread the beam radially at the target.
The neutrons are generated in a 50 mg/cm.sup.2 thick beryllium
metal layer on a 1 mm gold backing. Stopping the deuterons in gold
after they pass through the beryllium reduces the low energy
neutron flux. The gold is in turn mounted on a water cooled copper
backing. The power in the beam is as large as 5KW giving a power
density of .about. 1 KW/cm.sup.2 if the beam can be spread over 5
cm.sup.2. This is manageable but will require a carefully designed
coolant channel with a water flow of about 2 gal. per minute. Under
normal treatment conditions at 120 rad per one minute fraction, the
fractions will be repeated at 10- 15 minute intervals i.e., the
duty cycle is .about. 10%. The target assembly is mounted on a tube
extending from the outside of the shielding and arranged so that it
can be withdrawn into a vacuum lock for servicing.
In a typical design, a pair of coils with an inside diameter of
36.5 cm, a cross section of 12 cm x 12 cm and a separation of 13
cm, provides a field with approximately the correct magnitude and
shape. There are 1.8 x 10.sup.6 ampere turns in each coil. This
corresponds to an overall current density of 12,500 A/cm.sup.2,
which is within the state of the art for niobium-tin coils. The
maximum radial field is 3 Tesla. The increase near the central
region in the required coil field shown in curve B of FIG. 3
results from the flutter poles not extending to the cyclotron
centre.
Lorentz forces within the coils are large but tractable. The
maximum hoop stress, which occurs at the inside winding is
approximately 30,000 psi assuming no transfer of the radial force
to the outer windings. If such transfer is assumed so that a
uniform hoop stress is realized, then its value would be no greater
than 15,000 psi. The tensile strength of stainless steel is
considerably larger than this at low temperatures so no problems
are expexted here. Two types of superconductor may be used in
winding the coils, Nb.sub.3 SN and V.sub.3 Ga. The first of these
is commercially available and has a guaranteed short sample
critical current of 600 amperes at 10 Tesla and 4.2K. The conductor
has the Nb.sub.3 Sn vapor deposited on a Hastelloy substrate and is
clad on both sides with copper. Its dimensions are 0.5 inch wide
.times.0.0072 in. thick. With this conductor, a coil as illustrated
in FIG. 4 is made up of 9 pancake windings each with 400 turns. The
total length of superconductor to wind two coils is then 11,300
metres. Total weight of superconductor is 210 Kgm and the weight of
both coils with an inter coil bridge is 400 Kgm.
The coils are cooled by circulating supercritical liquid helium at
4.2 K around them. The advantages of supercritical helium at a
pressure above the critical value of 2.26 atm.,) as a cooling
medium have been known but have not yet been extensively exploited.
For the present application the advantages are twofold; first high
heat transfer coefficients are possible and second no difficulties
caused by two phase flow of the helium can occur, since it is
always maintained in the liquid state. The disadvantage includes
the requirement that the helium vessel around the coils must
support a pressure greater than 3 atmospheres - 4 atmospheres seems
to be a good design value.
* * * * *