Boron Neutron Capture Therapy (BNCT) technique needs a special design methodology. In this work, a new design method for a multilayer spectrum shifter is proposed in order to improve the design process of epithermal BNCT beams for research reactors. This design method is based on two concepts: stepwise spectrum shifting, and separation of the design process into two design stages, namely the material(s) selection and the thickness(es) determination. Java-based Nuclear Information Software (JANIS) and Monte Carlo N’Particle code ‘MCNP5’ have been used for cross sections computations and beam design, respectively. In order to validate the developed methodology, it has been applied on two cases studies using two different neutron spectra as neutron sources: a Watt fission neutron spectrum and a leakage neutron spectrum of a typical MTR-TYPE reactor. The different beam components have been designed but more concern has been given for the spectrum shifter. Various parametric studies have been done for the selected spectrum shifter materials using the MTR-TYPE reactor spectrum. For both cases studies, comparisons have been done between the proposed spectrum shifter and the well-known patent material ‘FLUENTALTM’. The results of comparisons confirm that the proposed design outperforms the patent material ‘FLUENTALTM’. These results indicate that the developed design method in this work could generate efficient and effective spectrum shifters for epithermal BNCT beams in research reactors. The proposed method has some advantages as: time saving, simplicity, effectiveness, and efficiency. Therefore, this method may be extended to be applied for other BNCT facilities such as accelerator-based BNCT as well as other epithermal beams used for different applications.
TABLE OF CONTENTS
TABLE OF CONTENTS 3
LIST OF FIGURES 5
LIST OF TABLES 7
Boron Neutron Capture Therapy (BNCT) 9
General requirements of BNCT Facility at a Nuclear Research Reactor 9
Desired Neutron Beam Parameters 14
General beam properties 14
Epithermal beam intensity 16
Incident beam quality 17
Beam size 19
REVIEW OF LITERATURE 20
Reactor And Beam Design Considerations 20
Approaches to Using Reactors for Epithermal Neutron
Performance of Some Current Epithermal Neutron Irradiation Facilities 31
BNCT beam configuration 34
Proposed spectrum shifter design methodology 36
First stage: Material selection 38
3.2.2 Second stage: Thickness determination 39
RESULTS AND DISCUSSION 41
BNCT beam design using Watt fission neutron spectrum 41
Reflector, Collimator, and Shield 41
Spectrum shifter 42
First stage: Material selection 45
Second stage: Thickness determination 46
Thermal neutron filter 47
Gamma filter 47
Optimum position of the thermal neutron filter 48
Validation of proposed spectrum shifter design methodology
using a Watt fission neutron spectrum source 49
BNCT beam design for a typical MTR-TYPE reactor 51
4.3.1 Reflector, Collimator, and Shield
First thermal neutron filter 53
Spectrum shifter 53
Second thermal neutron filter 54
Gamma filter 55
Collimator thickness 56
Validation of proposed spectrum shifter design methodology
in a typical MTR-TYPE reactor 57
Parametric analysis for the selected spectrum shifter materials 57
4.4.1 Verification of material sequence 65
CONCLUSIONS AND RECOMMENDATIONS 68
LIST OF FIGURES
Fig. 1.1 schematic for BNCT irradiation facility 10
Fig. 1.2 Comparison of flux-depth distributions for thermal and epithermal neutrons 16
Fig. 1.3 Typical spectrum shifting arrangement 21
Fig. 1.4 Example of an effort to minimize core-to-patient distance 22
Fig. 3.1 Schematic diagram BNCT beam configuration 35
Fig. 4.1. Group ratios for the studied materials 43
Fig. 4.2. Group ratios for the studied materials 44
Fig. 4.3. Schematic diagram of the final beam MCNP model for case 49
Fig. 4.4. Beam inlet normalized neutron energy spectrum for case 2 52
Fig. 4.5. Beam inlet gamma energy spectrum for case 2 52
Fig. 4.6. Schematic diagram of the final beam MCNP model for case 57
Fig. 4.7. Change in fast and epithermal currents for 7LiF with
Fig. 4.8. Change of fast neutron groups to epithermal current ratios for 7LiF with thickness 61
Fig. 4.9 Change in fast and epithermal currents for AlF3 with
Fig. 4.10. Change of fast neutron groups to epithermal current ratios for AlF3 with thickness 62
Fig. 4.11. Change in fast and epithermal currents for 60Ni with
Fig. 4.12. Change of fast neutron groups to epithermal current ratios for 60Ni with thickness 63
Fig. 4.13. Performance of selected materials: 7LiF, AlF3, and 60Ni at thicknesses 4-60 cm 63
Fig. 4.14. Performance of selected materials: 7LiF, AlF3, and 60Ni
at thicknesses 4-60 cm for the 3rd fast neutron group 65
Fig. 4.15A. Performance of selected materials: 7LiF, AlF3, and 60Ni
at thicknesses 4-60 cm for the 2nd fast neutron group 66
Fig. 4.15B. Performance of selected materials: 7LiF, AlF3, and 60Ni
at thicknesses 4-24 cm for the 2nd fast neutron group 66
Fig. 4.16. Performance of selected materials: 7LiF, AlF3, and 60Ni
at thicknesses 4-60 cm for the 1st fast neutron group 66
LIST OF TABLES
Table 1.1 Facility characteristics and beam performance figures of merit for various clinical NCT centers determined by in-air and in-phantom measurements using BPA and where available (in brackets) a hypothetical advanced boron compound 31
Table 3.1 The layers and fast groups numbers, energy groups, and corresponding fast neutron groups 37
Table 4.3 In air beam port quality parameters for the empty beam for
case 1 42
Table 4.4 Neutron in air beam port quality parameters after adding reflector, collimator, and shield for case 1 42
Table 4.5 The efficient & selected materials of each layer and corresponding neutron & energy groups for case 1 45
Table 4.6 In air beam port quality parameters after adding the first spectrum shifter layer for case 1 46
Table 4.7 In air beam port quality parameters after adding the second spectrum shifter layer for case 1 46
Table 4.8 In air beam port quality parameters after adding the third (last) spectrum shifter layer for case 1 47
Table 4.9 In air beam port quality parameters after adding the thermal filter for case 1 47
Table 4.10 In air beam port quality parameters after adding the gamma filter for case 1 48
Table 4.11Comparison between in air beam port quality parameters when placement of the thermal filter in two different positions for case 1 49
Table 4.23 Comparison between in air beam port parameters of the proposed spectrum shifter and FLUENTALTM for case 1 50
Table 4.12 In air beam port quality parameters for the empty beam for case 2 51
Table 4.13 Neutron in air beam port quality parameters after adding reflector, collimator, and shield for case 2 53
Table 4.14 In air beam port quality parameters after adding the first thermal filter for case 2 53
Table 4.15 The selected materials and optimum thickness for each spectrum shifter layer for case study two 54
Table 4.16 In air beam port quality parameters after adding the first spectrum shifter layer for case 2 54
Table 4.17 In air beam port quality parameters after adding the second spectrum shifter layer for case 2 54
Table 4.18 In air beam port quality parameters after adding the third (last) spectrum shifter layer for case 2 55
In air beam port quality parameters after adding the second thermal filter for case 2 55
In air beam port quality parameters after adding the gamma filter for case 2 56
Final in air beam port quality parameters after adding two Bismuth gamma filters for case 2 56
Comparison between in air parameters when using Pb collimator with thickness 10 and 20 cm for case 2 57
Comparison between in air beam port parameters of the proposed spectrum shifter and FLUENTALTM for case 2 59
Boron Neutron Capture Therapy (BNCT)
Boron neutron capture therapy (BNCT) is a biochemically targeted radiotherapy based on the nuclear capture and fission reactions that occur when non-radioactive boron-10, which is a constituent of natural elemental boron, is irradiated with low energy thermal neutrons to yield high linear energy transfer alpha particles and recoiling lithium-7 nuclei. Therefore, BNCT enables the application of a high dose of particle radiation selectively to tumour cells in which boron-10 compound has been accumulated .
General requirements of BNCT Facility at a Nuclear Research Reactor
The design and construction of a BNCT facility in a nuclear research reactor is a very demanding task. From the engineering and physical viewpoint, the general requirements to house a complete facility in the confines of a reactor building are :
i) there should be a sufficient and adequate fluence rate of neutrons that emanate through a large diameter beam tube, that itself preferably faces a large source area of the neutron source.
ii) there must be suitable space at the exit side of the beam, to accommodate a large working area for building the irradiation room.
iii) the facility should be freely accessible such that the patient, who may be unable to walk, can be readily and easily brought into the irradiation room from outside.
Furthermore, outside the reactor building confinement:
iv) there should be space allocated outside the reactor building to house facilities for the reception and preparation of the patient prior to entering the facility for treatment. However, this latter condition, providing that there is sufficient space within the reactor building, may apply to within the building.
The basic design can be broken down into specific components, which formed the basis of one of the 2 Working Groups held at the Prague Workshop, entitled: ‘How to build up an irradiation facility for BNCT’. The Working group defined the following 9 components, based on the schematic simplification, shown in figure 1.1, where each numbered component represents:
Fig. 1.1 schematic for BNCT irradiation facility
2. Beam tube
4. Irradiation room
5. Control room
6. Reception area
7. Beam dosimetry/characterisation
8. Beam monitors
9. Beam shutter
Following are some details for each component requirements:
1. The Reactor
In BNCT, there is no singly-available reactor-type that may be regarded as the ideal reactor to develop a BNCT facility. Over the decades a wide variety of reactor-types have been utilised, as the neutron source for a BNCT facility. The closest single design could be a TRIGA, however even this may vary in power and configuration, e.g. utilisation of the thermal column or beam tube.
2. The beam tube The neutrons, and gammas, emanating from the reactor, need to be moderated, filtered and/or attenuated, such that the required beam intensity and average neutron energy are within the requirements, which themselves can vary depending on the type of tumours to be treated. The recommended beam parameters will be discussed in details later.
3. The patient
The type of tumour to be treated will also influence the beam design, whether an epithermal beam or thermal beam is required. Also whether the patient can only be treated in a horizontal position or seated, will influence the design.
4. The irradiation room
If a patient is to be treated in a horizontal position, and especially if brain tumours are to be treated, there should be sufficient space, such that the treatment table can be rotated ”90 degrees about the beam axis. Preferably, the central beam axis should be at least 1m vertically above the floor. The room itself should contain most if not all facilities expected in a conventional radiotherapy room, e.g. cameras, laser positioning devices, ease-of-access, etc.
5. The control or observation room
The patient and the facility need to be monitored. Monitoring equipment such as, puls-oxymeters, TV monitors, microphones, radiation level instrumentation, radiation monitors measuring the beam characteristics, all need to be housed in an area or room, where the medical and physics staff can comfortably sit and observe the patient and radiation beam parameters. Such a room should be next to the irradiation room and be at least 10m2.
6. Reception area or building
Next to, or outside but close to the entrance to the reactor, an area or dedicated building should be available, where the patient can be received prior to treatment. It will be necessary to prepare the patient, e.g. infusion of the drug, change of clothing, medical examination. The area or building should also have the provisions of an office, for the medical staff, a waiting room, for accompanying relatives, as well as,, of course, a WC.
7. Beam dosimetry/characteristion
The radiation beam is a mixture of neutrons, of all energies, and gamma rays. The beam must be thoroughly characterized, as part of any quality assurance system, as well as, being absolutely necessary, in order that accurate and reliable calculations, especially for treatment planning can be performed. There are a variety of techniques available to characterize such a beam. The irradiation room should be designed such that any eventual measurements can be readily carried out.
8. The beam monitors
A beam monitoring system must be available that can measure the radiation beam during treatment and have the capacity to automatically close the beam when the required radiation dose is achieved or when an emergency situation occurs. As in conventional radiotherapy, all safety systems should be backed-up by an independent, second device acting in case of failure of the first. In BNCT, a beam monitoring system normally consists of four beam monitors: two 235U fission chambers (neutron counters) and two GM-tubes (gamma ray counters), which are located close to the beam line, usually in the wall of the beam tube or in the (fixed) collimator, downstream from any beam shutter. The automatic opening and closing of the beam should be controlled by the fission chambers, according to a pre-set number of monitor counts which correspond to the required boron dose delivered at the dose group identification point in a patient. Both fission chambers are pre-set to close the shutters, which are automatically triggered when the target counts are reached. The fission chambers, as well as the GM-tubes, should be monitored and the counts and count rates displayed on two independent computer systems. As an additional back-up for beam shutter closure, i.e. if the monitoring system fails, a timer should be available, with a pre-set time at 2% above the given irradiation time. If called into use, closure of the beam shutters is automatically triggered.
9. Beam shutter
Depending on the reactor-type utilized, a beam shutter or shutters are necessary, such that the radiation can be stopped or reduced sufficiently in intensity, such that staff can enter the irradiation room. Some reactors, such as the HFR at Petten, operate 24 hours a day, as such the beam shutters must open and close without any mutual effect on the operation of the reactor and to reduce radiation levels to levels where the staff can enter the room. At some facilities, such as low power, TRIGA reactors, the reactor itself is shut-down as part of the treatment procedure. Beam shutters are generally made of layers of lead and borated or lithiated polyethylene. Opening and closing of the beams must be performed remotely, with emergency back-up facilities to shut the shutters, if the power supply fails. The shutters must be equipped with interlock devices to enable automatic shut-down when the required dose is achieved or in emergencies. If necessary, it should be possible to close the beam shutter(s) manually, if any electrical failure occurs. As a last resort, the beam operator has the mandate to instruct the reactor operators to scram the reactor.
DESIRED NEUTRON BEAM PARAMETERS 
General beam properties
Before addressing how to modify reactors and how to condition beams, it is first necessary to establish the beam characteristics desired for BNCT.
For BNCT, an adequate thermal neutron field has to be created in the boron-labelled tumour cells within a prescribed target volume. This means that for target volumes well below the surface, epithermal beams will generally be best, while for target volumes near the surface, thermal beams will suffice.
Figure 1.2 shows that an epithermal beam entering tissue creates a radiation field with a maximum thermal flux at a depth of 2’3 cm, which drops exponentially thereafter. The penetration of the beam can be increased by increasing the average energy of the epithermal neutrons and by increasing the forward direction of the beam, especially with small beam sizes. In contrast to the epithermal beam, which shows a skin-sparing effect, the thermal flux falls off exponentially from the surface. Thus, thermal neutron irradiations have been used for melanoma treatments in the skin, as well as with open craniotomy for glioma treatments. In general, however, the current trend for treatment of patients with brain tumours is to use epithermal neutron beams.
Radiobiology research for NCT, on the other hand, requires access to both thermal and epithermal beams. Clinical facilities can be used to study the effects of epithermal irradiation, but when studying the effect of boron carrier compounds using cell cultures or small animals, a pure thermal neutron field is preferred.
Most epithermal beams are accompanied by, and produce, other radiations that are not selectively absorbed by labelled cells, and therefore contribute to both normal and tumour tissue damage. It is clearly desirable to reduce these radiations as much as possible in the incident neutron beam. Since the bulk of the report will focus on patient related aspects, it can be stated that the beam design objective is to deliver an epithermal neutron fluence within a reasonable treatment time and to produce the desired thermal neutron fluence at tumour depth with minimal other radiations present.
Fig. 1.2 Comparison of flux-depth distributions for thermal and epithermal neutrons.
The two principal beam characteristics of interest are intensity and quality. Beam intensity will be the main determinant of treatment time. Beam quality relates to the types, energies, and relative intensities of all the radiations present.
Epithermal beam intensity
For the purposes of reporting beam intensity, the common definition for an epithermal energy range should be used, namely 0.5 eV to 10 keV. If other energy limits are used, they should be clearly reported. Current experience shows that a desirable minimum beam intensity would be 109 epithermal neutrons cm’2 s’1. Beams of 5 x 108 n cm’2 s’1 are useable, but result in rather long irradiation times.
When aiming at higher intensities (>1010) the advantages of shorter irradiation times must be weighed against those of improved beam quality. Where there is a choice to be made, most practitioners would rather have better quality rather than more intensity, within the constraint of having a reasonable treatment time (possibly extending up to one hour). Requiring immobilization of patients for significantly longer times reduces the clinical acceptability of BNCT as a therapy.
Tumour boron concentration will affect the requirements for beam intensity. If the boron concentration can be raised from the currently values, the beam intensity requirement (or treatment time) will be reduced proportionately. On the other hand, if the beam intensity is too low, it may be difficult to maintain the necessary boron concentration in the tumour for the total irradiation time required. To avoid unduly lengthy irradiation times, fractionation may be considered as an alternative. It could also provide opportunity for boron compound retargeting.
Incident beam quality
Beam quality is determined by four parameters under free beam conditions. They are discussed below in order of importance.
The fast neutron component
In BNCT the energy range for fast neutrons is taken as >10 keV. Fast neutrons, which invariably accompany the incident beam, have a number of undesirable characteristics such as the production of high LET protons with a resulting energy dependence of their induced biological effects. Therefore, it is one of the main objectives of BNCT beam design to reduce the fast neutron component of the incident beam as much as possible.
Another major objective is clearly to have as high an epithermal flux as possible. In existing facilities the range of dose from this component is from 2.5 to 13 x 10’13 Gy cm2 per epithermal neutron. A target number should be 2 x 10’13 Gy cm2 per epithermal neutron.
The gamma ray component
Because of the energy range of the gamma radiation, it results in a non-selective dose to both tumour tissue and a large volume of healthy tissue. Hence it is desirable to remove as much gamma radiation from the beam as possible. Since there are also (n,”) reactions occurring inside the patient, the importance of this component in the incident beam is somewhat reduced. Nevertheless, a target number for this should be 2 x 10’13 Gy cm2 per epithermal neutron. The range in existing facilities is from 1 to 13 x 10’13 Gy cm2 per epithermal neutron.
The ratio between the thermal flux and the epithermal flux
To reduce damage to the scalp, thermal neutrons in the incident beam should be minimized. A target number for the ratio of thermal flux to epithermal flux should be 0.05.
The ratio between the total neutron current and the total neutron flux
This ratio provides a measure of the fraction of neutrons that are moving in the forward beam direction. A high value is important for two reasons: (1) to limit divergence of the neutron beam and thereby reduce undesired irradiation of other tissues, and (2) to permit flexibility in patient positioning along the beam central axis. A high ratio means that the epithermal neutron flux very close to the beam port opening will change only slightly with distance from the port. In cases where the body of the patient must be positioned perpendicular to the beam axis, this will permit a patient to be positioned somewhat farther from the port. This will increase the depth dose and facilitate patient positioning without seriously diminishing the available incident beam intensity. A target number for this ratio should be greater than 0.7.
Circular apertures of 12 to 14 cm diameter are being used in the present clinical trials. However, sizes of up to 17 cm have been proposed for irradiation of brain tumours. Other cancers in the body might require even larger apertures. These maximum sized apertures are reduced in accordance with the tumour size and position as determined by the treatment planning requirements.
REVIEW OF LITERATURE
REACTOR AND BEAM DESIGN CONSIDERATIONS 
It is important to know how the conversion of existing research reactors was done. Typically, this has meant modifying or adding components such as the reflector, a beam port or thermal column, shielding, collimators and filters in order to try to obtain a beam of the intensity and quality needed. Key aspects of reactor modification and beam conditioning are discussed. Much of the discussion is also relevant to the design of new reactors.
Most existing thermal research reactors have reflectors to optimize the core efficiency. Clearly, the need to provide a source of fast neutrons for the spectrum shifting moderator, or the filter, demands that the reflector must be removed from that part of the core. This means that a careful analysis of the core neutronics needs to be undertaken prior to this modification, and more fuel may be needed as a consequence.
Spectrum shift vs. filtered beam
While spectrum shifting using a moderator has proven to give a higher efficiency in producing an epithermal beam than filtering, the choice of which technique to use is typically determined by the existing reactor design. The former method requires the availability of a large opening in the shield such as that often used for a thermal column. Figure 1.3 shows a typical example. If the reactor does not have such a space then a higher powered reactor (>10 MW) with a beam port has the option of filtering the beam. Alternatively, part of the shielding can be opened up or removed to provide space for a spectrum shifter.
Fig. 1.3 Typical spectrum shifting arrangement
For spectrum shift facilities, the moderator has to be placed as close to the reactor core as possible to maximize the input of fast neutrons. A shorter distance from core to patient will thereby result in a higher epithermal flux at the dose point. In addition, it will allow the reactor core to subtend a larger angle allowing the production of a converging beam of higher intensity. However, the core-to-patient distance is often limited by the need to accommodate features such as a fission converter, moderator, filters, collimators, and shutters. Certainly, increasing the distance from the reactor to the patient beyond the thickness of the existing shield decreases the available flux and should be avoided if possible. Therefore, every effort should be made to fit all beam-conditioning components and shutters within the existing shielding dimensions (Figure 1.4). Some facilities have successfully opened up their existing reactor shielding in order to provide a larger beam aperture, and shorter core-to-patient distance.
Practically, the beam components, the moderator and collimator, need a length of about 1 to 2.5 meters. This gives the desired position for irradiating the patient supposing that the patient and the personnel can be shielded from the undesired radiation from the reactor core.
Fig. 1.4 Example of an effort to minimize core-to-patient distance
For filtered beam facilities the core to patient distance is usually dictated by the original design of the reactor and is not as critical because of the inherent higher current to flux ratio.
Beam intensity and current-to-flux ratio
Increasing beam intensity is achieved by surrounding the beam with an appropriate reflector and tapering it from a wide to a narrow aperture. Suitable reflector materials for this are those with high scattering cross section and high atomic mass (resulting in little energy loss). They include Pb, Bi, PbF2.
A forward-directed beam with a current-to-flux ratio of greater than 0.7 helps to deliver a higher intensity neutron beam at a distance from the reactor shield face. This allows greater flexibility in positioning the patient. The use of collimators can be used to improve the current to flux ratio of the final incident beam.
Increasing the distance between the core and the patient will improve the current to flux ratio for a given beam diameter. Hence, for reactors which use the filtering method rather than the spectrum shift method, a very forward directed beam is the natural result of a long, narrow penetration through the biological shield. The filtering components can be installed in the beam tube and the beam can then be transported long distances without further sacrifice in intensity. The longer distance between the core and the patient may offer additional space for beam shutters.
It is important to note that removing as many fast neutrons as possible, and using a beam delimiter to improve directionality, will not necessarily maximize the dose delivered at depth. MCNP modeling has shown that hardening the spectrum slightly by adjusting the thickness of the moderator results in better beam penetration. It also has shown that attempts to improve the directionality of the beam too much can remove so many neutrons that the intensity of the beam is reduced, lowering the dose delivered to the target volume. Optimal conditioning of the beam for a given case may be dependent on the detailed geometry of the target volume.
In the final analysis, the quest for high intensity is perhaps not as important as the production of a sharply defined, high quality epithermal beam, which limits the whole body dose to the patient. With small enough whole body doses, treatment in multiple fractions can be given, compensating for lower epithermal beam intensity.
Undesirable radiation components in the incident epithermal beam
One of the key aspects of reactor conversion and beam design is to maximize the desired epithermal neutrons while minimizing the healthy tissue dose from all other radiations in the incident beam.
Fast neutron contamination
In the spectrum shift type facilities, the objective is to moderate as many fast neutrons (>10 keV) as possible down to the desired epithermal energies. In the filter type facilities, fast neutrons are removed by filtration. Moderators and filters are discussed below
Materials such as Pb and Bi, which are relatively transparent to neutrons, may be placed in the beam to reduce gamma rays originating from the reactor core, but these will nonetheless somewhat reduce neutron beam intensity. Bismuth is nearly as good as lead for shielding gamma rays, while having a higher transmission of epithermal neutrons. However, caution is necessary in handling neutron-irradiated bismuth, because of the buildup of 210Po, a bone seeking alpha emitter created by neutron capture in 209Bi with subsequent beta decay of 210Bi. Encapsulation of the bismuth is highly recommended.
Outside of the neutron beam area, high-density concrete (mixed with iron minerals) can be used to reduce gammas. Steel and iron can be protected from neutrons by shielding containing 10B or 6Li, to prevent neutron activation of these components with subsequent emission of hard gamma rays. It should be noted that 10B emits a low energy capture gamma ray (478 keV) but 6Li does not and its use is to be preferred in locations close to the patient.
Thermal neutron contamination
For epithermal neutron beams, it is desirable to limit thermal neutron contamination by filtering. Filter materials for thermal neutrons require either elements with 6Li or 10B (1/v cross sections) or Cd (0.4 eV resonance). The 1/v cross section materials may deplete the lower energy part of the epithermal neutron spectrum, but Cd produces a high energy capture gamma ray which is difficult to control and cadmium oxide represents a health hazard.
Moderators (or spectrum shifters)
Moderation of fast neutrons is best accomplished by low atomic mass materials. Any moderator or filter materials chosen must not decompose in a high radiation field, nor produce moisture. Any neutron activation products from the materials should be short lived. Suitable candidates in the literature are Al, C, S, Al2O3, AlF3, D2O, and (CF2)n. Combinations of Al followed by Al2O3 or AlF3 downstream are very efficient because the O and F cross-sections fill in the valleys between the energy resonance peaks of Al. FLUENTALTM was developed by the technical Research Centre of Finland and stands up well to radiation, but is very expensive. TEFLONTM is susceptible to radiation damage, but even so, may be acceptable when exposed to the relatively modest neutron fluences projected for the facility over its anticipated lifetime.
Analyses have been done for various materials that may be helpful for reactor facilities desiring to use the filter methodology. The objective is to start with a very high intensity beam from a high power reactor and filter out all but the neutrons with energies of 0.01 to 10 keV from the reactor beam. This can be done with thick neutron filters of natural or isotopically enriched materials, for which an interference minimum in the total neutron cross section exists in this epithermal energy range.
The total cross section of 60Ni isotope has the deep and wide interference minimum in the energy range from several eV to 10 keV and therefore this material is useful for BNCT purposes. To suppress the neutron groups with energies above 10 keV a set of additional filter materials must be used. Materials such as the isotopes 32S, 10B and others may be used.
By using the 99.5% enriched 60Ni isotope (112 g cm’2) as the main filter component and 32S (54 g cm’2) and 10B (1.15 g cm’2) isotopes as additional filters, a beam with an energy range of 0.01 to 9 keV may be obtained with a purity in the main neutron group of about 92%.
If the above filter design is modified by the addition of some 99.7% enriched 54Fe (50 g cm’2) isotope, a filtered neutron beam may be obtained with approximately 96% of the neutrons in the energy range of 0.01 to 6 keV.
Collimators inside the shielding should reflect neutrons back into the beam. Therefore, neutron reflecting type materials are used. Collimators that are used near the beam exit are beam delimiters and should absorb rather than reflect neutrons. Interchangeable exit collimators having different size inner bore diameters can be used to delimit the final size and divergence of the beam delivered at the patient treatment position. These collimators are made with B4C or 6Li2CO3 dispersed in polyethylene. Epithermal neutrons striking the wall of the collimator are thermalized and captured with minimal emission of hard gamma rays, which could shower the patient.
A dedicated NCT reactor that can be started up and brought to full power quickly and reproducibly might not need a shutter. However, the need for continuous operation of the reactor or other characteristics of the reactor operation can dictate the installation of one or more beam shutters. Even with the reactor shut down, a shutter may be required to protect personnel working in the treatment area from radiation from the core and long lived radioactive components along the beam line.
The extra space required for a shutter needs to be taken into account in the design of the irradiation facility. To save space along the beam direction, the filter/moderator can be arranged to move into the space in the beam line vacated by the shutter when it moves into the open position. Another important consideration is the loading capacity of the building structure supporting the weight of the shutter and the available crane capacity required to assemble the shutter. This is not a trivial problem, since the dense materials comprising the shutter can weigh many tons.
A combination of fast acting and slow acting shutters achieve a balance between requirements for quick termination of high dose levels and reduction of low level residual dose at the patient position. The fast acting shutter can be thinner and lighter. Also, the use of a fission converter may require a separate shutter to prevent undesired burnup of the converter fuel.
One shutter design involves pumping water, containing boron, in or out of a tank placed in the beam. This has the advantages of remote storage capability in the open beam configuration, and mechanical simplicity compared with controlled movement of massive blocks. Under power failure, shutter mechanisms should fail in the closed position (i.e. use gravity to close the shutter).
Reactor beam design analysis
Beam design requires a great deal of safety analysis prior to any submission of the facility change to the licensing authority for approval.
The Monte Carlo code, MCNP, has been demonstrated to be very useful for the detailed design of a beam facility and gives excellent agreement with measured values of spectra and flux. However, it can be very laborious to use at the early stages when a variety of potential configurations are being studied. Therefore, for design optimization studies a 2 or 3 dimensional transport codes is more convenient.
The adaptability of the 2-D ordinate transport code DOT for the design of a neutron beam for BNCT was verified during the design of the JRR-4 neutron beam facility. The neutron spectra and neutron fluxes calculated by DOT were in good agreement with those measured by the foil activation method using Au, Au covered by Cd, Cu and Ni foils.
Approaches to Using Reactors for Epithermal Neutron BNCT
In the past, neutron beam facilities for NCT have not generally been part of the original design specifications for research or test reactors. The two exceptions are the Massachusetts Institute of Technology Research Reactor (MITR)  and the now decommissioned Brookhaven Medical Research Reactor (BMRR) , both of which were commissioned in the 1950s when interest in testing the concept of BNCT initially developed. The NCT facilities at these reactors were designed specifically for thermal neutron NCT and were used in the early clinical studies during the 1950s. During the 1990s, Brookhaven  and MIT  each constructed epithermal neutron beams at these reactors that were subsequently used in more recent trials of BNCT. More recently, a new, small (30 kW) reactor specifically designed for NCT has been constructed near a hospital site in Beijing, China . This is the first reactor constructed specifically for BNCT since the 1950s.
Research interest in BNCT grew rapidly in the 1990s, and as the feasibility of external beam irradiations became clear, a significant number of research or test reactors were modified to incorporate epithermal neutron beams. The most common approach for retrofitting these reactors is to use the reactor core directly as the source for the epithermal neutron beam. Reactors ranging in power from 100 kW to several MW have been successfully converted using this approach. Examples include the low-power (250 kW) Finnish reactor FiR-1, the 1 MW Washington State University Reactor, and the high-power (45 MW) test reactor, HFR, at Petten. Small, low-power ultrasafe reactors could also be built to obtain high-performance epithermal neutron beams using designs that are specifically optimized for BNCT. These reactors would require only 100’300 kW of fission power because core neutrons could be used directly as a source for the epithermal neutron beam. Several preliminary designs have been proposed for this type of special purpose reactor. These special purpose NCT reactors could be expected to meet the requirements for clinical investigations of BNCT as well as more routine clinical treatments .
Another approach to modifying existing reactors for epithermal NCT is to use a subcritical array of fuel called a fission converter originally proposed by Rief et al. that is located outside the reactor core and driven by thermalized neutrons from the moderator. The first such facility, known as the fission converter beam (FCB), was constructed at the MITR. A few other fission converter-based beams have been designed, one for the 3-MW BMRR and another for the 2 MW McClellan Air Force Base Reactor. A fission converter is particularly appropriate for higher power or multipurpose research reactors without a movable core that support a broad range of experiments .
Performance of Some Current Epithermal Neutron Irradiation Facilities 
Table 1.1 summarizes parameters and figures of merit published for most epithermal neutron beams that have been used in NCT clinical trials together with pertinent details about the corresponding irradiation facilities. The figures of merit are good first-order indicators of beam performance. More sophisticated analyses, such as treatment plans for identical targets showing tumor isodose contours as well as dose to nearby normal tissue, are beyond the scope of this chapter. The data in Table 1.1 are taken from an experimental study comparing seven different clinical epithermal neutron beams  as well as published reports on the performance and features of the irradiation facilities. Unless otherwise noted, the figures of merit (where available) are all derived using a common set of dose conversion parameters, boron concentrations, and
Table 1.1 Facility characteristics and beam performance figures of merit for various clinical NCT centers determined by in-air and in-phantom measurements using BPA and where available (in brackets) a hypothetical advanced boron compound
...(download the rest of the essay above)