Boron coated straw neutron detector

This invention relates to a large area, position sensitive, thermal neutron detector the most elemental component of which is a ¹⁰B₄C-lined straw tube.

Neutron scattering is a very valuable technique which is critically important to materials science and structural biology applications. Neutron scattering is an important source of information about the atomic positions, motions, and magnetic properties of solid materials. As a beam of thermal neutrons is directed at a target material some neutrons will interact directly with the atomic nuclei of such material and "bounce" away at various angles related to the atomic arrangements that define the structure of the target material. This behavior is referred to as neutron diffraction, or neutron scattering.

With appropriate detectors one can count the scattered neutrons, measure their energies and the angles at which they scatter, map their final position and thereby calculate the atomic positions of the atoms in the target material that caused such scatter pattern. In this way one can determine details about the nature of target materials ranging from liquid crystals to superconducting ceramics, from proteins to plastics, and from metals to micelles to metallic glass magnets. The knowledge gained with respect to such materials has resulted in far-reaching advances in engineering, pharmaceutical and biotechnology industries, to name a few.

New facilities for neutron generation at much higher flux (neutron events/second), such as the Spallation Neutron Source (SNS) facility due for completion at Oak Ridge in 2006, will greatly enhance the capabilities of neutron scattering. The benefits offered by superior neutron scattering techniques extend to many fields and include, for example, development of improved drug therapies and materials that are stronger, longer-lasting, and more impact-resistant. However, in order to fully realize this greater neutron scatter potential these higher flux rates must be met with improved neutron scatter detection capabilities; particularly higher count rate capability in large area size detectors, while maintaining practicality.

The SNS facility will enhance the available thermal neutron flux by at least an order of magnitude above that now achievable at any other neutron science facility. This higher neutron beam intensity, together with time of flight energy discrimination provided by pulsed operation of the SNS facility, will facilitate unprecedented capabilities, which will be exploited in more than 18 experiment stations (http://www.sns.gov/). The markedly increased deliverable neutron flux imposes extreme rate requirements on the neutron scattering detectors in many of these facilities, which cannot now be met without fundamental detector improvements.

Although present ³He pressurized area detectors can provide needed spatial resolution, sensitivity and gamma ray discrimination, this ³He detector technology cannot now achieve the needed high rate operation, which, for many SNS detector stations, can reach rates over 10⁸ neutrons/sec, for detectors of 1 m² in area. The expansive scope of the experimental stations of the SNS has also pushed the area requirements of neutron detectors to many square meters. The high resolution powder diffractometer facility, for example, calls for 47 m² of detectors. Achieving such large areas for neutron detection with pressurized ³He technology is extremely expensive and difficult because of the complexity of the pressure containing structures required for ³He neutron detection. Current neutron detectors, such as ³He tubes, have significant practical limitations, including high cost, substantial weight and bulkiness, and are dangerous in portable use due to the high pressure of ³He they require.

<patcit id="pcit0001" dnum="US6426504B"><text>US 6426504</text></patcit> describes a dual range neutron detector comprising a chamber, an insulator at either end of the chamber, an anode located within the chamber and supported by the insulators, and an electrical connector mounted on one of the insulators for transmission of a signal collected by the anode. The chamber is filled with ³He and an inner wall of the chamber which serves as the cathode is provided with a thin boron coating.

A compelling need exist for alternative detectors with more favorable characteristics. According to the invention, there is therefore suggested a neutron detector according to claim 1 and a method of constructing a neutron detector according to claim 13.

A neutron detector technology based on ¹⁰B thin film conversion of neutrons and detection of neutron capture reaction products in a counter gas within a thin straw tube detector body is described. This neutron scatter detector is based on gas-filled thin wall straw tubes, modified for the conversion of neutrons in a very thin coating, or layer, of ¹⁰B, applied for example as a sputtered-coated film of ¹⁰B₄C, that lines the interior, or inside of the straw tube surface; and the subsequent detection of the neutron reaction products in the counter gas.

Straw tube bodies for the neutron detector may be constructed by first applying a very thin coating/layer of a ¹⁰B⁴C containing substance to one surface of a thin substrate material which, after installation thereon of the ¹⁰B⁴C coating/layer, is of suitable mechanical properties such that the substrate material may be fabricated into straw tube structures. Further, the structural integrity of such substrate material should permit cutting and/or splitting into long lengths of ribbons without disturbance of the applied ¹⁰B₄C coating, this so as to provide long continuous lengths of ribbon raw material that can be fabricated into straw tube bodies by high speed winding techniques without frequent interruptions as are now known for straw tube construction. Alternatively, such substrate material may first be produced in long continuous ribbon form and thereafter the thin ¹⁰B₄C coating/layer applied to the ribbon material. The ¹⁰B₄C of the coating may be supplied by any compound or mixture containing boron, which is preferably enriched in ¹⁰B₄C, that can be applied as a coating to the substrate by a vapor deposition or sputter technique or other means, such as a sprayed on film which adheres to the substrate through a suitable binder. Such adherence must be capable of withstanding the rigors of the rapid winding process used in straw construction which imparts substantial small radius binding and abrasive forces as the material is forced against a cylindrical mandrel. The layer must further provide high electrical conductivity sufficient to provide efficient conduction of electrical signals from the gas proportional detector. The compound of boron together with other elemental constituents or binder materials by which the ¹⁰B₄C layer is applied to the substrate is utilized in a quantity so as to provide a coating layer to the substrate in an amount of from about 0.12 mg/cm² to about 1.2 mg/cm² and containing preferably the maximum fractional quantity of ¹⁰B. The preferable layer thickness providing maximum efficiency with a minimum of detector layers is 0.25 mg/cm^2. Such high concentration of boron is readily achieved by using the inexpensive boron compound of boron carbide, B₄C, containing a high level approaching 100% of ¹⁰B. Such a layer of 0.25 mg/cm² is the maximum thickness which permits efficient escape of both the alpha and ⁷Li fragments produced in the event of neutron capture in ¹⁰B assuring efficient detection of such capture event when the thin layer is employed as the inner lining of a gas straw detector.

Ribbons of such 10B₄C coated material may be helically wound with a second ribbon having no such ¹⁰B coating as an outer overlapping layer (i.e., one over the other) with application of a very thin fast setting adhesive layer onto a precision cylindrical mandrel, producing a strongly bonded and rigid, geometrically precise cylindrical detector body. Thereafter an anode wire is centrally positioned within each detector tube body out of electrical contact with the ¹⁰B₄C coating. Means for providing a flow of a counter gas (i.e., 98% Ar - 2% ethane) through the tube is connected to the tube and the anode wire and the ¹⁰B₄C film, which also serves as a cathode shell, are connected to appropriate electrical circuitry for detecting the occurrence of neutron capture by ¹⁰B within the tube. Such elemental tube detectors can achieve element sizes (straw diameter) of 4 mm or less in lengths up to 2 meters, and can withstand decades of operation in extremely harsh radiation and counting conditions. The material by which such ¹⁰B coating is provided, by sputter-coating, is B₄C with ¹⁰B in its natural abundance, but preferably of an enriched ¹⁰B content, up to 100% ¹⁰B content.

The preferred structure for detection of neutrons over a large area is a closely packed stack, or array, of such detector straws, of an overall stack or array thickness on the order of 5 cm or more for a diameter of 4 mm and with a length of 1 meter or more. Through use of a very thin ¹⁰B coating (about 0.5 µm to about 3 µm thicknesses) within the tube interior, in the form of a ¹⁰B₄C coating or film, the distributed stopping power of ¹⁰B per unit depth of such small tube detector array can easily exceed that of ³He contained at practical pressures within an area detector. Because the ¹⁰B is distributed within and as part of a very thin film i.e., as contained within a thin film of B₄C, on the inside of each straw tube detector wall, efficient escape of the reaction decay products of neutron capture by ¹⁰B can be achieved, providing a thermal neutron (1.8 Å) detection efficiency of 65% or greater.

Because of the linearly discrete nature of each straw tube detector element, area detectors composed of a closely packed array of a multiplicity of such straw tube detector elements can be arranged either in flat panel form, such as 1 m² area, or could be configured in curved sections to match the radius of the scattering geometry. Such arrangements are illustrated by views A and B, respectively, of <figref idref="f0001">Figure 1</figref>. In this manner, a target could be essentially surrounded by such detectors, which because of their large size could be positioned at a large separation from the target in order to improve scattering angle measurement.

With simple charge division longitudinal coordinate readout such arrays constructed of 4 mm diameter straws can provide spatial resolution of 4 mm FWHM or better (as a function of the I.D. of the straw tube body which carries such ¹⁰B layer/coating. Since an array of detectors of such form consists of several thousand individual straw elements per m², which can be read out independently in parallel, achievable event rates are very high. With appropriate electronics, a 1 m² detector could operate at an instantaneous count rate exceeding 3x10⁹ events/sec and each individual event could be time tagged with a time resolution of less than 0.1 µsec, allowing accurate identification of neutron energy by time of flight. Because neutron events produce markedly higher levels of counter gas ionization, gamma ray events can be essentially eliminated by simple application of a pulse height threshold. Monte Carlo simulations indicate that with such simple pulse height/energy level threshold technique the gamma ray discrimination factor can be more than 10⁸. Furthermore, considering basic elemental component cost, this new neutron imaging detector can be commercially produced economically, probably at a small fraction of the cost of a ³He detector of comparable performance.

The straw tube detector/array, as here described, provides a unique technique for imaging thermal neutrons with good spatial resolution and high sensitivity. This light weight detector, which operates at ambient pressure, affords a low cost technology with highly advantageous properties for use in neutron scattering experiments, neutron imaging for contraband and nuclear weapons detection and identification, and other field applications of neutron surveillance.

The described straw detector element can additionally provide a low cost efficient neutron detection device for survey applications such as might be required for detection of fissile materials which emit neutrons. In such applications it may be advantageous to employ coatigs which substantially deviate from the optimally efficient thickness described and which may for purposes of optimization of cost deviate as well from the optimum (maximum) concentration of ¹⁰B. In such applications the detection of neutron interaction position may be of little interest further reducing cost and complexity be eliminating complex position decoding circuitry. <ul id="ul0001" list-style="none"><li><figref idref="f0001">Figure 1</figref> illustrates a flat panel form of a closely packed array of straw tube neutron detectors (view A) and a curved panel form of such detector array (view B).</li><li><figref idref="f0002">Figure 2</figref> illustrates in sectional view the basic straw tube structure for detection of neutron capture reaction products.</li><li><figref idref="f0003">Figure 3</figref> illustrates a side view and a cross sectional slice of a twister element which is a component of the end fitting of each end of the basic straw tube detector of <figref idref="f0001">Figure 1</figref>. The cross sectional area view is taken across a width that encompasses 180° of turn of the helix of the twister. The circumference of the twister is also portrayed.</li><li><figref idref="f0004">Figure 4</figref> illustrates the straw tube end fitting sleeve and a twister that is securely locatable therein, this to form the end tube fitting for the straw tube of <figref idref="f0001">Figure 1</figref>.</li><li><figref idref="f0005">Figure 5</figref> illustrates the cathode board from a perspective of its outer face, which together with a anode board as per <figref idref="f0006">Figure 6</figref> form a module end fitting for an array of straw tube detectors.</li><li><figref idref="f0006">Figure 6</figref> illustrates the anode board from a perspective of its outer face, which together with an cathode board as per <figref idref="f0005">Figure 5</figref> form a module end fitting for an array of straw tube detectors.</li><li><figref idref="f0007">Figure 7</figref> illustrates in cross sectional view of one end of a module of a closely packed array of straw tube detectors as this array is assembled with the cathode board - anode board of <figref idref="f0005">Figures 5</figref> and <figref idref="f0006">6</figref> and as the anode wire of each straw tube detector is electrically connected to its corresponding terminal in the anode board.</li><li><figref idref="f0008">Figure 8</figref> illustrates an N x M matrix read out circuit for an N x M array of straw tube detectors assembled in a module.</li></ul>

<b>DETAILED DESCRIPTION OF THE INVENTION</b>

Embodiments of the invention provide a neutron detector of practical utility, of a large area detection capability, and in its preferred embodiment a capability of detecting neutron interaction within a reasonably precise spatial location of such large area neutron detector.

The detection of thermal neutrons is based on the generation of secondary radiations, following the capture of neutrons in a converter material. A converter material reaction commonly used is that of neutron (n) capture in boron-10 (¹⁰B), described by, <maths id="math0001" num=""><math display="block"><msup><mrow /><mn>10</mn></msup><mi mathvariant="normal">B</mi><mo>+</mo><mi>n</mi><mo>−</mo><msup><mo>→</mo><mn>7</mn></msup><mi>Li</mi><msup><mo>+</mo><mn>4</mn></msup><mi>α</mi><mfenced><mrow><mn>2.792</mn><mi> MeV</mi><mi>, ground state</mi></mrow></mfenced><msup><mi>and </mi><mn>7</mn></msup><mi>Li*</mi><msup><mo>+</mo><mn>4</mn></msup><mi>α</mi><mfenced><mrow><mn>2.310</mn><mi> MeV</mi><mi>, excited state</mi></mrow></mfenced><mo>.</mo></math><img file="EP1578375B1_D0001.tif" /></maths> The energy released by the reaction is 2.310 million electron volts (MeV) in 94% of all reactions (2.792 MeV in the remaining 6%), and equals the energy imparted to the two reaction products (the energy of the captured neutron is negligible by comparison). The reaction products, namely an alpha particle (α) and a lithium nucleus (⁷Li) are emitted isotropically from the point of neutron capture by ¹⁰B in exactly opposite directions and, in the case of the dominant excited state, with kinetic energies of 1.47 MeV and 0.84 MeV, respectively (dictated by the conservation of energy and momentum).

Boron-lined proportional detectors based on the general principle above described have been employed for many years, but such detectors achieve at most a few percent efficiency, due to the fact that the ¹⁰B foil thickness therein needed for a substantial capture of such neutrons exceeds the escape range of the neutron capture reaction products; so little escape of the capture reaction products occurs. Thus, only conversions of neutrons in a very thin layer near the surface of the ¹⁰B foil adjacent the counting gas are detected. Since this very thin layer of the ¹⁰B foil captures only a very small percentage of the incident neutrons, efficiency of a neutron detector of such design is low.

Preparing the ¹⁰B converter in the form of a very thin layer on the inside wall of a thin walled cylinder structure allows one of the two charged reaction particles (the one directed inward) to escape and be counted, through ionization of a counter gas. The opposing reaction particle cannot penetrate the wall of even a very thin cylinder and thus does not excite an adjacent detector cylinder. Therefore, when a thin cylinder structure with a very thin interior coating of B is utilized, highly accurate location of the interaction reaction can be obtained.

One embodiment of the new detector design of this invention employs a closely-packed array of ¹⁰B₄C-very thin layer lined straw tubes. This design hence removes the barrier to efficient neutron capture reaction product escape while still providing for efficient neutron capture by providing a plurality of very thin ¹⁰B converters, each individual converter layer providing efficient reaction product escape. Using multiple stacks of such straw tube detectors, a reasonable stack depth allows a high neutron detection efficiency to be achieved on the 1-10 Å wavelength range of thermal neutrons. Furthermore, the elemental component of such detector, the straw tube element, has very low sensitivity to gamma radiation, which is an essential attribute in many applications.

As a first material in the production of the elemental straw tube component of the neutron detector there is a requirement for production of a narrow film having one surface coated with a layer of ¹⁰B, preferably in the form of B₄C, preferably of a thickness of 0.5 to 3 µm. From such B₄C coated narrow film strips or ribbons there is produced small diameter straw tubes of up to 2 meters in length having the thin layer of B₄C preferably of a thickness of 0.5 to 3 µm on their interior surface. Such straws tubes are then installed with an end fitting means, affixed in each end of said straw tube, which is capable of receiving and positioning a wire centrally within said straw tube out of electrical contact with said coating of B₄C and capable of allowing a flow of gas through said tube. The straw tube is then provisioned with a wire positioned centrally within said straw tube out of electrical contact with said coating of B₄C. This structure is the elemental neutron detector structure which is illustrated in <figref idref="f0002">Figure 2</figref> in a cross section view. For convenience the end fitting are not illustrated in <figref idref="f0002">Figure 2</figref> itself. In <figref idref="f0002">Figure 2</figref> the straw tube detector element 10 comprises a straw tube body 12 the interior surface of which carries a B₄C coating 14 and an anode wire 16 which is centrally located in tube 12 out of electrical contact with the B₄C coating 14. Preferably the B₄C coating has an enriched content of ¹⁰B compared to the natural abundance of ¹⁰B in elemental B. The thickness of the B₄C layer is not greater than the thickness of B₄C through which the neutron capture reaction products can penetrate, or escape, i.e., no greater than about 3.5 µm. Although B₄C is a preferred material from which to form the thin ¹⁰B layer, alternative ¹⁰B containing substances can be used, such as elemental B, B₁₀H₁₄, BN, B₂O₃, B₆Si, B₃Si, B₂S₅, B₂S₃ and the like. Whatever boron compound is employed to provide the ¹⁰B content of the coating is applied to the substrate surface at a rate of 0.12 mg/cm² to about 1.2 mg/ cm², preferably at a rate of 0.19 mg/ cm² to about 0.5 mg/ cm², and most preferably at about 0.25 mg/ cm².

Aluminum and plastic were evaluated as backing film materials for a B₄C coating and as the base material for the straw tube body. Mylar® (C₅H₄O₂) plastic film was used. Since the density of aluminum (2.7 g/cm³) is nearly double that of Mylar® (1.4 g/cm³), the plastic is advantageous with respect to minimization of gamma ray sensitivity. Another criterion in selection of the straw material is maintenance of structural integrity. Any other plastic film forming material of a density less that aluminum may be employed. Such plastic materials for straw formation are well known to those skilled in the art.

Boron, in the form of boron carbide (B₄C), was vapor deposited on aluminum and Mylar® foils using a plasma deposition process. The B₄C coating was found to be extremely adherent and mechanically stable when applied to substances like aluminum and plastics. Deposition was accomplished by wrapping narrow 9.5 mm wide and 25 µm thick strips or ribbons of the respective materials around a cylindrical drum (16" diameter x 16" long), which thereafter was rotated adjacent to a sputtering head. A tape running down the side of the drum kept the strips in place (but also produced a 1 cm dead space every 50"). Using this process, a highly uniform boron carbide coating was achieved on continuous strips of material with a length up to 50 meters. Such continuous material is required in straw construction, which utilizes high speed automated equipment.

Straws are manufactured using a high speed winding technique in which narrow ribbons of plastic or metal-coated plastic film are helically wound around a cylindrical mandrel of precise dimension. Quickset adhesive may be applied to the film on the fly to instantly bond the multiple layers of plastic film together. In this manner, large quantities can be produced at high rates and thus low cost.

Samples of each B₄C coated material, in lengths of 10-25 meters, were sent to Lamina Dielectrics, Ltd., Sussex, England, and about 35 meters of 4 mm diameter straws were produced with various boron coating thicknesses and using both the Mylar® and aluminum backing. For several practical reasons, the Mylar® straws were found to be superior to the aluminum straws. During straw fabrication, the aluminum material was found to be susceptible to breakage, resulting in interruption of the winding process in order to splice in additional material, which is incompatible with efficient commercial production. The aluminum straws also were more difficult to work with during detector assembly. Greater care was required to avoid crushing the aluminum straws when cutting them into precise working segments. Finally, the aluminum straws demonstrated no advantage with regard to rigidity or basic performance. Mylar® backing was concluded to be the better structural choice. However because plastic contains high levels of hydrogen having a high elastic scattering cross section for thermal neutrons aluminum may nonetheless be the preferred material in many applications.

Ultimately an anode wire is precisely centered within the straw tube. Although laborious, this wire installation-centering could be done on a straw by straw basis and then the anode wire installed tubes could be assembled into an array of straws. However, the preferred assembly procedure for the large area thermal neutron detector is described below.

A large area detector based on this straw technology can be achieved practically by dividing the detector into a modest number of independent modules, each consisting of an N x M array of straws. For example, using 50 straw modules, a 5 cm thick, 1 m² detector, consisting of about 3600 straws, would consist of 72 modules in a close packed configuration. Thus, the feasibility of production of densely packed large-area arrays of such straw tube detectors depends upon practical and relatively low-cost methods of module fabrication and readout. Although each straw tube detector serves as an individual detection element, the module serves as a basic structural and readout unit. Therefore, techniques for volume fabrication and assembly in modules were considered from the outset and developed hand-in-hand with the straw tube detector design.

<figref idref="f0007">Figure 7</figref> shows a cross-sectional schematic of the module design. A "twister" component, shown in greater detail in <figref idref="f0003">Figure 3</figref>, was developed to permit precise wire centering. The twister component was machined from a high performance insulating polymer which provides excellent mechanical properties and can be machined and bonded easily, for example Ultem® 1000 (General Electric Company). As illustrated in <figref idref="f0003">Figure 3</figref> the twister 20 has a helical structure and its length may be about 0.110"-0.115". In any given thin cross-section of the twister over half of the cross-sectional area 21 of the twister is open space 24. When these cross-sections are summed together to form a complete helix there exist a cylindrical space 22 running down the entire length of the twister which approximates the diameter of the anode wire it will ultimately receive, for example 0.003" - 0.005" in diameter, and this cylindrical space 22 is precisely centered. Advantages of this twister design are convenience, low-cost fabrication and compatibility with precise anode wire centering through efficient air-driven wire placement.

As shown in <figref idref="f0004">Figure 4</figref> view B the twister 20 is housed inside a simple end fitting sleeve 30 (view A of <figref idref="f0004">Figure 4</figref>), which was manufactured on a bench lathe from a high performance insulating polymer which provides excellent mechanical properties and can be machined and bonded easily, for example from Ultem® 1000. The end fitting sleeve is open-ended 32 and 34 and has a bottle-shape and a central channel 36 that tapers in slightly to form a shoulder 38 so that the twister 20 can be seated firmly against shoulder 38. Using a micro drop of glue, the twister component may be permanently affixed inside the end fitting sleeve, and identical end fittings 50, comprising a sleeve and twister, are epoxies into each end of the straw. While the base 40 of the fitting sleeve 30 sits inside the straw, a 'neck' 42 of a reduced diameter protrudes out and serves as a male connector to a fiberglass board assembly that is mounted at each end of the array.

As shown in <figref idref="f0005 f0006 f0007">Figures 5-7</figref> a dual-board assembly was developed that permits electrical connection, gas flow, and wire tensioning. This assembly consists of two (fiberglass) circuit boards preferably of fiberglass, a cathode board 52 as shown in <figref idref="f0005">Figure 5</figref> and an anode board 54 as shown in <figref idref="f0006">Figure 6</figref>. The neck 42 of straw end fittings 50 plug into apertures 57 positioned through the outer face 56 of the cathode board, which is coated with a uniform conductive layer, facilitating a single common high voltage connection.

<figref idref="f0005">Figure 5</figref> illustrates the cathode board 52 from the perspective of its outer face 56 whereas the cathode board 52 in cross sectional view may be seen in <figref idref="f0007">Figure 7</figref>. Cathode board 52 is prepared with apertures 57 that are dimensioned to receive the neck 42 of the sleeve of a tube end fitting. Apertures 57 are positioned across the outer face 56 of cathode board 52 in a pattern that corresponds to a straw tube array that will plug into the cathode board. As <figref idref="f0007">Figure 7</figref> illustrates, the cathode board 52 incorporates a milled recess 66 and a protruding lip 68.

Since individual anode signal readout is required, as best seen in <figref idref="f0006">Figure 6</figref>, each pad on the anode board 54 leads to a unique termination 58 adjacent to the corresponding aperture, or hole, in the cathode board. The anode board also included a hole 60 for a gas entry/exit port and high voltage connection. As shown in <figref idref="f0007">Figure 7</figref> the dual board design when assembled is also compatible with efficient threading, tensioning, and attachment of the anode wire. Anode wires are held with tension in electrical contact to the anode board using a tapered pin 62 and a brass eyelet 64. This procedure is explained in further detail below. As seen in <figref idref="f0007">Figure 7</figref>, the anode board has an inset 70 milled around the edge of the anode board which corresponds to the protruding lip 68 of the cathode board. With this design, the two boards fit together snugly and form an interior cavity 72. The straw end fittings form an open connection to the resulting chamber, which is fed by a gas port 60 of the anode board. In this manner, effective gas flow is achieved through a single entry and exit port. Finally, this dual board assembly is designed to have apertures 57 in the cathode board 52 and terminations 58 in the anode board 54 that match precisely the cross sectional area of a straw array, so that modules of straws can then be closely packed to form a large area detector, if desired. Electrical contact between the anode wire and the anode board is achieved using a brass eyelet 64.

Using a jig support structure to prevent crushing, straws are cleanly cut into the desired lengths, for example 10 cm. Assembly of straws into closely packed honeycomblike arrays is quite feasible by constructing individual rows, which are then stacked vertically. After inserting the twister-sleeve end fitting inside the straws, straws are arranged adjacent to each other atop a granite slab to ensure linearity. An alignment device may be used to gently hold the tubes in place while they are bonded together using spot application of fast setting cement applied with a hypodermic syringe. Each layer is then positioned so that straws of one layer fit into grooves created by the top of an underlying row of straws. This assembly technique results in very sturdy straw tube arrays.

In the first step of module assembly, the preconstructed straw tube array is connected to the cathode board. Conductive epoxy is liberally applied to the inside surfaces of the apertures 57 in the cathode board 52, and the straw array is plugged into the cathode board. The displacement of excess epoxy provides a secure electrical connection between the ¹⁰B₄C interior cathode shells and the cathode board which is coated with a uniform conductive layer. This process is performed for both ends of the array. The connection provides a secure seal, preventing gas leakage, and helps to maintain structural integrity. Using a micro drop of conductive epoxy, an eyelet is then affixed in each termination hole 58 of the anode board 54, thereby establishing electrical connection and fixation. Non-conductive epoxy is then applied around the edge of each eyelet forming a tight seal to prevent gas leakage. The anode board is then mated to the cathode board through the lip 68 of the cathode board 52 and the inset 70 milled around the edges of the anode board 54, and the lip-inset connection is sealed using epoxy to prevent gas leakage. Two redundant high voltage connectors and a gas flow port through aperture 60 are also connected to the anode board. The final product is shown in <figref idref="f0007">Figure 7</figref>.

Another advantage offered by this module design is a greatly improved technique for wire threading and tensioning. This technique allowed these steps to be performed for an entire array of straw tube detectors in a quick and efficient manner. A module, consisting of a straw tube array with a dual-board assembly structure mounted on each end, is mounted in a jig. A gas-driven threading system is implemented by connecting compressed gas, such as argon, (about 15 psi) through a plastic hose to a length of small diameter copper tubing in which a small hole has been drilled that approximates the diameter of the anode wire to be used. For each straw in the array, the copper tubing is inserted through a termination hole 58 in the anode board, through the chamber 72, and down to the end fitting of the straw tube corresponding to that termination hole. When a high flow of gas is achieved, a columnar gas flow through the copper tubing and the straw tube is created. A spool of anode wire is positioned at one end of the setup, and anode wire is inserted through the small hole drilled in the copper tubing. The rapid gas jet produces sufficient tension in the wire so that it may be easily fed through the straw by simply rotating the source wire reel until the wire appears at the output end of the straw. For each straw, after spooling the wire through its length, the copper tubing is removed, and a small brass pin 62 as illustrated in <figref idref="f0007">Figure 7</figref> is inserted to hold the wire in place against the metal eyelet 64 in the anode board. A small weight (20 g) is attached to the other end of the wire to apply required tension, and a second brass pin is inserted in the anode board at the other end of the module. The pin 62 is slightly tapered and closed on one end and machined to match the diameter of the eyelet 64, thereby ensuring a snug fit. In this manner, secure electrical connection to the anode board, reliable wire tension, and gas containment are maintained. Because highly accurate longitudinal position measurement is desired in the neutron detector, a high resistivity wire (like 20 µm Stablohm 800, California Fine Wire Co.) is utilized.

<figref idref="f0007">Figure 7</figref> illustrates one end of a modular assembly of closely packed straw tube detectors with anode wires installed and electrical connections between the cathode board and the B₄C cathode interior shell coating and between the anode wires of each tube with its respective anode board termination 58 is shown as all elements thereof are in their operative assembly. Not illustrated in <figref idref="f0007">Figure 7</figref> is the operative connection of gas part 60 in one anode board end since means for such connection to a counter gas, as like Ar 98%-ethane 2%, would be readily apparent to one skilled in the art.

Gas mixtures comprised of varying ratios of argon and ethane (C₂H₆) were examined to determine if there was an optimum level of quench ethane gas for stable detector performance, long lifetime, and gamma discrimination. It was found that relatively low levels of C₂H₆ improved the energy resolution of a straw tube arrayed detector, 19% FWHM at 5.9 keV for a 98% Ar 2% C₂H₆ mixture at a charge gain of 100. At larger gains, the energy resolution obtained with the 98/2 mixture degraded to about 25% FWHM which is still sufficient to easily discriminate neutrons from gammas. A low percentage of ethane quench gas also reduced the level to which neutron events were saturated. Saturation of the neutron events results in a measured energy that is lower than is actually deposited in the detector, i.e. nonlinear performance.

To determine what benefits could be obtained by varying the mixing ratios of Ar and C₂H₆, neutron spectra were taken for four different gas mixtures: Ar/C₂H₆ (50/50), Ar/C₂H₆ (72/28), Ar/C₂H₆ (92/8), and Ar/C₂H₆ (98/2). The operating voltage for read out for each mixture was chosen so that spectra could be taken at an equivalent charge gain of about 104, as determined by the amplitude of pulses from 5.9 keV ⁵⁵Fe X-rays. By comparing the neutron spectra taken for the different gas mixtures dual advantages of an ameliorated saturation effect and improved energy resolution for a small admixture of ethane was seen to allow for better gamma discrimination at high detector gains. For a constant gain, a reduced amount of ethane quench gas improves the distinction between the rapidly falling, low deposited energy gamma events, and the neutron events that form a broad peak in the upper higher energy channels. Using a small percentage admixture of ethane quench is also beneficial from the perspective of lifetime extension since there will be less hydrocarbon available to polymerize and deposit onto the anode wire. A 98/2 mixture of Ar/C₂H₆ minimizes the saturation of neutron events seen at high gains and should increase the lifetime of the straw detector.

Because of the large number of detector elements, straw readout issues are of importance. Modularization of the arrays of elemental straw tube detectors greatly reduces the number of channels and, accordingly, the amount of electronics required, lowering cost but raising an issue of straw decoding (i.e. which straw tube detector in a module fired). Delay line readout offers a viable approach because signals from several inputs can be differentiated based on time of arrival. Taking advantage of the high anode resistivity and using additional termination resistors, straws can be connected along rows on one end and along columns on the other end, as illustrated in <figref idref="f0008">Figure 8</figref>. Thus, the module is read out as an N x M matrix using two multitap delay lines, one for rows and one for columns. Difference in time of signal arrival enable determination of which row and column fired. Because it requires a relatively low number of electronics components, the technique can be implemented quite cost effectively.

In order to achieve efficient, optimally cost effective readout of large arrays of boron straws, the electronics must combine straw identity decoding with resistive charge division readout of longitudinal position. Delay lines offer an attractive approach to straw identity decoding because signals from many straws can be fed into a single amplifier with a small delay separating each. A problem with this approach is that straw identity is lost in such passive addition of signals. Therefore, a method of decoding straw identity was developed through use of a simple delay line method. In this technique as illustrated by <figref idref="f0008">Figure 8</figref>, anode signals are connected through the termination resistors to successive taps on a multitap delay line. Since signals are typically very fast, a delay of 5-10 ηsec is quite sufficient to allow decoding based on time.

These small delays interfere very little with the integration of the charge delivered in the signal and thus resistive longitudinal readout is not impaired. This technique takes advantage of the additional powerful method of row-column organization to further achieve reduction of components. Such a system, applied to a 20-straw array of 2 mm diameter straws, is shown in <figref idref="f0008">Figure 8</figref>.

One end of the straw array module had only a passive U-connection board, where each pair of adjacent straw anodes were connected together, so that all readout electronics may be placed on the other end. This scheme also reduced the number of taps needed on the delay line. Thus, although the straw array illustrated by <figref idref="f0008">Figure 8</figref> is actually a 5 x 4 row-column matrix, it was read out as a 5 x 2 matrix of 20 cm long "effective", U-connected straws, which had both readout ends on the same end of the module.

The "effective straws" of the above example were connected together through resistors along rows on one end and along columns on the other end. Each row was connected to a tap in a 5-tap delay line, and each column was connected to a tap in a 2-tap delay line. Since each tap delays the signal by a fixed time, the difference in time of arrival at each end of the delay line indicated which row or column contained the firing straw. Events were histogrammed based on the differences in arrival times, and a count profile was generated for each delay line. For each delay line, the outputs from both ends of the delay line were also summed together to give A and B signals. Thus, spatial resolution was measured using an A/(A + B) charge division technique with A and B measurement performed with a charge integrating ADC.

In the above example of a readout system, "effective" straw anodes were connected along rows on one end and along columns on the other end. Thus, all anodes were in effect connected together, raising some concerns regarding degradation of spatial resolution. The success of this approach largely relies upon the inherent resistance of the system. The count profiles produced by each delay line were well-differentiated peaks, allowing exact straw identification. Use of delays lines with greater time delay per tap would enable peaks to be even better resolved. In addition, excellent charge division spatial resolution of 1.74 mm was achieved. This performance was equivalent to that displayed by the module without delay lines, which showed 1.83 mm resolution.

As described above, in order to accommodate the needs of a large variety of applications, a generic straw detector module may be constructed distinct from the electronics used to read it out. Varying electronics modules can then be used to process these signals according to the needs of the specific application.

In a 2D application, only the XY location of an event is desired while the Z-coordinate can be ignored. This low-cost solution, as compared to full 3D readout, will serve as a simple planar detector, enabled to decode events in a 2D plane. Position in one dimension is achieved using charge division along the length of the detector module, while the other coordinate can be determined by using a simplified version of the row-column readout. This information, along with longitudinal position, will give an XY coordinate of the event detected.

If a more accurate readout system is desired, a fully 3D readout system can be employed simply by using a more advanced plug-in electronics module. This 3D module will interface with the detector in the same way as the 2D module did, but will incorporate a delay-line readout system in both the row and column directions. This will allow the detector to read out an accurate 3D position. Both 2D and 3D plug-in electronics modules may be built that achieve high position resolution and can interface with a large array of neutron straws.

Another factor that can vary from application to application is the count rate desired from the detector. Current paralleling techniques that connect a large number of straws together into one amplifier cannot handle very large count rates due to the large dead time of such a system. In a paralleled system, only one straw in a module can fire at a given time without ambiguity. If a very high count rate is desired, a separate readout system can be used for each straw individually, instead of a module in aggregate. For a 50-straw detector, this will allow a 50-fold increase in count rate while still maintaining spatial resolution. If a 2D system is required, each row, as opposed to each straw, can be connected to a separate amplifier for a lower cost solution than the full 3D system.

The thermal neutron sensitivity of a 10 x 5 array of straws was measured. The array was populated by 34 active straws of the 4 mm diameter and 77 mm active length, and with a B₄C thickness of 2 µm. All anodes were tied together, such that events were registered from any one of the 34 straws. A neutron source was placed 10 inch away, and the flux was measured with a BF₃ tube. For a flux of 5.04 nv, there were 17.3 neutron counts per second in the array, resulting in a sensitivity of 3.4 cps/nv. This sensitivity predicts a single tube sensitivity of 0.10 cps/nv, which agrees very well with the value of 0.09 cps/nv measured for a single straw.

Fabrication of a rugged 50-straw, 10 cm long module was achieved utilizing a production technique as previously described. The end fittings and twister components were manufactured using low cost base materials and simple fabrication techniques, which permit both large-scale production and low per-unit expense. Since the large area neutron detectors will consist of several thousand straws, these factors are particularly important for these parts, which will be required in duplicate for each straw. Furthermore, the dual-board design provides a practical solution to module fabrication. Another key advantage of this design is that the main construction steps (wire centering and placement, gas flow, electrical connection) are met on the level of the module rather than for each straw element, and fabrication involves a series of operations that are performed for the array as a whole. Such an approach is much more practical than painstaking fabrication of individual 4 mm straw detectors.

In addition to the construction of a 50 straw module a 1 meter straw was constructed. The length of such a straw creates an issue of wire electrostatic instability. In order to maintain wire stability, an additional twister, with an Ultem housing sleeve, was positioned precisely at the center of the 1 m length. The twister was contained in a cylindrical shell, open on both ends, permitting the gas driven wire placement. The twister shell was tacked in place by piercing the straw body with a small gauge hypodermic needle and injecting a small drop of epoxy.

The successful construction of a 50-straw module and a 1 meter straw detector together provide proof that the proposed assembly techniques can be used to produce rugged large area detectors of 1 m² area.

The efficiency of the proposed detector is dictated both by the probability of slow neutron interactions in the B₄C lining, and the ability of the ¹⁰B (n, α) reaction products to escape the lining and enter the counting gas, where they can be counted. Whereas the probability of neutron interactions in boron increases for a thicker lining, the reaction products have a limited range and will reach the gas only if the B₄C layer is thin. An array of B₄C -lined straw tubes offers improved detection efficiency, by providing more target material for incident neutrons, while at the same time maintaining a thin lining.

The detection efficiency (ε) for neutrons in an array of the proposed detector is given by <maths id="math0002" num=""><math display="block"><mi>ε</mi><mo>=</mo><msub><mi mathvariant="normal">f</mi><mi mathvariant="normal">e</mi></msub><mfenced><mrow><mn>1</mn><mo>−</mo><msup><mi>e</mi><mrow><mo>−</mo><mi mathvariant="italic">Nσt</mi></mrow></msup></mrow></mfenced></math><img file="EP1578375B1_D0002.tif" /></maths> where σ is the neutron cross section for the ¹⁰B (n, α) reaction and <i>N</i> is the effective nuclear density of ¹⁰B in a detector array of depth <i>t</i>. The neutron cross section is a function of the wavelength λ of the neutron, given by σ = 2133σ (for λ in Å and σ in

barns). The nuclear density N is computed as N = N_B(<i>II</i> d)/(0.866 D) where D is the tube diameter, d the ¹⁰B₄C film thickness in each tube, and N_B the nuclear density of ¹⁰B in ¹⁰B-enriched B₄C (1.10 x 10²³∼atoms/cm³). The fraction of reaction products that ionize the counter gas, f_e, has been determined computationally, in Monte Carlo simulations, and as also a function of the B₄C layer thickness d.

The predicted efficiency for room temperature thermal neutrons (0.0253 eV) was plotted as a function of the B₄C film thickness, for array depths of 5 cm and 10 cm, and for a single straw tube 4 mm in diameter. Maximum efficiencies were observed at decreasing film thicknesses as the array deepened. Thus, a maximum efficiency of 65% was achieved for the 5 cm-deep array, at a B₄C film thickness of 1.0 µm; the 10 cm deep array can offer close to 78% efficiency, but only if the B₄C film is 0.7 µm thick. By comparison, a single, planar layer of B₄C, that lines a charged particle detector, has a maximum thermal neutron (1.8 Å) detection efficiency of only 4.7%, at a film thickness of 3.2 µm. This film thickness equals the average range of α particles in B₄C.

The improvement in efficiency offered by the proposed detector is due to the fact that, although the neutron interaction probability is benefited by the multiple layers of tubes, the B₄C lining remains thin enough to allow escape of the neutron reaction products.