WO1992009088A1

WO1992009088A1 - Improved multiple channel configurations for conditioning x-ray or neutron beams

Info

Publication number: WO1992009088A1
Application number: PCT/AU1991/000530
Authority: WO
Inventors: Stephen William Wilkins; Keith Alexander Nugent; Henry Nicholas Chapman
Original assignee: Commonwealth Scientific And Industrial Research Organisation; The University Of Melbourne
Priority date: 1990-11-20
Filing date: 1991-11-20
Publication date: 1992-05-29

Abstract

This invention concerned generally with x-ray and neutron beam instrumentation, and in particular relates to the focusing and collimation of x-rays or neutrons. In one embodiment the invention comprises an x-ray or neutron instrument incorporating x-ray or neutron lens means disposed in a path for x-rays or neutrons in the instrument, the lens means comprising multiple elongate open-ended channels arranged across the path to recieve and pass segments of an x-ray or neutron beam occupying said path, which channels have side walls reflective to x-rays or neutrons of said beam incident at a grazing angle less than the critical grazing angle for total external reflection of the x-rays or neutrons, whereby to cause substantial focusing or collimation of the thus reflected x-rays or neutrons, wherein said channels are of a quadrilateral cross-section.

Description

"IMPROVED MULTIPLE CHANNEL CONFIGURATIONS FOR CONDITIONING X-RAY OR NEUTRON BEAMS"

This invention is concerned generally with x-ray and neutron beam instrumentation, and in particular relates to the focusing and collimation of x-rays or neutrons.

X-ray mirrors of various types, generally relying on grazing angle incidence of the x-rays, have long been used in some x-ray scattering instruments to provide a means of focusing x-rays and improving flux and intensity, relative to pin-hole optics, by increasing the angular acceptance of the system with respect to the x-ray source. These methods for enhancing intensity have not found widespread application in x-ray scattering instruments because they lack spatial compactness, and flexibility in use, and are awkward to align. In the case of x-ray optical systems, simultaneous high-resolution in wavelength, angular collimation and spatial extent are usually achievable only at the expense of considerable loss in flux and intensity.

An early proposal for an x-ray collimator consisted of two glass plates facing each other at a small angle. This principle was extended in a conical x-ray guide tube proposed by Nozaki and Nakazawa [J.Appl.Cryst. (1986) 19,453] .

Another class of known focusing devices is the zone plate, which relies on diffractive optics, but these have not proven effective at shorter wavelengths.

In a recent paper, Yamaguchi et al [Rev. Sci. Instrum. 58(1), Jan 1987, 43], there has been proposed a two dimensional imaging x-ray spectrometer utilizing a channel plate or capillary plate as a collimator. It is apparent that Yamaguchi et al are treating the channel plate as a large aperture device acting solely as a set of Soller slits consisting of an array of channels surrounded by opaque walls.

The present applicant's international patent application PCT/AU87/00262 (WO 88/01428) discloses a focusing device which utilises multiple elongate open-ended channels. That application discusses only cylindrical channels and flat plate geometries in detail, and it has been appreciated, in accordance with the present invention, that there are other enhanced channel arrangements which provide improved performance, e.g. enhanced focussing efficiency.

In a first aspect, the invention accordingly provides, in an x-ray or neutron instrument incorporating x-ray or neutron lens means disposed in a path for x-rays or neutrons in the instrument, the lens means comprising multiple elongate open-ended channels arranged across the path to receive and pass segments of an x-ray or neutron beam occupying said path, which channels have side walls reflective to x-rays or neutrons of said beam incident at a grazing angle less than the critical grazing angle for total external reflection of the x-rays or neutrons, whereby to cause substantial focusing or collimation of the thus reflected x-rays or neutrons, wherein said channels are of a quadrilateral cross-section.

An advantageous embodiment of the invention, for enlianced focussing efficiency, comprises plural segments arranged about an axis, wherein each segment has a multiplicity of said channels arranged at a prescribed orientation with respect to said axis. A preferred such orientation is to have diagonals of rectangular-section channels parallel to the central such diagonal intersecting said axis. The invention moreover provides, in a second aspect, in an x-ray or neutron instrument incorporating x-ray or neutron lens means disposed in a path for x-rays or neutrons in the instrument, the lens means comprising multiple elongate open-ended channels arranged across the path to receive and pass segments of an x-ray or neutron beam occupying said path, which channels have side walls reflective to x-rays or neutrons of said beam incident at a grazing angle less than the critical grazing angle for total external reflection of the x-rays or neutrons, whereby to cause substantial focusing or collimation of the thus reflected x-rays or neutrons, wherein said channels are grouped in a plurality of segments arranged about an axis, each segment having a multiplicity of said channels arranged at a prescribed orientation with respect to said axis.

The invention still further provides in a third aspect, an x-ray or neutron instrument incorporating x-ray or neutron lens means disposed in a path for x-rays or neutrons in the instrument, the lens means comprising multiple elongate open-ended channels arranged across the path to receive and pass segments of an x-ray or neutron beam occupying said path, which channels have side walls reflective to x-rays or neutrons of said beam incident at a grazing angle less than the critical grazing angle for total external reflection of the x-rays or neutrons, whereby to cause substantial focusing or collimation of the thus reflected x-rays or neutrons, wherein said channels comprise a plurality of concentric annular channels.

The instrument will typically though not necessarily include a source of x-rays and may have one or more slit assemblies, a monochromator, a sample goniometer stage and/or adjustable x-ray detector. In one embodiment, the side walls are planar but their inclinations progressively change from channel to channel with respect to the optical axis of said path whereby to enhance focusing or collimation of said incident beam.

Preferably, the outer side wall of each channel itself varies in inclination along the length of the channel to further enhance said focusing and collimation.

The device is preferably such that these inclinations can be adjusted, at least finely, on installation of the device in the instrument.

As employed herein, the terms "focus" and "collimate" are not strictly confined to beams convergent to a focus or substantially parallel, but respectively include at least a reduction or increase in the angle of convergence or divergence of at least a part of the x-ray beam in question. The term "lens" embraces beam concentration devices generally. The term "channel", as employed in the art, does not specifically indicate an open-sided duct but also embraces wholly enclosed passages, bores and capillaries.

The channels are preferably hollow capillaries or other bores and may comprise collectively a micro-capillary or micro-channel plate. Advantageously, the channels are formed in an integral element, for example by a micromachining technique such as x-ray lithography, a LIGA method, or ion beam etching. The integral element may be e.g. glass or methyl methacrylate. Alternatively, the channels may be formed of multiple hollow optical fibres or multiple optical fibres from which the core has been etched out. In general, the interior of the channels can be air and should be of a higher refractive index for x-rays than the surrounds. This requirement is met by hollow air filled ducts or channels in a suitable glass.

For optimum focussing efficiency with square channels and only two reflections in each channel (one from each of adjacent walls), and where the x-radiation is harder, say in the wavelength range 0.2 to lOA, the channels should have a diameter to side length ratio d/t approximately equal to θ /-.2 where θ is the critical grazing angle for total external reflection. In general, d/t is preferably in the range half to one times θ . Where the radiation is softer, say in the wavelength range 200 to 1000A, the reflectivity decreases gradually with increasing glancing angle and reaches zero at an angle θ . In the case, optimum focussing efficiency occurs when d/t = θ /2.

By "focussing efficiency" herein is meant that proportion of the rays within the solid angle which enter the channels that are reflected into the focal rectangle

In an embodiment of enhanced focussing efficiency, the lengths of the channels are varied according to a predetermined thickness profile.

It will be appreciated that not all rays will necessarily intercept channel walls and that a substantial portion of the x-ray beam will typically be absorbed in the channel walls or pass undeviated through the focusing device.

In an advantageous application of the invention, the x-ray lens device comprises a micro-capillary plate which is curved so that the angular tilts of the reflecting side walls in the channels vary parabolically with distance perpendicular to the optical axis. By parabolic bending in one or two dimensions, appropriate focusing and collimating effects may be simultaneously produced in the two dimensions - and may well be different in the two dimensions.

The invention will now be further described, by way of example only, with reference to the accompanying drawings, in which:

Figure 1 comprises two diagrams which depict the dual reflection route of an x-ray beam traversing one rectangular section channel of a channel plate;

Figures la and lb are schematic diagrams of possible non-limiting arrangements in accordance with the invention.

Figure 2 is a schematic cross-sectional diagram of a simple focussing x-ray instrument according to the first aspect of the invention, showing ray lines for a single rectangular channel of the lens device incorporated therein;

Figure 3 is a diagrammatic front elevation of an— -ray focussing plate according to the second aspect of the invention, formed with micro-channels in segments arranged with 8-fold rotational symmetry; and

Figure 4 is a schematic cross-sectional diagram of an x-ray focussing plate according to the third aspect of the invention, having a central cylindrical channel and surrounding annular channels; and

Figure 5 is a diagrammatic front elevation of a further embodiment of x-ray focussing plate which combines the second and third aspects of the invention. International patent application PCT/AU87/00262 (WO 88/01428) explains the essential principles involved in multiple channel focussing instruments in accordance with the various aspects of the present invention. In the first aspect, a two-dimensional array of aligned channels of rectangular cross-section is formed in a plate, preferably by micromachining discrete bores in the plate.

As explained in the aforementioned international application, the channel plate can be flat, in which case all channels will be parallel, or it can be parabolically, spherically or cylindrically curved. X-rays are focussed by total external reflection from two adjacent orthogonal walls in each rectangular channel, similar to the "corner cube" effect. The channel lengths are arranged to be just long enough to prevent more than two reflections. Figure 1 depicts the dual reflection mechanism, in side and front diagrammatic perspective views of a single channel.

The channel plate can be made of glass, and could be coated with a material of high atomic number to increase the critical angle of reflection, when the radiation is in the hard x-ray regime, or to increase the reflectivity when the radiation is in the soft x-ray regime.

The channel plates may be manufactured using fibre- drawing techniques, in which a glass preform is drawn, bundled, then drawn again. The preform would be a glass rod of square cross-section, and could either be hollow or have a core of softer glass which is later etched away. Another method is to use photolithography. To obtain the high channel aspect ratios and tolerances required, an x-ray LIGA process may be implemented. These are processes of preference in that replicas can be formed from a master by injection moulding so that a great variety of materials may be used, including metals. LIGA is a German language acronym for lithography, galvanoforming and plastic moulding: The technique is described in Becker et al, Microelec. Eng. 4, 35(1986) "Fabrication of microstructures with high aspect ratios and great structural heights by synchrotron radiation lithography, galvanoforming, and plastic moulding".

The image of a point source focussed by a flat or spherically curved plate is a rectangular focal spot with side-lengths dx(M+l) and dy(M+l), where M is the magnification. The focus has a pyramid-like intensity profile. The image also contains two orthogonal line foci, one of width d (M+l), and one of width d (M+l) . These line foci have triangular intensity profiles. The intensity of the line foci is much less than the intensity of the focal rectangle. For imaging, it is desirable to make the focal rectangle as small as possible, which means that, at most, the channel cross-section must be as small as the x-ray source size. The relation between the source to channel plate distance

1s, and the channel plate to image distance 1X. is

1/1 - 1/1.= 2/R, where R is the radius of curvature of the plate. Collimation occurs when 1 = R/2. The transverse magnification is given by M = l./l , which is unity for flat plates.

The flat and curved plates produce imaging systems which are completely spatially invariant, and hence alignment is similar to that of an optical lens. Flat plates are extremely insensitive to misalignment due to the fact that they are invariant under translations.

Figures la and lb show possible arrangements in accordance with the invention. Figure la shows the use of a microchannel plate as an x-ray focusing element in an x-ray scattering instrument. Microchannel plate 52 focuses x-rays emanating from source 50. The x-rays are directed to sample 56 which is located on a translation/rotation stage via monochromator 54. X-rays scattered by the sample are detected by detector 60. Figure lb illustrates an arrangement in which microchannel plate focusing device 70 is used to focus x-rays emanating from source 50 onto sample 72 located on a translation/rotation stage or goniometer. X-rays scattered by sample 72 passing through slit 74 are detected by x-ray detector 76 which may be adjustable. It is to be emphasised that the present invention is not restricted to the particular arrangements shown in figures la and lb.

Figure 2 illustrates the case of a plate 10 with a two-dimensional array of micro-capillaries in the form of square channels 12 of side d. In use, a divergent beam 14 from source 5 may be focussed as a convergent beam 16 onto focal square F. The aspect ratio of the channels (side length d: channel length determined by plate thickness t) must be optimised for the type of radiation to be imaged or collimated. For x-radiation in the wavelength range 1 to 10 angstrom, the channel reflectivity vs. glancing angle has a very sharp cut-off at the critical angle, θc. In this case optimum focussing efficiency occurs when d/t= θ / 2. For softer radiation in the range 200 to 1000 angstrom, the reflectivity decreases gradually with increasing glancing angle, and reaches zero at an angle θ . In this case optimum focussing efficiency occurs when d/t = 0 /2. In the hard x-ray case θ is typically 5 mrad, giving an optimum aspect ratio of 280, while in the soft x-ray case θ is typically 0.1 rad, giving an optimum aspect ratio of 20. For perfectly smooth channel walls and perfect channel alignment, 34% of the rays within a solid

2 angle of (2M θ /(1+M) ) which enter the channels will be reflected into the focal square. This proportion is herein termed the focussing efficiency.

The focussing efficiency can be improved further by locally varying the thickness, locally varying the channel cross- section, or by segmenting the plate. For a flat plate with square channels, a thickness profile of the form t(x,y) = 2dl /[(|x| + |y|)x] where x = J_$(l+1/M) will increase the focussing efficiency to for the hard x-ray case and a source located a distance 1 from the channel plate. A focussing efficiency approaching 100% can be achieved for a plate with rectangular channels which vary as d_χ(x) = t|x|x/l_s; d_y= t|y|χ/l_s where d_χ and d are the channel side lengths parallel to the x- and y- axes respectively. This design approaches 100% efficiency only for one set of source and image distances. The focus will be square with a width M+l times the maximum channel width.

Figure 3 diagrammatically shows, in front elevation, a segmented plate 10' . Here the square-section channels 12' are grouped in segments 20 having eight fold rotational symmetry about an axis 22 at the centre line of the x-ray beam. Each segment 20 has a symmetrical kite shape of which the major axis is a radius from axis 22 and in which diagonals of the square-channel cross-sections are all parallel with this major axis. It will be appreciated that plate 10' may be viewed in plan as formed by superposing a square array of channels on another square array with a relative rotational displacement of 45°.

This segmented design gives an improved efficiency for any set of source and image distances. For the eight-fold rotationally symmetric square channel plate of Figure 3, the focussing efficiency is 48%. - li ¬ lt should be noted that the plate of varying thickness profile and the segmented plate of Figure 2 do not produce spatially invariant imaging systems. The efficiencies quoted for these designs apply only for the on-axis source point, or to be more precise, the square region of the source of side length d centered on the beam axis.

The tolerances preferred for all of the designs discussed thus far are quite strict. The degree of alignment of channels is preferably such that the average divergence

Δ§ of the channels leads to a spreading of the focal spot by a distance much less than d. This leads to the preferment that Δ§c<<d(l+M)/(21s).

The angle between adjacent channel walls is also important. The angular deviation Δ$ of channel walls from orthogonality is preferably constrained by

Δ§ <<d(l+M)/(21 ) so that the focal point is w s not spread by more than a distance (1+M)d. For a plate used with hard x-rays and parameters d=10μm, t=2.8 mm, and 1s= 100 mm, the above inequalities imply that over the length of the channels, for an optimum focus, the deviation of the channel walls from the desired direction should be less than O.lμm, and the angular deviation should be less than O.lmrad. Larger deviations will result in a broader, less intense focus.

Due to the high focussing efficiency and small image aberrations, extremely high intensities may be achieved in the focus. A detailed study has been made of the device performance using a theoretical model and computer simulations, the results of which are in excellent agreement. The parameters used in the calculations were

1s = 60mm, d = 25μm. For the hard x-ray case t=3.5 mm, θ = 5 mrad, and for the soft x-ray case t=0.5 mm, θm = 0.1 rad. The maximum g^ains recorded (at the 2 origin in a 1 μm pixel) were 58,000 for the soft x-ray case, and 580 for the hard x-ray case. It has been shown in these studies that the intensity profile is independent of 1 , and as such the gain increases with the square of 1 .

Figure 4 depicts a further embodiment 10" of the invention, , in which concentric cylinders 30 are assembled define a central cylindrical channel 12a and surrounding annular channels 12b of increasing radial width. Rays are focussed by reflecting once at the concave surfaces of the channels.

To maximise the reflecting efficiency, the radial width of a channel at a radius r from the cylinder axis is given, in simple terms, by:

d(r) = tr/1 . where t is the plate thickness (i.e. cylinder axial length), and 1s is the source to channel plate distance.

More specifically, the radius r of the nth cylinder is given by:

and the radial width d of the channel outwardly bounded by the nth cylinder is given by:

dn= n The smallest channel width will be determined by the manufacturing process, while the largest cylinder and the maximum number of channels N will be determined by the critical angle of reflection θ :

The focussing efficiency E of this embodiment is given by

Channel plate 10" has several advantages over designs based on arrays of channels:

* Very high collection efficiency. Greater than 80% of rays intersecting the plate will be reflected into the focus.

* A spherically curved plate should collimate rays into a very uniform beam.

* There is only one reflection per ray, resulting in lower losses than for plates based on square channels, which require two reflections. Rectangular channel plates are a class of devices with comparable collection efficiency to this design.

On the other hand, it needs to be appreciated that the concentric channel design is not spatially invariant and is of maximum efficiency from a point source at a distance 1 from the channel plate. It is only optimised for one point on the optical axis, and so may be harder to align. Nevertheless, this design is well suited for collimating rays emitting from a point source, or concentrating rays from a synchrotron source. A cylindrically curved plate could be used to collimate and focus in orthogonal planes.

If the concentric channels are not cylindrical but instead the surfaces of frustocones having a common imaginary apex, preferably at the point x-ray source, it can be demonstrated that the above formulae for r , d , N and n' n

E still apply if t = 1 - 1s/R is substituted for t. The term R is the radius of curvature of the respective plate, i.e. the side of the cone, positive if at the left (as seen in Figure 3) and negative if at the right.

It is to be emphasised that formulae set forth above generally apply to the case where the wall thicknesses of the channel plate between the channels (i.e. in the plane of the plate) can be or are assumed to be negligible. More accurate formulae would need to include a term w, the wall thickness.

By way of example, if the average wall thickness w is not assumed to be negligible, it can be shown that the efficiency E is given by:

where Ω = 1 /R, and

w 2d + w 2w _ε = 1 • a l

1sθc d + w 1sθc Figure 5 depicts in front plan, a microchannel plate design which combines annular and right angular channels, for the purpose of increasing the area of acceptance of the plate. The radius of the acceptance area (the area of the front of the plate for which rays entering will be focused) for a plate consisting of concentric annular channels is 1 θ whereas for a square channel plate the furthest active channel from the optic axis is located a distance Λ^"21 50C away from the optic axis. This is the channel located at the corner of the plate, and its diagonal is colinear with a line which intersects the optic axis. The limits of the acceptance area are greater for square channels than annular channels because the grazing angle that a ray makes with each of two orthogonal walls is less than θ for all square channels arranged in a regular array within a square of side-length 21 5θC, centered on the optic axis. Thus arranging channels arranging channels with cross-sections that are square, rectangular, or quadrilateral with at least one right-angle, in the annular region between the circles at a radius of at most 1sθc and at a radius of at most

-121s θc will increase the accep ^tance area of the plate by approximately a factor of 2.

The quadrilateral channels will be generally arranged to increase the efficiency of the plate. For example they may be square channels arranged in segments, as in Figure

3, with rotational symmetry and with the diagonals of the squares parallel to a radial line. In another arrangement, quadrilaterals could be arranged so that the diagonal of every quadrilateral points towards the centre of symmetry (optic axis) of the plate. The vertex furthest away from the optic axis will be a right-angle for every quadrilateral. The lengths of the two sides furthest away from the optic axis (d ) will be eqxιa.1 and the efficiency can be optimised by setting din = tr/1s, where r is the distance of the channel from the optic axis. Figure 5 shows an arrangement along those lines in which the distances between the concentric cylinders

(d ) has been optimised by setting d = tr/1 and where the distance dm obeys the above optimisation condition. This example has been chosen as a simple illustration: in most applications the number of channels will be much greater. This class of designs is practicable by the LIGA micromachining methods mentioned earlier.

The described arrangement has been advanced merely by way of explanation and many modifications may be made thereto without departing from the spirit and scope of the invention which includes every novel feature and combination of novel features herein disclosed.

Claims

1. An x-ray or neutron instrument incorporating x-ray or neutron lens means disposed in a path for x-rays or neutrons in the instrument, the lens means comprising multiple elongate open-ended channels arranged across the path to receive and pass segments of an x-ray or neutron beam occupying said path, which channels have side walls reflective to x-rays or neutrons of said beam incident at a grazing angle less than the critical grazing angle for total external reflection of the x-rays or neutrons, whereby to cause substantial focusing or collimation of the thus reflected x-rays or neutrons wherein said channels are of a quadrilateral cross-section.

2. An instrument according to claim 1 wherein said channels are of a rectangular cross-section.

3. An instrument according to claim 1 wherein said channels are of a square cross-section.

4. An instrument in accordance with cl-a-±fn 1 comprising plural segments arranged about an axis, wherein each segment has a multiplicity of said channels arranged at a prescribed orientation with respect to said axis.

5. An instrument according to claim 4 wherei said segments are arranged with rotational symmetry around said axis.

6. An instrument in accordance with claim 4 wherein said channels are rectangular and oriented such that the diagonals of said rectangular cross-section channels are parallel to the central such diagonal intersecting said axis.

7. An instrument in accordance with claim 4 wherein said channels are of quadrilateral cross-section and arranged so that the diagonal of each quadrilateral channel points towards the centre of symmetry (optic axis) of the lens.

8. An instrument according to claim 7 wherein the vertex of each channel furthest away from the optic axis is a right-angle for every quadrilateral channel and the lengths of the two sides of each channel furthest away from the optic axis are equal.

9. An instrument according to claim 8 wherein the length (^d _m) ^of each said channel is given by the relationship dm = tr/1s where t is the channel axial length, r is the distance of the channel from the optic axis and 1s is the source to lens distance,

10. An instrument according to claim 1 wherein said channels have a diameter to side length ratio d/t approximately equal to θ / 2 where θ is the critical grazing angle for total external reflection.

11. An instrument according to claim 1 wherein said channels have a diameter to side length ratio d/t approximately equal to θ /2 where θ is the angle at which reflectivity reaches zero.

12. An instrument according to claim 1 wherein said lens is a plate having channels of a square cross-section wherein the thickness profile of the plate is of the form t(x,y) = 2dl /[(|x| + |y| )x] where |x| and, |y| are the distances from the optic axis parallel to the x- and y- axes respectively and x is related to the magnification factor by x = _ (1+1/M).

13. An instrument according to claim 1 wherein said lens is a plate with rectangular channels which vary as d_χ(x) = t|x|χ/l_s; d_y=t|y|x/l_s where |x| and |y| are the distances from the optic axis, d and d are the channel side lengths parallel to the x- and y- axes respectively and x is related to the magnification factor M by x = _ (1 + 1/M) .

14. An instrument in accordance with any one of the preceding claims in which the degree of alignment of the channels is such that the average divergence ΔΦ of the channels leads to a spreading of the focal spot by a distance less than (1+M)d, preferably

Δ c<< (1+M)d/(21s), wherein d is the channel side length, (1+M)d the side length of the square focal region centred on the beam axis and 1 is the source to lens s distance.

15. An instrument in accordance with any one of the preceding claims wherein the angular deviation Δ$ of said channel walls from orthogonality is constrained by the relationship Δ$ << (1+M)d/(21 ).

16. An x-ray or neutron instrument incorporating x-ray or neutron lens means disposed in a path for x-rays or neutrons in the instrument, the lens means comprising multiple elongate open-ended channels arranged across the path to receive and pass segments of an x-ray or neutron beam occupying said path, which channels have side walls reflective to x-rays or neutrons of said beam incident at a grazing angle less than the critical grazing angle for total external reflection of the x-rays or neutrons, whereby to cause substantial focusing or collimation of the thus reflected x-rays or neutrons, wherein said channels comprise a plurality of concentric annular channels.

17. An instrument according to claim 16 wherein said concentric annular channels are assembled to define a central cylindrical channel and surrounding annular channels of increasing radial width.

18. An instrument according to claim 17 wherein the radial width of a channel at radius r from the cylinder axis is given by the relationship d(r)

where t is the cylindrical axial length and 1s is the x-ray source to channel plate distance.

19. An instrument according to claim 18 wherein the rraaddiiuuss rrn of the nth cylinder is given by the relationship

and the radial width d of the channel outwardly bounded by the nth cylinder is given by the relationship

20. An instrument according to claim 16 wherein said concentric channels are the surface of frustocones having a common imaginary apex.

21. An x-ray or neutron instrument incorporating x-ray or neutron lens means disposed in a path for x-rays or neutrons in the instrument, the lens means comprising multiple elongate open-ended channels arranged across the path to receive and pass segments of an x-ray or neutron beam occupying said path, which channels have side walls reflective to x-rays or neutrons of said beam incident at a grazing angle less than the critical grazing angle for total external reflection of the x-rays or neutrons, whereby to cause substantial focusing or collimation of the thus reflected x-rays or neutrons, wherein said channels are grouped in a plurality of segments arranged about an axis, each segment having a multiplicity of said channels arranged at a prescribed orientation with respect to said axis.

22. An instrument in accordance with claim 21 wherein said instrument includes a plurality of annular channels and a plurality of channels of quadrilateral cross-section.

23. An instrument in accordance with claim 22 wherein said quadrilateral channels are in accordance with any one of claims 2 to 15.

24. An instrument in accordance with claim 22 wherein said annular channels are in accordance with any one of claims 17 to 19.

25. An instrument in accordance with any one of claims 22 to 24 wherein said channels of quadrilateral cross-section in an annular region between circles at a radius of at most 1sθ and a radius of at most

26. An instrument in accordance with any one of the preceding claims wherein the side walls are planar but their inclinations progressively change from channel to channel with respect to the optical axis of said path whereby to enhance focusing or collimation of said incident beam.

27. An instrument according to any one of the preceding claims wherein an outer side wall of each channel varies in inclination along the length of the channel to further enhance focusing and collimation.

28. An instrument according to any one of the preceding claims further including means for adjusting the inclination of the side walls of said channels.

29. An instrument in accordance with any one of the preceding claims wherein said channels are hollow capillaries or other bores.

30. An instrument in accordance with any one of the preceding claims wherein said channels are defined collectively by a micro-capillary or micro-channel plate.

31. An instrument in accordance with claim 30 wherein said channel plate is flat or parabolically, spherically or cylindrically curved.

32. An instrument in accordance with any one of the preceding claims wherein the x-ray lens device comprises a micro-capillary plate which is curved so that the angular tilts of the reflecting side walls in the channels vary parabolically with distance perpendicular to the optical axis.

33. An instrument in accordance with any one of the previous assemblies further including a monochromator, a sample goniometer stage and/or adjustable x-ray detector.