High Resolution Full Material Fresnel Zone Plate Array and Process for its Fabrication

Abstract

The invention relates to a high resolution full material Fresnel Zone Plate array comprising a plurality of full material Fresnel Zone Plates on a common carrier, a method for producing a full material Fresnel Zone Plate precursor array by providing a common carrier comprising a plurality of micro-pillars and deposition of alternating layers of at least two different materials onto at least some of the micro-pillars and an apparatus for producing high resolution full material Fresnel Zone Plate (FZP) arrays. The apparatus for producing a full material Fresnel Zone Plate arrays, comprises a first device for providing a plurality of micro-pillars in a common carrier, a deposition device which applies alternating layers of at least two different materials onto at least some of the micro-pillars arranged on the common carrier and a slicing device which slices a full material Fresnel Zone Plate out of a pillar.

Claims

1-15. (canceled)

16. A full material Fresnel Zone Plate precursor array comprising a plurality of full material Fresnel Zone Plate precursors arranged on a common carrier.

17. The full material Fresnel Zone Plate precursor array according to claim 16, wherein the full material Fresnel Zone Plate precursors are spaced apart from each other on the common carrier.

18. The full material Fresnel Zone Plate precursor array according to claim 16, wherein central axes of the full material Fresnel Zone Plate precursors are arranged parallel to each other on the common carrier.

19. The full material Fresnel Zone Plate precursor array according to claim 16, wherein the central axes of the full material Fresnel Zone Plate precursors are arranged parallel to the surface normal of the common carrier.

20. The full material Fresnel Zone Plate precursor array according to claim 16, wherein the full material Fresnel Zone Plate precursors comprise controlled tapering angles.

21. The full material Fresnel Zone Plate precursor array according to claim 20, wherein the tapering angles differ between some of the Fresnel Zone Plate precursors arranged on the same common carrier.

22. A method for producing a full material Fresnel Zone Plate precursor array, comprising the steps: a) providing a common carrier comprising a plurality of micro-pillars, and b) depositing alternating layers of at least two different materials onto at least some of the micro-pillars.

23. The method according to claim 22, wherein the plurality of micro-pillars is produced by at least one process selected from the group comprising: focused ion beam (FIB) milling, plasma focused ion beam (Plasma-FIB) milling, spin-coating combined with Direct Laser writing or two photon photopolymerization or any other microprinting method, direct laser ablation, and a combination of any optical, electron, ion beam lithography or any directed-self-assembly or self-assembly method to form a mask layer followed by reactive ion etching (RIE), deep reactive ion etching (ORIE) or a chemical etching method.

24. The method according to claim 22, wherein depositing the alternating layers is performed via selective or non-selective atomic layer deposition (ALD).

25. The method according to claim 22, wherein depositing the alternating layers is performed simultaneously on a plurality of micro-pillars.

26. The method according to claim 22, further comprising the step of: c) detaching at least one micro-pillar from the common carrier after depositing the alternating layers on the at least one micro-pillar.

27. The method according to claim 26, wherein the detached micro-pillar is bound to a mount.

28. The method according to claim 27, wherein the mount is a TEM mount.

29. The method according to claim 27, wherein the detached micro-pillar is bound to the mount via ion beam induced metal deposition.

30. The method according to claim 27, wherein the detached micro-pillar is bound to the mount via ion beam induced Pt deposition.

31. The method according to claim 22, further comprising the step of: d) slicing a full material Fresnel Zone Plate out of a pillar after depositing the alternating layers on said pillar.

32. The method according to claim 31, wherein the slicing is performed by means of an ion beam.

33. The method according to claim 22, wherein at least some of the micro-pillars are produced to be parallel with respect to each other.

34. The method according to claim 22, wherein steps a) and b) are performed independently from each other.

35. An apparatus for producing a full material Fresnel Zone Plate precursor array, said apparatus comprising: a) a first device for providing a plurality of micro-pillars in a common carrier, b) a deposition device which applies alternating layers of at least two different materials onto at least some of the micro-pillars arranged on the common carrier; and c) a slicing device which slices a full material Fresnel Zone Plate out of a pillar.

Description

[0071] Further goals, advantages, features and applications of the invention arise out of the following description of embodiments of the invention on the basis of the figures. Thereby all features described and shown in the figures alone or in arbitrary reasonable combination provide the subject matter of the invention independent of their conclusion in the claims or its dependency.

[0072] It shows:

[0073] FIG. 1: an example of a micro-pillar array fabricated via Plasma FIB milling of single crystalline Si wafer,

[0074] FIG. 2: an individual micro-pillar in more detail,

[0075] FIG. 3: a Plasma FIB fabricated micro-pillar with 5° taper angle,

[0076] FIG. 4a): micro pillars fabricated on a glass substrate via spin coating of a photoresist,

[0077] FIG. 4b): micro pillars fabricated by laser writing via the KLOE Dilase laser writer,

[0078] FIG. 5: a schematic illustration of the fabrication steps of Fresnel Zone Plate arrays with area selective ALD,

[0079] FIG. 6: a schematic illustration of the possible single steps of an ALD-cycle,

[0080] FIGS. 7.1-7.9: illustrations of individual steps of producing an FZP from a Fresnel Zone Plate array,

[0081] FIG. 8: advantages of the high resolution full material Fresnel Zone Plate array with respect to other techniques,

[0082] FIGS. 9a) and 9b): SEM images of one ML-FZP,

[0083] FIG. 10a)-10d): illustrations of the extremely high accuracy of the zones of the multilayer FZP,

[0084] FIGS. 11 a) and 11b): a TEM image of an Al.sub.2O.sub.3—SiO.sub.2 Multilayer,

[0085] FIGS. 12a) and 12b): an SEM image of a ML-FZP of Al.sub.2O.sub.3—HfO.sub.2,

[0086] FIG. 13: an illustration of the linear correlation of the growth per cycle for the ALD,

[0087] FIGS. 14a) and 14b): examples for direct imaging results obtained at BESSY II,

[0088] FIGS. 15a) and 15b): illustration of results of diffraction efficiency calculations.

[0089] FIG. 1 shows an example of a micro-pillar array (1) fabricated via Plasma FIB milling of a common carrier (2), which is in this case a single crystalline Si wafer. Each micro-pillar (3) is located in a recess (4) formed in the common carrier (2). In the shown embodiment the recesses (4) are right circular cylinders which are arranged in lines and rows on the common carrier (2). In the center of each recess (4), a micro-pillar (3) is located. Each micro-pillar (3) stands on a bottom surface (5) of a recess (4) and is oriented along the central axis of the respective cylinder. The depth of each recess (4) is identical to the length of each micro-pillar (3). The top surface (6) of each micro-pillar (3) is arranged in the same plane as the upper surface (7) of the common carrier (2). The tapering angle (not indicated in FIG. 1) of each micro-pillar (3) is 1°. However, the micro-pillar arrays (1) can be fabricated with any other desired tapering angle via the Plasma or other FIB, and allow ML-FZPs to be fabricated with zones tilted to the Bragg angle, paving the way for highly efficient focusing to sub 10 nm resolution.

[0090] FIG. 2 shows an individual micro-pillar (3) in more detail. Such a pillar provides smooth (side) walls (8), high circularity and a tapering angle (a). FIG. 2 further illustrates that between the sidewalls of the micro-pillar (3) and the sidewall of the recess/cylinder (4) there is a certain distance that has to be at least two times larger than the deposition thickness and preferably between 3 or 4 times larger. Thus, material layers can be deposited between these sidewalls. In FIG. 2 also the common carrier (2) with its upper surface (7) is shown. Its upper surface (6) of the micro-pillar (3) is arranged on the same plane as the upper surface (7) of the common carrier (2). The pillars (3) produced via PFIB can be even more precise than glass fibers in terms of diameter vs circularity.

[0091] FIG. 3 shows another individual micro-pillar (3) in more detail which provides a different taper angle (α) of 5°. The sidewalls are inclined with respect to the central axis (L) by the taper angle (α). This pillar was also fabricated via Plasma FIB. Also the recess/cylinder (4) and the common carrier (2) with its upper surface (7) are shown.

[0092] FIG. 4a) and b) are showing micro-pillars (3) fabricated via other techniques than FIB. The pillars (3) shown in FIG. 4a) are fabricated on a glass substrate, which serves as a common carrier (2), via spin coating of a photoresist (KCL). Similar to the arrangement of the pillars shown in FIG. 2, the pillars (3) are arranged in lines and rows. However, the pillars are not located in recesses but located on the upper surface (7) of the common carrier (2). In this embodiment, each micro-pillar (3) is a right circular cylinder standing perpendicular on the upper surface (7) of the common carrier (2).

[0093] FIG. 4b) shows micro-pillars (3) produced by laser writing via the KLOE Dilase laser writer. Similar to the micro-pillars (3) shown in FIG. 4a) these pillars (3) are arranged in lines and rows and not located in recesses but on the upper surface (7) of the common carrier (2).

[0094] However, fabrication of micro pillars (3) via FIB is preferred because it easily allows for controlling the tapering angles (a) by varying sputter time, beam energy and beam current. Independent from the above-mentioned techniques but especially when FIB is used, the fabrication of micro pillars (3) can result in controlled circularity, diameter, tilt angle and high surface smoothness.

[0095] In a preferred embodiment, Plasma FIB is used. By using Plasma FIB the milling speed can be further enhanced. With Plasma FIB, milling out pillars is about 30 times faster, compared to non-plasma FIB.

[0096] In a preferred embodiment single crystalline Au is used as a substrate material for the common carrier (2). This material allows for improved milling rates. The sputter yield of Au is very high. In an even further preferred embodiment the substrate is (111) type Au single crystal. It has been shown that milling out (111) type Au single crystal is about twice as fast as using (100) Au single crystal.

[0097] Irrespective of whether pillars (3) or cylindrical holes (4) are milled out/or laser written, circular geometries are preferred. The circularity of the (etched) structures and longer crack propagation lengths due to the trenches around pillars (3) or the (cylindrical) holes (4) may hinder crack propagation, enabling thicker depositions to be done crack-free.

[0098] As mentioned above, the step of production of the full material Fresnel Zone Plate precursor array (1) (e.g. the array of micro pillars) is followed by the deposition of alternating material layers onto it. The deposition is preferably performed by ALD. The deposition of alternating layers can be performed via selective or non-selective atomic layer deposition (ALD) and the individual layers act as the zones of the Fresnel zone plate, which can be sliced in a later step as described in detail below.

[0099] Thus, this process allows for the first time production of full material FZP (widely also known as multilayer or as sputter sliced) that does not have a micro fiber core. Until now, the micro fiber substrates could be deposited by other techniques (e.g. PLD or magnetron sputtering) than ALD although with more difficulty. However, by the currently disclosed process, highly conformal deposition onto these full material Fresnel Zone Plate precursor arrays (1), e.g. mi-cro-pillar arrays (1) can only be achieved by the ALD. This feature provides big advantages with respect to other techniques (e.g. the sputtered sliced or PLD deposited FZP intermediate products and the resultant FZPs). Here, especially circularity, diameter and tapering angle can be controlled very precisely. If a sputtering technique is to be used for the zones of the FZP, one has to start with a wire because it has to be ultimately symmetric for the following sputter deposition of the material while rotating the substrate wire core. However, rotation always results in errors such as inhomogeneous deposition of the materials.

[0100] In a preferred embodiment the arrangement of the pillars (3) or the different arrays (1) on the substrate may not be symmetrical with respect to central axis of the carrier substrate (2). However, each individual pillar (3) can be symmetric to its central axis. If the pattern of pillars as a whole is not symmetric to the central axis (L), the possibilities for conformal deposition of layers are very limited. Since patterns which are non-symmetric with respect to a (global) central axis are often utilized, ALD offers the only possibility to deposit conformal layers (or zones) of the FZP to these substrates.

[0101] So far, deposition of full material FZPs onto any other material than smooth glass fiber cores resulted in poor quality layer structures. Glass fibers provide a high circularity and smooth surface compared to metallic or ceramic fibers by its amorphous nature therefore limiting the substrate usage mostly to glass fibers. One further advantage of the method described here is that it vastly expands the possibly available materials as the substrate. Polymeric, single crystalline silicon and gold substrates have been utilized for fabrication of the pillar arrays (1) described here. However, one can envision utilization of a vast number of heat resistant polymeric, metallic, amorphous metallic, ceramic, glass, glass-ceramic and diamond based materials for the realization of the micro-pillars (3). When utilized, some of the high Z materials such as gold or lead can also act as a central stop to improve imaging contrast thus adding further advantage by avoiding the need for an external beamstop or a beamstop deposition onto the center of the full material FZP.

[0102] In a preferred embodiment the deposition process comprises the step of application of area selective ALD. One possible way would be to apply a hydrophobic material on the substrate and the top (6) of the micro pillars (3) but not onto the sidewalls. This can be achieved, for instance, by selectively functionalizing all horizontal planes with a Self-Assembled Monolayer (SAM) via a use of a proper molecule. This will ensure no deposition on those surfaces, enabling the deposition to happen only on the side walls (8) of the micro pillars (3). FIG. 5 sche-matically shows 4 steps of the deposition procedure. In step 1 the micro pillars (3) are provided. They are arranged on a common carrier (2) which is preferably a substrate which is transparent for X-rays. In step 2 a coating (9) is applied onto the surfaces (6) of the micro-pillars (3) onto which no layers should be deposited by ALD. Some of these surfaces (6, 7) which usually are coated are the substrate surface (7) and the top surface (6) of the micro pillars (3). Depending on the precursor type, the coating (9) might be a hydrophobic material. By this process ALD becomes an area selective ALD process. Covering of the top (6) of the micro-pillars (3), and the bottom of the substrate (7), except for the walls (8) of the micro-pillars (3) with a hindering material prior to atomic layer deposition, will enable area selective deposition of the multilayer zones only on the walls (8) of the micro-pillars (3). So far all the methods for fabricating multilayer type or sputter sliced FZPs needed FIB slicing (or other more crude methods) and mounting on a grid. The selective deposition method eliminates the need for FIB slicing, and hun-dreds or even thousands of FZPs can be fabricated in a two-step process.

[0103] In step 3 the alternating layers are deposited by ALD onto all non-coated surfaces (8). ALD is known from the prior art and will be described in more detail later. Because of the coating, ALD can be performed area selectively and the multilayer zones of the FZP can be deposited selectively onto the side walls (8) of the micro pillars (3). After deposition of the layers by ALD is completed, the coating (9) can be removed as illustrated in step 4. It remains a Fresnel Zone Plate array (1). If desired single FZP can be sliced down from this array and handled/used individually.

[0104] FIG. 6 illustrates the single steps of an ALD-cycle. This technique is well known in the prior art. Thus, it is only described briefly. In step 1, a first precursor (10) is added to the reaction chamber (not shown) containing the material surface (the side walls (8) of the pillars (3)) to be coated. The first precursor (10) is added in large excess into the reaction chamber. Thus, it is adsorbed on the entire surface. Any excess of the first precursor (10) is removed from the reaction chamber in step 2. The molecules of the first precursor (10) react on the surface. Then, a second precursor (11) is added in step 3. It reacts with the new surface to create another layer on the surface. The second precursor (11) is then removed from the reaction chamber and a homogeneous layer remains on the original surface. This process is repeated until a desired thickness for the respective layer is achieved. Then, the precursors are changed and the process is repeated for the next layer of another material. The deposition of the layers of different materials is repeated until the desired number of layers is deposited.

[0105] A tremendous advantage of ALD over other methods is that it is the only method for applying the individual layers onto the sidewalls of the pillars without rotation of the pillar (3) itself with respect to its central axis (L), hence making the process described here possible. Furthermore, because of the large excess of the precursor and their gas state, the individual molecules can easily reach every surface area, even within the holes or recesses (4) in which each individual pillar (3) is located (by using the FIB). Since the excess of precursor molecules is purged away (FIG. 6, steps 2 and 4) a homogeneous new surface layer remains even in areas with sharp edges or narrow corners. This method allows fabricating FZPs having outermost zone widths smaller than 10 nm, preferably smaller than 5 nm and most preferred smaller than 1 nm with extremely high aspect ratios compared to the FZPs fabricated by lithography techniques.

[0106] Thus, ALD allows [0107] precise and conformal deposition through complementary, self-limiting, surface reactions; [0108] uniform thickness and conformal coating; and [0109] highly precise thickness control through controlling the number of cycles [0110] large selection of materials relevant for X-ray optics such as oxides, nitrides and pure metals Preferably plasma enhanced ALD is used for fast deposition. It allows a PE-SiO.sub.2 growth rate and/or deposition rate of 2.6 μm/day.

[0111] If it is not desired to use the array of FZPs as a whole and/or if single FZPs are needed, single full material FZPs with optimum tilt angels can be sliced from the array and transferred on grids. A preferred embodiment of this process is illustrated in FIG. 7.

[0112] As shown in FIG. 7.1 a micro manipulating needle (12) is bound to the micro-pillar (3). Then the pillar (3) is detached from the substrate (2) as illustrated in FIG. 7.2. After providing a TEM grid (13) to mount, the sliced pillar (3) is bound to it (FIGS. 7.3 and 7.4). Binding of the sliced pillar (3) can e.g. be performed via ion beam induced Pt deposition. After the sliced pillar (3) is bound to the mount (13), the needle can be detached. If necessary, a further material (e.g. Pt) can be deposited between the sliced pillar (3) and the mount (13) to improve the stability of the sliced pillar (3) on the grid (13) (FIG. 7.5).

[0113] As shown in FIG. 7.5 the sliced pillar (3) is still far away from the ideal geometry of a FZP with parallel and planar surfaces. Thus, it has to be thinned further. A preferred method for slicing an FZP out of the pillar (3) via an ion beam is shown in FIGS. 7.6-7.9. Therefore, the FZP has to be protected from ion beam damage. Thus, the sides of the FZP are covered by a material layer (14) in the desired length as illustrated in FIG. 7.6. Then, the pillar (3) is sliced to the desired length. As mentioned, ion beam techniques can preferably be used. FIG. 7.7 shows a pillar (3) which has been already sliced (on the left side), but is still much longer than the desired FZP which is protected by the covering (14) of the sidewalls. Slicing by ion beam techniques is preferred because ion beams can also be used for polishing the parallel and planar surfaces of the FZP (15) after it has been sliced to the desired length. A polished surface (16) is shown in FIG. 7.8. FIG. 7.9 shows a full material FZP (15) mounted on a grid (13). This FZP (15) can be used to focus X-rays. Since the length of the FZP can be adapted to specific need during or after its mounting on the grid, this method allows not only to fabricate FZPs (15) having sub nm outermost zone widths but also having extremely high aspect ratios.

[0114] Some advantages of the high resolution full material Fresnel Zone Plate array (column ALD ML-FZP) and the process for its fabrication are summarized in FIG. 8.

EXAMPLES

[0115] An SEM image of one ML-FZPs (15) is shown in FIGS. 9a) and 9b). Being monolithic optics, ML-FZPs (15) are easy to align, have a point focus, and they can be efficient in a very wide X-ray energy range. The layers are deposited on a glass substrate. The central glass fiber core (17) will be deposited with Pt to act as a beam stop. The shown FZP (15) has a diameter d=44 μm. The outermost zone (18) has a width Δr=20 nm and a deposition thickness of 6.5 μm. With such an ML-FZP (15), a half-pitch imaging resolution of 15.5 nm could be achieved. This is currently the highest imaging resolution reported for a multilayer type FZP (15). However, by the fabrication route via FZP arrays (1) or controlled tilt angles (α), resolutions below 10 nanometer for focusing of hard and soft X-rays to through ML-FZPs, are possible.

[0116] In addition to the FZP (15) shown in FIG. 9 also the example of FIG. 10 illustrates the extremely high accuracy of the zones (19) of the multilayer FZP (15). FIG. 10a) is an SEM image of a single FZP (a) cut-out of an ALD deposit. The areas marked with rectangles in FIG. 10a) are shown in more detail in FIGS. 10b)-d). FIG. 10b) is a detailed view of the uppermost rectangle (red), FIG. 10c) is a detailed view of the right rectangle (blue) and FIG. 10d) is a detailed view of the lowermost rectangle (green). In all these detailed FIGS. 10 b)-d) the individual layers (19n) and their diminishing width with larger distance to the central axis can be seen.

[0117] The sharp edges between individual layers (e.g. 19n and 19m or 19a and 19b) and the low roughness are illustrated even better in FIG. 11a) and b). These figures are TEM images of an Al.sub.2O.sub.3—SiO.sub.2 multilayer. With current processes 20 μm multilayers of Al.sub.2O.sub.3—SiO.sub.2 could be deposited within less than 9 days.

[0118] Since the accuracy of each single layer (e.g. 19n, 19m, 19a or 19b) is extremely high, a very large number of layers in the range of thousands (19) can be applied onto a substrate without propagation of irregularities to further layers (19). Thus, the deposition thickness can be improved.

[0119] FIG. 12a) shows an ML-FZP (15) of Al.sub.2O.sub.3—HfO.sub.2 Δr=20 nm with an overall deposition (D) of 6.5 μm.

[0120] FIG. 12b) is a detailed view of the ML-FZP (15) of Al.sub.2O.sub.3—HfO.sub.2 shown in FIG. 12a) and illustrates that even at this extreme thickness (D), the accuracy of the outer layers (19x-19z) and the outermost layer (18) is still extremely high.

[0121] Since ALD is a cycle based deposition, it could be expected that the growth per cycle follows a linear curve. However, the correlation has been checked. The results are illustrated in FIG. 13. The linear correlation could be confirmed.

[0122] Direct imaging results which have been obtained at BESSY II are shown in FIG. 14a) and b). Figure a) shows an STXM image (20S) of a Siemens star taken at 1 keV, 10 nm step size and 5 ms pixel dwell time and figure b) an STXM image (20B) of the BAM L-200 sample taken at BESSY UE46-PGM2. Step size 4 nm. Dwell time 30 ms. The full pitch cut-off resolution is 31 nm corresponding to 15.5 nm half pitch resolution. In Fig. a) an FZP with Δr=35 nm is used whereas the Fig b) is taken by using an FZP with an outermost zone width of 25 nm.

[0123] Furthermore diffraction efficiency calculations have been performed. For illustration purposes a selected FZP of different materials, the results are illustrated in FIGS. 15a) and b). FIG. 15a) shows the expected diffraction efficiency vs thickness graphs calculated by using the coupled wave theory (CWT) of Al.sub.2O.sub.3—SiO.sub.2 ML-FZP with tilted zones for 1 keV X-rays. FIG. 15b) shows the expected diffraction efficiency vs thickness graphs calculated by using the coupled wave theory (CWT) of Al.sub.2O.sub.3—SiO.sub.2 ML-FZP with tilted zones for 10 keV X-rays. Both calculations were performed with the following parameters: Diameter D=40 μm, Outermost zone width Δr=10 nm, d(Al.sub.2O.sub.3)=2.99 g\cm.sup.3, d(SiO.sub.2)=2.11 g\cm.sup.3.

[0124] For X-rays from 1 to 1.6 keV, an imaging resolution of 15.5 nm could be achieved at a measured efficiency from 3 to 12.5% for the material system Al.sub.2O.sub.3—HfO.sub.2/Ta.sub.2O.sub.5.

[0125] For X-rays of 7.9 keV, a focal spot of 30 nm could be achieved at a measured efficiency of 15.6% for the material system Al.sub.2O.sub.3—Ta.sub.2O.sub.5.

[0126] All the features disclosed in the application documents are claimed as being essential to the invention, insofar as they are novel either individually or in combination as compared with the prior art.