X-RAY TARGET Technical field The present invention relates to targets for X-ray generation. More particularly, the invention relates to solid targets for electron-impact X-ray sources, the solid targets comprising target features that generate X-ray radiation upon exposure to an electron beam. Background Electron-impact X-ray sources are generally known. In such X-ray sources, X- ray radiation is generated by interaction between atoms in a target material and incident electrons from an electron beam source. The X-ray radiation is primarily generated as bremsstrahlung, although characteristic emission lines also contribute. Only a relatively small fraction of the energy contained in the incident electron beam is converted into X-ray radiation. Excess heat generated in the target must therefore be handled. To this end, it is known in the art to have X-ray targets comprising an X-ray generating material, such as tungsten, deposited on a substrate having high thermal conductivity, such as diamond. Attempts have been made to embed the X-ray generating material in the substrate in order to increase thermal conduction into the substrate. In this context, US 2011/058655 discloses an X-ray target comprising a target portion (e.g. made from tungsten, gold, platinum, or the like) deposited in a bottomed hole formed in a diamond substrate. However, it has proven to be notoriously difficult to embed e.g. tungsten in a diamond substrate. Forming holes of sufficient depth in the diamond substrate is not straightforward, and a comparatively thin embedded target feature may result in less X-ray radiation than desired. Furthermore, the difference in thermal expansion between the target portion (e.g. tungsten) and the diamond substrate may increase the risk of defects/cracks forming in the target portion and/or the substrate during operation. Also, more generally, a good thermal contact between the target portion and the substrate may be impeded leading to overheating of the target portion.
Hence, there is a need for improved X-ray targets that address the above issues. An X-ray target according to the present invention comprises a substrate made from a material with good thermal properties, preferably diamond. The substrate is selected to have a sufficiently high thermal conductivity so that heat generated by an incident electron beam may be distributed from the point of impact to avoid local overheating and ensuing target damage. In general, the substrate material should have a low X-ray yield, to not contribute with background noise, and low X-ray absorption to ensure that the generated radiation is emitted. However, for some applications selective X-ray absorption in the substrate may be preferred, e.g., because a material that absorbs low energy X-ray radiation will have a monochromatizing effect, a beneficial effect for dose sensitive samples. In this case substrates made from e.g. aluminum may be considered. Furthermore, the target comprises at least one target feature in which X-ray radiation is generated upon exposure to the electron beam. As will be understood, a plurality of target features may conveniently be provided in the target. The target features are made from a material with a high X-ray yield, i.e., a comparatively large fraction of electrons impacting on the target feature contributes to generation of X-ray radiation, either as bremsstrahlung or characteristic emission lines. Typically, an element with higher atomic number (Z) will have a higher X-ray yield. A preferred material for the target features is tungsten but other elements like copper, rhenium, rhodium, palladium, molybdenum, vanadium, and niobium, or alloys comprising these elements may also be considered. Provided that the electron beam impacts the entire target feature (i.e., that the electron beam focus at the target feature is equal to or larger than the target feature), the size of the target feature will define the size of the X-ray spot. To increase the thermal load that the target feature is able to withstand, it has been proposed, e.g., in US 2011/058655, to embed the target feature within the substrate, thus providing for more surface contact between the target feature and the substrate than would be the case if the target feature were provided on a top face of the substrate. However, in practice it is difficult to provide embedded target features with the preferred material choices.
The present invention therefore proposes an improved solution for a transmission-type X-ray target, wherein an embedding layer is provided on top of a flat top face of the substrate and the, or each, target feature is provided in, and preferably embedded within, this embedding layer without reaching into the substrate. As generally understood in the art, a transmission-type X-ray target is a target which is thin enough to allow the generated X-ray radiation to pass through the target and be emitted downstream of the target relative to the incoming electron beam. The material in the embedding layer can be selected to provide good adhesion to the substrate and the target material, to enable formation of suitable holes or indentations for the target features, and to provide for good thermal conduction from the target feature(s) to the substrate. The preferred choice of material for the embedding layer is silicon carbide, a material that is used extensively in the semiconductor industry enabling embedding of target features with desired shapes and sizes. Alternatively, materials like beryllium, boron nitride, boron carbide, aluminum nitride, silicon nitride, and silicon boride may be used. In embodiments of the present invention, the materials used for the substrate, the embedding layer, and the target feature(s) are all different. The present invention also provides a method of producing an X-ray target. A substrate having a top face is provided, an embedding layer is formed on the top face of the substrate, and one or more target features are formed in the embedding layer. The, or each, target feature is embedded in the embedding layer without reaching into the substrate, thereby avoiding the need to make holes in the substrate. In embodiments, a target feature has an extension in a direction orthogonal to the top face of the substrate that is larger than an extension of the target feature in a direction parallel to the top face. Herein, the extension of the target feature in a direction orthogonal to the top face of the substrate may be referred to as the first extension of the target feature, and the extension of the target feature in a direction parallel to the top face of the substrate may be referred to as the second extension of the target feature. Various geometries for the target feature are described herein, such as conical/pyramid or frustum, for which the extension of the target feature is different for different slices thereof parallel and orthogonal to the top face of the substrate. The extension of the target feature should thus be understood in its
natural sense to mean the largest overall extension of the target feature parallel and orthogonal, respectively, to the top face of the substrate. The invention is particularly advantageous for target features that have an axial (first) extension that is long in relation to a radial (second) extension of the target feature. In this case thermal transport out of the target feature will, to a large extent, be in a radial direction and hence benefit from being provided in a material with a high heat conduction ability. For a “flat” target feature, i.e., a target feature where the axial (first) extension is small in relation to the radial (second) extension, most of the thermal transport will be in the axial direction and embedding the lateral part of the target feature may not contribute significantly to the thermal load that the target feature is able to withstand. Another way of stating this condition is that the lateral surface area of the target feature should be larger than its top or base area, or more preferable be larger by some factor e.g., 2. To avoid risking thermal failure with the introduction of an embedding layer, it is advantageous that coefficients of thermal expansion for the materials are matched. This is particularly the case for the interface between the target feature and the embedding layer, where the thermal load is expected to be at its highest. Not matching the coefficients of thermal expansion would imply that stress is created at the interface whenever the temperature changes. In a preferred embodiment, the embedding layer comprises silicon carbide and the target feature comprises tungsten, and it can be noted that these two materials advantageously have similar thermal expansion coefficients of about 4 · 10-6 K-1. A further advantage may be attained by providing a plurality of target features on a common substrate. Should one feature, for some reason, be worn out or otherwise fail, the electron beam may be moved to another feature and operation may be continued without extensive maintenance work. It is also conceivable to have features with different characteristics, e.g., different lateral extensions, to provide for a wider range of X-ray source performance using different selected target features. Features of different lengths in the propagation direction of the electron beam may be provided to achieve optimum results for each applied acceleration voltage. To understand the performance, and limits, of this type of target some simplified physical models may be considered. From a general perspective, to
increase the X-ray flux generated by an incoming electron beam more target material should be provided. In the case of a transmission type target (i.e., a target which is thin enough to allow X-ray radiation to pass through the target and be emitted downstream of the target relative to the incoming electron beam) the thickness of the target layer is limited both by the ability of the electrons to penetrate the target material and by the ability of the generated X-ray radiation to escape from within the target layer. Furthermore, it is well known that an electron beam will broaden due to scattering when propagating through a material. Thus, although it may be desired to have a target layer that covers a larger surface area than the actual desired X-ray spot size, which may promote a higher X-ray flux, increasing also the target layer thickness entails an increase of the spot size due to broadening of the electron beam in the target material. A solution to this, according to some embodiments of the present invention, is to provide a target feature with a lateral extension equal to the desired X-ray spot size. In other words, the target feature is made sufficiently small relative to the electron beam such that the spot size is determined by the extension of the target feature rather than by the extension of the electron beam. In this way the electrons comprised in the broadened part of the electron beam do not contribute to the production of X-ray radiation. In principle, electrons will to some extent interact with any material they travel through and produce some X-ray radiation, but elements with low atomic number have a low X-ray yield so this contribution will only create some background noise when compared to the X-ray radiation created by interactions between the electrons and the target feature. Nevertheless, the electrons will still scatter within the target feature and some electrons will scatter out of the target feature and not contribute to further X-ray radiation. Making the target feature thicker thus only makes sense up to some value, after which additional target feature material will not result in more X-ray flux due to the diminishing density of the electron beam. The optimal thickness will typically be a function of the electron energy, where a higher acceleration voltage enables use of a thicker target feature. For the case where a target layer with an area considerably larger that the electron beam spot is provided, the optimal target layer thickness may be calculated either from Monte Carlo simulations or from a simple model where the incoming electron beam is exponentially attenuated over a first characteristic length as it
progresses into the target layer and the generated X-ray radiation is attenuated in the target layer over a second characteristic length. The optimal target layer thickness as a function of energy may be approximated as linear for low electron energies and as a power law if larger energies are considered. For the case where the X-ray spot is defined by the target feature, electrons scattered out of that feature do not contribute to the generated X-ray radiation and neither do electrons impacting outside of the target feature. For a given total available power, it thus makes sense to provide an electron beam spot equal in size to the target feature since otherwise some of the available power will not contribute to X-ray production. However, when the electron beam spot size and the target feature extension are comparable, broadening of the electron beam in the target material will reduce the number of electrons that contribute to X-ray production. The result of this is that the optimal target thickness for this case will be smaller than the corresponding value for a target layer with an extension considerably larger than the electron beam spot size. If the electron beam spot size is considerably larger than the target feature, broadening of the beam may not impact the amount of X-ray radiation produced but then again electrons impacting outside the target feature do not contribute to X-ray production. To illustrate these limitations a simple model based on a paper by Gauvin and Rudinsky from 2016 (“A universal equation for computing the beam broadening of incident electrons in thin films”, Ultramicroscopy, 167, pp.21-30) is employed. A typical thickness of a continuous target layer of tungsten (W) is about 0.5 µm. Spot sizes used for such a target are typically larger than 300 nm. Target features with a diameter of 300 nm (in a plane orthogonal to the electron beam propagation direction) and thickness up to 3 µm will be considered below, i.e. a 10:1 ratio between axial extension and diameter. This is for illustrative purposes only and should not be considered as limiting the scope of the invention. As evident from the disclosure herein other target feature diameters and/or thicknesses may be used to an advantage. The diameter may be less than 1 µm, such as less than 500 nm, such as less than 300 nm. For a non-cylindrical, e.g. conical, pyramid, or otherwise tapered target feature, the diameter may be understood as the largest diameter of the target feature in a plane parallel to the top face of the substrate along the length
of the target feature. The thickness may be less than 10 µm, such as less than 5 µm, such as less than 3 µm. The diameter of an electron beam traversing a material may be written as ^(^) = ^^ ^ ^ + ^(^)^ (1) where D0 is the diameter of the incoming electron beam, and b(^) is the broadening of the electron beam at the depth ^ into the material. The broadening of the electron beam may, in the limit where the thickness of the film is considerably larger than the mean free path of the electrons, be written as
where b and ^ are in m, Z is the atomic number of the target material, ρ is the density in g/cm3, A is the atomic weight in g/mol, and E0 is the electron energy in keV. The mean free path of electrons (λ) may be calculated in cm from
where θ0 is the characteristic scattering angle, e is the electron charge in CGS units, and N0 is Avogadro’s number. For 160 kV electrons impacting on W this computes to about 330 nm. Thus, it seems justified to use the above equation for electron beam broadening, considering that only thicknesses above 0.5 µm are of interest. By considering an incoming electron beam having a diameter of 300 nm, the beam diameter as a function of propagation depth into the target material, i.e. the material of the target feature, may thus be estimated from the equations above. This is an approximation in the sense that only scattering of electrons from W is considered. In a more realistic setting, electrons leaving the target feature should scatter against the atoms in the embedding layer. However, capturing this effect in the calculations would require Monte Carlo simulations or the like, which goes beyond the needs and purposes here. The general characteristics of this simplified model are expected to hold also for a more complete calculation. Fig.1 shows a plot
of the electron beam diameter as a function of propagation depth in tungsten for three different electron energies as described by equation (1) above. As expected, scattering is more pronounced for low energy electrons and the beam broadening is consequently larger. Since electrons scattered out of the target feature will not contribute to the useful X-ray production, there is an energy dependent limit beyond which increasing the target thickness does not increase the useful X-ray output. This is illustrated in Fig. 2, where the fractional yield of X-ray generation is plotted as a function of propagation depth into tungsten for three electron energies. The data has been normalized with the theoretical maximum that the entire electron beam is within the target feature up to 3 µm depth (i.e., no beam broadening). As can be seen, the curves flatten out well before the 10:1 ratio is reached, i.e. when the depth equals 3 µm for the incoming electron beam diameter of 300 nm. Although a more complete calculation would consider the electron beam intensity profile and not just the area, the general characteristics of the present model are expected to be preserved also in a more elaborate scheme. One reason to let the target feature define the X-ray spot, or in other words to provide a target feature that is smaller than the electron beam spot size, is to attain a small X-ray spot. However, for a cylindrical target feature, even a small misalignment between the viewing direction and the target will result in an apparent enlargement of the X-ray spot. This is because X-ray radiation will be emitted not only from the end surface of the cylindrical target feature but also from the lateral surface. At perfect alignment only the radiation emitted from the end surface will reach an observer, whereas if there is any misalignment also radiation emitted from the lateral surface will reach the observer. If instead a tapered, e.g. conical, target feature is provided this enlargement effect may be suppressed. For applications where only one operational X-ray spot at the time is desired, embodiments having more than one X-ray generating target feature are preferably designed such that electrons directed to one target feature do not result in X-ray generation from neighboring features. This may be accomplished by providing sufficient distance between the target features. A lower limit on this distance can be expressed as the mean free path of the electrons according to Equation (3) above. However, electrons travelling one mean free path will typically have undergone one
scattering event and thus still carry sufficient energy to generate X-ray radiation. Therefore, a more stringent requirement would be to have a distance between target features at least equal to the broadening according to Equation (2). Other applications may rely on X-ray radiation being emitted from more than one target feature at the time, e.g., multiple X-ray source points used for Talbot-Lau interferometry or linear accumulation. In such cases the distance between target features may be smaller than the limits imposed by Equation (2) or (3). In embodiments where target features having different depths are desired, the target may be configured with multiple layers of SiC on top of a diamond substrate. In such cases, the embedding layer thus comprises multiple sub-layers. Since the proposed materials for the embedding layer are typically semiconductors, charge accumulation may present itself as a potential issue. To prevent charge accumulation, a conducting layer on top of the embedding layer and/or between the embedding layer and the substrate may be provided. If this layer is sufficiently thin, the interference with the electron beam and/or the generated X-ray radiation can be made negligible. Another approach may be to dope the embedding layer to provide a conductive path for removal of charge carriers. Brief description of the drawings In this description, reference is made to the accompanying drawings, on which: Fig.1 shows a plot of electron beam broadening as a function of propagation depth into tungsten for three different electron energies; Fig. 2 shows a plot of the fractional yield of X-ray radiation as a function of propagation depth into tungsten for three different electron energies; Fig. 3a-c schematically show X-ray targets having a target feature provided in an embedding layer on top of a substrate; Fig.4 schematically shows an X-ray target having a plurality of target features provided in a multilayered embedding layer; Fig.5 schematically shows an X-ray source including a target according to the present invention. On the drawings, like parts are designated by like numerals throughout.
Detailed
X-ray targets according to the principles disclosed herein are schematically shown in Fig.3a-c. All X-ray targets shown in Fig.3 comprise a substrate 30, an embedding layer 32 provided on a top face of the substrate, and a target feature 34 provided in the embedding layer. The target feature 34 is provided in the embedding layer 32 without reaching into the substrate 30, which means that no holes or recesses need to be formed in the substrate. Although the target feature 34 may protrude from the embedding layer on a side facing away from the substrate, it is generally preferred to have the target feature 34 flush with or buried in the embedding layer 32. In a preferred embodiment, the substrate 30 is made from diamond, which is a durable material having excellent thermal properties. Furthermore, diamond has a low X-ray yield (thereby causing only low background noise) and has good transparency for the X-ray radiation generated in the target feature 34. In a preferred embodiment, the embedding layer 32 is made from silicon carbide and the target feature is made from a material having an atomic number above 20, most preferably tungsten. Conveniently, tungsten and silicon carbide have similar thermal expansion coefficients of about 4 · 10-6 K-1 which leads to low thermal stress at the interface between the target feature and the embedding layer. As will be understood, if the X-ray spot defined by the target feature 34 is viewed from an inclined angle, there will be an apparent broadening of the X-ray spot since some radiation exiting the target feature through the lateral sides then also contribute. Therefore, it may be advantageous to implement the target feature to have a conical or tapered shape, as schematically shown in Fig.3b. As long as the X-ray spot is viewed from an angle that is smaller than the taper angle, there will be no apparent broadening of the X-ray spot. Such conical or tapered target feature may be truncated, as shown in Fig. 3b, or end at an apex. Although the target feature in Fig. 3b has a decreasing cross section towards the substrate, the opposite configuration where the target feature is tapered away from the substrate is also conceivable. The latter configuration may be preferable from a thermal point of view whereas the former may be more attractive from a manufacturing point of view. Such conical/tapered shape can be used in any embodiment, as will be readily understood.
Fig.3c schematically shows an X-ray target comprising a conducting layer 36 provided on top of the embedding layer 32. Such conducting layer is useful for preventing charge build-up caused by the incident electron beam. Other implementations of conducting layers for carrying away excess charges are also possible, and may also be provided, for example, between the target feature 34 and the embedding layer or between the embedding layer and the substrate. It is also conceivable to have an embedding layer that is doped, such that charge carriers can be removed through the embedding layer. In this latter case, a separate conducting layer may not be required. Typically, the cross section of the target feature in a plane parallel to the top face of the substrate will be circular, although other cross-sectional shapes are also conceivable. The target features are conveniently formed in connection with forming the embedding layer. Thus, in embodiments where target features having different depths are required, the target may be configured with multiple layers of the embedding material, for example silicon carbide, as schematically shown in Fig.4. As shown, the embedding layer comprises a plurality of sub-layers 32-1, 32-2, 32-3 provided on the substrate 30. A first group of target features 34-1 have the largest depth and reach through all of the sub-layers of the embedding layer; a second group of target features 34-2 have a smaller depth than the first group, and reach through the two top-most sub-layers 32-2, 32-3, and a third group of target features 34-3 have the smallest depth of the groups and only reach through the top-most sub-layer 34-3. In this example, each group of target features comprises three individual target features, but it is understood that this is merely an example, and that more or fewer individual target features can be provided. The advantage of enhanced cooling capacity in embodiments of the present invention is most pronounced for a target feature where an extension (thickness) of the target feature along a direction perpendicular to the top face of the substrate is longer than an extension (diameter) of the target feature in the lateral direction, i.e. a direction parallel to the top face. Hence, the present invention proposes that the thickness of the target feature is at least as large as its diameter. By instead applying the condition that the lateral surface area should be larger than the top surface area a less strict requirement may be imposed on the minimum thickness, and for a
cylindrical target feature the minimum thickness may in this case be one quarter of the diameter, or one half of the diameter if the lateral surface area should be at least twice the top surface. For a given diameter of the target feature (below denoted d0) the thickness of the target feature (below denoted t) may be selected so that the amount of produced X-ray radiation is sufficiently large. As the electrons penetrate into the target the electron beam broadens, and thus fewer electrons are available for interaction with the target feature. A low thickness of the target feature means that electrons available for interaction with the target feature penetrate into the substrate material without further contributing to the produced X-ray radiation. A reasonable engineering option may be to select the thickness of the target feature such that at least two thirds of the electrons have scattered out of the target feature before the electron beam has penetrated to a depth corresponding to the thickness of the target feature. The thickness should in this case be larger than a limit thickness that may be written as
Another option could be to ensure that a larger fraction of the electrons contributes to the production of X-ray radiation, and the criterion may then instead be set so that 90% of the electrons have scattered out of the target feature. The expression for the limit thickness may then instead be written as
It is noted that the limit thickness depends on the electron energy. Thus, a plurality of target features may be provided with different thicknesses adapted to corresponding different electron energies. A controller provided in an X-ray source may be configured to direct the electron beam to a target feature suitable for the set acceleration voltage.
Making the target feature considerably thicker than the limits introduced in Equations (4) and (5) above may not increase the X-ray output significantly as indicated in Fig.2. Increasing the thickness of the target feature will also result in more self-absorption, i.e. reabsorption of X-ray radiation in the target feature, and thus a decrease in X-ray output for thicknesses in excess of an energy and geometry dependent optimum thickness. An upper limit for the thickness of the target feature may be imposed by considering the upper limit for the penetration depth of electrons into the target material. The upper limit for electron penetration depth may written as
(6) Making the target feature thicker than this will only result in more self-absorption and not in more X-ray production. It should be acknowledged, however, that some degree of self-absorption (which is preferential to low-energy photons) may in some instances be desired due to the monochromatizing effect it has on the generated X- ray radiation. Taking all the above aspects into consideration, the requirements on the thickness of the target feature for a given diameter thereof to reach the level of two thirds of the beam being scattered out of the target feature may be written as
where d0 and t are in m. The corresponding range for the 90% scattering level may be written as
Considering that there will always be some self-absorption, it may be preferable for implementations to operate in the lower part of the ranges allowed according to Equations (7) and (8). For the particular case where the target feature comprises tungsten, Equations (7) and (8) yield the following allowed ranges for the thickness of the target feature 5.84 ⋅ 10^^ ⋅ ^^.^^ (9)
max ^ ^^ ^ < ^ < 5.84 ⋅ ^^ ^.^^ 7.38 ∙ 10^^ ∙ (^^^)^⁄ ^ 10 ⋅ ^ (10) where Equation (9) corresponds to the two thirds scattering level and Equation (10) to the 90% scattering level. For target features with diameters 100 nm, 300 nm, and 500 nm corresponding thickness ranges for electron energies equal to 80 keV, 160 keV, and 240 keV are tabulated below. d0 E 80 keV 160 keV 240 keV 100 nm (2/3) 180 nm < t < 9 µm 280 nm < t < 28 µm 370 nm < t < 55 µm 100 nm (90%) 300 nm < t < 9 µm 470 nm < t < 28 µm 610 nm < t < 55 µm 300 nm (2/3) 370 nm < t < 9 µm 590 nm < t < 28 µm 770 nm < t < 55 µm 300 nm (90%) 610 nm < t < 9 µm 980 nm < t < 28 µm 1.28 µm < t < 55 µm 500 nm (2/3) 520 nm < t < 9 µm 820 nm < t < 28 µm 1.10 µm < t < 55 µm 500 nm (90%) 860 nm < t < 9 µm 1.37 µm < t < 28 µm 1.80 µm < t < 55 µm As a comparison the optimal thickness (topt) for a continuous tungsten target layer may, according to Sofiienko et al ("Electron range evaluation and X-ray conversion optimization in tungsten transmission-type targets with the aid of wide electron beam Monte Carlo simulations." 11th European Conference on Nondestructive Testing. ECNDT.2014.), be written as
^ = 4.8 ∙ ^^ ^.^^ ^^^ 10 ^ m. (11) This translates to 3.1
for 80 keV, 8.8
for 160 keV, and 16
for 240 keV respectively. As discussed above the optimum thickness will be smaller for the case where the target feature and the electron beam spot are comparable in size. Furthermore, as discussed above, it may be beneficial to use a somewhat thicker target feature than that corresponding to maximum X-ray production due to the beam hardening effect that may be achieved. For the particular case where the target feature comprises tungsten of a thickness t the diameter of the target feature should be less than 107000^^⁄ ^ /^ m or less than the target feature thickness, whichever is smallest, to achieve a beam broadening such that two thirds of the electrons have scattered out of the target feature. To reach a situation where 90% of the electrons scatter out of the target feature the diameter should be less than 50000^^⁄ ^ /^ m or less than the target feature thickness, whichever is smallest. As a set of non-limiting examples consider electron energies of 80 keV, 160 keV, and 240 keV and thicknesses of 0.5 µm, 1 µm, and 2 µm. This would imply corresponding upper limits of the diameter of the respective target feature at the respective scattering level according to the table below. t E 80 keV 160 keV 240 keV 0.5 µm (2/3) 470 nm 240 nm 160 nm 0.5 µm (90%) 220 nm 110 nm 70 nm 1 µm (2/3) 1 µm 670 nm 450 nm 1 µm (90%) 620 nm 310 nm 210 nm 2 µm (2/3) 2 µm 1.89 µm 1.26 µm 2 µm (90%) 1.76 µm 880 nm 590 nm Fig.5 schematically shows an example of an X-ray source including a target according to the present invention. The X-ray source comprises an electron source 110 with a cathode 112, coupled to a power supply 114. Electrons emitted by the cathode 112 are accelerated towards a ring-shaped anode 116 to form an electron beam, indicated by I in Fig. 5. Beam aligning means 118 and electron optics 120 are
provided to shape and direct the electron beam towards the target 130. As described above, the target 130 comprises a substrate 132 and one or more target features provided in an embedding layer provided on a top face 134 of the substrate facing the incoming electron beam. As shown, the electron source (and the cathode/anode), the aligning means, the electron optics and the target are provided in a low-pressure (vacuum) environment enclosed by a housing 140. Further, at least the aligning means and the electron optics are operatively connected to a controller 150. In conclusion, a target for an X-ray source is disclosed herein. The X-ray target comprises a substrate having a top face, and an embedding layer provided on the top face of the substrate. A target feature operative to generate X-ray radiation upon electron impact is at least partly provided in the embedding layer without reaching into the substrate. Preferably, an extension of the target feature in a direction orthogonal to the top face is larger than an extension of the target feature in a direction parallel to the top face. An X-ray source comprising such X-ray target is also disclosed. Arrangements, sources, and methods according to the invention may be used for different types of X-ray imaging such as X-ray microscopy, radiography, fluoroscopy, laminography, ptychography, or CT scanning.