Escolar Documentos
Profissional Documentos
Cultura Documentos
a r t i c l e
i n f o
Article history:
Received 7 May 2009
Received in revised form 29 July 2009
Available online 19 August 2009
PACS:
52.38.Ph
52.59.Px
87.59.e
Keywords:
X-ray tube
X-ray target
Microstructure
Microfocus
a b s t r a c t
We present a novel concept to develop a microfocus X-ray tube based on a microstructured X-ray target
that is irradiated with a nonfocused electron beam. X-ray emissions from the microstructured targets
with various morphologies were studied using Monte-Carlo simulation code MCNP5. The calculations
revealed that the microstructured targets are quite capable of minimizing the effective X-ray spot size
compared with those of conventional transmission-type X-ray targets. Based on the simulation results
of X-ray brightness, optimum geometric parameters were derived for the microstructured targets with
different morphologies. Moreover, the stability of the microstructured target against heat loads delivered
by an electron beam was also investigated under both the continuous and pulsed operation modes. From
the analysis, the limitations of the maximum allowable electron beam currents for the stable operation of
the X-ray targets are presented. The combination of the microstructured targets and nonfocused electron
beam allows the miniaturization of a microfocus X-ray tube by eliminating the needs for massive and
complex focusing devices.
2009 Elsevier B.V. All rights reserved.
1. Introduction
Microfocus X-ray tubes have now become a robust and powerful tool for high quality X-ray imaging because they provide minute image details at the micron level [1,2]. The high image
resolution rendered by microfocus X-ray sources is due to their
small effective X-ray focal spot size [3,4]. The term, effective Xray focal spot size, stems from the fact that the interactions between
the impinging electrons and the atoms of an X-ray target result in
signicant electron beam broadening [5]. The phenomenon of the
electron beam broadening happens even if a well focused and collimated electron beam strikes an X-ray target. Minimization of the
effective X-ray spot size is desirable in order to enhance the spatial
resolution of an X-ray image. In this sense, a reection-type X-ray
target composed of a bulk metal is not feasible to obtain smaller
X-ray focal spot size because of serious electron beam broadening
inside the target. Recently, a transmission-type X-ray target made
of a thin metal layer on an X-ray window has been widely used to
obtain a small X-ray focal spot size, especially for microfocus X-ray
tubes [6]. Since the thickness of a metal layer in a transmissiontype X-ray target is of the order of a few micrometers, electron
beam broadening inside the target can be limited [7]. Thus, most
of the present microfocus X-ray tubes are operated with a transmission-type X-ray target and a micrometer-sized electron beam
* Corresponding author. Tel.: +82 42 350 3823; fax: +82 42 350 3810.
E-mail address: socho@kaist.ac.kr (S.O. Cho).
0168-583X/$ - see front matter 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.nimb.2009.08.012
A. Ihsan et al. / Nuclear Instruments and Methods in Physics Research B 267 (2009) 35663573
3567
microstructured X-ray targets of different morphologies, we propose the optimum dimensions of the targets to achieve the brightest possible X-rays. In addition, we discuss the thermal stability of
the microstructured X-ray targets and the limitations of electron
beam power loading for the stable operation of the target.
2. Materials and methods
2.1. The concept of a microfocus X-ray tube with a microstructured
target
In X-ray radiography, microfocus X-ray tubes are considered
state of the art for rendering high spatial resolution of an image because of their small focal spot sizes. Generally, a very ne-focused
and small-sized electron beam is considered an essential element
of a microfocus X-ray system. However, once the electrons strike
an X-ray target, the effective focal spot size becomes larger than
the actual size of the electron beam due to beam broadening. A
transmission-type X-ray target, which is composed of a thin metallic layer supported by a thicker X-ray window layer, is very helpful
in minimizing the effective focal spot size by restricting the electron beam broadening. A conventional microfocus X-ray tube is
generally composed of a well-focused electron beam and a thin
transmission-type X-ray target, as illustrated in Fig. 1a.
Alternatively, we propose another approach to form a microfocus X-ray tube, as shown in Fig. 1b. In our concept, a metallic
microstructure is attached to a substrate that acts as an X-ray
window. The microstructured X-ray target is irradiated with a
nonfocused electron beam with a much larger size than the
micrometer-sized X-ray target. Although the beam size is sufciently large, X-rays are produced only at the microstructured
target and electrons passing beyond the lateral dimensions of the
micro-particle cannot contribute to X-ray production. In addition,
even if beam broadening occurs inside the microstructured target,
the electrons crossing over the lateral extent of the target do not
produce X-rays. Consequently, the focal spot size of the generated
X-ray is determined only by the lateral size of the microstructures
and independent of the incident electron beam size. A
3568
A. Ihsan et al. / Nuclear Instruments and Methods in Physics Research B 267 (2009) 35663573
beams cannot spread out beyond the lateral dimensions of the targets. However, despite the preclusion of electron beam broadening,
the effective focal spot size of the X-ray produced from a microstructured target may not be comparable with the real dimensions
of the target. It is because that the X-ray emission takes place isotropically throughout the length of a microstructured target, and
hence, the effective size of the X-ray focal spot at the end of the target is larger than the target size.
In order to dene the effective X-ray focal spot size, we can consider an imaginary plane normal to the target central axis on the
backside surface of the target. Then, the effective X-ray spot size
can be dened as the diameter of the region on the imaginary
plane through which the X-rays are emanating from. A prevalent
approach for the estimation of an effective X-ray spot size is based
on the distribution of X-ray intensity obtained from the targets
[2022]. These reports indicate full width at half maximum
(FWHM) of the Gaussian as an established criterion for the estimation of X-ray spot size from the distribution of X-ray intensity. In
this work, we calculated the lateral distribution of X-ray intensity
on the backside of the target. Simulations were carried out in
coupled electron-photon mode. In order to calculate the intensity
distribution, an MCNP point detector tally was employed. The
effective spot size was then estimated from the FWHM of the
X-ray intensity distribution at the imaginary plane for each of
the target geometry.
2.4. X-ray brightness
Although a smaller focal spot is desirable to obtain a better spatial resolution of an X-ray image, such a small focal spot can be a
drawback from the standpoint of X-ray production. A very small
X-ray spot size might not be useful because of its poor productivity. In this sense, X-ray spot size and produced X-ray intensity are
considered to be competing parameters. To evaluate the X-ray
source both in size and in X-ray intensity, we considered the
brightness of the produced X-ray as a gure of merit. Brightness
takes into account both X-ray spot size and X-ray intensity, and
is dened as [22]:
I
;
DA DX
The results of the effective X-ray spot size and the X-ray brightness produced from each of the target are enclosed in the following
subsections. The X-ray brightness produced from each of the
microstructured targets is also compared with that produced from
a foil target with a thickness of 3 lm. The foil thickness of 3 lm
was chosen because it is optimum for the maximization of brightness for Mo targets irradiated by a 50 keV electron beam [11]. The
geometric dimensions (in lm) of the microstructured targets used
in the simulations are indicated within the legends of the respective intensity proles (Figs. 35, 7 and 9). The FWHM of the Gaussian for each of the intensity prole is also included in the same
legend and its value is underlined. The acronyms used for the geometric dimensions are mentioned in Fig. 2.
3.1. Effective focal spot size and brightness
3.1.1. Spherical micro-particles (MSP)
Six different cases of spherical micro-particles were simulated
with diameters ranging from 3 to 10 lm. Fig. 3 shows the intensity
proles for the estimation of effective focal spot size. As expected,
the effective spot size was found to increase with an increase in the
particle diameter. The result of the X-ray brightness produced from
these particles is shown in Fig. 6. A comparison of these results
with the brightness produced from a foil target with a thickness
of 3 lm is also shown with the help of a dotted horizontal line. A
comparison of the result with that of a foil target indicates a decrease in brightness for the particles with the diameter exceeding
4 lm. This is because of the increase in the focal spot size and the
reduction in the X-ray intensity. The latter can be explained in that
the increase in the particle size increases the penetration thickness
of the X-ray inside the particle, while X-ray attenuation occurs during the penetration. This implies that spherical morphologies with
small diameter can yield higher X-ray brightness.
where I is the X-ray intensity emitted into the solid angle DX and
DA is the effective focal spot size of X-rays. The brightness of X-rays
as dened in the above equation was evaluated for all of the morphologies of the microstructured targets shown in Fig. 2.
3. Results and discussion
The evaluation of X-ray brightness was carried out through a
two-step calculation process. In the rst step, the effective focal
spot size was estimated by MC simulations for different morphologies of microstructured targets. The morphologies of microstructured geometries include the following: spherical and
hemispherical micro-particles, micro-shells, micro-rods and micro-cones. An electron beam with a size of 10 lm and energy of
50 keV was employed throughout the simulations for the calculation of effective spot size and X-ray intensity. Molybdenum (Mo)
was considered to be the target material. In the second step, the
X-ray intensity generated from each target type was calculated.
In the X-ray intensity calculation, we considered X-rays produced
into the solid angle of 0.8415 sr, which corresponds to the cone angle of 60. Finally X-ray brightness was evaluated from the calculated values of the effective focal spot size and X-ray intensity
using Eq. (1). In these calculations, the number of particles histories for each simulation case was 5 107, which was sufcient to
keep the statistical error within the reliable limits.
Fig. 4. Lateral distribution of X-ray intensity emitted from hemispherical microparticles with different diameters.
A. Ihsan et al. / Nuclear Instruments and Methods in Physics Research B 267 (2009) 35663573
3569
Fig. 5. Lateral distribution of X-ray intensity emitted from micro-shells with inner diameters of (a) 2 lm, (b) 3 lm, (c) 4 lm and (d) 5 lm.
the electron beam size was xed as 10 lm. The intensity proles
are shown in Fig. 5, whereas the brightness results are shown in
Fig. 6. The results indicate that for a xed inner shell diameter,
the effective spot size increases with an increase in the shell thickness. Consequently, a decreasing trend in brightness is observed.
For cases like 1MS3 and 2MS3, brightness still decreases despite
the spot size not being larger than that of foil. This is because of
the decrease in the intensity, which is caused by the larger thickness of the shell. Micro-shells with thicknesses of up to 2 lm are
found to produce brighter X-rays.
3570
A. Ihsan et al. / Nuclear Instruments and Methods in Physics Research B 267 (2009) 35663573
Fig. 7. Lateral distribution of X-ray intensity emitted from micro-rods with diameters of (a) 3 lm, (b) 4 lm, (c) 5 lm, (d) 6 lm, (e) 8 lm and (f) 10 lm.
A. Ihsan et al. / Nuclear Instruments and Methods in Physics Research B 267 (2009) 35663573
3571
Fig. 9. Lateral distribution of X-ray intensity emitted from micro-cones with base diameters of (a) 3 lm, (b) 4 lm, (c) 5 lm, (d) 6 lm, (e) 8 lm and (f) 10 lm.
The irradiation of an intense electron beam onto the microstructured target can induce melting of the microstructure, thereby
result in operational instability. In addition to the production of Xrays, an electron beam produces heat inside an X-ray target. Almost 99% of the electron energy is converted into heat. For stable
operation, the heat must be dissipated quickly and efciently so
that the maximum temperature of a target/substrate assembly
can be kept much lower than the melting limits of the materials
involved.
Thermal loading scenario in the case of a micro-particle is a bit
different from that of a foil target. In the case of a foil target, a
micrometer-sized electron beam strikes the target and looses a
portion of its energy while penetrating through the foil. The beam
then enters the substrate layer with greatly reduced energy and
deposits the rest of the energy in the substrate. The substrate is
3572
A. Ihsan et al. / Nuclear Instruments and Methods in Physics Research B 267 (2009) 35663573
Q qC p
@T
rkrT
@t
Q qC p
Fig. 11. Axis-symmetric temperature prole of a micro-cone (beam current 3.4 A,
pulse size 1 ls).
@T 1 @
@T
@
@T
kr
k
;
@t r @r
@r
@z
@z
Fig. 12. Axial and radial temperature proles for the target (micro-cone) and substrate regions (beam current 3.4 A, pulse size 1 s).
A. Ihsan et al. / Nuclear Instruments and Methods in Physics Research B 267 (2009) 35663573
the temperature. The temperature of the substrate backside and circular periphery of target/substrate assembly was assumed to be
kept at 300 K, which can be realized by water cooling. Convection
heat transfer was not taken into account, in order to testify the
model for operational stability for even adverse and unfavorable
conditions.
3.3.2. Temperature prole and maximum beam current
The intensity and consequently the X-ray brightness depends
on the electron beam current impinging on an X-ray target.
According to our calculations, in the case of a foil target incident
upon by 50 keV electron beam, a beam current of about 17 lA is
needed to achieve a high level of brightness, such as 109 photons
s1 lm2 msr1 (eld of view: 60) [11]. This implies that the target composed of micro-particles should also be able to withstand a
beam current of at least 17 lA in order to produce bright X-rays.
We investigated the capability of the micro-cone target to withstand high beam currents. At rst, we carried out a steady-state
analysis with the conditions described in the previous subsection
using COMSOL Multiphysics. Under steady-state conditions, when
the target/substrate assembly was irradiated by a 50 keV electron
beam 1 mm in size, the maximum value of the beam current that
the target/substrate assembly could withstand without being
melted was found to be 13.5 mA. Only a fraction of this beam
current strikes against the target, and the effective value of beam
current that is responsible for X-ray production is just 0.67 lA. This
value is much smaller than 17 lA needed for a foil target to obtain
a brightness of 109. This clearly indicates that the microstructured target cannot produce such high X-ray brightness under a
steady-state operation.
Therefore, instead of a steady-state operation, transient conditions when the electron beam is operated in the pulse mode were
investigated. The calculations show that the target/substrate
assembly could withstand a beam current of 3.4 A when an electron beam with a pulse width of 1 ls is incident upon. The beam
current incident upon the microstructured target is then calculated
to be 85 lA, which is much larger than the 17 lA needed in the
case of foil. Figs. 11 and 12 exhibit the results of a particular scenario in pulse operation with the pulse a width of 1 ls and a beam
current of 3.4 A. Fig. 11 shows an axis-symmetric temperature prole of the target (micro-cone) soon after the heat source is
switched off (after 1 ls). Interestingly, the maximum beam current
under both the steady-state and pulse operation conditions is limited by the melting temperature of the Be substrate (1560 K), and
the maximum temperature of the target remained considerably
lower than its melting limit (2896 K). Fig. 12a and b show the temperature prole along the cone axis and targetsubstrate interface.
While axial and radial temperature proles for Be substrate are
exhibited in Fig. 12c and d. The results indicate that the temperature decreases sharply in the radial direction for the regions beyond the beam diameter. In the axial direction, temperature falls
off beyond the depth equal to the range of 50 keV electrons in
the substrate (35 lm).
According to the analysis, a target composed of a microstructure
is found to be promising in producing high X-ray brightness under
3573
pulse operation, without inducing the melting of any target/substrate assembly material.
4. Conclusion
We have demonstrated a new concept to develop a microfocus
X-ray tube, which utilizes a microstructured target that is irradiated by a nonfocused electron beam. Monte-Carlo simulation results demonstrate that microstructured targets are very effective
to obtain microfocus X-rays. Moreover, the calculation results
show that certain morphologies of the microstructured targets
with optimized dimensions are promising in producing high Xray brightness. However, the microstructured targets are not capable of withstanding continuous heat loads. As a result, the utility of
the microstructured targets is limited to pulse operation. The proposed concept requires no focusing devices in order to obtain a
micrometer-sized X-ray focal spot, and thus, we believe that this
would be very useful for the realization of miniaturized microfocus
X-ray tubes.
Acknowledgement
This work was supported by the Korea Science and Engineering
Foundation (KOSEF) grant funded by the Korea Ministry of Education, Science and Technology (MEST) (No. 2008-03983).
References
[1] W. Kai-Ge, W. Lei, L. Wen-Qing, N. Han-Ben, Chin. Phys. Lett. 23 (9) (2006)
2368.
[2] J. Jakubek, T. Holy, M. Jakubek, D. Vavrik, Z. Vykydal, Nucl. Instr. Meth. A 563
(2006) 278.
[3] M. Salamon, R. Hanke, P. Kruger, N. Uhlmann, V. Voland, Nucl. Instr. Meth. A
591 (2008) 50.
[4] S. Krimmel, J. Stephen, J. Baumann, Nucl. Instr. Meth. A 542 (2005) 399.
[5] J.K. Hyun, P. Ercius, D.A. Muller, Ultramicroscopy 109 (2008) 1.
[6] P.M. Jenneson, R.D. Luggar, E.J. Morton, O. Gundogdu, U. Tuzun, J. Appl. Phys.
96 (5) (2004) 2889.
[7] A. Yoshioka, Y. Yamaguchi, K. Tamura, R. Shimizu, Surf. Interface Anal. 37
(2005) 356.
[8] S. Senda, Y. Sakai, Y. Mizuta, S. Kita, F. Okuyama, Appl. Phys. Lett. 85 (23)
(2004) 5679.
[9] S.H. Heo, A. Ihsan, S.O. Cho, Appl. Phys. Lett. 90 (18) (2007) 183109.
[10] A. Hirakimoto, S. Ohnishi, H. Maeda, T. Kishi, T. Shiota, T. Tamura, M. Ukita, S.
Fujita, M. Kamegawa, Spectrochim. Acta B 59 (2004) 1101.
[11] A. Ihsan, S.H. Heo, S.O. Cho, Nucl. Instr. Meth. B 264 (2007) 371.
[12] L.R. Harriott, Nucl. Instr. Meth. B 55 (1991) 802.
[13] A.R. Rodriguez, F.H. Ramirez, Microelectron. Eng. 84 (2007) 789.
[14] M.J. Vasile, R. Nassar, J. Xie, J. Vac. Sci. Technol. B 16 (4) (1998) 2499.
[15] P.R. Munroe, Mater. Charact. 60 (2009) 2.
[16] S.K. Tripathi, V.N. Kulkarni, Nucl. Instr. Meth. B 267 (2009) 1381.
[17] A.S. Schneider, B.G. Clark, C.P. Frick, P.A. Gruber, E. Arzt, Mater. Sci. Eng. A 508
(2009) 241.
[18] Los Alamos National Lab., MCNP-A General Monte Carlo N-Particle Transport,
version 5, 2003.
[19] COMSOL Multiphysics version 3.3, COMSOL.INC, 2005.
[20] J.U. Madsen, NDT Int. 22 (5) (1989) 292.
[21] L.L.W. Wang, K. Leszczynski, Med. Phys. 34 (2) (2007) 485.
[22] Y. Yamaguchi, R. Shimizu, Surf. Interface Anal. 39 (2007) 343.
[23] D.R. Lynch, L. Berman, P. Montanez, S. Pjerov, P. Stefan, M. Woodle, SPIE 2855
(1996) 119.