10) Máy đo phổ tia X - Tụ điện HVC, Tụ gốm HV để chế tạo Tất cả các loại máy X-quang.
rithm that is based on the comparison of two separately recorded
images; one of them is recorded from a phantom sample and the
other one is a reconstructed image of the real sample, and the
correlation coefficients are calculated. The phantom sample is used
to calculate the set of projections and to compare between the
phantom and the reconstructed image. For the image reconstruc-
tion, the authors applied the wavelet transform theory and the
Fourier back projection and algebraic reconstruction technique.
The paper introduces the reader into the principal ideas of the
above listed reconstruction algorithm developed for XRF tomog-
raphy. Absorption X-ray microtomography gives an excellent tool
for analysis of micro- and nanoparticles. Jenesson et al. published
(E13 ) on a complete system of tomography spectrometer with a
spatial resolution of 5 ím using a microfocal spot X-ray tube, a
high-resolution complementary metal oxide semiconductor flat
panel pixellated detector with 2400 2400 pixels (size of one pixel
is 50 ím), and a mechanical rotation stage for sample movement.
The system was developed for tomographic analysis of dynamic
sintering processes; therefore, a special sample chamber was
added to the system in which the temperature and atmospheric
pressure could be varied. The authors outlined in their paper that
their developed spectrometer can be utilized in fields where
nanoparticles are handled, as in the pharmaceutical and biotech-
nological industry. Resonant X-ray Raman scattering (RRS)
informs about the threshold structure of the atomic shells because
it becomes dominant when the monoenergetic X-ray excitation
energy has an extremely narrow energy bandwidth and its energy
is precisely close to a near-resonance energy. Szlachetko et al.
described ( E14 ) RRS experiments on Si, SiO
2
, Al, and Al
2O3
, Nơi
the X-ray beam was monochromatized by means of a 20-Å Ni/
B4
C multilayer with suppression of residual higher energy photons
by a Ni mirror. The achieved incident photon number per second
was about 10
12
-10
13
on the target surface, with an energy
resolution of 6 eV. Several energy levels were excited between
1540 and 1900 eV, depending on the target material, and precise
beam energy calibration was performed by the K-absorption
edges. The authors outlined in their final conclusions that RRS
spectroscopy helps to observe the influence of the chemical state
of the atoms on the energy level structure in the electron shell.
A new confocal XRF technique was published by Vekemans
et al. (E15 ) for 3D quantitative elementary analysis of natural
diamond samples in order to characterize the inclusions inside
the diamond bulk, at the ID18F beamline in the ESRF. These
inclusions are potential information sources about the geophysical
and geochemical origin of the diamond. In the confocal technique,
the exciting X-ray beam is focused with a polycapillary lens to
the analyzed microvolume (several tens of ím
3
) of the sample
and the emitted characteristic lines from this volume are guided
and defocused by another polycapillary onto the whole surface of
the semiconductor detector. The excitation X-ray energy was set
at 25 keV, the depth resolution of the confocal analysis, i.e., the
focal spot of the primary polycapillary lens, was 20 ím (fwhm),
and the 3D map was constructed from 31 31 2D mappings at
30 different individual depths. The measuring time of one single
measuring point was 5 s.
The recent advances in the instrumentation of automated
cryogenic sample holders and high spatial resolution X-ray optics
offer a new possibility in X-ray absorption tomography for 3D
imaging of single cells. Le Gros et al. published ( E16 ) their recent
experimental setup for bending magnet and undulator insertion
devices for both polychromatic and monochromatic X-ray beams;
the spectrometer provided a resolution of 50 nm. In case of
bending magnets, two FZPs were used for focusing the primary
beam and to form the final images that were detected by a CCD
camera. When the undulator radiation was used, the first FZP
was substituted by an off-axis zone plate monochromator and
rotating plane mirrors. The sample was involved in a silicon nitrate
cone of 100 nm thick, separating the sample zone and the vacuum
space, where the X-ray beams were formed by X-ray optical
elements as described above. The sample holder containing the
cells was mounted on a x,y,z,£ stage for translation moving and
for rotating the sample. The authors demonstrated the capability
of the microtomography spectrometer with some 3D pictures of
different cells, showing the subcell structures.
Castro et al. published (E17 ) a simple method to characterize
different human tissues based on the measurement of the angular
distribution of scattered radiation. The primary beam was mono-
chomatized with a double-crystal Si(111) device, and the X-ray
energy was 8 keV. The scattered beams were detected by NaI(Tl)
detector in an angle interval from 5 ° to 50° with 0.2° steps. The
authors investigated different urine, renal, and breast tissues in
both healthy and cancerous forms and found significant differ-
ences between the scattering angle-resolved spectra, which acted
as “fingerprints”.
TOTAL REFLECTION X-RAY FLUORESCENCE
PHÂN TÍCH
The simplest XRF method is offered by the TXRF analytical
technique due to the simplicity in sample preparation, quantifica-
tion procedures, and attractive low detection limit down to 10
8
nguyên tử / cm
2
. The greatest motivation of the TXRF method is
semiconductor research; in this realm, TXRF offers a most suitable
technique for testing the surface impurities on Si wafers. The other
field of interest is the analysis of the biological, medical, and
environmental samples due to the multielement character. Pajek
(F1) et al. published a review article on the essential problem of
the detection limits in TXRF analysis and their statistical proper-
ties. The authors offered a new censoring procedure for estimation
of the mean concentrations below the mean level of detection
limits that is based on the Kaplan-Meier statistical approach, and
they also concluded that these results should be of interest for
semiconductor research. For testing the proposed statistical
method, they performed Monte Carlo simulations to estimate the
mean and median concentrations from the detection limits
censored data for which the calculation was based on the
assumption of a log-normal distribution of concentrations and
detection limits.
A critical step in the sample preparation in TXRF analysis is
the sample dropping onto the surface of the quartz or Si holder,
because the sample drying under normal laboratory conditions
(temperature and air ventilation) takes generally about 2 -3hin
the case of 1- íL dropping volume. Miller and co-workers reviewed
(F2) the possibility of a nanoliter (10-50 nL) dropping technique
in single element and multielement standardization. In their
experiments, an X-ray tube operating at 30-keV high voltage and
100-mA current was used, and the typical detection limit for Fe
was found to be10
10
nguyên tử / cm
2
for a 0.8-cm
2
X-ray excitation
beam area for both nanoliter dried spot and the vapor-phase
decomposition (VPD) technique. The authors emphasized that,
by using nanoliter droplets, the required sample volume is
drastically reduced resulting in a higher sample throughput than
with the conventional VPD method.
Kurunczi and Sakurai (F3) also studied the reduced sample
quantity problem in TXRF analysis for natural water specimens
using a Johansson-type WD spectrum spectrometer. The Si wafer
was cleaned with HF etching in order to obtain a hydrophobic
surface on which they arranged an 80- ím sample diameter after
drying the droplet spot. The motivation of this research was that
the WDXRF spectrometer was equipped with focusing optics for
secondary X-ray radiation emitted by the droplet on the sample
holder, and this device requires a very small size of the sample
nơi.