**2. Experimental**

94 Radioisotopes – Applications in Physical Sciences

external magnetic field, the magnetic field produces a torque on the magnetic dipole. The torque is tending to align the dipole with the field, associated with this torque; there is a

*μl* is the orbital magnetic dipole moment of an electron. According to the quantum theory, all spectral lines arise from transitions of electrons between different allowed energy levels within the atom and the frequency of the spectral line is proportional to the energy difference between the initial and final levels. The slight difference in energy is associated with these different orientations in the magnetic field. In the presence of a magnetic field, the elementary magnetic dipoles, whether permanent or induced, will act to set up a field of

Today investigations of magnetic effects on X-ray spectra became actual both from theoretical and experimental points of view. The numbers of works on this subject deal with magnetic circular dichroism (MCD) in X-ray absorption spectroscopy (XAS), that gives information on empty electron states in a valence band and their spin configurations (Thole et al., 1992, Stöhr&Wu, 1994). Several experiments have been performed on the external magnetic field effect on the *K* shell X-ray emission lines. Demir et al., (2006a) determined how the radiative transitions and the structures of the atoms in a strong magnetic field are affected, K*α* and K*β* X-ray production cross sections, the *K*-shell fluorescence yields and *I*(*Kβ/Kα*) intensity ratios for ferromagnetic Nd, Gd, and Dy and paramagnetic Eu and Ho were investigated using the 59.5 keV incident photon energy in the external magnetic fields intensities ± 0.75 T. On the other hand, Demir et al., (2006b) measured *L*3 subshell fluorescence yields and level widths for Gd, Dy, Hg and Pb at 59.5 keV incident photon energy in the external magnetic field of intensities ± 0.75 T. Porikli et al. (2008a; 2008b; 2008c) conduct measurements using pure Ni, Co, Cu and Zn and their compounds. Characteristic quantities such as position of line maxima, full widths at half maximum (FWHM), indices of asymmetry and intensity ratio values were determined in the values of external magnetic field 0.6 T and 1.2 T. Several experiments have been performed on the external magnetic field effect on the *K* shell X-ray emission lines. Commonly, experimental *L* X-ray intensities are measured using radioisotopes as excitation sources (Han et al., 2010; Porikli, 2011b). They have the advantages of stable intensity and energy and of small sizes, which allow compact and efficient geometry, and

Our motivation in performing this experiment has been two fold. First, with the aim of a better understanding of the chemical effect and external magnetic field effect, we conduct measurements using pure yttrium (Y) and its compounds. Characteristic quantities such as position of line maxima, full widths at half maximum (FWHM), indices of asymmetry and *Kβ*1/*Kα*, *Kβ*2/*Kα*, *Kβ*2/*Kβ*1 and *Kβ*/*Kα* intensity ratio values are determined in the values of external magnetic field 0.6 T and 1.2 T. In the present work, the measurements were done using a filtered 22.69 keV from Cd-109 and 59.54 keV from Am-241 point source and Si(Li) detector. Particle size effects were circumvented. Peak areas were determined using Gaussian fitting procedures and the errors in various corrections such as self-absorption and detector efficiency were minimized. The measured values were compared due to the external magnetic field and chemical effect. The measured values for B=0 were compared with other experimental and theoretical results. To our knowledge, these intensity ratio values of Y in the external magnetic field have not been reported in the literature and appear

∆*E*=-*μlB (*1)

potential energy of orientation:

induction of their own that will modify the original field.

they operate without any external power.

#### **2.1 Experimental set up (EDXRF)**

Yttrium compounds can serve as host lattices for doping with different lanthanide cations and they used as a catalyst for ethylene polymerization. As a metal, it is used on the electrodes of some high-performance spark plugs. Yttrium is also used in the manufacturing of gas mantles for propane lanterns as a replacement for thorium, which is radioactive. Developing uses include yttrium-stabilized zirconia in particular as a solid electrolyte and as an oxygen sensor in automobile exhaust systems. Yttrium is used in the production of a large variety of synthetic garnets. Small amounts of yttrium (0.1 to 0.2%) have been used to reduce the grain sizes of chromium, molybdenum, titanium, and zirconium. It is also used to increase the strength of aluminium and magnesium alloys. The addition of yttrium to alloys generally improves workability, adds resistance to high-temperature recrystallization and significantly enhances resistance to high-temperature oxidation (see graphite nodule discussion below).

The studied elements were Y, YBr3, YCl3, YF3, Y(NO3)3.6H2O, Y2O3, YPO4, Y(SO4)3.8H2O and Y2S3. The purity of commercially obtained materials was better than 99%. For powdered samples, particle size effects have a strong influence on the quantitative analysis of infinitely thick specimens. Even for specimens of intermediate thickness, in which category the specimens analyzed in the present study fall, these effects can be significant. Therefore, to circumvent particle size effects all samples were grounded and sieved through a -400 mesh (<37 μm) sieve. The powder was palletized to a uniform thickness of 0.05-0.15 g cm-2 range by a hydraulic press using 10 ton in-2 pressure. The diameter of the pellet was 13 mm.

All of the lines were excited using a 100 mCi Am-241 annular radioactive source and Cd-109 point source of 10 mCi strength (providing 5.0x103 steradian-1 photon flux of Ag Xradiation). The fluorescent X-rays emitted from the targets were analyzed by a Si(Li) detector (effective area 12.5 mm2, thickness 3 mm, Be window thickness 0.025 mm).

For each sample three separate measurements have been made just to see the consistency of the results obtained from different measurements agreed with a deviation of less than 1%. The experimental setup consist of a Si(Li) detector and Cd-109 radioactive source as shown in Fig. 1. The mechanical arrangement to house the source-sample-detector combination in a definite geometry was shown in Fig. 1. An Al, Pb conical collimator was used between the sample and the detector for the excitation to obtain a large beam of emergent radiation and to avoid the interaction of the X-rays emitted by the component elements of the radioactive capsule and detector.An Al, Pb conical collimator was used between the sample and the detector for the excitation to obtain a large beam of emergent radiation and to avoid the interaction of the X-rays emitted by the component elements of the radioactive capsule and detector. This collimator has an external diameter of 13 mm and it was placed in the internal diameter of the radioactive source (8 mm). A graded filter of Pb, Fe and Al to obtain a thin beam of photons scattered from the sample and to absorb undesirable radiation shielded the

Determination of Chemical State and External Magnetic Field

X-ray intensity (normalized)

0

0

spectra were plotted after smoothing.

1

2

3

X-ray intensity (normalized)

4

5

The spectra were plotted after smoothing.

1

2

3

4

Effect on the Energy Shifts and X-Ray Intensity Ratios of Yttrium and Its Compounds 97

Kα

*K*α Energy (keV)

Fig. 2. A typical *K* X-ray spectrum of the Y target in B=0, B=0.6T and B=1.2 T magnetic field.

14.8 15.6 16.4 17.3

Kβ

*K*β

> Y2 (SO4 ) 3 .8H2 O

14.8 15.6 16.4 17.3

Energy (keV)

Fig. 3. Measured *Kα*, *Kβ*1,3 and *Kβ*2,4 spectra of Y, YBr3, YCl3, Y(SO4)3.8H2O and Y2S3. The

B=0.6 T

B=1.2 T

*K*β2,4

*K*β1,3

16.5 16.7 16.9 17.1 17.3

B=0 T

Y

YCl3

YBr3

Y2 S3

detector. The sample-detector and excitation source-sample distances were optimized to get maximum count rate in the fluorescent peaks. The sample was placed approximately at 45° to the source-plane as well as to the detector-plane so that the intensity of scattered radiation could be minimized (Giauque et al., 1973). The count rate kept below 1000 counts s-1 in order to avoid peak broadening, energy shift and non-linearity. The data were collected into 16384 channels of a digital spectrum analyzer DSA-1000. The energy per channel was adjusted as 4 eV to determine the peak centroits and to discriminate the overlapped peaks. The samples were mounted in a sample holder placed between the pole pieces of an electromagnet capable of producing the magnetic field of approximately 2.66T at 2 mm pole range. During the study, the magnetic field intensities of, 0.6 T and 1.2 T were applied to the samples. An ammeter monitored the continuity and stability of the currents feeding the electromagnet. A typical *K* X-ray spectrum of Y at the 0.0 T, 0.6 T and 1.2 T is shown in Fig. 2. A typical *Kα*, *Kβ*1,3 and *Kβ*2,4 spectrum of Y, YBr3, YCl3, Y(SO4)3.8H2O and Y2S3 are shown in Fig. 3.

Fig. 1. Experimental set-up.

detector. The sample-detector and excitation source-sample distances were optimized to get maximum count rate in the fluorescent peaks. The sample was placed approximately at 45° to the source-plane as well as to the detector-plane so that the intensity of scattered radiation could be minimized (Giauque et al., 1973). The count rate kept below 1000 counts s-1 in order to avoid peak broadening, energy shift and non-linearity. The data were collected into 16384 channels of a digital spectrum analyzer DSA-1000. The energy per channel was adjusted as 4 eV to determine the peak centroits and to discriminate the overlapped peaks. The samples were mounted in a sample holder placed between the pole pieces of an electromagnet capable of producing the magnetic field of approximately 2.66T at 2 mm pole range. During the study, the magnetic field intensities of, 0.6 T and 1.2 T were applied to the samples. An ammeter monitored the continuity and stability of the currents feeding the electromagnet. A typical *K* X-ray spectrum of Y at the 0.0 T, 0.6 T and 1.2 T is shown in Fig. 2. A typical *Kα*,

*Kβ*1,3 and *Kβ*2,4 spectrum of Y, YBr3, YCl3, Y(SO4)3.8H2O and Y2S3 are shown in Fig. 3.

Fig. 1. Experimental set-up.

Fig. 2. A typical *K* X-ray spectrum of the Y target in B=0, B=0.6T and B=1.2 T magnetic field. The spectra were plotted after smoothing.

Fig. 3. Measured *Kα*, *Kβ*1,3 and *Kβ*2,4 spectra of Y, YBr3, YCl3, Y(SO4)3.8H2O and Y2S3. The spectra were plotted after smoothing.

Determination of Chemical State and External Magnetic Field

the samples, a sample spinner facility was used in all cases.

measured for same conditions to determine these errors.

**2.4 Spectral profile analysis (WDXRF)** 

**2.3 Experimental set up (WDXRF)** 

Effect on the Energy Shifts and X-Ray Intensity Ratios of Yttrium and Its Compounds 99

A commercial WDXRF spectrometer (Rigaku ZSX 100e) was used for analysis of the different samples. This instrument is usually equipped with a 3 kW Rh-anode tube working at a voltage range of 20–50 kV and a current from 20 to 50 mA. It is possible to use primary beam filters (made of Zr, Al, Ti or Cu) between the primary radiation and the sample holder to reduce the background continuum and to improve the signal-to-noise ratio. Energy resolution and efficiency for each analytical line also depend on the collimator aperture and the analyzer crystal in use. Several different collimators can be used to reduce the step/scan resolution, as well as up to ten analyzer crystals, to better enhance spectral data for a specific element. Detection can be performed using a flow proportional counter (light elements) or a scintillation counter (heavy elements). In this work, analyses were made in vacuum atmosphere. Moreover, to avoid possible problems with inhomogeneity when measuring

To investigate the spectrometer sensitivity in measuring of intensity and energy shift, one sample at same conditions was measured for three times. Because of the use of instruments such as sieve weight and hydraulic press, errors are caused in the results of analysis. These errors were called manual and instrumental errors. Three samples were prepared and

The common method for evaluation of spectra in WDXRF is by the use of net peak line intensity. This is due to the high efficiency in the analytical results from the scintillation and/or the flow counter detectors. These detectors can receive up to 2×106 cps. In contrast, the common spectra evaluation in EDXRF is based on integration of the gross or net peak area due to a lower efficiency in the solid state detectors, usually limited to a maximum count of 5×104 cps. Taking into account these facts and to improve the sensitivity of the signal, the spectral data obtained by the WDXRF equipment were treated using the deconvolution software (Microcal Orgin 7.5), traditionally used in EDXRF spectrometry, to obtain the peak areas. The total number of counts increases considering the total peak area instead of only the analytical line. This leads also to an improvement of sensibility and detection limits. Once samples were analyzed, the identification of elements from the WDXRF spectra was done by using the qualitative scanning mode linked to the equipment, which includes automatic peak and element identification. The principle of WDXRF spectrometry is the use of different analyzer crystals to diffract and separate the different characteristic wavelengths of the elements present in the sample. For that reason, in WDXRF measurements, a multi-spectrum was obtained

resulting from the use of different analyzing crystals, excitation conditions, etc.

line represents the overall fit. The background is shown as a blue line.

ultra-light elements), varying atmospheres, impurities and different sample sizes.

Rigaku has improved their semi-quantitative software package further with the introduction of SQX. It is capable of automatically correcting for all matrix effects, including line overlaps. SQX can also correct for secondary excitation effect by photoelectrons (light and

The obtained multispectra were split into the different individual spectra and were converted to energies by inversion of the channels to be treated using the means of the SQX software to perform spectral deconvolution and fitting and to evaluate element net peak areas from the spectra. Peak fitting was done by iteration to better adjust the peak and the background to minimize the chi-square of the fitting on each spectra. Fig. 4 shows the spectrum of Y. Measured numbers of counts are shown as solid black circles, while the red

Spectrum evaluation is a crucial step in X-ray analysis, as much as sample preparation and quantification. As with any analytical procedure, the final performance of X-ray analysis is determined by the weakest step in the process. The processing of ED spectra by means of computers has always been more evident because of their inherent digital nature. Due to relatively low resolving power of the employed Si(Li) detector, the process of evaluating XRF spectra is prone to many errors and requires dedicated software. For this purpose a software package called ORGIN was used for peak resolving background subtraction and determination of the net peak areas of *K* X-rays which is based on the non-linear least squares fitting of a mathematical model of the XRF spectrum.

#### **2.2 Data analysis (EDXRF)**

The *Kβ*/*Kα* X-ray intensity ratio values have been calculated by using the relation

$$\frac{I(K\beta)}{I(K\alpha)} = \frac{N(K\beta)}{N(K\alpha)} \frac{\varepsilon(K\alpha)}{\varepsilon(K\beta)} \frac{\beta(K\alpha)}{\beta(K\beta)}\tag{2}$$

where *N*(*Kα*) and *N*(*Kβ*) are the net counts under the *Kα* and *Kβ* peaks, respectively. *β*(*Kα*) and *β*(*Kβ*) are the self-absorption correction factor of the target and ε(*Kα*) and ε(*Kβ*) are the detector efficiency for *Kα* and *Kβ* rays. The values of the factors, *I*0*Gε* which contain terms related to the incident photon flux, geometrical factor and the efficiency of the X-ray detector, were determined by collecting the *Kα* and *Kβ* X-ray spectra of Ti, As, Br, Sr, Y, Zr and Ru with the mass thickness 0.02-0.17 g/cm2 in the same geometry and calculated by using the following equation

$$I\_0 \mathbf{G} \boldsymbol{\varepsilon}\_{Ki} = \frac{N\_{Ki}}{\sigma\_{Ki} \beta\_{Ki} \mathbf{t}\_i} \tag{3}$$

where *NKi* and *βKi* (*i*=*α*, *β*) have the same meaning as in Eq. (2). *σKi* is X-ray fluorescence crosssection, *G* is a geometry factor and *t* is the mass of the sample in g/cm2.

The self absorption correction factor *β* is calculated for both *Kα* and *Kβ* separately by using the following expression

$$\beta\_{\rm Ki} = \frac{1 - \exp\{-\[\mu(E\_0)\sec\theta\_l + \mu\_{\rm Ki}(E)\sec\theta\_2]t\}}{[\mu(E\_0)\sec\theta\_l + \mu\_{\rm Ki}(E)\sec\theta\_2]t} \tag{4}$$

where *µ*(*E*0) and *µKi*(*E*) are the total mass absorption coefficients taken from WinXCOM programme which is the Windows version of XCOM. XCOM is the electronic version of Berger and Hubbell's Tables (Berger et al., 1987). The angles of incident photons and emitted X-rays with respect to the normal at the surface of the sample *θ*1 and *θ2* were equal to 45○ in the present setup.

The term *σKi* represents the *K* X-ray fluorescence cross-sections and is given by

$$
\sigma\_{Ki} = \sigma\_K^P w\_K f\_{Ki} \tag{5}
$$

*P* σ *<sup>K</sup>* is the *K* shell photo ionization cross-section (Scofield, 1973), *wK* is the fluorescence yield (Krause et al., 1979) and *fKi* is fractional X-ray emission rate (Scofield, 1974a).

#### **2.3 Experimental set up (WDXRF)**

98 Radioisotopes – Applications in Physical Sciences

Spectrum evaluation is a crucial step in X-ray analysis, as much as sample preparation and quantification. As with any analytical procedure, the final performance of X-ray analysis is determined by the weakest step in the process. The processing of ED spectra by means of computers has always been more evident because of their inherent digital nature. Due to relatively low resolving power of the employed Si(Li) detector, the process of evaluating XRF spectra is prone to many errors and requires dedicated software. For this purpose a software package called ORGIN was used for peak resolving background subtraction and determination of the net peak areas of *K* X-rays which is based on the non-linear least squares fitting of a mathematical model of the XRF

The *Kβ*/*Kα* X-ray intensity ratio values have been calculated by using the relation

β

α

0

section, *G* is a geometry factor and *t* is the mass of the sample in g/cm2.

μ

The term *σKi* represents the *K* X-ray fluorescence cross-sections and is given by

σ

(Krause et al., 1979) and *fKi* is fractional X-ray emission rate (Scofield, 1974a).

*Ki*

β *Ki*

ε

μ

*<sup>N</sup> I G*

σ β

where *NKi* and *βKi* (*i*=*α*, *β*) have the same meaning as in Eq. (2). *σKi* is X-ray fluorescence cross-

The self absorption correction factor *β* is calculated for both *Kα* and *Kβ* separately by using

 θμ

where *µ*(*E*0) and *µKi*(*E*) are the total mass absorption coefficients taken from WinXCOM programme which is the Windows version of XCOM. XCOM is the electronic version of Berger and Hubbell's Tables (Berger et al., 1987). The angles of incident photons and emitted X-rays with respect to the normal at the surface of the sample *θ*1 and *θ2* were equal to 45○ in

*P*

*<sup>K</sup>* is the *K* shell photo ionization cross-section (Scofield, 1973), *wK* is the fluorescence yield

 σ

( ) ( )( ) ( ) ( ) ( )( ) ( ) *IK NK K K IK NK K K*

where *N*(*Kα*) and *N*(*Kβ*) are the net counts under the *Kα* and *Kβ* peaks, respectively. *β*(*Kα*) and *β*(*Kβ*) are the self-absorption correction factor of the target and ε(*Kα*) and ε(*Kβ*) are the detector efficiency for *Kα* and *Kβ* rays. The values of the factors, *I*0*Gε* which contain terms related to the incident photon flux, geometrical factor and the efficiency of the X-ray detector, were determined by collecting the *Kα* and *Kβ* X-ray spectra of Ti, As, Br, Sr, Y, Zr and Ru with the mass thickness 0.02-0.17 g/cm2 in the same geometry and calculated by

ββ β

*Ki*

*Ki Ki i*

*t*

01 2

*E Et*

*Ki*

 θ  θ

*Ki K K Ki* = *w f* (5)

−− + <sup>=</sup> + (4)

01 2 1 exp{ [ ( )sec ( )sec ] } [ ( )sec ( )sec ]

 θμ

*E Et*

*Ki*

<sup>=</sup> (2)

<sup>=</sup> (3)

 β ε α β α

 αε

spectrum.

**2.2 Data analysis (EDXRF)** 

using the following equation

the following expression

the present setup.

*P* σ

A commercial WDXRF spectrometer (Rigaku ZSX 100e) was used for analysis of the different samples. This instrument is usually equipped with a 3 kW Rh-anode tube working at a voltage range of 20–50 kV and a current from 20 to 50 mA. It is possible to use primary beam filters (made of Zr, Al, Ti or Cu) between the primary radiation and the sample holder to reduce the background continuum and to improve the signal-to-noise ratio. Energy resolution and efficiency for each analytical line also depend on the collimator aperture and the analyzer crystal in use. Several different collimators can be used to reduce the step/scan resolution, as well as up to ten analyzer crystals, to better enhance spectral data for a specific element. Detection can be performed using a flow proportional counter (light elements) or a scintillation counter (heavy elements). In this work, analyses were made in vacuum atmosphere. Moreover, to avoid possible problems with inhomogeneity when measuring the samples, a sample spinner facility was used in all cases.

To investigate the spectrometer sensitivity in measuring of intensity and energy shift, one sample at same conditions was measured for three times. Because of the use of instruments such as sieve weight and hydraulic press, errors are caused in the results of analysis. These errors were called manual and instrumental errors. Three samples were prepared and measured for same conditions to determine these errors.

#### **2.4 Spectral profile analysis (WDXRF)**

The common method for evaluation of spectra in WDXRF is by the use of net peak line intensity. This is due to the high efficiency in the analytical results from the scintillation and/or the flow counter detectors. These detectors can receive up to 2×106 cps. In contrast, the common spectra evaluation in EDXRF is based on integration of the gross or net peak area due to a lower efficiency in the solid state detectors, usually limited to a maximum count of 5×104 cps. Taking into account these facts and to improve the sensitivity of the signal, the spectral data obtained by the WDXRF equipment were treated using the deconvolution software (Microcal Orgin 7.5), traditionally used in EDXRF spectrometry, to obtain the peak areas. The total number of counts increases considering the total peak area instead of only the analytical line. This leads also to an improvement of sensibility and detection limits. Once samples were analyzed, the identification of elements from the WDXRF spectra was done by using the qualitative scanning mode linked to the equipment, which includes automatic peak and element identification. The principle of WDXRF spectrometry is the use of different analyzer crystals to diffract and separate the different characteristic wavelengths of the elements present in the sample. For that reason, in WDXRF measurements, a multi-spectrum was obtained resulting from the use of different analyzing crystals, excitation conditions, etc.

Rigaku has improved their semi-quantitative software package further with the introduction of SQX. It is capable of automatically correcting for all matrix effects, including line overlaps. SQX can also correct for secondary excitation effect by photoelectrons (light and ultra-light elements), varying atmospheres, impurities and different sample sizes.

The obtained multispectra were split into the different individual spectra and were converted to energies by inversion of the channels to be treated using the means of the SQX software to perform spectral deconvolution and fitting and to evaluate element net peak areas from the spectra. Peak fitting was done by iteration to better adjust the peak and the background to minimize the chi-square of the fitting on each spectra. Fig. 4 shows the spectrum of Y. Measured numbers of counts are shown as solid black circles, while the red line represents the overall fit. The background is shown as a blue line.

Determination of Chemical State and External Magnetic Field

*Kα*, *Kβ*1,3 and *Kβ*2,4 FWHM values.

Element External Magnetic

Effect on the Energy Shifts and X-Ray Intensity Ratios of Yttrium and Its Compounds 101

The FWHM values for all compounds investigated for Cd-109 and Am-241 radioactive sources are given in Tables 1 and 2. Among all the Y compounds, the Y(NO3)3.6H2O *Kα* emission line shows large widths for both lines. Both the *Kα*, *Kβ*1,3 and *Kβ*2,4 lines become narrow when the compounds crystal system is monoclinic. In all cubic compounds such as Y2O3 and Y2S3 the lines FWHM values come closer to pure Y FWHM value. As can be seen from Table 1 and 2, for Zn compounds, the variation in the values of FWHM is relatively small. But when we compare the Y compounds with pure forms, we realize changes in both

The nonmonotonic behavior of the *Kα* widths is probably due to the behavior of the *L* levels widths rather than the *K* level ones. The smaller overlap of the *M* and *K* wave functions, as compared to the *K* and *L* ones, may reduce the relative influence of possible similar sized nonmonotonic contributions originating in the final state level widths. Since experimental results FWHM and the index of asymmetry for B≠0 cannot be found in the literature, the comparison is not made with the other experimental values. As can be seen from Table 1 and 2, all FWHM values systematically decrease with increasing magnetic field intensity. Table 1 and 2 show that the experimental values of FWHM are unity within experimental

uncertainties, suggesting the absence of chemical and external magnetic field effects.

 *Kα Kβ*1,3 *Kβ*2,4 *Kα Kβ*1,3 **Y B=0 1.125 1.136 1.008 3.232 4.333** B=0.6T 1.079 1.112 1.006 3.223 4.268 B=1.2T 0.991 0.109 0.996 3.057 4.016 **Y(NO3)3.6H2O B=0 1.122 1.127 1.011 3.211 4.228** B=0.6T 0.997 1.103 1.009 3.170 4.220 B=1.2T 0.968 1.002 0.994 3.109 4.103 **YCl3 B=0 1.110 1.119 0.994 3.197 4.267** B=0.6T 1.017 1.077 0.991 3.133 4.209 B=1.2T 0.984 1.004 0.985 3.110 4.111 **YPO4 B=0 1.114 1.111 0.987 3.199 4.337** B=0.6T 0.983 1.071 0.967 3.139 4.259 B=1.2T 0.980 1.002 0.956 3.062 4.058 **YBr3 B=0 1.098 1.078 0.971 3.201 4.284** B=0.6T 1.007 1.006 0.970 3.187 4.065 B=1.2T 1.000 0.989 0.944 3.110 3.943 **Y2O3 B=0 1.071 1.05 0.938 3.266 4.121** B=0.6T 1.010 1.001 0.930 3.133 4.043 B=1.2T 0.994 0.983 0.915 3.109 3.997 **YF3 B=0 1.099 1.077 0.933 3.245 4.264** B=0.6T 1.022 1.011 0.929 3.120 4.001 B=1.2T 0.988 0.984 0.886 3.100 3.966 **Y(SO4)3.8H2O B=0 1.055 1.035 0.921 3.189 4.166** B=0.6T 1.003 1.004 0.917 3.166 3.976 B=1.2T 0.969 0.966 0.910 3.037 3.874 **Y2S3 B=0 1.024 1.031 0.910 3.337 4.494**  B=0.6T 1.012 0.994 0.883 3.266 4.441 B=1.2T 0.985 0.981 0.867 3.190 4.284 Table 1. Full width at half maximum (FWHM) and asymmetry index values of *Kα, Kβ*1,3 and

*Kβ*2,4 emission lines in Y compounds for Cd-109 radioactive source.

Field Asymmetry Index (eV) FWHM (eV)

Fig. 4. Solid circles: Measured spectrum of Y *K* X-rays. Lines: Overall fitting function (red) and its components (green).

#### **3. Results and discussion**

A frequently used and convenient (but not very accurate) quantification of the line shape is by its full width at half maximum (FWHM) and index of asymmetry. These characteristics of lines (line parameters) are sensitive to change: line position (chemical shift), line shape (full width at half maximum (FWHM) and index of asymmetry) and additionally mutual ratios of line intensities. The peak position was determined at the center point of the 9/10 intensity of the smoothed line shape as illustrated in Fig. 5. It was known from our experience that the standard deviation of the peak position was determined using the peak top. Parameters such as FWHM and asymmetry index, defined in Fig. 5, were evaluated using the smoothed data. The Savitzky-Golay smoothing method was iteratively processed one time. Spectral smoothing was important for reducing the standard deviation of these parameters.

Fig. 5. Definition of asymmetry index, FWHM and peak position determined from 9/10 intensity.

0,0 2,0x101 4,0x101

Kβ2,4

Energy (keV) 16.5 16.7 16.9 17.1 17.3

Kβ1,3

6,0x101 8,0x101 1,0x102

Intensity (kcps)

Fig. 4. Solid circles: Measured spectrum of Y *K* X-rays. Lines: Overall fitting function (red)

A frequently used and convenient (but not very accurate) quantification of the line shape is by its full width at half maximum (FWHM) and index of asymmetry. These characteristics of lines (line parameters) are sensitive to change: line position (chemical shift), line shape (full width at half maximum (FWHM) and index of asymmetry) and additionally mutual ratios of line intensities. The peak position was determined at the center point of the 9/10 intensity of the smoothed line shape as illustrated in Fig. 5. It was known from our experience that the standard deviation of the peak position was determined using the peak top. Parameters such as FWHM and asymmetry index, defined in Fig. 5, were evaluated using the smoothed data. The Savitzky-Golay smoothing method was iteratively processed one time. Spectral

smoothing was important for reducing the standard deviation of these parameters.

Fig. 5. Definition of asymmetry index, FWHM and peak position determined from 9/10

14,5 14,6 14,7 14,8 14,9 15,0 15,1 15,2 15,3

Energy (keV)

Kα

0 1x102 2x102 3x102 4x102 5x102 6x102

and its components (green).

**3. Results and discussion** 

Intensity (kcps)

intensity.

The FWHM values for all compounds investigated for Cd-109 and Am-241 radioactive sources are given in Tables 1 and 2. Among all the Y compounds, the Y(NO3)3.6H2O *Kα* emission line shows large widths for both lines. Both the *Kα*, *Kβ*1,3 and *Kβ*2,4 lines become narrow when the compounds crystal system is monoclinic. In all cubic compounds such as Y2O3 and Y2S3 the lines FWHM values come closer to pure Y FWHM value. As can be seen from Table 1 and 2, for Zn compounds, the variation in the values of FWHM is relatively small. But when we compare the Y compounds with pure forms, we realize changes in both *Kα*, *Kβ*1,3 and *Kβ*2,4 FWHM values.

The nonmonotonic behavior of the *Kα* widths is probably due to the behavior of the *L* levels widths rather than the *K* level ones. The smaller overlap of the *M* and *K* wave functions, as compared to the *K* and *L* ones, may reduce the relative influence of possible similar sized nonmonotonic contributions originating in the final state level widths. Since experimental results FWHM and the index of asymmetry for B≠0 cannot be found in the literature, the comparison is not made with the other experimental values. As can be seen from Table 1 and 2, all FWHM values systematically decrease with increasing magnetic field intensity. Table 1 and 2 show that the experimental values of FWHM are unity within experimental uncertainties, suggesting the absence of chemical and external magnetic field effects.


Table 1. Full width at half maximum (FWHM) and asymmetry index values of *Kα, Kβ*1,3 and *Kβ*2,4 emission lines in Y compounds for Cd-109 radioactive source.

Determination of Chemical State and External Magnetic Field

the others.

**Element External** 

in Y compounds for Cd-109 radioactive source.

Effect on the Energy Shifts and X-Ray Intensity Ratios of Yttrium and Its Compounds 103

When the crystal system of Y compounds is cubic, the *Kα* emission lines are almost symmetric for Y2O3and Y2S3. It is clear from results that, except for Y and Y(NO3)3.6H2O ,the asymmetry are generally larger for the *Kα* peak. So we can say that, the line shapes of *Kα* are not symmetric. It is also found from the Table 2, when the crystal system of Y compounds is trigonal (YBr3), the *Kα*, *Kβ*1,3 and *Kβ*2,4 emission lines are almost symmetric too. For Am-241, line shapes are more symmetric than the Cd-109. As seen from Table 1 and 2, in the presence of an external magnetic field, the asymmetry index of the Y compounds change. The *Kα*, *Kβ*1,3 and *Kβ*2,4 emission lines asymmetry indices values decrease with external magnetic field. However, a more asymmetric structure is encountered for the elements of which their crystal symmetry is cubic. Also, it is observed that monoclinic group is more symmetric than

**Magnetic Field Chemical shift (***ΔE***) (eV) Energy shift (***δE***) (eV)** 

 *Kα Kβ*1,3 *Kβ*2,4 *Kα Kβ*1,3 *Kβ*2,4 **Y B=0 0 0 0 0 0 0**  B=0.6T 0.074 0.109 0.101 B=1.2T 0.161 0.260 0.237 **Y(NO3)3.6H2O B=0 -0.333 -0.441 -0.216 0 0 0**  B=0.6T 0.086 0.098 0.115 B=1.2T 0.141 0.137 0.227 **YCl3 B=0 -0.238 -0.309 -0.115 0 0 0**  B=0.6T 0.481 0.235 0.288 B=1.2T 0.612 0.368 0.339 **YPO4 B=0 -0.121 -0.235 -0.103 0 0 0**  B=0.6T 0.335 0.279 0.336 B=1.2T 0.455 0.355 0.399 **YBr3 B=0 -0.033 -0.065 -0.065 0 0 0**  B=0.6T 0.444 0.131 0.338 B=1.2T 0.657 0.4 0.551 **Y2O3 B=0 0.135 0.017 0.056 0 0 0**  B=0.6T 0.111 0.124 0.543 B=1.2T 0.185 0.303 0.441 **YF3 B=0 0.164 0.191 0.198 0 0 0**  B=0.6T 0.167 0.166 0.112 B=1.2T 0.533 0.529 0.144 **Y(SO4)3.8H2O B=0 0.609 0.387 0.33 0 0 0**  B=0.6T 0.089 0.12 0.051 B=1.2T 0.354 0.293 0.237 **Y2S3 B=0 0.899 0.872 0.808 0 0 0**  B=0.6T 0.173 0.204 0.206 B=1.2T 0.199 0.307 0.333 Table 3. Chemical shift (*ΔE*) and energy shift (*δE*) values of *Kα, Kβ*1,3 and *Kβ*2,4 emission lines

To obtain more definite conclusions on FWHM dependency of the external magnetic field, more experimental data are clearly needed. The experimental uncertainties are always <0.05 eV for the FWHM.

According to Allinson (1933), the index of asymmetry of an X-ray emission line is defined as the ratio of the part of the FWHM lying to the long-wavelength side of the maximum ordinate to that on the short-wavelength side. In Table 1 and 2, the index of asymmetry for *Kα*, *Kβ*1,3 and *Kβ*2,4 emission lines are presented. The experimental uncertainties in the values cited in the table were determined taking into account multiple measurements and multiple fits of each spectrum. The errors for the index of asymmetry are ≤0.1 eV for *Kα*, *Kβ*1,3 and *Kβ*2,4.


Table 2. Full width at half maximum (FWHM) and asymmetry index values of *Kα, Kβ*1,3 and *Kβ*2,4 emission lines in Y compounds for Am-241 radioactive source.

To obtain more definite conclusions on FWHM dependency of the external magnetic field, more experimental data are clearly needed. The experimental uncertainties are always <0.05

According to Allinson (1933), the index of asymmetry of an X-ray emission line is defined as the ratio of the part of the FWHM lying to the long-wavelength side of the maximum ordinate to that on the short-wavelength side. In Table 1 and 2, the index of asymmetry for *Kα*, *Kβ*1,3 and *Kβ*2,4 emission lines are presented. The experimental uncertainties in the values cited in the table were determined taking into account multiple measurements and multiple fits of each spectrum. The errors for the index of asymmetry are ≤0.1 eV for *Kα*, *Kβ*1,3 and

 *Kα Kβ*1,3 *Kβ*2,4 *Kα Kβ*1,3 **Y B=0 1.120 1.113 1.022 3.229 4.297** B=0.6T 1.089 1.100 1.016 3.212 4.203 B=1.2T 1.009 0. 989 0.984 3.050 4.100 **Y(NO3)3.6H2O B=0 1.113 1.110 1.020 3.114 4.116** B=0.6T 1.028 1.104 1.013 3.017 4.111 B=1.2T 0.996 0.993 1.001 3.091 4.075 **YCl3 B=0 1.109 1.105 1.017 3.077 4.122** B=0.6T 1.003 1.007 1.009 3.041 4.110 B=1.2T 0.978 0.987 0.987 3.000 4.004 **YPO4 B=0 1.111 1.075 0.993 3.201 4.177** B=0.6T 0.989 1.032 0.984 3.126 4.170 B=1.2T 0.969 1.008 0.953 3.013 4.005 **YBr3 B=0 1.101 1.031 0.988 3.421 4.136** B=0.6T 1.017 1.004 0.980 3.234 4.065 B=1.2T 0.994 0.971 0.965 3.048 3.993 **Y2O3 B=0 1.056 1.006 0.974 3.555 4.008** B=0.6T 1.031 0.999 0.972 3.229 3.989 B=1.2T 0.987 0.977 0.954 3.096 3.974 **YF3 B=0 1.077 1.001 0.970 3.301 4.123** B=0.6T 1.005 0.984 0.954 3.137 4.012 B=1.2T 0.993 0.975 0.950 3.009 3.940 **Y(SO4)3.8H2O B=0 1.064 0.991 0.964 3.202 4.109** B=0.6T 1.001 0.990 0.966 3.113 3.989 B=1.2T 0.978 0.946 0.956 3.074 3.866 **Y2S3 B=0 1.011 0.995 0.949 3.441 4.301**  B=0.6T 1.010 0.994 0.937 3.368 4.039 B=1.2T 0.974 0.961 0.921 3.222 3.974 Table 2. Full width at half maximum (FWHM) and asymmetry index values of *Kα, Kβ*1,3 and

*Kβ*2,4 emission lines in Y compounds for Am-241 radioactive source.

Magnetic Field Asymmetry Index (eV) FWHM (eV)

eV for the FWHM.

Element External

*Kβ*2,4.

When the crystal system of Y compounds is cubic, the *Kα* emission lines are almost symmetric for Y2O3and Y2S3. It is clear from results that, except for Y and Y(NO3)3.6H2O ,the asymmetry are generally larger for the *Kα* peak. So we can say that, the line shapes of *Kα* are not symmetric. It is also found from the Table 2, when the crystal system of Y compounds is trigonal (YBr3), the *Kα*, *Kβ*1,3 and *Kβ*2,4 emission lines are almost symmetric too. For Am-241, line shapes are more symmetric than the Cd-109. As seen from Table 1 and 2, in the presence of an external magnetic field, the asymmetry index of the Y compounds change. The *Kα*, *Kβ*1,3 and *Kβ*2,4 emission lines asymmetry indices values decrease with external magnetic field. However, a more asymmetric structure is encountered for the elements of which their crystal symmetry is cubic. Also, it is observed that monoclinic group is more symmetric than the others.


Table 3. Chemical shift (*ΔE*) and energy shift (*δE*) values of *Kα, Kβ*1,3 and *Kβ*2,4 emission lines in Y compounds for Cd-109 radioactive source.

Determination of Chemical State and External Magnetic Field

which crystal structure different are needed.

measurement from the average value was 0.1 eV.

**Element Differences between FWHM values** 

emission lines in Y compounds obtained for WDXRF.

the *Kβ*1,3 and *Kβ*2,4 peak shifts are large, show prominent asymmetry.

discuss the values of these ratios as obtained in our measurements.

Effect on the Energy Shifts and X-Ray Intensity Ratios of Yttrium and Its Compounds 105

The chemical shift was the difference between the center point of the 9/10 peak intensity of a compound and that of pure Y measured before and after the measurement of the compound. When the environment of the emitting atom is changed, there are changes in the position of emission lines with respect to those in the pure metal. These changes are called chemical shifts. They are presented in Tables 3 and 4. Both the 4*d* electron configuration and crystal structure affect the chemical shift and energy shift with applied external magnetic field. There is a clear relationship with the external magnetic field values and the energy shift values of Y compounds, as is found in Table 3 and 4's last column. For higher values of external magnetic field, the values of energy shift increases systematically. But we do not find any relationship between external magnetic field and crystal structure of the compound. To obtain more definite conclusion, more experimental data for 4*d* compounds

The errors of the chemical and energy shifts, originate mainly from the limited precision of our measurements, determined by repetitive measurements of all pure targets, which were prepared and placed in the same experimental geometry. The precision in the position of *Kα*, *Kβ*1,3 and *Kβ*2,4 lines was determined as 0.05 eV, whereas the maximum deviation of a single

*Kα Kβ***1,3** *Kβ***2,4** *Kα Kβ***1,3** *Kβ***2,4 Y** 0 0 0 0 0 0 **Y(NO3)3.6H2O** 0.155 0.031 0.012 0.132 0.126 0.125 **YCl3** 0.091 0.058 0.033 0.268 0.238 0.284 **YPO4** -0.223 -0.087 -0.107 -0.454 -0,444 -0.415 **YBr3** 0.114 0.071 0.067 0.144 0.179 0.156 **Y2O3** -0.077 -0.068 -0.021 -0.096 -0.126 -0.085 **YF3** -0.095 -0.034 -0.048 -0.115 -0.127 -0.167 **Y(SO4)3.8H2O** 0.054 0.023 0.181 0.183 0.145 0.106 **Y2S3** -0.013 -0.064 -0.079 -0.301 -0.397 -0.385 Table 5. Chemical shift *(ΔE)* and differences between FWHM values of *Kα, Kβ*1,3 and *Kβ*2,4

It is also seen from Table 5 that the *Kα* line width of YPO4 compound is wider than that of the other compounds. Compare with the *Kα* peak of the pure Y, that of cubic crystal structure Yttrium compounds shifted to lower energy, and the peak shift ordering was YF3<Y2O3<Y2S3<YPO4. The line shapes of *Kα* are generally symmetric. Y2S3 and Y2O3, where

The accurate knowledge of the *Kβ*1/*Kα*, *Kβ*2/*Kα*, *Kβ*2/*Kβ*1 and *Kβ*/*Kα* intensity ratios is required for a number of practical applications of X-rays, e.g. molecular and radiation physics investigations, in non-destructive testing, elemental analysis, medical research etc. Therefore, these ratios depend sensitively on the atomic structure. Thus they have been widely used also for critical evaluation of atomic structure model calculations. We now

The relevant information in a spectrum is contained in its peaks whose position and area are linked respectively to the photon energy and the activity of the connected radionuclide. The peak areas can also be used to determine emission probabilities. In this work, peak areas

**[ΔFWHM=FWHMcom.-FWHMpure] (eV) Chemical shift (ΔE) (eV)** 

The more unpaired 3d or 4d electrons the atom possesses, the more asymmetric will be the line observed. This kind observation led Tsutsumi (1959) to consider that the interaction between the hole created in the 2p3/2 or 2p1/2 shell (due to the transition of an electron from this shell to the 1s level) and the electrons in the incomplete 3d shell in the transition metal atoms is responsible for asymmetric nature of *Kα* lines. They proposed a theoretical model based on this idea to account for the asymmetry in the X-ray emission lines in the firs-row transition metal compounds. However, this is not the only consideration which can explain the origin of the asymmetry of the line; there are other considerations which are based on the relaxation effect of the inner state proposed by Parratt (1959) or on the interactions between 2p hole and electrons in the Fermi sea as proposed by Doniach and Sunjic (1970).


Table 4. Chemical shift (*ΔE*) and energy shift (*δE*) values of *Kα, Kβ*1,3 and *Kβ*2,4 emission lines in Y compounds for Am-241 radioactive source.

The more unpaired 3d or 4d electrons the atom possesses, the more asymmetric will be the line observed. This kind observation led Tsutsumi (1959) to consider that the interaction between the hole created in the 2p3/2 or 2p1/2 shell (due to the transition of an electron from this shell to the 1s level) and the electrons in the incomplete 3d shell in the transition metal atoms is responsible for asymmetric nature of *Kα* lines. They proposed a theoretical model based on this idea to account for the asymmetry in the X-ray emission lines in the firs-row transition metal compounds. However, this is not the only consideration which can explain the origin of the asymmetry of the line; there are other considerations which are based on the relaxation effect of the inner state proposed by Parratt (1959) or on the interactions between 2p hole and electrons in the Fermi sea as

**Magnetic Field Chemical shift (***ΔE***) (eV) Energy shift (***δE***) (eV)** 

 *Kα Kβ*1,3 *Kβ*2,4 *Kα Kβ*1,3 *Kβ*2,4 **Y B=0 0 0 0 0 0 0**  B=0.6T 0.125 0.121 0.132 B=1.2T 0.179 0.219 0.269 **Y(NO3)3.6H2O B=0 -0.401 -0.347 -0.316 0 0 0**  B=0.6T 0.091 0.111 0.158 B=1.2T 0.104 0.169 0.201 **YCl3 B=0 -0.023 -0.109 -0.153 0 0 0**  B=0.6T 0.226 0.254 0.259 B=1.2T 0.559 0.472 0.318 **YPO4 B=0 0.231 0.304 0.221 0 0 0**  B=0.6T 0.553 0.602 0.436 B=1.2T 0.598 0.556 0.511 **YBr3 B=0 -0.133 -0.227 -0.194 0 0 0**  B=0.6T 0.342 0.313 0.387 B=1.2T 0.600 0.499 0.458 **Y2O3 B=0 0.440 0.316 0.241 0 0 0**  B=0.6T 0.447 0.423 0.505 B=1.2T 0.682 0.613 0.553 **YF3 B=0 0.206 0.391 0.190 0 0 0**  B=0.6T 0.367 0.276 0.211 B=1.2T 0.533 0.590 0.438 **Y(SO4)3.8H2O B=0 0.194 0.276 0.133 0 0 0**  B=0.6T 0.194 0.146 0.115 B=1.2T 0.402 0.281 0.296 **Y2S3 B=0 0.575 0.503 0.421 0 0 0**  B=0.6T 0.197 0.340 0.247 B=1.2T 0.331 0.349 0.366 Table 4. Chemical shift (*ΔE*) and energy shift (*δE*) values of *Kα, Kβ*1,3 and *Kβ*2,4 emission lines

proposed by Doniach and Sunjic (1970).

**External** 

in Y compounds for Am-241 radioactive source.

**Element** 

The chemical shift was the difference between the center point of the 9/10 peak intensity of a compound and that of pure Y measured before and after the measurement of the compound. When the environment of the emitting atom is changed, there are changes in the position of emission lines with respect to those in the pure metal. These changes are called chemical shifts. They are presented in Tables 3 and 4. Both the 4*d* electron configuration and crystal structure affect the chemical shift and energy shift with applied external magnetic field. There is a clear relationship with the external magnetic field values and the energy shift values of Y compounds, as is found in Table 3 and 4's last column. For higher values of external magnetic field, the values of energy shift increases systematically. But we do not find any relationship between external magnetic field and crystal structure of the compound. To obtain more definite conclusion, more experimental data for 4*d* compounds which crystal structure different are needed.

The errors of the chemical and energy shifts, originate mainly from the limited precision of our measurements, determined by repetitive measurements of all pure targets, which were prepared and placed in the same experimental geometry. The precision in the position of *Kα*, *Kβ*1,3 and *Kβ*2,4 lines was determined as 0.05 eV, whereas the maximum deviation of a single measurement from the average value was 0.1 eV.


Table 5. Chemical shift *(ΔE)* and differences between FWHM values of *Kα, Kβ*1,3 and *Kβ*2,4 emission lines in Y compounds obtained for WDXRF.

It is also seen from Table 5 that the *Kα* line width of YPO4 compound is wider than that of the other compounds. Compare with the *Kα* peak of the pure Y, that of cubic crystal structure Yttrium compounds shifted to lower energy, and the peak shift ordering was YF3<Y2O3<Y2S3<YPO4. The line shapes of *Kα* are generally symmetric. Y2S3 and Y2O3, where the *Kβ*1,3 and *Kβ*2,4 peak shifts are large, show prominent asymmetry.

The accurate knowledge of the *Kβ*1/*Kα*, *Kβ*2/*Kα*, *Kβ*2/*Kβ*1 and *Kβ*/*Kα* intensity ratios is required for a number of practical applications of X-rays, e.g. molecular and radiation physics investigations, in non-destructive testing, elemental analysis, medical research etc. Therefore, these ratios depend sensitively on the atomic structure. Thus they have been widely used also for critical evaluation of atomic structure model calculations. We now discuss the values of these ratios as obtained in our measurements.

The relevant information in a spectrum is contained in its peaks whose position and area are linked respectively to the photon energy and the activity of the connected radionuclide. The peak areas can also be used to determine emission probabilities. In this work, peak areas

Determination of Chemical State and External Magnetic Field

 B=1.2T *Kβ*1,3/*Kα* 0.2322±0.010 *Kβ*2,4/*Kα* 0.0324±0.008

 *Kβ***/***Kα* **0.1818±0.010**  YPO4 **B=0** *Kβ*1,3/*Kα* 0.2355±0.010

 *Kβ*2,4/*Kβ*1,3 0.1999±0.010 *Kβ***/***Kα* **0.1840±0.009**  B=0.6T *Kβ*1,3/*Kα* 0.2336±0.007 *Kβ*2,4/*Kα* 0.0320±0.009 *Kβ*2,4/*Kβ*1,3 0.1966±0.011

 B=1.2T *Kβ*1,3/*Kα* 0.2322±0.011 *Kβ*2,4/*Kα* 0.0304±0.009

 *Kβ***/***Kα* **0.1738±0.011**  YBr3 **B=0** *Kβ*1,3/*Kα* 0.2359±0.008

 *Kβ*2,4/*Kβ*1,3 0.2004±0.009 *Kβ***/***Kα* **0.1848±0.010**  B=0.6T *Kβ*1,3/*Kα* 0.2342±0.008 *Kβ*2,4/*Kα* 0.0321±0.009 *Kβ*2,4/*Kβ*1,3 0.1987±0.010

 B=1.2T *Kβ*1,3/*Kα* 0.2321±0.009 *Kβ*2,4/*Kα* 0.0303±0.011

 *Kβ***/***Kα* 0.1794±0.009 Y2O3 **B=0** *Kβ*1,3/*Kα* 0.2364±0.011

 *Kβ*2,4/*Kβ*1,3 0.2008±0.010 *Kβ***/***Kα* 0.1856±0.011 B=0.6T *Kβ*1,3/*Kα* 0.2351±0.008 *Kβ*2,4/*Kα* 0.0340±0.008 *Kβ*2,4/*Kβ*1,3 0.1998±0.010

 B=1.2T *Kβ*1,3/*Kα* 0.2302±0.010 *Kβ*2,4/*Kα* 0.0307±0.010

Intensity Ratio

Element

External Magnetic Field

Effect on the Energy Shifts and X-Ray Intensity Ratios of Yttrium and Its Compounds 107

Scofield (1974a)

Manson &Kennedy (1974)

Ertuğral et al. (2007)

This Work

*Kβ***/***Kα* **0.1821±0.011** 

*Kβ*2,4/*Kβ*1,3 0.1972±0.009

*Kβ*2,4/*Kα* 0.0333±0.010

*Kβ***/***Kα* **0.1773±0.008** 

*Kβ*2,4/*Kβ*1,3 0.1693±0.009

*Kβ*2,4/*Kα* 0.0341±0.007

*Kβ***/***Kα* 0.1818±0.010

*Kβ*2,4/*Kβ*1,3 0.1976±0.010

*Kβ*2,4/*Kα* 0.0349±0.010

*Kβ***/***Kα* 0.1831±0.010

*Kβ*2,4/*Kβ*1,3 0.1985±0.010

were determined after the *Kα*, *Kβ*1,3 and *Kβ*2,4 areas were separated by fitting the measured spectra with multi-Gaussian function plus polynomial backgrounds using Microcal Orgin 7.5 software program. Details of the experimental set up and data analysis have been reported earlier (Porikli et al., 2011b).

Table 6 lists the theoretical values which were calculated by Scofield (Scofield 1974a; Scofield, 1974b). Addition to this, the measured values of the *Kβ*1/*Kα*, *Kβ*2/*Kα*, *Kβ*2/*Kβ*1 and *Kβ*/*Kα* intensity ratios in Y, and previous experimental and the other theoretical values of these ratios for pure elements and their compounds are listed in Table 6.


were determined after the *Kα*, *Kβ*1,3 and *Kβ*2,4 areas were separated by fitting the measured spectra with multi-Gaussian function plus polynomial backgrounds using Microcal Orgin 7.5 software program. Details of the experimental set up and data analysis have been

Table 6 lists the theoretical values which were calculated by Scofield (Scofield 1974a; Scofield, 1974b). Addition to this, the measured values of the *Kβ*1/*Kα*, *Kβ*2/*Kα*, *Kβ*2/*Kβ*1 and *Kβ*/*Kα* intensity ratios in Y, and previous experimental and the other theoretical values of

> This Work

*Kβ*2,4/*Kα* 0.0317±0.008 0.02902 *Kβ*2,4/*Kβ*1,3 0.1981±0.011 0.19220

Scofield (1974a)

*Kβ***/***Kα* **0.1822±0.008 0.16960 0.1685 0.1856±0.009** 

Manson &Kennedy (1974)

Ertuğral et al. (2007)

these ratios for pure elements and their compounds are listed in Table 6.

*Kβ*2,4/*Kα* 0.0311±0.008 *Kβ*2,4/*Kβ*1,3 0.1956±0.011 *Kβ***/***Kα* **0.1753±0.011**

*Kβ*2,4/*Kα* 0.0304±0.008 *Kβ*2,4/*Kβ*1,3 0.1941±0.011 *Kβ***/***Kα* **0.1712±0.011** 

*Kβ*2,4/*Kα* 0.0325±0.010 *Kβ*2,4/*Kβ*1,3 0.1987±0.011 *Kβ***/***Kα* **0.1829±0.006** 

*Kβ*2,4/*Kα* 0.0316±0.008 *Kβ*2,4/*Kβ*1,3 0.1977±0.010 *Kβ***/***Kα* **0.1796±0.011**

*Kβ*2,4/*Kα* 0.0310±0.011 *Kβ*2,4/*Kβ*1,3 0.1954±0.011 *Kβ***/***Kα* **0.1742±0.010** 

*Kβ*2,4/*Kα* 0.0329±0.008 *Kβ*2,4/*Kβ*1,3 0.1992±0.009 *Kβ***/***Kα* **0.1836±0.010** 

Intensity Ratio

Y **B=0** *Kβ*1,3/*Kα* 0.2307±0.010 0.22910

B=0.6T *Kβ*1,3/*Kα* 0.2289±0.007

B=1.2T *Kβ*1,3/*Kα* 0.2275±0.007

B=0.6T *Kβ*1,3/*Kα* 0.2320±0.008

B=1.2T *Kβ*1,3/*Kα* 0.2315±0.008

H2O **B=0** *<sup>K</sup>β*1,3/*Kα* 0.2339±0.008

YCl3 **B=0** *Kβ*1,3/*Kα* 0.2341±0.010

 B=0.6T *Kβ*1,3/*Kα* 0.2337±0.006 *Kβ*2,4/*Kα* 0.0305±0.009 *Kβ*2,4/*Kβ*1,3 0.1952±0.009

reported earlier (Porikli et al., 2011b).

External Magnetic Field

Element

Y(NO3)3.6


Determination of Chemical State and External Magnetic Field

(0.5-3.0%) and the other systematic errors (1.0-2.0%).

environment.

compounds.

**4. Conclusion** 

characteristic X-rays.

Effect on the Energy Shifts and X-Ray Intensity Ratios of Yttrium and Its Compounds 109

When you look at the Table 6, a serious difference between the *Kβ*1,3/*Kα* experimental and theoretical values can be seen. This situation is mainly because of the limited resolution of the detector. The *Kα*1 and *Kα*2 X-ray components appear as one line. In the most general case, chemical speciation is preferably performed via the analysis of the *Kβ*1,3 or *Kβ*2,4 lines. These lines, emitted after transition of valance electrons are more sensitive to the chemical

As can be seen from Table 6, the *Kβ*/*Kα* ratios of Y in all Y compounds are in close agreement with the ratios of corresponding pure metals. The greatest increase of the *Kβ*/*Kα* ratio has been observed for Y2S3. We found a general increase of the *Kβ*/*Kα* intensity ratios for different compounds. This situation is more complex because the *Kβ*/*Kα* intensity ratio is affected by the chemical bonding type, (ionic, metallic, covalent), the individual characteristics of the structure of molecules, complexes and crystals (polarity, valency and

We found that the chemical effect on the *Kβ*/*Kα* ratios for 4d elements is small but the dependence of the *Kβ*2/*Kα* ratios on the chemical environments is appreciable. This can be understood by the fact that in 4d elements the valance state consists of the 4d, 5s and 5p electrons and the influence of the chemical state on the *Kβ*1,3 (3p→1s) X-ray emission is negligible. Yamoto et al. (1986) found similar results for compounds involving Tc isotopes and Mukoyoma et al. (2000) found similar results theoretically for Mo and Tc

The overall error in the present measurements is estimated to be 3-8%. This error is attributed to the uncertainties in different parameters used to determine the *Kβ*1/*Kα*, *Kβ*2/*Kα*, *Kβ*2/*Kβ*1 and *Kβ*/*Kα* values; such as, *I*0*Gε* product (1.0-2.5%), in the absorption correction factor (0.3-1.5%), the error in the area evaluation under the *Kα*, *Kβ*1, *Kβ*2 and *Kβ* X-ray peak

There has been increasing interest in chemical speciation of the elements in recent years which can be attributed to the great alterations in the chemical and biological properties of the elements depending on their oxidation state, the type of chemical bonds etc. Usually, the influence of the chemical environment results in energy shifts of the characteristic X-ray lines, formation of satellite lines and changes in the emission linewidths and relative X-ray intensities. High resolution X-ray spectroscopy, employing crystal spectrometers of a few eV resolutions, can be applied to probe these phenomena efficiently, exploiting them for chemical state analysis. Measurements of the shapes and wavelengths of certain X-ray lines have been made by previous investigators with both EDXRF and WDXRF. It has been shown that the WDXRF spectrometer is capable of measuring X-ray wavelengths with a precision equal to or greater than that attained with the EDXRF system. Both EDXRF and WDXRF technique has been used to study the effect of chemical state of an element on

We have presented and discussed the effect of chemical composition and external magnetic field on the *Kβ*1,3/*Kα*, *Kβ*2,4/*Kα*, *Kβ*2,4/*Kβ*1,3 and *Kβ*/*Kα* intensity ratios for some Yttrium compounds. The experimental measurements have been performed with a Si(Li) detector. The observed spectral features, namely the asymmetry indices, FWHM values, chemical shifts, energy separations between *Kα* and *Kβ* lines and *Kβ/Kα* intensity ratio values show an interesting correlation with crystal symmetries. Furthermore, these values change

electronegativity of atoms, co-ordination number, ionicities of covalent bond etc.).


Table 6. *Kβ*1,3/*Kα, Kβ*2,4/*Kα, Kβ*2,4/*Kβ*1,3 and *Kβ*/*Kα* X-ray intensity ratios of pure Y their compounds.

When you look at the Table 6, a serious difference between the *Kβ*1,3/*Kα* experimental and theoretical values can be seen. This situation is mainly because of the limited resolution of the detector. The *Kα*1 and *Kα*2 X-ray components appear as one line. In the most general case, chemical speciation is preferably performed via the analysis of the *Kβ*1,3 or *Kβ*2,4 lines. These lines, emitted after transition of valance electrons are more sensitive to the chemical environment.

As can be seen from Table 6, the *Kβ*/*Kα* ratios of Y in all Y compounds are in close agreement with the ratios of corresponding pure metals. The greatest increase of the *Kβ*/*Kα* ratio has been observed for Y2S3. We found a general increase of the *Kβ*/*Kα* intensity ratios for different compounds. This situation is more complex because the *Kβ*/*Kα* intensity ratio is affected by the chemical bonding type, (ionic, metallic, covalent), the individual characteristics of the structure of molecules, complexes and crystals (polarity, valency and electronegativity of atoms, co-ordination number, ionicities of covalent bond etc.).

We found that the chemical effect on the *Kβ*/*Kα* ratios for 4d elements is small but the dependence of the *Kβ*2/*Kα* ratios on the chemical environments is appreciable. This can be understood by the fact that in 4d elements the valance state consists of the 4d, 5s and 5p electrons and the influence of the chemical state on the *Kβ*1,3 (3p→1s) X-ray emission is negligible. Yamoto et al. (1986) found similar results for compounds involving Tc isotopes and Mukoyoma et al. (2000) found similar results theoretically for Mo and Tc compounds.

The overall error in the present measurements is estimated to be 3-8%. This error is attributed to the uncertainties in different parameters used to determine the *Kβ*1/*Kα*, *Kβ*2/*Kα*, *Kβ*2/*Kβ*1 and *Kβ*/*Kα* values; such as, *I*0*Gε* product (1.0-2.5%), in the absorption correction factor (0.3-1.5%), the error in the area evaluation under the *Kα*, *Kβ*1, *Kβ*2 and *Kβ* X-ray peak (0.5-3.0%) and the other systematic errors (1.0-2.0%).
