Nuclear Instruments and Methods in Physics Research B 268 (2010) 2010–2014
Contents lists available at ScienceDirect
Nuclear Instruments and Methods in Physics Research B journal homepage: www.elsevier.com/locate/nimb
CdTe detector use for PIXE characterization of TbCoFe thin films P.C. Chaves a,b,c,*, A. Taborda a,b, N.P. Barradas a,d, M.A. Reis a,b a
Instituto Tecnológico e Nuclear, EN10, Apartado 21, 2686-953 Sacavém, Portugal Centro de Física Atómica da Universidade de Lisboa, Av. Prof. Gama Pinto, 2,1649-003 Lisboa, Portugal c Instituto Superior Técnico da Universidade Técnica de Lisboa, Av. Rovisco Pais, 1049-001 Lisboa, Portugal d Centro de Física Nuclear da Universidade de Lisboa, Av. Prof. Gama Pinto, 2,1649-003 Lisboa, Portugal b
a r t i c l e
i n f o
Article history: Available online 2 March 2010 Keywords: Microcalorimeter CdTe detectors PIXE analysis TbCoFe
a b s t r a c t Peltier cooled CdTe detectors have good efficiency beyond the range of energies normally covered by Si(Li) detectors, the most common detectors in PIXE applications. An important advantage of CdTe detectors is the possibility of studying K X-rays lines instead the L X-rays lines in various cases since CdTe detectors present an energy efficiency plateau reaching 70 keV or more. The ITN CdTe useful energy range starts at K-Ka (3.312 keV) and goes up to 120 keV, just above the energy of the lowest c-ray of the 19 F(p, p’c)19F reaction. In the new ITN HRHE-PIXE line, a CdTe detector is associated to a POLARIS microcalorimeter X-ray detector built by Vericold Technologies GmbH (an Oxford Instruments Group Company). The ITN POLARIS has a resolution of 15 eV at 1.486 keV (Al-Ka) and 24 eV at 10.550 keV (PbLa1). In the present work, a TbCoFe thin film deposited on a Si substrate was analysed at the HRHE-PIXE system. The good efficiency of the CdTe detector at 45 keV (Tb-Ka), and the excellent resolution of POLARIS microcalorimeter at 6.403 keV (Fe-Ka), are presented and the new possibilities open to the IBA analysis of systems with traditionally overlapping X-rays and near mass elements are discussed. Ó 2010 Elsevier B.V. All rights reserved.
1. Introduction IBA analysis of thin films is normally done using Rutherford Backscattering Scattering (RBS). This is frequently not enough to solve the whole problem, particularly when elements with similar atomic number are present in the film, in which cases, PIXE may be the ideal complementary technique [1–6]. Still cases emerge where elements having similar X-ray energies are present in the samples limiting the applicability of PIXE. In this work the study of a TbCoFe film is discussed as example of such cases and the new ITN High Resolution High Energy-PIXE setup (HRHE-PIXE) shows to be a very good tool to solve this type of problems. In this case, the applicability of the PIXE technique is questioned due to the overlapping of the X-ray energies emitted by the sample. The TbCoFe thin film studied deposited on a Si substrate has structure as represent in Fig. 1. The main problem being related to the overlapping of X-rays energies from the TbCoFe thin film, itself Tb-L X-rays energies (Tb-La = 6.273 keV, Tb-Lb1 = 6.978 keV, Tb-Lb215 = 7.367 keV) are very close to the energies of Fe K X-rays (Ka = 6.403 keV and Kb = 7.060 keV) and Co (Ka = 6.930 keV and * Corresponding author. Address: Instituto Tecnológico e Nuclear, EN10, Apartado 21, 2686-953 Sacavém, Portugal. Tel.: +351 219946065; fax: +351 219941525. E-mail address:
[email protected] (P.C. Chaves). 0168-583X/$ - see front matter Ó 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.nimb.2010.02.120
Kb = 7.652 keV). Using a standard 150 eV resolution Si(Li) detector, these X-ray lines strongly overlap. Properly applying PIXE to such a system thus requires either the quantification of Tb-K X-rays or detecting Tb-L X-ray lines with a detector having much better resolution than the available using Si(Li) detectors. 2. Materials and methods In the present work the sample of Si/Ta/TbCoFe/Ta was irradiated at the 3.0 MV ITN Tandetron accelerator. PIXE spectra were obtained in the new ITN HRHE-PIXE setup [7] using 1.0 and 2.5 MeV proton beams, and two different incidence angles: 30° and 60°. RBS spectra were obtained in the 2.5 MV Van de Graaff accelerator using 2 MeV 4He+ beams and two different incidence angles, 7° and 70°. The RBS data analysis was done with DataFurnace NDF v9.2 l [8]. In the HRHE-PIXE setup two different detectors are used simultaneously. A CdTe detector and a POLARIS microcalorimeter X-ray detector. The CdTe detector is an Amptek Peltier Cooled 3 3 1 mm CdTe detector placed at 35° in respect to the beam direction (see Fig. 2), has a 250 lm Beryllium window and an overall small size: 95 44 29 mm. Its useful energy range goes from K–Ka1 (3.312 keV) up to above 110 keV, the energy of the lowest c-ray from the 19F(p, p0 c)19F reaction. The positioning of CdTe
P.C. Chaves et al. / Nuclear Instruments and Methods in Physics Research B 268 (2010) 2010–2014
2011
Fig. 1. Description of sample that contains a TbFeCo thin film on a Si substrate. Sample also has two very thin layers of Ta. The thickness of the layers is also presented in the figure.
detector is made using a Perspex sleeve, which fits into an Aluminum flange. This flange is open ended and the vacuum seal is made by a NylonÒ window shared with the POLARIS sleeve (see Fig. 2). X-ray collimators and absorber foils are introduced outside the vacuum, on the top of detector. The POLARIS is a Oxford Instruments/Vericold Technologies GmbH liquid cryogens free, pulse tube cooled, transition edge superconducting (TES) microcalorimeter [9–11] having 2503 lm3 Au cube absorber [12–14] collimated to 0.01 mm2. Like the CdTe detector, the POLARIS is fixed to the analysis chamber although the connection is in this case mediated by the chamber support. POLARIS detector snout fits, at 90° with beam direction, in an open ended Aluminum sleeve, which is closed by a Nylon cup shared with CdTe detector flange. In front of the POLARIS, The Nylon cup has a hole closed by a 6 lm Mylar window. In Fig. 2, the schematics of the HRHE-PIXE chamber (top view) are presented. In the case of the POLARIS detector, filters are placed inside the chamber, in vacuum. This configuration was optimized to have both detectors working at same time and the closest possible to the samples. Spectra were collected simultaneously with the CdTe and the POLARIS X-ray detectors. The experimental conditions used in RBS and PIXE analyses can be found in Table 2. PIXE spectra were deconvoluted using the ORIGIN software, which can be used due to the slight peak overlap between two X-ray lines. 3. Results and discussion Concerning the complete characterization of the TbCoFe thin film, the combination of RBS and PIXE techniques is fundamental. In this case, the RBS technique offers consistent results about Tb. However, Co only can be estimated using PIXE and high resolution is needed due to the presence of Tb X-rays. In Fig. 3, we present the spectra collected with the CdTe and the POLARIS X-ray detectors emitted during irradiations with H+ beams of 2.5 MeV. Two different spectra are shown. In the first one we present the spectrum collected with the CdTe, where we can observe the K X-ray lines of Tb and Ta. Ta-L X-rays are also measurable with the CdTe detector with good statistics. No effort was made here to obtain good statistics for the K lines since Tb and Ta were quantified by RBS. However, it is important to note, that although in presence of Ta thin layers, the K lines, can be considered for quantification, if necessary. The POLARIS spectrum is shown in the second graph. Spectrum is shown in full scale, which permits us to make a first observation of the X-rays lines and to have a very good idea of the resolution of the system. In this spectrum it is clear that the traditional overlapping observed with Si(Li) detectors, the common detectors used in PIXE, are resolved, since Tb-La1,2, Fe-Ka and Co-Ka, Tb-Lb1, Fe-Kb and finally the Tb-Lb2,15 X-ray are completely separated. This capacity turns it possible to determine a value for the Fe/Co ratio, which is impossible to obtained otherwise. We found a Fe/Co ratio equal to 33.6 (with an error of 13%) and a Fe/Tb ratio equal to 6.1 (with an error of 5%).
Fig. 2. HRHE-PIXE chamber. POLARIS and CdTe detector are placed at 90° and 35°, respectively to the beam entrance. Both detectors are located on the atmosphere and separated from vacuum by a single Nylon seal piece.
Finally in the last spectrum of this figure, we present a vertical zoom of the POLARIS spectrum, with the aim of showing that almost no overlapping exists between the X-rays lines shown. It can be seen, that for X-rays energies below the X-ray energy of Tb-La1,2 we also have the Tb-Ll, Tb-Lt and Tb-Ls and also the radiative auger Fe-KLL X-rays lines. The X-ray energies are in Table 1. On the other hand, for higher energies than the Tb-Lb215 X-ray we can also found the Co-Kb X-ray and the L X-rays of Ta. The Ta-La1/Ta-La2, the Ta-Lb1/Ta-Lb3/Ta-Lb215 and the Ta-Lc1/ Ta-Lc2 that also traditional overlaps are very well resolved using the POLARIS X-ray detector. Although the Ta-L X-rays lines presenting show low statistics, they are sufficient to observe the very good resolution of the system, and they can be used for quantification if necessary, even though we are in face of two very shallow layers of Ta having just 30 Å each and medium high X-ray energies. Fe/Co and Fe/Tb ratios were obtained after correcting the ratio of the areas of Fe-Ka, Co-Ka and Tb-La1,2 to the value of the production cross section from Paul [16] for K X-rays and of Reis and Jesus [17] for the Tb-La12 X-ray. These calculations were made using DATTPIXE [18]. However in the case of Fe and Co the value of production cross section is total for the K shell so we also need to correct them to the branching ratios. After these first corrections, we also need to correct the values to the absorptions in all the windows until the X-rays reach the detector, for which we have calculated the transmission in the Mylar, chamber window plus filter (6 lm + 150 lm), in the windows of detector, 21 lm of Be and 300 nm de Al, and finally considered the air between the chamber window and the entrance in the detector window. In Fig. 4, we present the RBS spectra and the corresponding fits to the experimental data under the experimental conditions described before. The signal from the layers of Ta are clearly observed, as well as that for the layer of Tb. However, concerning Fe and Co, RBS cannot separate the contributions of these two elements, reason why the fit takes only into consideration the Fe. The results obtained from both the RBS and PIXE experiments are presented in Table 3. From the RBS results we determine a 14.4 ± 0.5 at.% of Tb in the TbCoFe film, which gives a result of 85.6 at.% for Fe. Concerning the PIXE results, two groups of results
2012
P.C. Chaves et al. / Nuclear Instruments and Methods in Physics Research B 268 (2010) 2010–2014
Fig. 3. Spectrum of TbCoFe thin film obtained simultaneously with CdTe detector and POLARIS microcalorimeter X-ray detector at 2.5 MeV H+ beam with a incidence angle of 30°. The high resolution allows resolve the Tb-La/Fe-Ka and Co-Ka/Tb-Lb1/Fe-Kb traditional overlaps.
are shown. The first one was obtained using the Tb-L3 fluorescence yield from Krause and Oliver [19] and the second group of PIXE results was obtained using the Tb-L3 fluorescence yield from McGeorge et al. [20]. These two authors published two different values for the fluorescence yield, namely 0.164 and 0.188. The use of the two coefficients was necessary since upon the calculation of the Tb concentration in the layer using the fluorescence yield of Krause, we found that the Tb at.% is not in agreement with the Tb RBS results. However, using the value of McGeorge, we obtained exactly the same concentration as in the RBS results. We thus conclude that Krause and Oliver [19] fluorescence yield for Tb is not correct, and that the McGeorge et al. [20] fluorescence coefficient should be used instead. The recommended value of Campbell [21] of 0.175 will provide a better result than the Krause one, but also not in agreement to the RBS result. Concerning the Co and Fe concentration ratio in the sample, we found a very low concentration of Co in the layer, 2.2% at not possible to determine neither by RBS nor by standard Si(Li) based PIXE.
Table 1 K X-ray energies of Si, Fe, Co, Tb and Ta. L X-rays energies of Ta and Tb are also presented [15]. Element (energy – keV)
K X-rays
L X-rays
K a1 Kb KLL Ll Lt Ls L a2 L a1 Lb1 Lb3 Lb2,15 Lc 1 Lc 2
Si
Fe
Co
Tb
Ta
1.739 1.835
6.403 7.060 5.698
6.930 7.651 6.154
44.482 50.383
57.533 65.223
5.547 5.750 5.906 6.238 6.273 6.978 7.096 7.367 8.102 8.398
7.173 7.412 7.686 8.088 8.146 9.343 9.487 9.652 10.895 11.217
2013
P.C. Chaves et al. / Nuclear Instruments and Methods in Physics Research B 268 (2010) 2010–2014
Table 2 Description of the experimental conditions used in PIXE experiments. The filter conic plus plastic consist a 2 mm Perspex conic filter (less than 1 mm in center) plus a 1 mm plastic filter is placed on top of CdTe detector in way to reduce the low-energy X-ray intensity (). PIXE Ep = 1.0 MeV Incidence angle (°) Beam collimator (mm) Filter (POLARIS) Filter (CdTe) Beam current (nA) Count rate (POLARIS) counts/s Acquisition time (s) Solid angle – CdTe (msr) Solid angle – POLARIS (msr) RBS – 2 MeV 4He+ Incidence angle (°) Scattering angle (°)
Ep = 2.5 MeV
15 7 n.a. S/Abs 100 35 2570 5.1 0.13
30 7 n.a. Conic + plastic () 70 70 240
7 160
70 160
60 3 n.a. Conic + plastic () 120 150 1400
30 7 Mylar 150 Conic + plastic () 250 15 4700
60 3 Mylar 150 Conic + plastic () 350 15 3400
Fig. 4. RBS spectra of Si/Ta/TbFeCo/Ta sample collected at 7° and 70°. The best fit for both spectra are also presented in figure.
Table 3 RBS and PIXE results for Fe, Co and Tb. Results are in at.%. The corresponding thickness is also presented. Errors are showed in %. Fluorescence yield from McGeorge must be used to obtained agreement with RBS data. Element
RBS
PIXE TbCoFe (Krause et al.)
Fe Co Tb
TbCoFe (McGeorge et al.)
at.% ± Error (%)
at.% ± Error (%)
Thickness (Å)
at.% ± Error (%)
Thickness (Å)
85.6 ± 0.5 (Fe)
81.7 ± 1 2.2 ± 4 16.1 ± 1
637 17 126
83.4 ± 1 2.2 ± 4 14.4 ± 1
651 17 112
14.4 ± 0.5
4. Conclusions In this work, very important conclusions were obtained. The first is related to the HRHE-PIXE setup, which have completely solve the traditional problem of X-ray energies overlapping. The other important conclusion is related to the Tb fluorescence yield. Tb fluorescence yield from Krause [19] is 15% below the value of McGeorge et al. [20] which must be taken into consideration for reaching an agreement with the RBS results.
Acknowledgements This work was partially supported by the Portuguese Foundation for Science and Technology, FCT fellowships SFRH/BD/27557/
2006 and SFRH/BD/43379/2008. The HRHE-PIXE system was installed at ITN under co-funding of FCT project REEQ/377/FIS/2005. Production and characterization of samples here presented was carried out under the scope of FCT project PTDC/CTM/66321/2006. References [1] P.C. Chaves, O.R. Oliveira, V. Corregidor, N. Barradas, O. Vigil Galán, A. Arias Carbajal, M.A. Reis, CdTe detector use in PIXE characterization of Sn dopped CdO thin films, in: M. Budnar, M. Kavcic (Eds.), Proceedings of the 10th International Conference on Particle Induced X-ray Emission and Its Analytical Application, 4 a 8 Junho, 2004, Portoroz – Slovenia (CD-Edition). [2] M.A. Reis, P.C. Chaves, V. Corregidor, N. Barradas, E. Alves F. Dimroth, A. Bett, Grazing detection geometry for PIXE characterization of thin films, in: M. Budnar, M. Kavcic (Eds.), Proceedings of the 10th International Conference on Particle Induced X-ray Emission and Its Analytical Application, 4 a 8 Junho, 2004, Portoroz – Slovenia (CD-Edition).
2014
P.C. Chaves et al. / Nuclear Instruments and Methods in Physics Research B 268 (2010) 2010–2014
[3] M.A. Reis, N.P. Barradas, C. Pascual-Izarra, P.C. Chaves, A.R. Ramos, E. Alves, G. González-Aguilar, M.E.V. Costa, Holistic RBS–PIXE data reanalysis of SBT thin film samples, in: Proceedings of the 19th International Conference on the Application of Accelerators in Research and Industry, August 20–25, Fort Worth, Texas, USA, 2006, Nucl. Inst. Meth. Phys. Rev. B 261 (2007) 439–442. [4] M.A. Reis, P.C. Chaves, V. Corregidor, N. Barradas, E. Alves, F. Dimroth, A. Bett, Detection angle resolved PIXE and the equivalent depth concept for thin films characterization, X-Ray Spectrom. 34 (2005) 372–375. [5] V. Corregidor, N.P. Barradas, E. Alves, N. Franco, L.C. Alves, P.C. Chaves, M.A. Reis, Ion beam analysis of GaInAsSb films grown by MOVPE on GaSb, in: Proceedings of the 18th International Conference on the Application of Accelerators in Research and Industry, Nucl. Inst. Meth. Phys. Rev B (2005) 326–330. [6] V. Corregidor, P.C. Chaves, M.A. Reis, C. Pascual-Izarra, E. Alves, N.P. Barradas, Combination of IBA techniques for composition analysis of GaInAsSb films, Mater. Sci. Forum 514–516 (2006) 1603–1607. [7] P.C. Chaves, M.A. Reis, E. Alves, New High Resolution and High Energy Lisbon pixe set-up, in: J. Miranda, J.L. Rucalva-Sil, O.G. de Lucio (Eds.), Proceedings of the XI International Conference on PIXE and Its Analytical Applications, 25–29 May, Puebla, Mexico, Digitally Published, 2007. [8] N.P. Barradas, C. Jeynes, R.P. Webb, Simulated annealing analysis of Rutherford backscattering data, Appl. Phys. Lett. 71 (1997) 291. [9] C. Hollerith, B. Simmnacher, R. Weiland, F.V. Feilitzsch, C. Isaila, J. Jochum, W. Potzel, J. Hohne, K. Phelan, D. Wernicke, T. May, Energy calibration of superconducting transition edge sensors for X-ray detection using pulse analysis, Rev. Sci. Instrum. 77 (2006) 053105. [10] S.H. Moseley, J.C. Mather, D. McCammon, Thermal detectors as X-ray spectrometers, J. Appl. Phys. 56 (5) (1984).
[11] D. McCammon, W. Cui, M. Juda, P. Plucinsky, J. Zhang, R.L. Kelley, S.S. Holt, G.M. Madejski, S.H. Moseley, A.E. Szymkowiak, Nucl. Phys. A527 (1991) 821c–824c. [12] D.A. Wollman, K.D. Irwin, G.C. Hilton, L.L. Dulcie, D.E. Newbury, J.M. Martinis, High-resolution energy-dispersive microcalorimeter spectrometer for X-ray microanalysis, J. Microsc. 188 (Pt. 3) (1997) 196–223. [13] D. Redfern, J. Nicolosi, J. Hohne, R. Weiland, B. Simmnacher, C. Hollerich, The microcalorimeter for industrial applications, J. Res. Natl. Inst. Stand. Technol. 107 (2002) 621–626. [14] D.A. Wollman, S.W. Nam, D.E. Newbury, G.C. Hilton, K.D. Irwin, N.F. Bergren, S. Deiker, D.A. Rudman, J.M. Martinis, Superconducting transition-edgemicrocalorimeter X-ray spectrometer with 2 eV energy resolution at 1.5 keV, Nucl. Inst. Meth. Phys. Res. A 444 (2000) 145–150. [15] Günter Zschornack, Handbook of X-Ray Data, Springer, 2007. [16] H. Paul, An analytical cross-section formula for K X-ray production by protons, Nucl. Inst. Meth. Phys. Res. B 3 (1984) 5–10. [17] M.A. Reis, A.P. Jesus, Semiempirical approximation to cross sections for L X-ray production by proton impact, Atom. Data Nucl. Data Tables 63 (1996) 1–55. [18] M.A. Reis, L.C. Alves, DATTPIXE, a computer package for TTPIXE data analysis, Nucl. Inst. Meth. Phys. Res. B 68 (1–4) (1992) 300–304. [19] M.O. Krause, J.H. Oliver, Natural widths of atomic K and L levels, K X-ray lines and several KLL Auger lines, J. Phys. Chem. Ref. Data 8 (1979) 329. [20] J.C. McGeorge, H.U. Freund, R.W. Fink, Decay of 159Dy: L-subshell X-ray fluorescence and Coster–Kronig yields at Z = 65; branching ratio, and Kconversion of the 58 keV transition in 159Tb, Nucl. Phys. A 154 (3) (1970) 526– 538. [21] J.L. Campbell, Fluorescence yields and Coster–Kronig probabilities for the atomic L subshells, Atom. Data Nucl. Data Tables 85 (2003) 291–315.