Open Access
Add to BibSonomyAbstract
The surface chemical state and local electronic structure of AlxGa1-xN (x = 0~0.45) epi-layers have been systematically investigated by X-ray photoelectron spectroscopy (XPS) and extended X-ray absorption fine structure (EXAFS) spectroscopy. The results show that the surface of AlxGa1-xN is a composite of oxide and nitride of gallium and aluminum. In addition, it was identified that the Ga-O components were converted to Al-O components when the AlxGa1-xN sample was exposed to air. The EXAFS analysis also reveals that the Ga-N and Ga-Al bond lengths are independent of the Al composition, whereas the Ga-Ga bond length is a function of Al composition.
© 2014 Optical Society of America
1. Introduction
In recent years, AlN, GaN, and their alloys have gained a lot of interests for the development of optoelectronic devices with ultraviolet (UV) light emitting wavelength [1–3]. Moreover, these nitrides are of extremely high hardness, very large heterojunction band offsets, high thermal conductivity, and high melting-point temperature, which make them promising candidates for the fabrication of high temperature and high power devices. The optical properties of III- nitride-based light emitting diodes (LEDs) and their relationship with growth parameters, such as substrate temperature, precursor flow rates, and dopants employed have been studied by many researchers using photoluminescence (PL) and electroluminescence (EL) spectra [4–6]. However, some important issues like the surface chemical and electronic properties and their relationship with growth conditions as well as associated light emitting features, have not been paid sufficient attention inspect of the fact that most semiconductor devices are extremely sensitive to their surface properties [7, 8]. In particular, understanding and control of AlGaN surface are essential for the development of stable Schottky and ohmic contacts fabrication, etching, and passivation processes [9, 10].
X-ray photoelectron spectroscopy (XPS) is a powerful analytical method in elemental identification because of its high sensitivity to chemical states. However, XPS is essentially a core-level spectroscopy and its chemical-state sensitivity is based on the indirect effect on shielding and deshielding of core-levels due to the X-ray-induced changes in the density and the distribution of the electrons involved in atomic bonding. The binding energy (BE) of a core-level is thus significantly affected by the electron distribution. Although the the changes in formal oxidation state are easily detected by XPS [11], the changes in geometrical parameters such as the variation in coordination geometries are difficult to resolve in XPS measurement. Therefore, it is necessary to find a characterization method which can provide more sensitivity than XPS to the surface chemical and local electronic properties. Synchrotron radiation X-ray absorption fine structure (SR-XAFS) spectroscopy has been proven to be useful to obtain the information on surface chemical and local electronic properties [12, 13]. The SR-XAFS spectra are measured by scanning the photon energy across the absorption edge for several hundred eV. The extended X-ray absorption fine structure (EXAFS) region, typically starting at the position of >50 eV above the absorption edge, is generally used to obtain various kinds of information about the microstructure, such as the nearest neighboring atomic distances and coordination numbers (CN) around the X-ray absorbing atom. Unfortunately, up to date a detailed analysis in atomic scale on the III-nitrides especially for the local electronic properties is still less reported.
In this work, the surface chemical and local electronic properties of AlxGa1-xN epi-layers were systematically investigated by XPS and SR-XAFS spectra. In particular, the EXAFS has been employed to study the bond length and coordination number around the Ga atom, which turned out to be a unique and superior characterization method to all the other currently available tools. Since the information obtained by EXAFS is complementary to that obtained by XPS, the combination of XPS and EXAFS is very useful to collect the information necessary for the subsequent device processing.
2. Experiments
The AlxGa1-xN (x = 0, 0.15, 0.35, 0.45) epi-layers were grown on c-plane sapphire substrates in a low pressure (40 Torr) metal-organic chemical vapor deposition (MOCVD) system. Trimethyl-aluminum (TMA), trimethyl-gallium (TMG), and ammonia (NH3) were used as the precursors for Al, Ga, and N, respectively. Prior to the growth, the sapphire substrates were heated to 1100 °C in H2 ambience to remove surface contamination. A 20 nm-thick low-temperature (LT)-grown AlN nucleation layer with a V/III ratio of 3000 was firstly deposited on the sapphire substrate at 600 °C. The temperature was then raised to 1040 °C to grow a high-temperature (HT) AlN interlayer. Finally a 600 nm-thick AlxGa1-xN epi-layer was grown on HT-AlN interlayer at a growth temperature of 1140 °C.
The chemical bonding states on the AlxGa1-xN surface were characterized by XPS using a VG ESCALAB 250 system with a monochromated Al Kα line at 1486.6 eV. The XPS spectra were collected at a photoelectron take-off angle of 75° with respect to the sample surface. The adventitious hydrocarbon C 1s binding energy at 284.5 eV was used as reference to calibrate the energy shift of the Ga 3d, Al 2p, and N 1s shallow core-levels. The EXAFS measurement was performed at National Synchrotron Radiation Research Center (NSRRC) in Hsinchu, Taiwan, using the fluorescence mode. The beam spot size was 6 × 2 mm2. A double crystal monochromator equipped with Si (111) crystal was used in the EXAFS measurement, and the wave-vector of incident X-ray was always arranged to be parallel to the c-axis of AlxGa1-xN epi-layer samples.
3. Results and discussion
In order to investigate the surface chemical composition and thermal stability of the AlxGa1-xN epi-layers, Ga 3d-, Al 2p-, and N 1s-related XPS features of AlxGa1-xN epi-layers without Ar+ sputtering and with Ar+ sputtering for 50 s have been measured.
The Ga 3d XPS features of AlxGa1-xN epi-layers without Ar+ sputtering and with Ar+ sputtering for 50 s are shown in Fig. 1(a) and 1(b), respectively. As indicated in Fig. 1(a), the major Ga 3d XPS peak is consisted of two components that locate at ~21.0 ± 0.1 eV and ~19.3 ± 0.1 eV corresponding to the BE of Ga-O and Ga-N, respectively [14–16]. This feature is due to the fact that the surface of the air-exposed AlxGa1-xN epi-layer is usually covered by a thin natural oxide layer. No any peak at BE lower than 18 eV was detected, which means that there are no Ga-clusters formed in AlxGa1-xN epi-layers since the BE of Ga-Ga bond is ~17.5 ± 0.1 eV. It is evident that the intensity of Ga 3d peak decreases monotonously with increasing the Al composition in AlxGa1-xN epi-layers. It was also found that the intensity of the Ga-O-related peak in the Ga 3d spectra decreased significantly with increasing the Al composition, implying that more and more Ga-O bonds were converted to Al-O bonds when the surfaces of AlxGa1-xN epi-layers with higher Al composition were exposed to air. The higher reactivity of Al and the higher formation heat of Al-O bond than those of Ga-O bond were considered to be responsible for this phenomenon [17]. This conclusion was supported by the Al 2p spectra that will be described below. Approximately 1~2 nm-thick AlxGa1-xN top layer including the surface contaminants and the thin natural oxide layer could be removed by 1 KeV Ar+ bombardment. In fact, as shown in Fig. 1(b), it was noted that the major Ga 3d XPS peak position for the AlxGa1-xN epi-layer is nearly independent of Al composition. Or in other words, the Ga-O bond was completely broken and transformed to Ga-N bond after Ar+ sputtering for 50 s.
Fig. 1 Ga 3d core-level XPS spectra of AlxGa1-xN epi-layers as a function of Al composition without Ar+ sputtering (a) with Ar+ sputtering for 50 s (b).
Gaussian-Lorentzian line shapes were used to simulate the Ga 3d spectra after standard Shirley background subtraction. Figure 2 exhibits the XPS fitting results for Ga 3d core-level spectra of (a) GaN and (b) Al0.45Ga0.55N without Ar+ sputtering. When the Ga 3d spectra for the AlxGa1-xN epi-layer samples were simulated as the superposition of the Ga-N- and Ga-O-related components, both line shape and peak position are in excellent agreement with the experimental data. As shown clearly in Fig. 2, the intensity of the Ga-O-related peak decreases whereas the intensity of the Ga-N-related peak increases accordingly with the increase of Al composition in AlxGa1-xN epi-layers. These results suggest that the existence of Al can give rise to a significant influence on the surface oxidation of AlGaN alloy. The fitted peak positions and related results for Ga 3d spectra of GaN and Al0.45Ga0.55N samples obtained from Fig. 2 are summarized in Table 1. It can be seen that there are remarkable difference in either peak position or peak width (full width at half maxium, FWHM) for both Ga-N- and Ga-O-related peaks of GaN and Al0.45Ga0.55N epi-layer samples. The ratio of Ga-N/Ga-O peak intensity is much higher in Al0.45Ga0.55N (~3.52) than that in GaN (~0.33). The ratio increases with increasing the Al composition, indicating a strong interaction between O atoms and Al atoms. The intensity of the Ga-O-related peak decreases accordingly with the increase of Al composition in AlxGa1-xN epi-layers. The Ga exists in the form of Ga-N mainly on the surface of the epi-layer. Compared with Ga, Al is in a dominant position in the competitive relationship of oxidation. Based on the analysis above, we can conclude that the oxidation of GaN can be greatly suppressed by the incorporation of Al and the concentration of oxygen is higher near the surface than that inside the AlxGa1-xN epi-layers.
Fig. 2 XPS fitting results for Ga 3d core-level spectra of (a) GaN and (b) Al0.45Ga0.55N without Ar+ sputtering.
Table 1. The fitted peak positions and related results for Ga 3d spectra of GaN and Al0.45Ga0.55N epi-layers.
Figure 3 shows Al 2p core-level XPS spectra of AlxGa1-xN as a function of Al content without Ar+ sputtering (a) and with Ar+ sputtering for 50 s (b). In Fig. 3(a), there is no significant Al 2p XPS signal for AlxGa1-xN with low Al content (x = 0, 0.15). On the other hand, for AlxGa1-xN samples with high Al content (x = 0.35, 0.45), only one peak corresponding to Al-O was observed at the location of BE = 75.0 ± 0.2 eV in Fig. 3(a). This means that Al exists on the surface of AlxGa1-xN samples with high Al content in the form of oxide rather than nitride. After Ar+ sputtering for 50 s, however, the Al 2p peak position for AlxGa1-xN (x = 0.15, 0.35, 0.45) shifts towards to a location with low BE of ~73.6 ± 0.2 eV, which corresponds to the Al-N bonding [18]. Moreover, a shoulder located at the left side of the major peak could be clearly observed and was identified to be due to metallic Al. This fact implies that Al-clusters (Al-Al, at ~72.8 ± 0.2 eV) were formed in the AlxGa1-xN epi-layers as well after Ar+ sputtering. This feature provides a direct evidence that O atom is easily incorporated with Al by occupying a nitrogen site. Moreover, the incorporation of oxygen into AlxGa1-xN is much more pronounced for the sample with higher Al composition than the sample with lower Al composition.
Fig. 3 Al 2p core-level XPS spectra of AlxGa1-xN epi-layers as a function of Al composition without (a) and with Ar+ sputtering for 50 s (b).
Figure 4 exhibits N 1s core-level XPS spectra for AlxGa1-xN as a function of Al composition without Ar+ sputtering (a) and with Ar+ sputtering for 50 s (b). The dotted line demonstrates the fitting result obtained by taking into account N-Ga bond-, N-Al bond-, and Ga Auger transition-related components. In the case of AlxGa1-xN (x = 0.15) shown in Fig. 4(a), the peak appearing at 397.8 ± 0.2 eV is originated from N-Ga bond, whereas the peak associated with N-Al bond was detected at 396.3 ± 0.2 eV which agrees well with the Al 2p XPS spectra [19]. The shoulder centered at 393.8 ± 0.2 eV was identified to be due to the Ga Auger transition [20]. Ga and Al oxides were generated on the surface during the oxidation process of AlxGa1-xN epi-layers resulting the quenching of the Ga Auger-related component. In fact, it was observed that with the increasing of Al composition in AlxGa1-xN epi-layers, the Ga Auger-related shoulder disappeared, and only N-Ga bond- and N-Al bond-related peaks were detected at 397.5 ± 0.2 eV and 396.1 ± 0.2 eV, respectively for AlxGa1-xN epi-layers with relatively high Al composition (x = 0.35, 0.45). Furthermore, the ratio of N-Ga/N-Al signal intensity increased quickly from ~3.61 for Al0.35Ga0.65N to ~7.28 for Al0.45Ga0.55N epi-layer. This feature is slightly different from the Al 2p spectra where no N-Al bond-related signal could be detected on the surface of AlxGa1-xN epi-layer without Ar+ sputtering. We conclude hence that the bonding strength of O-Al bond is much larger than that of N-Al bond on the surface. As a result, the N-Al bond-related signal was gradually quenched with the increase of Al composition in the AlxGa1-xN epi-layers. On the other hand, as shown in Fig. 4(b), after Ar+ sputtering for 50 s, N-Ga bond-, N-Al bond-, and Ga Auger transition-related peaks are all evident in the N 1s spectra regardless of Al composition in the AlxGa1-xN epi-layers. Since Ga Auger transition-related peaks are associated with O-Ga bond or the Ga oxide, the feature shown in Fig. 4(b) indicates clearly that the surface of AlxGa1-xN is a composite of oxides and nitrides of gallium and aluminum.
Fig. 4 N 1s core-level XPS spectra for AlxGa1-xN epi-layers as a function of Al composition without Ar+ sputtering (a) and with Ar+ sputtering for 50 s (b).
EXAFS spectra were measured to study the local electronic structure of AlxGa1-xN materials at the Ga K-edge. The commercial ATHENA program was used to remove the background [21], and to extract the K-space oscillation signal from the EXAFS spectra shown in Fig. 5(a). The R-space signal shown in Fig. 5(b) by black solid line, is Fourier transformed (FT) from the K-space signal shown in Fig. 5(a) in the K range from 2.638 to 11.524 Å−1. As shown clearly in Fig. 5(b), the variation in either peak shape or position of the first shell Ga-N-related peaks is not significant for all the AlxGa1-xN epi-layer samples. In contrast, remarkable variation in peak amplitude was observed for the second shell Ga-Ga (Al)-related peak with the increase in Al composition. This phenomenon is attributed primarily to the cation-cation second shell backscattering. The decrease in the cation-cation backscattering intensity with the increase in Al composition indicates that Al atom causes a smaller backscattering effect as compared to Ga atom. This is reasonable since Al atom has a smaller radius than that of Ga atom. The reduction in the EXAFS and FT amplitudes can also be attributed to the increase in the Debye-Waller factor (the relative mean square deviation in bond length, σ2) for the Ga-Al bond with increasing Al composition. To extract the interatomic distance between a Ga atom and its neighbors, a structural model based on the wurtzite GaN structure was built to fit the EXAFS data in the R range from 1.007 to 3.335 Å by using the IFEFFIT software package [22]. In the case of X-ray normal incidence, the N atom along the c-axis does not have any contribution to the EXAFS spectra, while the other three N atoms have an equal contribution. Therefore, the Ga-N first shell was fitted based on the model wherein only single contribution was considered.
Fig. 5 (a) The Ga K-edge oscillation signal k2χ(k) of AlxGa1-xN epi-layer samples; (b) R-space signal of the Fourier transformed k2χ(k) (indicated with black solid line) as a function of spacing R together with the best fitted results (indicated by blue circle).
The best fitting results plotted by the blue circles in Fig. 5(b) are in good agreement with the experiment data. The interatomic distances (Ga-N, Ga-Ga, and Ga-Al) extracted from Fig. 5(b) are summarized in Table 2. For the Al composition ranged from 0 to 0.45, it was found that the first shell Ga-N bond length is nearly composition-independent. However, the second shell Ga-Ga bond length was significantly larger than that of Ga-Al bond. In other words, the second shell interatomic distances are strongly dependent on atomic bond type and Al composition. The composition-dependent Ga-Ga bond length obtained from the experimental data is very close to that predicted by Vegard’s Law [23], while the Ga-Al bond length is essentially composition independent. It was also noted that the bond length of Ga-Ga has a much greater composition dependence than that of Ga-N and Ga-Al bonds. On the other hand, the second shell CN obtained for both Ga-Ga and Ga-Al bonds agreed well with the calculated results by assuming a random site occupancy of Al in the cation sublattice [24]. This fact implies that there is no specific alloy ordering in the AlxGa1-xN epi-layer samples and therefore, the local structural parameters were representatives of random AlxGa1-xN epi-layers.
Table 2. Extracted structural parameters from the EXAFS data for all the AlxGa1-xN epi-layer samples. R-factor is the parameter used to evaluate the fitting quality.
4. Conclusions
The surface chemical and local electronic properties of AlxGa1-xN epi-layers grown by MOCVD were systematically investigated by XPS and EXAFS. By analyzing the Ga 3d, Al 2p and N 1s core-level XPS spectra of AlxGa1-xN epi-layers without Ar+ sputtering and with Ar+ sputtering for 50 s, it was verified that the surface of AlxGa1-xN is a composite of oxides and nitrides of gallium and aluminum. Furthermore, it was identified that the Ga-O components were converted to Al-O components when the AlxGa1-xN sample was exposed to air. The EXAFS measurement results for the AlxGa1-xN epi-layers demonstrate that the first shell Ga-N bond length is nearly Al composition-independent. However, the second shell Ga-Ga bond length is strongly dependent on Al composition and is significantly longer than that of Ga-Al bond. On the other hand, the Ga-Al bond length was essentially composition independent. Moreover, the second shell interatomic distances were found to be bond type- and composition-dependent. These results achieved should be helpful to optimize the growth and fabrication processes to make excellent AlGaN-based Schottky and ohmic contacts for potential applications in UV-LEDs, UV detectors, and other AlGaN-based optoelectronic devices.
Acknowledgments
The work at National Taiwan University was supported by NSC 98-2221-E-002-015-MY3 and 98-3114-E-005-002-CC2, and by NTU Excellent Research Project (10R80908).
References and links
1. F. Qian, M. Brewster, S. K. Lim, Y. Ling, C. Greene, O. Laboutin, J. W. Johnson, S. Gradečak, Y. Cao, and Y. Li, “Controlled synthesis of AlN/GaN multiple quantum well nanowire structures and their optical properties,” Nano Lett. 12(6), 3344–3350 (2012). [CrossRef] [PubMed]
2. S. Wang, X. Zhang, H. Guo, H. Yang, M. Zhu, L. Cheng, X. Zeng, and Y. Cui, “Enhanced performance of GaN-based light-emitting diodes by using a p-InAlGaN/GaN superlattice as electron blocking layer,” J. Mod. Opt. 60(21), 2012–2017 (2013). [CrossRef]
3. M. Shatalov, W. Sun, A. Lunev, X. Hu, A. Dobrinsky, Y. Bilenko, J. Yang, M. Shur, R. Gaska, C. Moe, G. Garrett, and M. Wraback, “AlGaN deep-ultraviolet light-emitting diodes with external quantum efficiency above 10%,” Appl. Phys. Express 5(8), 082101 (2012). [CrossRef]
4. J. Zhang and N. Tansu, “Improvement in spontaneous emission rates for InGaN quantum wells on ternary InGaN substrate for light-emitting diodes,” J. Appl. Phys. 110(11), 113110 (2011). [CrossRef]
5. Y. D. Ko, K. C. Kim, and Y. S. Kim, “Effects of substrate temperature on the Ga-doped ZnO films as an anode material of organic light emitting diodes,” Superlattices Microstruct. 51(6), 933–941 (2012). [CrossRef]
6. S. O. Jeon, S. E. Jang, H. S. Son, and J. Y. Lee, “External quantum efficiency above 20% in deep blue phosphorescent organic light-emitting diodes,” Adv. Mater. 23(12), 1436–1441 (2011). [CrossRef] [PubMed]
7. G. Ashkenasy, D. Cahen, R. Cohen, A. Shanzer, and A. Vilan, “Molecular engineering of semiconductor surfaces and devices,” Acc. Chem. Res. 35(2), 121–128 (2002). [CrossRef] [PubMed]
8. Y. Lei, S. Yang, M. Wu, and G. Wilde, “Surface patterning using templates: concept, properties and device applications,” Chem. Soc. Rev. 40(3), 1247–1258 (2011). [CrossRef] [PubMed]
9. Y. W. Lian, Y. S. Lin, J. M. Yang, C. H. Cheng, and S. S. H. Hsu, “AlGaN/GaN Schottky barrier diodes on silicon substrates with selective Si diffusion for low onset voltage and high reverse blocking,” IEEE Electron Device Lett. 34(8), 981–983 (2013). [CrossRef]
10. P. S. Park, K. M. Reddy, D. N. Nath, Z. Yang, N. P. Padture, and S. Rajan, “Ohmic contact formation between metal and AlGaN/GaN heterostructure via graphene insertion,” Appl. Phys. Lett. 102(15), 153501 (2013). [CrossRef]
11. D. Kochubey, V. Kaichev, A. Saraev, S. Tomyn, A. Belov, and Y. Voloshin, “Combined X-ray absorption near-edge structure and X-ray photoelectron study of the electrocatalytically active cobalt(I) cage complexes and the clathrochelate cobalt(II)- and cobalt(III)-containing precursors and analogs,” J. Phys. Chem. C 117(6), 2753–2759 (2013). [CrossRef]
12. L. S. Cavalcante, N. C. Batista, T. Badapanda, M. G. S. Costa, M. S. Li, W. Avansie, V. R. Mastelaro, E. Longo, J. W. M. Espinosa, and M. F. C. Gurgel, “Local electronic structure, optical bandgap and photoluminescence (PL) properties of Ba(Zr0.75Ti0.25)O3 powders,” Mater. Sci. Semicond. Process. 16(3), 1035–1045 (2013). [CrossRef]
13. S. C. Ray, H. C. Hsueh, C. H. Wu, C. W. Pao, K. Asokan, M. T. Liu, H. M. Tsai, C. H. Chuang, W. F. Pong, J. W. Chiou, M. H. Tsai, J. M. Lee, L. Y. Jang, J. M. Chen, and J. F. Lee, “Local atomic and electronic structures and ferroelectric properties of PbZr0.52Ti0.48O3: An x-ray absorption study,” Appl. Phys. Lett. 99(4), 042909 (2011). [CrossRef]
14. L. Q. Zhang, C. H. Zhang, J. Gou, L. H. Han, Y. T. Yang, Y. M. Sun, and Y. F. Jin, “PL and XPS study of radiation damage created by various slow highly charged heavy ions on GaN epitaxial layers,” Nucl. Instrum. Meth. B 269(23), 2835–2839 (2011). [CrossRef]
15. T. Fujishima, S. Joglekar, D. Piedra, H. S. Lee, Y. Zhang, A. Uedono, and T. Palacios, “Formation of low resistance ohmic contacts in GaN-based high electron mobility transistors with BCl3 surface plasma treatment,” Appl. Phys. Lett. 103(8), 083508 (2013). [CrossRef]
16. M. Higashiwaki, S. Chowdhury, B. L. Swenson, and U. K. Mishra, “Effects of oxidation on surface chemical states and barrier height of AlGaN/GaN heterostructures,” Appl. Phys. Lett. 97(22), 222104 (2010). [CrossRef]
17. R. D. Long and P. C. McIntyre, “Surface preparation and deposited gate oxides for gallium nitride based metal oxide semiconductor devices,” Materials 5(12), 1297–1335 (2012). [CrossRef]
18. T. Hashizume, R. Nakasaki, S. Ootomo, S. Oyama, and H. Hasegawa, “Surface characterization of GaN and AlGaN layers grown by MOVPE,” Mater. Sci. Eng. B 80(1-3), 309–312 (2001). [CrossRef]
19. J. Dumont, E. Monroy, E. Muñoz, R. Caudano, and R. Sporken, “Investigation of metal-GaN and metal-AlGaN contacts by XPS depth profiles and by electrical measurements,” J. Cryst. Growth 230(3-4), 558–563 (2001). [CrossRef]
20. C. L. Perkins, B. Egaas, I. Repins, and B. To, “Quantitative analysis of graded Cu(In1-x,Gax)Se2 thin films by AES, ICP-OES, and EPMA,” Appl. Surf. Sci. 257(3), 878–886 (2010). [CrossRef]
21. B. Ravel and M. Newville, “ATHENA, ARTEMIS, HEPHAESTUS: data analysis for X-ray absorption spectroscopy using IFEFFIT,” J. Synchrotron Radiat. 12(4), 537–541 (2005). [CrossRef] [PubMed]
22. M. Newville, “IFEFFIT: interactive XAFS analysis and FEFF fitting,” J. Synchrotron Radiat. 8(2), 322–324 (2001). [CrossRef] [PubMed]
23. T. Mattila and A. Zunger, “Predicted bond length variation in wurtzite and zinc-blende InGaN and AlGaN alloys,” J. Appl. Phys. 85(1), 160–167 (1999). [CrossRef]
24. M. Ferhat and F. Bechstedt, “First-principles calculations of gap bowing in InxGa1-xN and InxAl1-xN alloys: relation to structural and thermodynamic properties,” Phys. Rev. B 65(7), 075213 (2002). [CrossRef]
