Open Access
Add to BibSonomy
Abstract
We investigate here terahertz enhancement effects arising from micrometer and nanometer structured electrode features of photoconductive terahertz emitters. Nanostructured electrode based emitters utilizing the palsmonic effect are currently one of the hottest topics in the research field. We demonstrate here that even in the absence of any plasmonic resonance with the pump pulse, such structures can improve the antenna effect by enhancing the local d.c. electric field near the structure edges. Utilizing this effect in Hilbert-fractal and grating-like designs, enhancement of the THz field of up to a factor of ∼ 2 is observed. We conclude that the cause of this THz emission enhancement in our emitters is different from the earlier reported plasmonic-electrode effect in a similar grating-like structure. In our structure, the proximity of photoexcited carriers to the electrodes and local bias field enhancement close to the metallization cause the enhanced efficiency. Due to the nature of this effect, the THz emission efficiency is almost independent of the pump laser polarization. Compared to the plasmonic effect, these effects work under relaxed device fabrication and operating conditions.
Published by The Optical Society under the terms of the Creative Commons Attribution 4.0 License. Further distribution of this work must maintain attribution to the author(s) and the published article's title, journal citation, and DOI.
1. Introduction
Terahertz time-domain spectroscopy (THz-TDS) mainly relies on photoconductive terahertz (THz) emitters [1–3]. Many key parameters of a THz-TDS setup like signal-to-noise ratio, data acquisition rate, etc., depend on the efficiency of the THz emitters being used. In a photoconductive emitter, charge carriers are generated by exciting the semiconductor substrate by sub-picosecond optical pulses. These photoexcited charge carriers are accelerated by an already applied d.c. electric field in the semiconductor and emit the THz pulse [4–6]. Several techniques to improve the THz emission efficiency have been invented [7–13]. Many of them use different types of nanostructures ranging from grating lines to different types of nanodots [8,9,14–17]. The plasmonic-electrode technique which claims an improvement of several times in THz field amplitude has been invented by Jarrahi’s group [9,14]. This technique utilizes plasmon excitation on the grating-like structure in resonance with the NIR pump pulse resulting in confinement of most of the pumped optical energy close to metallic lines that are connected to the electrodes. Such nanostructure-based techniques offer great improvement if the dimensions of the nanostructures are well optimized for the pump-pulse wavelength [9]. However, for practical applications it will be of great importance if a THz enhancement technique is developed which does not require very precise structural dimensions and the emitter is compatible with flexible operating parameters. There have been successful attempts to make the efficiency of photoconductive THz emitters independent of the polarization of the pump pulse [18,19]. However, these devices also need very fine-tuned fabrication parameters to match the plasmonic resonance to the pump wavelength. In particular, the uniformity of the resonant feature size over the whole device is essential. The effect of plasmonic resonance with NIR pump should not be confused with the effect of plasmonic resonance of the THz waves with the emitter antenna/electrode [20]. The latter one depends on larger dimensions of the antenna/electrode, and is not part of this study. In this work we study effects which are different from plasmonic confinement of the NIR pump pulse and which can enhance the THz emission efficiency using comparatively relaxed fabrication and/or operating conditions.
It is now a well-established fact that the electric field distribution between the two contact electrodes in a GaAs photoconductor is locally enhanced close to the electrode edges. Hence, the excitation of the charge carriers close to the anode edge (due to higher electron mobility compared to hole mobility) enhances the THz emission from GaAs photoconductive emitters [21]. There have been several successful attempts to increase the THz emission based on this fact [22,23]. A drawback in this technique is that, when the pump pulse is focused close to the anode edge, a significant fraction of the pump pulse also falls onto the anode and cannot excite charge carriers in the photoconductor. As our simulations will show, the applied d.c. field is also enhanced below the electrode with maximum component in the direction perpendicular to the electrode surface. The field near the electrodes is expected to be enhanced further after the photoexcited electron and hole clouds separate and screen the field in the central region of the photoconducting medium between the two electrodes [24].
The THz emission mechanism in a photoconductive emitter can be attributed to two effects- 1) THz emission due to acceleration of photoexcited charge carriers within the photoconductor and 2) antenna effect from electrodes [11]. In the first mechanism, the polarization of the emitted THz pulse is parallel to the applied electric field direction in the photoconductor. Hence, maximum field lines due to applied bias should be in one direction (i.e. along the y-axis in our case) to get an efficient emission of linearly polarized THz pulse. However, in the antenna effect, charges approaching from any direction to the electrode will generate a displacement current in the electrode whose direction is restricted to the electrode/antenna length, which is also along the y-axis in our case. Thus, any type of electrode structure with a lot of edges to cause the local field enhancement can have efficient THz emission if we are able to excite charge carriers near the edges. Using a grating-like structure instead of a continuous surface for electrode structure will increase the edge length. A Hilbert-fractal design will also feature long edges and also it is expected to have a good and uniform transmittance for all the pump polarizations [25]. Here we study the effect of Hilbert-fractal (nanostructured electrode) and grating–line-like (nano- and micro-structured) electrodes on the THz emission efficiency of the photoconductive emitters.
2. Results on nanostructured electrodes
The THz emitters are fabricated on a semi insulating GaAs (SI-GaAs) by standard electron-beam lithography Ti (5 nm) and Au (25 nm) are used to deposit the electrodes. A schematic diagram of the emitter is shown in Fig. 1(a). Three emitters, #HF1, #HF2, and #HF3 with Hilbert-fractal design are fabricated with linewidths of 120 nm, 130 nm, and 140 nm respectively. The electrode structure close to the active area as enclosed in a rectangle in Fig. 1(a) is different for each emitter. The scanning electron microscope (SEM) images of one of the Hilbert-fractal electrodes is shown in Fig. 1(b) and (c). A reference bowtie emitter #RB with uniform metallization and an emitter #GL with grating line structures like plasmonic emitters with a linewidth of ∼130 nm and grating period of ∼ 450 nm, are also fabricated. The SEM images of emitter #GL are shown in Fig. 1(d) and (e).
Fig. 1. (a) Schematic diagram of the emitter electrode. (b-e) SEM images of a Hilbert-fractal emitter (b & c) and grating line (#GL) emitter (d & e). In (b & d), area A is completely open and area B is the patterned/semi-transparent part of the electrode.
The electric field amplitude distribution in the photoconductive medium due to the applied bias is simulated using the COMSOL Multiphysics software. The simulated field distribution in GaAs is for the dark condition i.e. before the arrival of pump pulse. In this condition the free-carrier density in GaAs and thus its conductivity is negligible compared to metal electrodes. Hence, GaAs can be considered as perfect dielectric material for the simulation. The result for emitter #GL is shown in Fig. 2. Results for other emitters are also similar and hence they are not shown. As expected, electric field amplitudes ((Ex2+Ey2+Ez2)1/2) are higher close to the electrode edges. Less obviously, this locally enhanced field also spans few µm into the region below the electrodes. This part of the bias field is responsible for the enhanced THz emission demonstrated for our emitters. The electric field in the middle between the two electrodes decreases to approximately 50% of the average field between the two electrodes. The field lines show that the field direction below the electrode is mainly along the z-axis since the metal surface defines an equipotential line. Thus any charge carrier excited in this region (denoted area B in Figs. 1(b) and (d)) will be accelerated mainly along the z-direction. Hence, the THz emission directly caused by their acceleration will not contribute in the emission of the main THz pulse, which is polarized in y-direction due to most of the charge carriers being excited between the two electrodes where the applied field is along the y-axis. The THz antenna (i.e., the metal dipole) is excited via photocurrents generated (i) in the gap between the electrodes (denoted as area A in Figs. 1(b) and (d)); and, (ii) below the semi-transparent parts of the electrodes (area B). All carriers reaching the electrodes (along the field lines) in the semiconductor cause displacement currents in the metal, all of which contribute to the THz emission. Thus the movement of the charge carriers excited in the area B (i.e. below the structured electrode) also induce a displacement current in the electrode due to the movement of corresponding image charges in the electrode. The displacement current will be in the electrodes and hence its direction is restricted to the electrode/antenna length, i.e., along the y-axis. The direction of the displacement current is indicated by the solid arrows in Fig. 2. Thus, the excitation of charge carriers below the electrodes near the edge becomes important for THz emission. Note that carriers excited close to the edge diffuse a few 100 nm on the timescale relevant for the THz emission [26]. Thus, it is possible to exploit efficiently the fields below metallized areas of these dimensions.
Fig. 2. Simulation of electric field amplitude ((Ex2+Ey2+Ez2)1/2 in V/m) distribution on a slice passing through the center of the emitter in yz-plane in the photoconductor when 10 V bias is applied to the electrodes. Field lines are drawn to show the direction of the electric field. A white dashed line is drawn at a depth (∼ 1 µm) of penetration depth of 800 nm in GaAs.
The emitters are pumped with a modelocked Ti:Sa laser operating at a central wavelength of ∼ 800 nm, 78 MHz repetition rate, and ∼ 100 fs pulse width. The NIR pump is focused between the two electrodes of the emitter using a lens of 5 cm focal length. The emitter position relative to the pump focus is optimized to get the maximum THz signal. We expect that a spot diameter of ∼ 15 µm centred in the middle of electrode gap should give the maximum THz emission from our nanostructured emitters. Current vs voltage (I-V) measurements of the emitters at the average pump power of 4 mW for two perpendicular pump polarizations are shown in Fig. 3. The pump polarization parallel to the y-axis of the emitter shown in Fig. 1(a) is called y-polarized, and polarization parallel to the x-axis is called x-polarized. The results of Hilbert-fractal emitters are plotted in Fig. 3(a); it turns out that for both pump polarizations the photocurrent of the emitters is almost the same. This is expected because Hilbert fractals are designed to result in a polarization-independent absorption [25]. I-V measurement results for the reference bowtie emitter #RB and the grating-line based emitter #GL are plotted in Fig. 3(b). There is no significant difference in the current for different emitters or different pump polarization. The absence of any polarization dependence on the photocurrent through emitter #GL implies that this set of samples does not show enhanced absorption due to plasmonic resonance of the NIR pump with grating lines. The grating linewidth in our device is close to the values of linewidths in plasmonic resonant devices reported by others [13–18,25], but there are some differences in our device like- absence of a dielectric layer on top, different metal thickness, etc. Thus, it is not surprising that we do not see a plasmonic resonance effect. Reflectance measurements in the 530 nm- 950 nm spectrum range on the grating lines also do not show any significant change due to plasmonic resonance (data not shown in the manuscript).
Fig. 3. (a) Current vs voltage characteristics of the three Hilbert-fractal emitters under illumination of pump pulse for two polarization orientations. The photocurrent through the devices is almost independent of pump pulse polarization. (b) Current vs voltage characteristics of #RB and #GL emitters illuminated under both pump polarization conditions. The photocurrents through all the devices under both conditions are similar without any trend of significant change.
The measured current in these experiments is time-averaged under a pulsed pumping condition. The THz emission intensity depends on the amplitude of the current spike for the initial ∼ 1 ps duration after pulsed excitation, but the photocurrent keeps flowing for longer times depending on the carrier lifetime of the semiconductor (∼ 100 ps in SI-GaAs) after each pump pulse. Hence the THz emission from the emitters does not necessarily need to be proportional to the average current in the emitters.
The THz emission performances of the emitters are compared by recording the emitted THz pulses in the time domain using a standard THz-TDS setup based on electro-optic detection using 1 mm thick ZnTe. Emitters are pumped with 2 mW pump power under two different polarization directions as explained for I-V measurements. Recorded THz pulses are shown in Fig. 4(a) for x-polarized pump. Compared to the regular bowtie emitter #RB, the THz electric field amplitudes from Hilbert-fractal emitters #HF1 and #HF2, and from emitter #GL are significantly higher (almost twice). The THz signal from #HF3 is low as compared to #HF1 and #HF2. The reason could be the wider line widths (140 nm) in #HF3. This will reduce the openings in the structured electrodes leading to reduced transmission of the pump light through the electrodes. The comparison of the peak electric field of the THz pulses emitted from 5 emitters under both pump polarizations is shown in Fig. 4(b). A polarization-independent performance of Hilbert-fractal emitters is expected. Interestingly, the performance of the grating line emitter #GL is also almost independent of pump pulse polarization. This rules out any significant plasmonic effect contribution in the THz emission enhancement in the grating line emitter #GL. We attribute the observed THz enhancement to the additional displacement current generated in the antenna/electrodes due to charge carriers approaching towards the electrodes from the photoconducting medium below the nanostructured part of the electrode. The fraction of charge carriers being generated below the structured part of the electrode could be much less as compared to the totally excited charge carriers but they will cause significant THz emission due to the antenna effect which they are able to create due to their close vicinity to the electrodes. This effect would enhance the THz emission only till the electrodes have enough charges to neutralize the charge carriers reaching the electrodes. At sufficiently high pump pulse energy, the number of photoexcited charge carriers will be large enough to completely discharge the electrodes before the complete THz pulse is emitted. This behavior is experimentally observed at pump power of 20 mW and results are shown in Fig. 4(c).
Fig. 4. (a) THz pulses emitted from five different emitters under 10 V applied bias and 2 mW, x-pol optical pumping. (b & c) Comparison of the electric field amplitude of the THz pulses emitted from five emitters under both pump polarization conditions and pump power of 2 mW (b) and 20 mW pump power (c). Amplitudes are normalized relative to the amplitude of emitter #RB under y-pol, 2 mW pump.
The observation of almost equal average photocurrent through the emitters but significant enhancement in the THz signal implies that the enhancement only occurs in the initial current spike just after the pumping for a very short time of the order of, or less than ∼ 1 ps. On a longer time average, the total photocurrent depends on the total number of photogenerated charge carriers, which is almost the same for all the emitters due to the same pump power. This means that enhanced THz emission will not cause any excessive Joule’s heating in the device.
3. Results on microstructured electrodes
In a different set of experiments on emitters with bowtie electrodes having finger-like structures with micrometre-sized features, we varied finger widths from 0.3 µm to 2.0 µm. The gap between two adjacent fingers is equal to the finger widths. A comparatively larger minimum feature size of the device simplifies the fabrication process. Microscope images of emitter electrodes of the regular emitter used as reference and the emitter with 1 µm wide fingers are shown in Fig. 5(a) and (b), respectively. Simulated electric field amplitude ((Ex2+Ey2+Ez2)1/2) distribution due to the applied bias in a xy-plane 200 nm below the top surface of GaAs is shown in Fig. 5(c). The electric field amplitude is higher, not only in the completely open area between the two electrodes (area A in Figs. 1(b) and (d)), but also in the region below the patterned part of the electrode (area B in Figs. 1(b) and (d)). Electric-field amplitudes are locally enhanced at the side edges of micro fingers and below the micro fingers for the regions close to the finger tips. As explained earlier, the direction of the bias field close to the electrode or electrode fingers is not important for the direction of the displacement current to be generated in the antenna.
Fig. 5. Optical microscope image of bowtie emitters with (a) regular electrode and (b) electrode having fingers of width 1 µm. The gap between the two electrodes is 10 µm. (c) A simulation of the electric field amplitude ((Ex2+Ey2+Ez2)1/2 in V/m) distribution on a slice in the xy-plane at a depth of 200 nm in the GaAs.
The recorded THz pulses emitted from emitters with different finger widths are shown in Fig. 6(a) and (b) for two different pump polarizations. Again, for both pump polarizations we observe that the emitters with finger-like structure have higher THz emission efficiency as compared to the emitter without finger-like structure. The amplitude increases as the finger width (and period) decreases. The plasmonic effect can be completely ruled out for the pump polarization parallel to the finger lines, but still we observe a clear enhancement in the THz amplitude. For the pump polarization perpendicular to the finger lines the enhancement is more profound and it also follows the same trend of increasing amplitude with decreasing finger width (and period). As the width and gap of the finger lines decreases, the average distance of photogenerated charge carriers from the finger edges decreases in the area between the fingers. The charge carriers closer to the electrode edges cause a more efficient antenna effect. An enhancement factor of ∼ 2 is observed in the THz electric field amplitude for the emitters with 0.3 µm and 0.4 µm finger-widths as shown in Fig. 6(a).
Fig. 6. THz pulses emitted from bowtie emitters, with regular electrode and with electrodes having fingers of different width. (a) The pump polarization is perpendicular to the fingers. (b) The pump polarization is parallel to the fingers. (c) Peak-to-peak field of THz pulses from emitters with varying finger widths.
4. Conclusion
In conclusion, our studies on photoconductive THz emitters having different types of nano- and micro-structures attached to the electrodes indicate that such structures can cause significant enhancement up to a factor of ∼ 2 in THz emission even in the absence of plasmonic resonance effect with the NIR pump pulse. This approach is useful for low pump power operation of THz emitter. The THz enhancement effect comes from decreased average distance of the photoexcited charge carriers from the electrode edges, which have locally enhanced d.c. field. Although the THz emission efficiency still depends on structural dimensions, it does not need to fulfil any resonance condition with the pump pulse and hence dimensions do not need to be precise. This property drastically simplifies the emitter fabrication process. The maximum enhancement factor of 2 observed in our emitter is less than what have been reported by plasmonic emitters [9], but it provides a very important fact that the THz enhacement from nanostructures are not always due to plasmonic effects. Since the THz emission is increased without any significant increase in the total average photocurrent through the emitters, it does not have any harmful effect on the thermal breakdown of the emitter. This effect does not depend much on the pump polarization relative to the emitter structures, and hence the user has the freedom to use any polarization of the pump light.
Funding
Helmholtz Association (PD 321).
Acknowledgment
Support by the Nanofabrication Facilities Rossendorf at IBC is gratefully acknowledged. Authors thank Mr. Bernd Scheumann for metal deposition for the fabrication of the emitters. Authors also thank Tommaso Venanzi for the reflection measurements.
Disclosures
The authors declare that there are no conflicts of interest related to this article.
References
1. S. S. Dhillon, M. S. Vitiello, E. H. Linfield, A. G. Davies, M. C. Hoffmann, J. Booske, C. Paoloni, M. Gensch, P. Weightman, G. P. Williams, E. Castro-Camus, D. R. S. Cumming, F. Simoens, I. Escorcia-Carranza, J. Grant, S. Lucyszyn, M. Kuwata-Gonokami, K. Konishi, M. Koch, C. A. Schmuttenmaer, T. L. Cocker, R. Huber, A. G. Markelz, Z. D. Taylor, V. P. Wallace, J. A. Zeitler, J. Sibik, T. M. Korter, B. Ellison, S. Rea, P. Goldsmith, K. B. Cooper, R. Appleby, D. Pardo, P. G. Huggard, V. Krozer, H. Shams, M. Fice, C. Renaud, A. Seeds, A. Stöhr, M. Naftaly, N. Ridler, R. Clarke, J. E. Cunningham, and M. B. Johnston, “The 2017 terahertz science and technology Roadmap,” J. Phys. D: Appl. Phys. 50(4), 043001 (2017). [CrossRef]
2. N. M. Burford and M. O. El-Shenawee, “Review of terahertz photoconductive antenna technology,” Opt. Eng. 56(1), 010901 (2017). [CrossRef]
3. C. A. Schmuttenmaer, “Exploring dynamics in the far-infrared with terahertz spectroscopy,” Chem. Rev. 104(4), 1759–1780 (2004). [CrossRef]
4. D. H. Auston, K. P. Cheung, and P. R. Smith, “Picosecond photoconducting Hertzian dipoles,” Appl. Phys. Lett. 45(3), 284–286 (1984). [CrossRef]
5. P. R. Smith, D. H. Auston, and M. C. Nuss, “Subpicosecond photoconducting dipole antennas,” IEEE J. Quantum Electron. 24(2), 255–260 (1988). [CrossRef]
6. C. Fattinger and D. Grischkowsky, “Point source terahertz optics,” Appl. Phys. Lett. 53(16), 1480–1482 (1988). [CrossRef]
7. A. Dreyhaupt, S. Winnerl, T. Dekorsy, and M. Helm, “High-intensity terahertz radiation from a microstructured large-area photoconductor,” Appl. Phys. Lett. 86(12), 121114 (2005). [CrossRef]
8. S. G. Park, K. H. Jin, M. Yi, J. C. Ye, J. Ahn, and K. H. Jeong, “Enhancement of terahertz pulse emission by optical nanoantenna,” ACS Nano 6(3), 2026–2031 (2012). [CrossRef]
9. C. W. Berry, N. Wang, M. R. Hashemi, M. Unlu, and M. Jarrahi, “Significant performance enhancement in photoconductive terahertz optoelectronics by incorporating plasmonic contact electrodes,” Nat. Commun. 4(1), 1622 (2013). [CrossRef]
10. D. R. Bacon, T. B. Gill, M. Rosamond, A. D. Burnett, A. Dunn, L. Li, E. H. Linfield, A. G. Davies, P. Dean, and J. R. Freeman, “Photoconductive arrays on insulating substrates for high-field terahertz generation,” Opt. Express 28(12), 17219–17231 (2020). [CrossRef]
11. A. Singh, M. Welsch, S. Winnerl, M. Helm, and H. Schneider, “Improved electrode design for interdigitated large-area photoconductive terahertz emitters,” Opt. Express 27(9), 13108–13115 (2019). [CrossRef]
12. A. Singh, A. Pashkin, S. Winnerl, M. Helm, and H. Schneider, “Gapless broadband terahertz emission from a germanium photoconductive emitter,” ACS Photonics 5(7), 2718–2723 (2018). [CrossRef]
13. S. Yang, M. R. Hashemi, C. W. Berry, and M. Jarrahi, “7.5% optical-to-terahertz conversion efficiency offered by photoconductive emitters with three-dimensional plasmonic contact electrodes,” IEEE Trans. Terahertz Sci. Technol. 4(5), 575–581 (2014). [CrossRef]
14. M. Jarrahi, “Advanced photoconductive terahertz optoelectronics based on nano-antennas and nano-plasmonic light concentrators,” IEEE Trans. Terahertz Sci. Technol. 5(3), 391–397 (2015). [CrossRef]
15. S. Lepeshov, A. Gorodetsky, A. Krasnok, N. Toropov, T. A. Vartanyan, P. Belov, A. Alu, and E. U. Rafailov, “Boosting terahertz photoconductive antenna performance with optimized plasmonic nanostructures,” Sci. Rep. 8(1), 6624 (2018). [CrossRef]
16. S. Lepeshov, A. Gorodetsky, A. E. Krasnok, E. Rafailov, and P. A. Belov, “Enhancement of terahertz photoconductive antennas operation by optical nanoantenna,” Laser Photonics Rev. 11(1), 1770001 (2017). [CrossRef]
17. A. Jooshesh, V. Bahrami-Yekta, J. Zhang, T. Tiedje, T. E. Darcie, and R. Gordon, “Plasmon-enhanced below bandgap photoconductive terahertz generation and detection,” Nano Lett. 15(12), 8306–8310 (2015). [CrossRef]
18. X. Li, N. T. Yardimci, and M. Jarrahi, “A polarization-insensitive plasmonic photoconductive terahertz emitter,” AIP Adv. 7(11), 115113 (2017). [CrossRef]
19. P. Maraghechi and A. Y. Elezzab, “Enhanced THz radiation emission from plasmonic complementary Sierpinski fractal emitters,” Opt. Express 18(26), 27336–27345 (2010). [CrossRef]
20. A. Singh, S. Winnerl, J. C. König-Otto, D. R. Stephan, M. Helm, and H. Schneider, “Plasmonic efficiency enhancement at the anode of strip line photoconductive terahertz emitters,” Opt. Express 24(20), 22628–22634 (2016). [CrossRef]
21. S. E. Ralph and D. Grischkowsky, “Trap-enhanced electric fields in semi-insulators: The role of electrical and optical carrier injection,” Appl. Phys. Lett. 59(16), 1972–1974 (1991). [CrossRef]
22. J. H. Kim, A. Polley, and S. E. Ralph, “Efficient photoconductive terahertz source using line excitation,” Opt. Lett. 30(18), 2490–2492 (2005). [CrossRef]
23. A. Singh, H. Surdi, V. V. Nikesh, S. S. Prabhu, and G. H. Döhler, “Improved efficiency of photoconductive THz emitters by increasing the effective contact length of electrodes,” AIP Adv. 3(12), 122106 (2013). [CrossRef]
24. E. Castro-Camus, J. Lloyd-Hughes, and M. B. Johnston, “Three-dimensional carrier-dynamics simulation of terahertz emission from photoconductive switches,” Phys. Rev. B 71(19), 195301 (2005). [CrossRef]
25. F. Afshinmanesh, A. G. Curto, K. M. Milaninia, N. F. Van Hulst, and M. L. Brongersma, “Transparent metallic fractal electrodes for semiconductor devices,” Nano Lett. 14(9), 5068–5074 (2014). [CrossRef]
26. M. B. Johnston, D. M. Whittaker, A. Corchia, A. G. Davies, and E. H. Linfield, “Simulation of terahertz generation at semiconductor surfaces,” Phys. Rev. B 65(16), 165301 (2002). [CrossRef]
