Open Access
Add to BibSonomyAbstract
We have observed a remarkable decrease in photoluminescence (PL) from a blue-light emitting InGaN single-quantum-well (SQW) structure under the radiation of a green laser due to the stimulated emission depletion (STED) phenomenon. By extending the observed STED effect, super-resolution imaging of the blue-light emission lateral distribution was demonstrated for the InGaN-SQW structure through co-irradiation using a doughnut-shaped green light beam and a Gaussian-shaped violet excitation light beam. We measured point-spread functions (PSFs) to evaluate the spatial resolution of the system by imaging a small emission area. A lateral PSF size of ~150 nm was confirmed, which was approximately 40% smaller than that without the STED beam. This demonstrates that the STED technique is applicable for PL imaging of semiconductor quantum structures. The present approach may make possible a new strategy for characterizing and investigating the spatial inhomogeneity of emission properties and carrier dynamics in InGaN-based quantum wells, as well as in other semiconductor materials exhibiting quantum confinement effects.
© 2014 Optical Society of America
1. Introduction
Semiconductor devices based on an InGaN quantum-well (QW) structure are now broadly and commercially used, especially for blue- and violet-light-emitting diodes (LEDs) and laser diodes (LDs) [1]. In general, indium compositional fluctuation formed in the QW structure is known to affect the device performance for InGaN-based LDs and LEDs. This leads to localized states in the QWs [2, 3], resulting in effects such as significant inhomogeneous broadening in LDs [4]. Several techniques, including cathodoluminescence microscopy [5–7] and scanning near-field optical microscopy [8–10], have been employed to investigate the spatially-resolved, optical characteristics of QWs. Although these methods have the potential to provide a spatial resolution of several tens of nanometers [5], the resolution is practically restricted by instrumental limitations and the carrier diffusion length inside the QWs. Spatially-resolved photoluminescence (PL) mapping, i.e., far-field laser microscopy with confocal detection [10, 11], is another simple and reliable method for characterizing the local behavior of light emission in QWs. Owing to a confocal pinhole, this method, based on confocal laser microscopy, enables PL mapping with high spatial resolution even at room temperature, where the influence of carrier diffusion in PL emissions is considerable [11]. Moreover, this method is advantageous for the investigation of carrier dynamics in QWs. However, in principle, the spatial resolution of confocal laser microscopy depends on the size of a focused excitation spot as well as the size of the pinhole [12]; therefore, due to the diffraction limit of light, the spatial resolution is typically in the range of 200-300 nm for the excitation wavelength of UV and the blue-violet region.
Stimulated emission depletion (STED) microscopy [13, 14], which was originally developed in the area of biological imaging, is a well-known super-resolution technique that realizes dramatically enhanced spatial resolution in laser microscopy. STED microscopy has recently been applied to imaging of nitrogen-vacancy color centers in diamond and has demonstrated a spatial resolution of ~6 nm [15]. In STED microscopy, a second laser beam (referred to as a STED beam), which is overlapped with an excitation beam, is focused onto the fluorescent sample. When the wavelength of a STED beam is properly tuned to a relatively longer wavelength within the fluorescence spectrum of an excited molecule, the STED beam is able to induce stimulated emission that prevents fluorescence emission from the molecule. By using a doughnut-shaped STED beam, the lateral spatial resolution of laser microscopy is remarkably enhanced because the effective fluorescence area is restricted to the tiny dark region in the center of the doughnut-shaped beam.
An InGaN-QW having considerable potential fluctuation is characterized by highly localized states that exhibit a large Stokes shift in PL emission. Therefore, in analogy with STED microscopy, the radiation of a laser beam (corresponding to a “STED” beam) tuned to the long-wavelength tail of the PL emission spectrum is expected to induce stimulated emission for carriers localized at deeper potential minima, causing the reduction of carriers and the eventual depletion of PL emission. This is confirmed by observing the variation in PL emission with and without STED beam irradiation. Hence, based on the concept of STED microscopy, an enhanced spatial resolution in PL mapping is possible for InGaN-QWs, provided that the depletion of PL emission indeed takes place in the QWs.
To the best of our knowledge, this is the first study reporting the observation of PL emission depletion in a blue-light emitting InGaN single-quantum well (SQW) by the radiation of a laser beam with a wavelength of 532 nm, which is considerably longer than that of the peak wavelength of the PL spectrum. By converting the 532-nm laser beam into a doughnut-shaped beam, we demonstrated the enhanced spatial resolution of the PL images. In addition, we measured point-spread functions (PSFs) to evaluate the spatial resolution of the system by imaging a small emission area fabricated on the InGaN-SQW covered by a metallic mask. The lateral PSF size was reduced by approximately 40% compared to the conventional PL images without the STED beam. The mechanism for depletion of the PL emission under STED beam irradiation for the InGaN-SQW is discussed.
2. Experimental setup
Figure 1(a) shows a schematic diagram of the experimental setup. The sample was a 3-nm-thick InGaN-SQW grown through molecular beam epitaxy on a sapphire (0001) substrate. To identify and restrict the region of PL emission from the SQW, the surface of the sample was masked by gold thin film having periodic square aperture arrays of 2 × 2 μm2 and 200 × 200 nm2. Since the active layer having the SQW is not located immediately below the metal mask, the influences of the metal layer such as carrier transfer to the metal layer or the enhancement of the electric field due to the mask structure can be negligible. The imaging system adopted either a 405-nm LD with continuous wave (CW) operation or a 375-nm LD with pulsed operation to excite the SQW. The 405-nm laser was used predominantly in this study. A CW green laser with a wavelength of 532 nm was used as the STED beam. The excitation and STED beams were combined by a dichroic mirror and focused onto the same point on the sample by an objective lens (MPLAPON100 × , NA = 0.95, Olympus). The doughnut-shaped intensity distribution for the STED beam was generated by using a transmissive, liquid-crystal-based vortex phase plate (LC-VPP). The LC-VPP produces a spiral phase variation from 0 to 2π on the beam cross section. When the appropriate driving voltage is applied to the LC-VPP, the intensity distribution of the STED beam is transformed from a Gaussian into a doughnut shape. In the present study, the PL emission from the SQW sample was collected by the same objective lens and detected by a PMT (H10721-20, Hamamatsu Photonics) through a focusing lens with a focal length of 60 mm. An optical band pass filter was placed before the detector to block the excitation and STED beams. To eliminate stray light, we used a pinhole with a diameter of 100 μm in front of the detector, which is relatively larger than 1 Airy unit (~17 μm) calculated for the confocal system. Hence, barring the effect of the diffusion length of carriers created in the SQW, the lateral spatial resolution of this system depends directly on the focal spot size of the excitation beam. A spectrometer (PMA-12 C10027, Hamamatsu Photonics) was used to measure the PL spectra from the sample. The experiment was performed at room temperature.
Fig. 1 (a) Schematic diagram of the experimental setup. DM: dichroic mirror, LC-VPP: liquid crystal vortex phase plate, BPF: band-pass filter. The excitation and STED laser beams are focused on the same position on the InGaN-SQW. The photoluminescence (PL) from the sample, which peaked at ~460 nm, was reflected by a DM and detected through a BPF. (b)-(d) Measured intensity distributions at the focus of the laser beams with 405 nm (b), 532 nm without the LC-VPP (c), and 532 nm with the LC-VPP (d). The scale bar is 400 nm.
The measured intensity distributions of the excitation beam, the Gaussian-shaped STED beam, and the doughnut-shaped STED beam at the focus are shown in Figs. 1(b) and 1(c). A 100-nm gold bead mounted on a glass plate was used as a probe. We measured the back-scattered light intensity from the gold bead at each stage position. The full-width at half-maximum (FWHM) values of the focal spot sizes along the horizontal direction for the excitation beam [Fig. 1(b)] and the Gaussian STED beam [Fig. 1(c)] were 259 nm and 344 nm, respectively. When a driving voltage was applied to the LC-VPP, the doughnut-shaped STED beam was obtained as shown in Fig. 1(d). Note that the reflectivity of gold beads for 405 nm is lower than that for 532 nm, which results in a higher background level as shown in Fig. 1(b).
3. Results
Figure 2 shows the measured PL spectra excited by a 375-nm LD with and without the Gaussian STED beam. The excitation power of the excitation beam at the focus was Pexc. = 70 nW. The excitation and STED beams were focused by an objective lens with NA = 0.4 to characterize a spatially-averaged emission from a relatively larger region (~6 μm in diameter) of the SQW. The PL peak wavelength was observed at approximately 460 nm without the STED beam. When a Gaussian STED beam with an incident power PSTED of 10.8 mW (at the focus) and the excitation beam were focused on the same position, the PL intensity almost disappeared, as shown in Fig. 2. This result indicates that the 532-nm STED beam induced the elimination of the PL emission. On the other hand, we also confirmed that similar and effective decrease of PL emission was not observed for STED beam irradiations in the wavelength range of 550 – 560 nm (data not shown), which suggests that thermal processes such as substrate heating due to STED beam irradiation are not the main cause of the depletion effect. Furthermore, the PL depletion caused by the 532-nm STED beam irradiation was repeatedly observed for iteration of excitation without and with STED beam irradiations. This implies that the decrease of the PL intensity for 532-nm STED beam irradiation was not caused by the deterioration of the sample due to an intense STED beam or the effect of photobleaching, which are almost ignorable for semiconductor-based light-emitting devices.
Fig. 2 Typical example of PL spectra indicating the STED effect. PL spectra (i) without and (ii) with a Gaussian STED beam are shown. Pexc. = 70 nW. PSTED = 10.8 mW.
Following the result of the PL spectral measurement, we mapped the PL intensity in a 2 × 2-μm2 square aperture using the scanning system and a 405-nm LD with Pexc. = 3.7 μW. Figure 3(a) shows a typical PL image obtained using this sample. The broken line in Fig. 3(a) represents the edge of the 2 × 2-μm2 aperture. Within the aperture, an inhomogeneous intensity distribution, which can be attributed to the composition fluctuation in the InGaN-QW structure, was obtained. The PL intensity mappings for the same region with the Gaussian and doughnut-shaped STED beams were acquired for different incident STED beam power levels (0.2, 5.0, and 17.1 mW at the focus), as shown in Figs. 3(b)–3(g). For both STED beams, we found that the PL intensity decreased depending on the position. In particular, the PL emission in the lower portion of the measured region drastically decreased for higher STED beam powers. By contrast, the PL emission in the top right of the region remains almost unchanged under the STED beam irradiation. It should also be noted that, on the right-hand side of the region for both the Gaussian and doughnut-shaped STED beams, we observed the nonlinear increase of the PL intensity subsequent to the decrease of the PL intensity when the STED beam power was increased, which implies the presence of the two-photon excitation by the intense STED beam itself. Thus, the PL emission behavior under the radiation of a STED beam depends greatly on the position and the incident power of the STED beam.
Fig. 3 (a) PL intensity mapping for the InGaN-SQW measured using a 2 × 2 μm2 aperture. The same region was imaged with a Gaussian STED beam (b)-(d) and a doughnut-shaped STED beam (e)-(g) with different incident powers. Pexc. = 3.7 μW. PSTED = 0.2 [(b) and (e)], 5.0 [(c) and (f)], and 17.1 mW [(d) and (g)]. (h)-(j) are the magnified images acquired for the areas marked by the red squares in (a), (d), and (g), respectively. The false-color scales used in (a)-(g) and (h)-(j) are normalized by the maximum intensities in (a) and (h), respectively. The intensity profiles along the dotted lines in (h)-(j) are plotted in (k). The spot-shaped distributions observed when the doughnut-shaped STED beam was used are indicated by arrows in (j) and (k).
The region where the PL intensity markedly decreased for STED beam irradiation was acquired as shown in Figs. 3(h)–3(j), which correspond to the regions denoted by red squares in Figs. 3(a), 3(d), and 3(g), respectively. Without the STED beam, the PL image became blurred and broadened in structure. While the co-irradiation with a Gaussian STED beam merely weakened the PL emission in this area, the finer structure having three lobes (depicted by arrows) along the dotted line in Fig. 3(j) was clearly resolved by using a doughnut-shaped STED beam. Figure 3(k) plots the intensity profiles along the dotted lines indicated in Figs. 3(h)–3(j). These intensity profiles also illustrate that the single blurred peak of the PL image obtained without a STED beam splits into three peaks under doughnut-shaped STED beam irradiation. In addition, the left bottom spot shown in Fig. 3(j) was not highly effective for STED beam irradiation compared to other two spots, which is also clearly found in the Fig. 3(k), revealing the position-dependent PL depletion behavior. Nonetheless, these results prove the enhancement of the spatial resolution in PL intensity mapping using a doughnut-shaped STED beam. The behavior observed is similar to that of STED microscopy applied to the bio-imaging using organic dyes and fluorescent proteins.
To evaluate the spatial resolution of the PL imaging with STED beam irradiation, we measured a PSF by mapping PL intensity through a 200 × 200-nm2 aperture, which can be seen as a quasi-point source. Since the PL emission and its depletion behavior for the InGaN-SQW sample are position dependent, we selected a region exhibiting efficient PL-depletion under STED beam irradiation. Figure 4 shows the PL intensity measured in the 200 × 200-nm2 region as a function of the incident power of the Gaussian STED beam. Under the radiation of the STED beam, the PL intensity reduced to nearly 10% of the initial intensity measured without STED beam irradiation. The depletion of PL emission was almost saturated at the incident STED power of ~4 mW. However, as shown in the inset of Fig. 4, the PL intensity was again increased slightly with an STED power of greater than 10 mW. This was likely due to two-photon excitation caused by the excess intensity of the STED beam as mentioned above. Therefore, the available STED power in this region was limited to about 10 mW.
Fig. 4 Measured PL intensity in a 200 × 200-nm2 region as a function of the STED beam power. Pexc. = 6.2 μW. A Gaussian STED beam was used. The inset shows a magnification of the plot between 4 and 18 mW.
Figures 5(a) and 5(b) show the corresponding PL images of the 200 × 200-nm2 region measured in Fig. 4 with and without the doughnut-shaped STED beam, respectively. The incident power of the STED beam was 9.6 mW. In both cases, a circular and bright spot shape was obtained as an image of a rectangular aperture having a size of 200 × 200 nm2. The intensity profiles of the PL images along the horizontal and vertical axes are shown in Figs. 5(c) and 5(d). We estimated the lateral size of the spot image by Gaussian curve-fitting of the intensity profiles. Without the STED beam, the FWHM of the lateral size of the spot image was approximately 270 nm, which is close to the size of the focal spot (259 nm) measured for the excitation beam. This implies that the PL emission within the small square aperture behaved as a quasi-point source. Accordingly, the PSF size in this system without the STED beam is nearly identical to the spot size of the excitation beam. With the doughnut-shaped STED beam, the lateral size of the PL image was reduced to 145 nm and 160 nm in the horizontal and vertical directions, respectively. The lateral size of the PL image decreased by about 40% when the doughnut-shaped STED beam was used. This is remarkably smaller than the spatial resolution of laser microscopy using a 405-nm excitation beam and an objective lens with NA = 0.95. These results indicate that super-resolution imaging is able to be obtained for PL mapping of an InGaN-SQW using a doughnut-shaped STED beam.
Fig. 5 Measured spot-shaped image of the same region as Fig. 4 without a STED beam (a) and with a doughnut-shaped STED beam (b). The intensity profiles along the horizontal and vertical axes for both (a) and (b) are shown in (c) and (d), respectively. Pexc. = 6.2 μW. PSTED = 9.6 mW.
4. Discussion
The present study demonstrated the STED effect on the PL emission of an InGaN-SQW, which improved the spatial resolution of the PL intensity mapping. This implies that the PL depletion observed in the InGaN-SQW was caused by a mechanism similar to that of the fluorescence depletion induced by stimulated emission in organic dyes or fluorescent proteins [13]. Although more detailed and elaborate studies are necessary in order to fully understand the present mechanism, the principle of the PL depletion and the resolution improvement of PL imaging can be briefly explained as follows.
The wavelength of the STED beam (532 nm) used in this study corresponds to the lower energy tail in the PL spectrum of the InGaN-SQW. The PL depletion observed for the combination of the STED and excitation beams indicates that the carriers contributing to the emission around 532 nm formed population inversion in the SQW. The population inversion in such a lower energy level will occur for an even lower carrier density since the carriers will readily occupy the localized states corresponding to the emission of ~532 nm, which appear as tail-states with a lower density of states and can be seen as quantum dots (QDs) in the InGaN-SQW due to the compositional fluctuation [2,3]. Therefore, while the use of the STED beam induces the STED effect in the QDs, these QDs can also behave as sinks for higher-energy carriers. Through the assistance of thermal escape and carrier diffusion driven by thermal energy at room temperature [16], higher-energy carriers may transfer to the lower-energy QDs where carrier recombination is induced by the STED beam. As a result, under the weak excitation condition, PL depletion is observed in the InGaN-QW. This also implies that the STED effect occurs only when photo-excited carriers exist near the QDs behaving as a sink within the diffusion length.
The STED process discussed above will be further explored by the combined use of more advanced techniques, including time-resolved measurements and spectroscopic analyses. In addition, this method offers the possibility to reveal carrier dynamics such as localization, diffusion, and redistribution, as well as the visualization of potential fluctuation in InGaN-SQWs with significantly enhanced spatial resolution.
5. Conclusions
We have demonstrated a novel technique that improves the spatial resolution of PL imaging of a blue-light-emitting InGaN-SQW based on the concept of STED microscopy. The PL emission from the InGaN-SQW, which peaked at 460 nm, decreased due to the stimulated emission depletion process when a 532-nm STED beam was used in combination with the excitation beam. Utilizing the STED effect, it was also confirmed that the spatial resolution of the PL image was remarkably improved by using a doughnut-shaped STED beam; the lateral PSF size for the PL images acquired with the STED laser beam was reduced by approximately 40% compared to that obtained without the STED laser beam. Our present results indicate that further improvements in the super-resolution imaging of the InGaN material system can verify the carrier localization scale, which is dominated by the material composition fluctuation in InGaN quantum structures. Furthermore, this super-resolution imaging may also enable carrier dynamics such as trapping and escaping in the potential well, as well as carrier diffusion, to be seen.
Acknowledgments
The authors wish to thank A. Sato for technical assistance. The authors also wish to thank Citizen Holdings Co., Ltd. for providing the liquid crystal device. This work was supported in part by Core Research for Evolutional Science and Technology (CREST), the Japan Science and Technology Agency (JST), and by KAKENHI Grant Number 20104004 from MEXT.
References and links
1. S. Nakamura and G. Fasol, The Blue Laser Diode (Springer, 1997).
2. S. Chichibu, T. Azuhata, T. Sota, and S. Nakamura, “Spontaneous emission of localized excitons in InGaN single and multiquantum well structures,” Appl. Phys. Lett. 69(27), 4188–4190 (1996). [CrossRef]
3. S. F. Chichibu, A. Uedono, T. Onuma, B. A. Haskell, A. Chakraborty, T. Koyama, P. T. Fini, S. Keller, S. P. Denbaars, J. S. Speck, U. K. Mishra, S. Nakamura, S. Yamaguchi, S. Kamiyama, H. Amano, I. Akasaki, J. Han, and T. Sota, “Origin of defect-insensitive emission probability in In-containing (Al,In,GaN)N alloy semiconductors,” Nat. Mater. 5(10), 810–816 (2006). [CrossRef] [PubMed]
4. K. Kojima, U. T. Schwarz, M. Funato, Y. Kawakami, S. Nagahama, and T. Mukai, “Optical gain spectra for near UV to aquamarine (Al,In)GaN laser diodes,” Opt. Express 15(12), 7730–7736 (2007). [CrossRef] [PubMed]
5. S. Chichibu, K. Wada, and S. Nakamura, “Spatially resolved cathodoluminescence spectra of InGaN quantum wells,” Appl. Phys. Lett. 71(16), 2346–2348 (1997). [CrossRef]
6. F. Bertram, S. Srinivasan, L. Geng, F. A. Ponce, T. Riemann, and J. Christen, “Microscopic correlation of redshifted luminescence and surface defects in thick InxGa1-xN layers,” Appl. Phys. Lett. 80(19), 3524–3526 (2002). [CrossRef]
7. S. Sonderegger, E. Feltin, M. Merano, A. Crottini, J. F. Carlin, R. Sachot, B. Deveaud, N. Grandjean, and J. D. Ganière, “High spatial resolution picosecond cathodoluminescence of InGaN quantum wells,” Appl. Phys. Lett. 89(23), 232109 (2006). [CrossRef]
8. M. S. Jeong, J. Y. Kim, Y.-W. Kim, J. O. White, E.-K. Suh, C.-H. Hong, and H. J. Lee, “Spatially resolved photoluminescence in InGaN/GaN quantum wells by near-field scanning optical microscopy,” Appl. Phys. Lett. 79(7), 976–978 (2001). [CrossRef]
9. A. Kaneta, T. Hashimoto, K. Nishimura, M. Funato, and Y. Kawakami, “Visualization of the local carrier dynamics in an InGaN quantum well using dual-probe scanning near-field optical microscopy,” Appl. Phys. Express 3(10), 102102 (2010). [CrossRef]
10. K. P. O’Donnell, M. J. Tobin, S. C. Bayliss, and W. Van Der Stricht, “Confocal microscopy and spectroscopy of InGaN epilayers on sapphire,” J. Microsc. 193(2), 105–108 (1999). [CrossRef]
11. K. Okamoto, A. Kaneta, Y. Kawakami, S. Fujita, J. Choi, M. Terazima, and T. Mukai, “Confocal microphotoluminescence of InGaN-based light-emitting diodes,” J. Appl. Phys. 98(6), 064503 (2005). [CrossRef]
12. T. Wilson and A. R. Carlini, “Size of the detector in confocal imaging systems,” Opt. Lett. 12(4), 227–229 (1987). [CrossRef] [PubMed]
13. S. W. Hell and J. Wichmann, “Breaking the diffraction resolution limit by stimulated emission: stimulated-emission-depletion fluorescence microscopy,” Opt. Lett. 19(11), 780–782 (1994). [CrossRef] [PubMed]
14. V. Westphal, S. O. Rizzoli, M. A. Lauterbach, D. Kamin, R. Jahn, and S. W. Hell, “Video-rate far-field optical nanoscopy dissects synaptic vesicle movement,” Science 320(5873), 246–249 (2008). [CrossRef] [PubMed]
15. E. Rittweger, K. Y. Han, S. E. Irvine, C. Eggeling, and S. W. Hell, “STED microscopy reveals crystal colour centres with nanometric resolution,” Nat. Photon. 3(3), 144–147 (2009). [CrossRef]
16. S.-W. Feng, Y.-C. Cheng, Y.-Y. Chung, C. C. Yang, Y.-S. Lin, C. Hsu, K.-J. Ma, and J.-I. Chyi, “Impact of localized states on the recombination dynamics in InGaN/GaN quantum well structures,” J. Appl. Phys. 92(8), 4441–4448 (2002). [CrossRef]
