Design, Fabrication and Characterization of Very Small Aperture Lasers

372 Progress In Electromagnetics Research Symposium 2005, Hangzhou, China, August 22-26 Design, Fabrication and Characterization of Very Small Aperture Lasers Jiying Xu, Jia Wang, and Qian Tian Tsinghua University, China Abstract An L-shaped aperture with 15nm ultra-high optical resolution and strong field enhancement is designed. The very small aperture lasers (VSAL) with L- and rectangular aperture are fabricated, and their confined optical fields are detected using an apertured scanning near-field optical microscope (SNOM). For L-aperture VSAL, when the drive current is 28mA, the power output can reach 1.049µW, while the power output of the rectangular aperture VSAL at most reaches 0.078µW. In the experiment based on the apertures in Au film on glass substrate, the peak intensity of C aperture is about 10 times over the square aperture. The experimental results indicate that the unconventional L- and C-aperture have stronger field enhancement than rectangular and square aperture. Introduction Very small aperture laser (VSAL) is one effective method to obtain a promising nanometric light source. It has great potential applications in high-density optical data storage, scanning near-field optical microscopes (SNOMs), nano-lithography and so on. The first VSAL was demonstrated by Partovi[1] in 1999, which has a 250nm square aperture and can be used in near-field optical storage experimentally. But the power delivered through the nano-aperture rapidly decays with its size decrease. Then some unconventional apertures with extraordinary field enhancement, such as Shi s C-aperture[2], Sendur s bow-tie slot antenna[3], Tanaka s I-shaped aperture[4] and Thio s nano-aperture with periodic surface topography[5], were proposed. Experimentally Chen[6] fabricated a VSAL with C-aperture and detected its optical field confinement using an apertureless SNOM. Shinada[7] fabricated a VSAL with a 200nm square aperture with periodic concentric circles. Under the constant injection current, its peak intensity was eight times larger than that without periodic topography, but it cannot obtain laser oscillation. In this paper, besides the square aperture and C-aperture proposed previously, an L-shaped aperture with 15nm optical resolution is designed and fabricated on VSAL. The variation of power output with working current proves that it can obtain laser oscillation. The confined optical fields from the apertures on the VSALs and the glass substrate are measured directly using an apertured SNOM. The strong field enhancement from the unconventional L- and C-aperture is demonstrated experimentally and quantificationally. Simulation and Design of Nano-apertures Figure 1: Schematic geometry of three nano-apertures. As shown in Fig.1, the square aperture s side length is denoted by M. One C aperture consists of five same square apertures with side length M. In the situation, M is set to be 70nm and both the Au film thicknesses are 150nm. For L-aperture, W1=L2=80nm, W2=70nm, L1=210nm, and the Au film thickness is 130nm. The calculation results in our previous paper[8] indicate that for the C- and L-aperture, the spot sizes are smaller, while both normalized peak intensity and power throughput are about 1000-fold higher than the

Progress In Electromagnetics Research Symposium 2005, Hangzhou, China, August 22-26 373 conventional square aperture. And for the L-aperture, there is an advantage that a spot with resolution 15nm is formed when light just exit from the aperture, so that the control of working distance is released in that it can be applied in contact mode and the distance need not to be set an in-between value like C-aperture. Fabrication of Nano-apertures and VSALs Fabrication of Nano-apertures The square and C-apertures with various sizes are fabricated in 110nm-thickness Au film on a glass substrate with focused ion beam (FIB) method as the verification of indispensable fabrication process and characterization. The scanning electronic microscope (SEM) image is shown in Fig.2. Apertures labeled by 1-3 are square apertures with unit lengths M =100nm, 200nm and 300nm respectively, and 4-6 are C-apertures with unit lengths as the former. It can be seen that the fabrication of nano-apertures is not perfect and especially C-apertures are disfigured to some extent. Figure 2: SEM image of square and C-apertures (1, 4: M=100nm, 2, 5: M=200nm, 3, 6: M=300nm) Fabrication of VSALs with Edge-emitting Laser Diodes Some commercial edge-emitting laser diodes (LD) with wavelength λ=650nm are used to fabricate VSALs. Firstly, a dielectric antireflection layer, SiNx, is deposited on the light exit facet of the LD to maximize the power output and prevent the p-n junction short circuit electrically. Then, Au layer is coated on the dielectric layer with designed thickness. Optimum multilayer structure above mentioned will benefit the throughput of the VSAL. Corresponding to the active region of LD, the nano-apertures are located exactly and etched carefully by FIB method. The SEM images of the VSALs with a 100 300nm rectangular aperture and an L-aperture with the same size in Fig.1 are respectively shown in Fig.3. Figure 3: SEM images of a rectangular aperture VSAL and an L-aperture VSAL Measurement and Characterization The near-field distribution from nano-aperture primarily consists of evanescent component. Optical distributions from the nano-apertures based on the glass substrate and VSALs are probed with a modified SNOM with an apertured fiber probe. The experimental setup has been described in the previous paper[9]. Nano-aperture Measurement As shown in Fig.4, the optical distributions corresponding to various apertures in Fig.2 are measured. It is noticeable that the optical distribution from the C-aperture is confined as an integrated spot, independent of the aperture shape. But there is a tail in every spot, which is a kind of artifact and probably is caused by the imperfect fiber probe. To compare clearly the optical field of C-aperture and square aperture with the

374 Progress In Electromagnetics Research Symposium 2005, Hangzhou, China, August 22-26 same unit length, intensity analyses are performed along lines I and II across the apertures. It reveals that the light spot sizes from the C-apertures and square apertures with same unit length are comparable in size, but both are expanded comparing with the physical size. One important reason is that they are probed at 1µm separation beyond the near field, and the other reason is due to the scattering of the probe tip. The intensity maximum from the C-aperture is about 10 times higher than the square aperture. It is hard to only attribute to the increase of the aperture area instead of field enhancement effect. Although the experimental data is not exactly consistent with the theoretical result, the enhancement effect still is distinct. It can be deduced that the enhancement effect should be stronger in near field for the rapid decay of evanescent field. Intensity maximum / a.u. 7 6 5 4 3 2 1 0 U=100nm I Square aperture II C-aperture U=200nm -1 0 2000 4000 6000 8000 10000 12000 14000 Transversal line I or II /nm U=300nm Figure 4: The optical field distributions at 1µm separation (1 and 4: M=100nm, 2 and 5: M=200nm, 3 and 6: M=300nm); Field intensity in cross section along white lines I and II in. Figure 5: P-I performance curves of VSALs with rectangular aperture and L-aperture VSAL Measurement Firstly, emission power from the VSAL is measured using a laser power meter in far field. The P-I performance curves of the VSALs with rectangular and L-aperture are shown in Fig.5a and b respectively. Y-coordinate P is power output in µw from the VSAL and x-coordinate I is drive current in ma. The background power from the back facet of the VSAL and other has been subtracted by a 40 microscopic objective and a spatial filter, so the measured power is the net emission power from the nano-aperture of the VSAL. From the P-I curves, it can be concluded that the two VSALs both work well in the laser oscillation state and the threshold currents are respectively about 15mA and 13mA for rectangular aperture VSAL and L-aperture VSAL. For rectangular aperture VSAL, when the drive current is 29mA, the power output only reaches 0.078µW. While for L-aperture VSAL, when the drive current is 28mA, the power output can reach 1.049µW. The power output from the L-aperture VSAL is increased about 13 times than the rectangular aperture VSAL, from which we can deduce that in near field the L-aperture VSAL can have higher power than the rectangular aperture VSAL. It indicates that the L-aperture is of stronger field enhancement effect and higher power output or power throughput. Secondly, the optical distributions from the VSALs are measured using the SNOM. As shown in Fig.6, VSAL with a rectangular aperture is measured in near field and at 1.7µm distance apart from the light exit facet respectively. According to the measurement results, the near-field spot size is 250 180nm, and the spot size becomes 580 418nm when measured at 1.7µm separation.

Progress In Electromagnetics Research Symposium 2005, Hangzhou, China, August 22-26 375 (c) Figure 6: Optical distributions of the rectangular aperture VSAL in Fig. 3a. and is near-field result; (c) and (d) is 1.7µm separation result. The optical distribution of the VSAL with an L-aperture is shown in Fig.7. It is measured at 1.7µm distance apart from the light exit facet. According to the experimental results, the spot size is 752 748nm. Due to some reasons, the near-field result cannot be measured, but the spot size should become much smaller than 1.7µm separation. (d) Figure 7: Optical distribution of the L-aperture VSAL in Fig. 3b. (Measurement distance is 1.7µm.) The above experimental results based on the square and C-aperture on the glass substrate and rectangular and L-aperture on VSALs display an uptrend of optical field intensity with C-aperture than the square aperture and higher power output and power throughput of L-aperture VSAL than rectangular aperture VSAL. Although the experimental data do not agree exactly with the theoretical results, there are several possible reasons to be considered. The first is that perfect Au film is used in the theoretical calculation, while in the experiment the Au film may be non-uniform and grainy. The second is that the aperture shape is not so perfect as in the theoretical simulation due to the fabrication deviation of the apertures. The third is that the measurement distance is difficultly controlled in the experiments so that the difference between the measurement separation and theoretical separation exists. The last one is the perturbation to the optical field exist as the probe enters the evanescent field and it is of complicated physical mechanism.

376 Progress In Electromagnetics Research Symposium 2005, Hangzhou, China, August 22-26 Conclusion In this paper, besides a rectangular aperture, an L-aperture with optical resolution 15nm and strong field enhancement is designed and fabricated on VSALs, and they can work well over the threshold current. The confined optical distributions from the VSALs and square and C-apertures in Au film on glass substrate are directly measured using an apertured SNOM. The experimental results quantificationally indicate that the unconventional C-aperture and L-aperture have stronger intensity and throughput than the conventional square aperture or rectangular aperture with a comparable spot size. Acknowledgement The research is supported by the Hi-Tech research and development program of China (863 program), Project No.2003AA311132 and the key basic research foundation of Tsinghua University. REFERENCES 1. Partovi, A., D. Peale, M. Wuttig, et al, Appl. Phys. Lett., Vol. 75, No. 11, 1515-1517, 1999. 2. Shi, X. L., R. L. Thornton, et al., ODS2001/Proc. SPIE., Vol. 4342, 320-326, 2002. 3. Sendur, K., W. Challener, Journal of Microscopy, Vol. 210, 279-283, 2003. 4. Tanaka, K., M. Tanaka, Journal of Microscopy, Vol. 210, 294-300, 2003. 5. Thio, T., K. M. Pellerin, et al, Optics Letters, Vol. 26, No. 24, 1972-1974, 2001. 6. Chen, F., A. Itagi, et al., Appl. Phys. Lett., Vol. 83, No. 16, 3245-3247, 2003. 7. Shinada, S., J. Hashizume and F. Koyama, Appl. Phys. Lett., Vol. 83, No. 5, 836-838, 2003. 8. Xu, Jiying, Tiejun Xu, Jia Wang, Qian Tian, Optical Engineering, Vol. 44, No. 1, 018001, 2005. 9. Xu, Tiejun, Jiying Xu, Jia Wang, Qian Tian, Chinese Physics Letters, Vol. 21, No. 8, 1644-1647, 2004.