We report on the fabrication of three-dimensional (3D) high-Q whispering gallery microcavities on a fused silica chip by femtosecond laser microfabriction, enabled by the 3D nature of femtosecond laser direct writing. The processing mainly consists of formation of freestanding microdisks by femtosecond laser direct writing and subsequent wet chemical etching. CO2 laser annealing is followed to smooth the microcavity surface. Microcavities with arbitrary tilting angle, lateral and vertical positioning are demonstrated, and the quality (Q)-factor of a typical microcavity is measured to be up to 1.07 × 106, which is currently limited by the low spatial resolution of the motion stage used during the laser patterning and can be improved with motion stages of higher resolutions.
© 2012 OSA
Whispering gallery mode (WGM) optical microcavities, that trap light via the total internal reflection at the circular boundary formed between the dielectric cavity and surroundings, exhibit very high quality (Q)-factors and very small volume. These excellent properties lend them very attractive for a variety of fields, ranging from fundamental science to engineering applications . Due to the nature of WGM, the light emission from the cavity is mainly in the plane parallel to the optical mode, i. e., for the on-chip WGM microcavities such as microdisks , microtoroids , and deformed microcavities  fabricated by lithography method, parallel to the substrate. However, this makes an obstacle for many applications that prefer out-of-plane light coupling or extraction. Though Levi et. al. had already started an early approach  shortly after their demonstration of the first microdisk lasers, till now, the realization of out-of-plane light output from microdisk lasers is still limited to vertical emission by adding grating structures either at the boundaries  or on the top surfaces  of microdisks.
Recently, femtosecond laser micromachining has been proved as a promising solution for high-precision and flexible fabrication of three dimensional (3D) microstructures, such as microoptics [7–8], microfluidics [9–11], optical waveguides [12–14], microbiochips [15–18] and polymer-based microcavities [19–21]. In comparison with polymer, fused silica is considered to be a very attractive substrate due to its wide transparency range and extremely low intrinsic material absorption loss . 3D microoptics in fused silica can be fabricated by femtosecond laser direct writing, followed by wet chemical etching and a postannealing method .
In this article, by applying femtosecond laser direct writing method, we demonstrate a new way to realize 3D high-Q microcavities on fused silica wafer that may have the light output from optical mode on an arbitrary plane respect to the substrate plane. Either the tilting angles or the heights of the microcavities are free of limitations, which, to the best of our knowledge, cannot be realized with any planar lithographic fabrication techniques. After the CO2 laser annealing process, the Q-factor of a microcavity is measured to be above 106, which can be further improved by replacing the low-resolution (~1 μm) motion stage used in the process of femtosecond laser fabrication with a better one, as explained in details in main context.
In this work, commercially available fused silica glass substrates (UV grade fused silica JGS1 whose upper and bottom surfaces are polished to optical grade) with a thickness of 1 mm are used. The process flow for fabrication of the high-Q microcavity mainly consists of two steps: (1) femtosecond laser exposure followed by selective wet etching of the irradiated areas to create the microdisk structures; and (2) selective reflow of the silica cavities by CO2 laser annealing to improve the quality factors, as illustrated in Fig. 1 . The laser system consists of a Ti: sapphire oscillator (Coherent, Inc.) and a regenerative amplifier, which emits 800 nm, ~58 fs pulses with maximum pulse energy of ~5 μJ at 250-kHz repetition rate. The initial 8.8-mm-diameter beam is reduced to 5 mm by passing through a circular aperture to guarantee a high beam quality. Power adjustment is realized using neutral density (ND) filters. The glass samples can be arbitrarily translated in 3D space with a resolution of 1 μm by a PC-controlled XYZ stage. In the femtosecond laser direct writing, a 100 × objective with a numerical aperture (NA) of 0.9 is used to focus the beam down to a ~1 μm-dia. spot, and the average femtosecond laser power measured before the objective is ~0.05 W. To form the microdisk supported by a thin pillar, a layer-by-layer annular scanning method with the lateral scanning step set to be 1 μm is adopted, and the femtosecond laser scanning speed is chosen to be ~600 μm/s. The scanning is designed to modify the regions surrounding the areas which form a disk with a radius of 29 μm, a thickness of ~7 μm, tilted with respect to the substrate at 24°, and an underneath pillar with a radius of 12 μm, as the modified regions will be preferentially etched away in hydrofluoride (HF) acid. In addition, the angle between the disk and the pillar is set to be 57°.
After the laser exposure, the sample is subjected to a ~20 min bath in a solution of 5% HF diluted with water, until the laser-modified part is etched away, leaving an on-chip microdisk structure [Fig. 2(a) ]. To facilitate the further coupling of light into the microtoroid with a fiber taper, the microdisk is located close to one corner of substrate, and both the adjacent sidewalls next to the microdisk are polished to have the edge of the micodisk exposed. To improve the surface smoothness and achieve the desirable high cavity Q factor, we smooth the surface by introducing surface-normal-irradiation with a CO2 laser (Synrad Firestar V30). The CO2 laser beam is focused by a lens to a circular spot approximately 100 µm in diameter. Because of the strong absorption of fused silica in the infrared band, the microdisk can be heated up by the laser to melting temperature which causes reflow. The laser is operated with a repetition rate of 5 kHz, and the time-averaged beam intensity is controlled by adjusting the duty ratio. As the disk diameter thermally shrinks during the reflow, surface tension induces a collapse of the fused silica disk, leading to a toroid-shaped boundary . During this process the disk is monitored by a charge coupled device (CCD) from the side with a 60 × objective lens. The total reflow process takes merely ~4 s, with a duty ratio of 5.0%. Due to the surface tension, the smoothness of the microtoroid surface is excellent, as shown in optical micrograph (Fig. 2(b)) and scanning electron microscope (SEM) image (Fig. 2(c)). The overall disk diameter is reduced to 43 μm with a 9-μm-thick toroid-shaped boundary. The freedom in controlling the lateral and vertical positions of the microcavities is demonstrated in Fig. 2(d), where two microtoroids are located close by yet at different heights. The heating CO2 laser beam is focused to a relatively small spot so that the reflow process on either of the microcavities will not affect the other. Thus the shape of each microtoroid can be controlled by adjusting the exposure time individually. The left microtoroid in Fig. 2(d) is achieved with 4-second longer CO2 laser annealing time in the reflow process than the right one in Fig. 2(d).
3. Results and discussion
To characterize the mode structure and Q factor of the microtoroidal cavity, resonance spectra are measured via the optical fiber taper coupling method . For facilitating a convenient coupling, a microtoroidal cavity parallel to the substrate with a diameter of ~39 μm and a thickness of ~9 μm is fabricated using the technique mentioned in Sec. 2, as shown by its SEM image in Fig. 3(a) . A swept-wavelength tunable external-cavity diode Laser (New Focus, Model: 6528-LN) and a swept spectrometer (dBm Optics, Model: 4650) are used to measure the transmission spectrum from the fiber taper with a resolution of 0.1pm. A periodic pulse signal with a power of 1 dBm is used to continuously sweep from 1530 to 1565 nm. The fiber taper formed by heating and stretching a section of a commercial optical fiber (Corning, SMF-28) has a minimum waist diameter of approximately ~1 μm, providing an evanescent excitation of WGMs of the cavity. The microcavity sample is fixed on a three-axis nanopositioning stage with a spatial resolution of 50 nm in the XYZ directions, so that the critical coupling may be realized by carefully adjusting the relative position between the cavity and the fiber taper. We use dual CCD cameras to simultaneously image microcavity and fiber taper from the side and the top, as shown in Fig. 3(b) [side view image not shown]. Please note that the thinnest portion of the fiber taper (dia. ~1 μm) is not very clear in Fig. 3(b) because of the limited resolution of the optical microscope (a 50X objective with a NA of 0.3 is used for the imaging in this case to obtain a sufficiently long working distance).
A resonance transmission spectrum of a fiber taper coupled to the microtoroidal cavity with various excited WGMs is depicted in Fig. 3(c). The experimentally measured free spectral range (ΔλFSR = 13.65 nm, defined as the wavelength spacing between modes with successive angular mode number) agrees well with the numerical calculation based on the experimentally measured cavity diameter, which is given approximately by the well-known expression ,Figure 3(d) shows an individual WGM located at 1534.72 nm with a Lorentzian shaped dip. The linewidth is measured as 1.44 pm, and the Q factor for the mode is calculated to be 1.07 × 106, inferred from the Lorentzian fit of the spectrum, as shown in Fig. 3(d). This indicates that femtosecond laser micromachining on fused silica enables fabrication of smooth cavity surfaces with low surface-scattering loss of the WGMs..
It is noteworthy that microtoroids reported in previous publications can frequently achieve ultra-high-Q in the order of over 108 , which is two orders of magnitude higher than the Q achieved in the current experiment. We believe that the major reason for this relatively lower-Q measured in our case is the imperfection in the fabrication of the microdisk with our motion stage of limited resolution (~1 μm) prior to the annealing by the CO2 laser. As we have seen in Fig. 2(a), the edge of the microdisk shows some irregular features prior to the annealing by the CO2 laser. After the reflow process, though the small scale roughness has been removed, as we can see very smooth surface for all the samples under SEM with high magnifications, on some of the microcavities we have observed clear large scale deformation so that boundary of the cavities deviates from an ideal circle. Such deformation is introduced by the non-uniformity of the disk radius or/and thickness after wet chemical etching. This problem can be solved by replacing the current motion stage with a higher resolution one. Actually for many applications (i. e., on-chip microcavity lasers or some sensors), the currently achieved Q value in the order of 106 will already be sufficient.
In the current experiment, it took ~6 hours for fabricating the microcavity by femtosecond laser direct writing. This long fabrication time is mostly due to the low performance of our X-Y-Z motion stage. If a rotary stage could be used, we expect that the fabrication time can be greatly shortened to less than ~30 min. Still, our technique is slower than the traditional photolithography-based technology, however, it should be stressed here that it is very difficult, if not completely impossible, to realize such flexible 3D microcavities presented here via planar lithography method such as photolithography or electron beam lithography. Though the microtoroids still exhibit isotropic properties of the confined WGM, the femtosecond laser direct writing can easily create preferred deformation into the cavity and thus introduce directional light output into any desired solid angle. This would open a broad spectrum of applications such as micro-lasers, sensors, and so on.
To summarize, we demonstrate the fabrication of 3D microcavities on the fused silica chip by femtosecond laser direct writing. The typical Q factors of the microcavities are over 106. Further improvement of the Q-factor should be achievable by replacing the XYZ stage of relatively low resolution in the current writing system with a high-resolution XYZ stage. Though the cavity shown here is a passive device, with such high Q factors, lasing can be easily achieved by coating the cavity surface with gain medium such as semiconductor colloidal quantum dots, or by using rare-earth-doped substrate to provide optical gain. Therefore, our technique opens a new avenue for constructing high-Q microcavities with non-inplane geometries which may find important uses in both fundamental research and biological and chemical sensing applications.
We thank Prof. Yunfeng Xiao of Peking University, Dr. Xiaoshun Jiang of Nanjing University for the helpful discussion. The work is supported by National Basic Research Program of China (No. 2011CB808100), NSFC (Nos. 11134010, 60825406, 61008011, 10974213, 11104245) and Fundamental Research Funds for the Central Universities.
References and links
1. V. S. Ilchenko and A. B. Matsko, “Optical resonators with whispering-gallery modes—part II: applications,” IEEE J. Sel. Top. Quantum Electron. 12(1), 15–32 (2006). [CrossRef]
2. S. L. McCall, A. F. J. Levi, R. E. Slusher, S. J. Pearton, and R. A. Logan, “Whispering-gallery mode microdisk lasers,” Appl. Phys. Lett. 60(3), 289–291 (1992). [CrossRef]
4. C. Gmachl, F. Capasso, E. E. Narimanov, J. U. Nöckel, A. D. Stone, J. Faist, D. L. Sivco, and A. Y. Cho, “High-power directional emission from microlasers with chaotic resonators,” Science 280(5369), 1556–1564 (1998). [CrossRef] [PubMed]
5. A. F. J. Levi, R. E. Slusher, S. L. McCall, J. L. Glass, S. J. Pearton, and R. A. Logan, “Directional light coupling from microdisk lasers,” Appl. Phys. Lett. 62(6), 561–563 (1993). [CrossRef]
6. L. Mahler, A. Tredicucci, F. Beltram, C. Walther, J. Faist, B. Witzigmann, H. E. Beere, and D. A. Ritchie, “Vertically emitting microdisk lasers,” Nat. Photonics 3(1), 46–49 (2009). [CrossRef]
7. Y. Cheng, H. L. Tsai, K. Sugioka, and K. Midorikawa, “Fabrication of 3D microoptical lenses in photosensitive glass using femtosecond laser micromachining,” Appl. Phys., A Mater. Sci. Process. 85(1), 11–14 (2006). [CrossRef]
8. E. Brasselet, M. Malinauskas, A. Žukauskas, and S. Juodkazis, “Photopolymerized microscopic vortex beam generators: precise delivery of optical orbital angular momentum,” Appl. Phys. Lett. 97(21), 211108 (2010). [CrossRef]
9. A. Marcinkevicius, S. Juodkazis, M. Watanabe, M. Miwa, S. Matsuo, H. Misawa, and J. Nishii, “Femtosecond laser-assisted three-dimensional microfabrication in silica,” Opt. Lett. 26(5), 277–279 (2001). [CrossRef] [PubMed]
10. Y. Bellouard, A. Said, M. Dugan, and P. Bado, “Fabrication of high-aspect ratio, micro-fluidic channels and tunnels using femtosecond laser pulses and chemical etching,” Opt. Express 12(10), 2120–2129 (2004). [CrossRef] [PubMed]
13. M. Ams, G. Marshall, D. Spence, and M. Withford, “Slit beam shaping method for femtosecond laser direct-write fabrication of symmetric waveguides in bulk glasses,” Opt. Express 13(15), 5676–5681 (2005). [CrossRef] [PubMed]
14. F. He, J. Lin, and Y. Cheng, “Fabrication of hollow optical waveguides in fused silica by three-dimensional femtosecond laser micromachining,” Appl. Phys. B 105(2), 379–384 (2011). [CrossRef]
15. K. Sugioka and Y. Cheng, “Integrated microchips for biological analysis fabricated by femtosecond laser direct writing,” MRS Bull. 36(12), 1020–1027 (2011). [CrossRef]
16. Y. Hanada, K. Sugioka, I. Shihira-Ishikawa, H. Kawano, A. Miyawaki, and K. Midorikawa, “3D microfluidic chips with integrated functional microelements fabricated by a femtosecond laser for studying the gliding mechanism of cyanobacteria,” Lab Chip 11(12), 2109–2115 (2011). [CrossRef] [PubMed]
17. A. Crespi, Y. Gu, B. Ngamsom, H. J. W. M. Hoekstra, C. Dongre, M. Pollnau, R. Ramponi, H. H. van den Vlekkert, P. Watts, G. Cerullo, and R. Osellame, “Three-dimensional Mach-Zehnder interferometer in a microfluidic chip for spatially-resolved label-free detection,” Lab Chip 10(9), 1167–1173 (2010). [CrossRef] [PubMed]
19. Z.-P. Liu, Y. Li, Y.-F. Xiao, B.-B. Li, X.-F. Jiang, Y. Qin, X.-B. Feng, H. Yang, and Q. Gong, “Direct laser writing of whispering gallery microcavities by two-photon polymerization,” Appl. Phys. Lett. 97(21), 211105 (2010). [CrossRef]
21. T. Grossmann, S. Schleede, M. Hauser, T. Beck, M. Thiel, G. von Freymann, T. Mappes, and H. Kalt, “Direct laser writing for active and passive high-Q polymer microdisks on silicon,” Opt. Express 19(12), 11451–11456 (2011). [CrossRef] [PubMed]