Open Access
Add to BibSonomy
Abstract
Micro- and nano-optomechanics has attracted broad interest for applications of mechanical sensing and coherent signal processing. For nonpiezoelectric materials such as silicon or silicon nitride, electrocapacitive effects with metals patterned on mechanical structures are usually adopted to actuate the mechanical motion of the micro- or nanomechanical devices. However, the metals have deleterious effects on the mechanical structures because they add an additional weight and also introduce considerable mechanical losses. To solve these problems, we have proposed and experimentally demonstrated a new scheme of electro-optomechanical integration on a silicon-on-insulator platform by using single-layer graphene as a highly conductive coating for electromechanical actuation. Mechanical modes of different groups were electrically actuated and optically detected in a micromechanical resonator, with the mechanical Q > 1000 measured in air. Compatible with CMOS technology, our scheme is suitable for large-scale electro-optomechanical integration and will have wide applications in high-speed sensing, communication, and signal processing.
© 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Cavity optomechanics provides an efficient platform for coupling mechanical and optical degrees of freedom [1], which has found a broad range of applications, such as coherent information processing [2,3], acceleration sensing [4], and mass sensing [5]. Most of these applications require efficient actuation of mechanical motion. Mechanical motion can be actuated by electrical, optical, and thermal methods. Among the three, electrical methods are the most popular because optical forces are intrinsically weak and thermal forces have limited frequency range. By electrical methods, the mechanical motion can be actuated by the piezoelectric effect for structures made in piezoelectric materials [6,7], or by electrocapacitive force between deposited metal films for structures made in nonpiezoelectric materials [8,9]. However, deposition of metals on micro- and nanoscale devices degrades the mechanical performance significantly. For example, a 20-nm-thick metal film results in degradation of the quality factor of a Si3N4 micromechanical resonator from 2000 to 500 [10], which is a disadvantage for applications that require a high mechanical quality factor such as nonlinear phononics and integrated phononic circuits. One way to avoid metals in mechanical devices made in nonpiezoelectric materials is ion implantation in selective regions [11,12]. However, this method only applies to semiconductor materials but not dielectrics such as Si3N4 and diamond, which are excellent mechanical materials due to low mechanical loss [13,14] but have limited applications in electromechanics due to insulating nature. Therefore, it is highly desired to develop new actuation schemes independent of the substrate materials for broader applications.
Graphene has high electron mobility, high Young’s modulus, and low mass [15,16]. These properties make it an excellent material for electromechanical integration. The excellent electrical conductivity of graphene has been exploited in various graphene–dielectric structures to obtain label-free sensors [17], photodetectors [18], and photoconductive antennas [19]. For micro- and nanoelectromechanics, the high electron mobility of graphene promises large electrocapacitive force, and its single-atom thickness and low mass ensure minimal influence on the performance of the mechanical structure [20]. It has been proven effective to use graphene for mechanical actuation in microelectromechanical systems with high mechanical Q factors [21,22]. Here, we introduced for the first time to our knowledge single-layer graphene onto an undoped silicon electro-optomechanical system for effective mechanical actuation. A hybrid graphene/Si micromechanical resonator has been experimentally demonstrated with graphene-assisted electromechanical actuation and optomechanical detection on a silicon-on-insulator (SOI) platform. The single-layer graphene plays a crucial role in electrically actuating the mechanical modes of the micromechanical resonator, with little effect on the quality factors and frequency values of the mechanical modes.
2. Device description
Figure 1 shows the schematic of the hybrid graphene/Si electro-optomechanical resonator. The device consists of two suspended microwheel resonators connected by a nanowire: the right one with coated graphene is an electromechanical resonator for actuating the mechanical modes, and the left one is an optomechanical resonator for detecting the mechanical modes. These two microwheel resonators have the same geometry. A single layer of graphene was applied onto the right microwheel resonator. The graphene was patterned such that it has substantial contact with the S electrode for optimal electrical conduction. The graphene on the microwheel and the surrounding metal pads connected to the G electrodes formed a capacitor. When a combination of ac and dc voltages (Vac and Vdc) was applied to the capacitor, an electrocapacitive force F = $\frac{\textrm{1}}{\textrm{2}}$ (dC/dg)(Vac + Vdc)2 was produced between the microwheel and surrounding metal, where C is the capacitance and g is the width of gap between the patterned graphene and the surrounding metal pads. Under the condition of Vac ${\ll} $ Vdc, the actuation force (dC/dg)VacVdc produced the mechanical motion of the right microwheel resonator, which excited the same mechanical mode of the left optomechanical resonator due to the mechanical coupling through the connecting nanowire. The excited mechanical mode was then detected by an optical method. We sent a probe light beam via a grating coupler into the optical waveguide with its optical frequency tuned to one of the cavity resonances of the optomechanical resonator. By optomechanical coupling, the mechanical vibration was transduced into an intensity modulation of the probe light, which was coupled out of the chip and then detected by a photodetector.
Fig. 1. Schematic of the graphene-coated microwheel electromechanical resonator (right) coupled to an optomechanical resonator (left). The graphene on the electromechanical resonator is of a single layer, and is connected to the S electrode via a microribbon. A combination of dc and ac voltage is applied to the electrodes to actuate the mechanical motion of the electromechanical resonator, which transfers to the optomechanical resonator for detection.
3. Device fabrication
The devices were fabricated on a standard SOI wafer (220-nm high-resistivity silicon device layer on 3-µm buried oxide) with CMOS-compatible processes as shown in Fig. 2. First, the patterns of the double microwheels with the optical waveguides and grating couplers were defined by high-resolution electron-beam lithography in ZEP520A, and then transferred to the top silicon device layer by plasma dry etching. Second, another step of electron-beam lithography in PMMA followed by gold deposition and lift-off processes were adopted to make the electrodes. Third, a single layer of chemical-vapor-deposited graphene was transferred onto the substrate, and patterned by an additional step of electron-beam lithography in PMMA and oxygen plasma ashing. Then, the oxide underneath the double microwheels was removed by a buffered oxide etchant to release the microwheels from the substrate. Finally, the devices were dried in a critical point dryer to prevent stiction. To maximize the optical quality factors of the microwheel resonators, we optimized the electron-beam lithography and dry etching processes. In electron-beam lithography, the electron beam for exposure was carefully adjusted and the dosage was well controlled to minimize roughness at edges of the patterns. Chlorine-based chemistry was adopted and the recipe was fine tuned for plasma dry etching to transfer the patterns in the ZEP520A resist to the silicon device layer with smooth sidewalls [23].
4. Device characterization
Figure 3(a) presents a scanning electron microscope (SEM) image of an entire fabricated device, which includes the optical waveguide, grating couplers, double microwheels, and electrodes. The optical waveguide is suspended and anchored to the substrate with tethers so that the gap between the waveguide and the optomechanical resonator can be precisely controlled at ∼200 nm. Figure 3(b) presents a zoomed-in image of the double microwheel resonators. The outer and inner radii of the microwheels were designed to be 15 and 13 µm, respectively. Figure 3(c) presents a zoomed-in image of the nanowire connecting the two microwheels, which has dimensions of 200 nm width and 2 µm length. The gap between the patterned graphene on the electromechanical resonator and the surrounding metal pads is 500 nm, whose value was chosen as a balance between a high efficiency of electromechanical actuation and a large fabrication tolerance for misalignment during electron-beam lithography for different layers. The graphene on the electromechanical resonator is not visible in the images due to its low contrast with the underlying silicon layer.
Fig. 3. (a) Top-view scanning electron microscope (SEM) image of an entire fabricated device. (b) Close-up of the coupled double microwheels. (c) Close-up of the nanowire connecting the electromechanical resonator and optomechanical resonator. (d) Experimental setup: VOA, variable optical attenuator; FPC, fiber polarization controller; DUT: device under test; PD, photodetector; DAC, data acquisition card; DC: dc power supply; bias-T: bias tee; RF-PA: radio-frequency power amplifier; RF-SA: radio-frequency signal amplifier; VNA: vector network analyzer.
Figure 3(d) shows the experimental setup for device characterization. Light from a tunable semiconductor laser was attenuated, polarization-adjusted, and then coupled into the device under test via a grating coupler. Light coupled out of the optomechanical resonator was collected by a fiber and then split into two paths by a 99/1 fiber splitter. The path with 1% of optical power was used for monitoring the optical transmission and controlling the laser wavelength. The other path with 99% of optical power was sent to a high-speed photodetector. The driving ac voltage (Vac) from a vector network analyzer (VNA) was amplified, combined with a dc bias (Vdc), and then applied to the electrodes of the device under test to actuate the mechanical motion. The mechanical signal was transduced by the optomechanical resonator into the optical domain, received by the high-speed photodetector, and then sent back to the VNA.
Our optimized fabrication processes rendered high optical Q factors of optomechanical resonators. For a single microwheel resonator shown in the inset of Fig. 4(a), we obtained a loaded optical Q factor of 8.5 × 105 near the critical-coupling condition [Fig. 4(a)] and over 1.0 × 106 in an under-coupling condition [Fig. 4(b)]. The optical Q factor obtained near the critical-coupling condition is among the highest experimental values reported to date for disk, wheel, or ring microcavities fabricated on 220-nm SOI platform without additional thermal processes [24–30]. Figures 4(c) and 4(d) show the optical transmission spectra of a double-microwheel resonator. Figure 4(c) shows a cavity resonance of the optomechanical resonator before graphene transfer and underlying-oxide removal, revealing a loaded optical Q factor of 1.2 × 105. Compared with the single-microwheel device in Fig. 4(a), the major loss of the double-microwheel resonator is from optical scattering due to the nanowire connecting the two microwheels. Figure 4(d) presents a cavity resonance after graphene transfer, patterning, and underlying-oxide removal. Because the oxide beneath the grating coupler and waveguide was also removed, the working wavelength of the grating couplers shifted by ∼100 nm, and the extinction ratio of the cavity resonance dropped. Only one family of cavity resonances could be measured, with a typical loaded optical Q factor of 9.0 × 103. The additional optical loss of device after all fabrication processes is due to contamination during the graphene transfer and surface damage during the graphene patterning.
Fig. 4. (a) Normalized optical transmission spectrum of a single microwheel resonator with radius of 20 µm, showing an optical resonance with loaded optical Q factor as high as 8.5 × 105 and ~20 dB extinction ratio. Inset: top-view SEM image of the measured resonator (scale bar: 10 µm). (b) An undercoupled optical resonance of the device in (a) showing a loaded optical Q factor over 1 million. (c) Normalized optical transmission spectrum of a double-microwheel resonator before the step of graphene transfer. Inset: top-view SEM image of the measured resonator (scale bar: 10 µm). (d) An optical resonance of the device in (c) after the steps of graphene transfer, patterning, and underlying-oxide removal. In all the plots, the blue dots represent the experimental data, and the red lines are the corresponding Lorentzian fits.
By using the optical resonance in Fig. 4(d), we measured the mechanical response of the device to electrical drive in the ambient condition, with the results shown in Figs. 5(a) and 5(b). It should be noted that the graphene played a dominant role in the electrocapacitive actuation, because the resistance of graphene was measured to be at least 5 orders of magnitude lower than that of silicon with the same in-plane geometry. The S21 spectra in Fig. 5(a) show that increasing the dc voltage Vdc from 0 to 25 V leads to substantial enhancement of the signal-to-noise ratio (e.g., 20 dB enhancement for mode No. 14). Most of the measured mechanical modes could be identified by finite-element simulation as the in-plane mechanical modes, including the stretching modes (No. 1–5 and 7–9), wine-glass modes of different orders (No. 10–13 and 16–18), and radial-contour modes (No. 14 and 15). Due to the mechanical coupling by the nanowire connecting the two microwheels, the originally degenerate radial-contour mode splits into an anti-phase mode at 91.4 MHz [No. 14, see Fig. 5(c)] and an in-phase mode at 93.5 MHz [No. 15, see Fig. 5(c)]. According to the finite-element simulation, the other mechanical modes (No. 6 and 19–26) in the measured S21 spectrum are out-of-plane modes. Although it is difficult to verify their modal profiles experimentally by using the integrated optomechanical detection method in this work, other approaches such as free-space optical interferometry can be employed for this purpose [31].
Fig. 5. (a) Electrical transmission S21 spectra showing the mechanical resonances of the graphene-coated double-microwheel resonator (Device 1). (b) Zoomed-in spectra of the mechanical resonances in (a) (scale bar: 1 MHz). The blue lines are the spectral traces collected in the experiment, and the magenta lines are the Lorentzian fits for each mechanical resonance. (c) Simulated displacement fields of some representative mechanical modes with their resonant frequencies.
The stretching modes suffer from higher mechanical loss with their mechanical Q factors ranging from 80 to 300, while the other types of mechanical modes have Q factors ranging from 250 to 600. The dominant mechanical loss of the device is the clamping loss at the graphene microribbon connecting the electromechanical resonator and the S electrode, and thus is highly dependent on the geometry of the graphene microribbon. It should be noted that the above measured mechanical Q factors were obtained from a device with a relatively wide graphene microribbon (4 µm width), which was designed to achieve a high device fabrication yield. By reducing this width to 2 µm, the measured mechanical Q factor of the radial-contour mode increased beyond 1000 [see the results from Device 2 in Fig. 6(a)], which is close to the Q value (1300) obtained from a similar silicon double-disk microresonator in [11]. Figure 5(c) also shows that the modal frequencies measured from the device are close to their numerical values obtained from simulation without including the graphene in the model. Therefore, with a proper design, adopting graphene as a conductor on a micromechanical resonator for electromechanical actuation would not affect the mechanical performance.
Fig. 6. (a) Electrical transmission S21 spectrum showing the radial-contour mode with a mechanical Q factor of 1001 (Device 2). The red line is a Lorentzian fit of the measured data. (b) Normalized mechanical intensity of the actuated radial-contour mode under different dc voltages Vdc (Device 3). (c) Electrical transmission S21 spectra showing the mechanical response of the radial-contour mode under different dc voltages Vdc (Device 3).
Figure 6(b) shows the measured mechanical intensity (blue squares) of the radial-contour mode under different dc voltages Vdc of another device (Device 3). As mentioned in Sec. 2, the actuation force is proportional to Vdc, yielding a quadratic dependence of the mechanical intensity on Vdc. The red line is a quadratic fit of the measured data, showing very good agreement with the theoretical prediction. Figure 6(c) shows the spectral evolution of the mechanical response of the radial-contour mode under different dc voltages Vdc. As Vdc is increased up to 28 V, the mechanical resonance maintains a Lorentzian line shape, and the mechanical frequency does not shift. It is clear that a higher dc voltage helps to improve the efficiency of electro-optomechanical transduction.
5. Conclusion
In conclusion, we have proposed and experimentally demonstrated graphene-assisted electro-optomechanical integration on a silicon-on-insulator platform. More specifically, we have demonstrated for the first time to our knowledge a hybrid graphene/silicon electro-optomechanical resonator that relies on a single layer of graphene for electromechanical actuation. In contrast to the conventional scheme where metals are used as the conductor for electromechanical actuation, graphene has its intrinsic advantages such as high electron mobility, high Young’s modulus, and low mass, which enable large electrocapacitive force and ensure minimal influence on the performance of the mechanical structure. By using optomechanical detection method, we have experimentally observed various types of the actuated mechanical modes in the very high frequency range from our hybrid graphene/silicon electro-optomechanical resonator, with the highest mechanical Q factor beyond 1000 measured in the ambient condition. Although high-resistivity silicon was adopted as the platform for demonstrating the role of graphene in electro-optomechanical integration, this technique can also be applied to other material platforms, such as Si3N4, germanium, and diamond. It is envisioned that the graphene-assisted electro-optomechanical integration method shall find wide applications in integrated optoelectronic, optomechanical, and phononic circuits for high-speed sensing [32,33], telecommunication [34], and signal processing [35,36].
Funding
Research Grants Council of Hong Kong (14206318, 14208717, 24208915, N_CUHK415/15, AoE/P-02/12); CUHK Group Research Scheme; CUHK Impact Postdoctoral Fellowship.
Disclosures
The authors declare no conflicts of interest.
References
1. M. Aspelmeyer, T. J. Kippenberg, and F. Marquardt, “Cavity optomechanics,” Rev. Mod. Phys. 86(4), 1391–1452 (2014). [CrossRef]
2. J. T. Hill, A. H. Safavi-Naeini, J. Chan, and O. Painter, “Coherent optical wavelength conversion via cavity optomechanics,” Nat. Commun. 3, 1196 (2012). [CrossRef]
3. R. W. Andrews, R. W. Peterson, T. P. Purdy, K. Cicak, R. W. Simmonds, C. A. Regal, and K. W. Lehnert, “Bidirectional and efficient conversion between microwave and optical light,” Nat. Phys. 10(4), 321–326 (2014). [CrossRef]
4. A. G. Krause, M. Winger, T. D. Blasius, Q. Lin, and O. Painter, “A high-resolution microchip optomechanical accelerometer,” Nat. Photonics 6(11), 768–772 (2012). [CrossRef]
5. W. Yu, W. C. Jiang, Q. Lin, and T. Lu, “Cavity optomechanical spring sensing of single molecules,” Nat. Commun. 7, 12311 (2016). [CrossRef]
6. L. Fan, X. Sun, C. Xiong, C. Schuck, and H. X. Tang, “Aluminum nitride piezo-acousto-photonic crystal nanocavity with high quality factors,” Appl. Phys. Lett. 102(15), 153507 (2013). [CrossRef]
7. S. Ghosh and G. Piazza, “Piezoelectric actuation of aluminum nitride contour mode optomechanical resonators,” Opt. Express 23(12), 15477–15490 (2015). [CrossRef]
8. X. Sun, X. Zhang, M. Poot, C. Xiong, and H. X. Tang, “A superhigh-frequency optoelectromechanical system based on a slotted photonic crystal cavity,” Appl. Phys. Lett. 101(22), 221116 (2012). [CrossRef]
9. P. L. Yu, T. P. Purdy, and C. A. Regal, “Control of material damping in high-Q membrane microresonators,” Phys. Rev. Lett. 108(8), 083603 (2012). [CrossRef]
10. L. Sekaric, D. W. Carr, S. Evoy, J. M. Parpia, and H. G. Craighead, “Nanomechanical resonant structures in silicon nitride: fabrication, operation and dissipation issues,” Sens. Actuators A 101(1-2), 215–219 (2002). [CrossRef]
11. S. Sridaran and S. A. Bhave, “Electrostatic actuation of silicon optomechanical resonators,” Opt. Express 19(10), 9020–9026 (2011). [CrossRef]
12. X. Sun, K. Xu, and H. X. Tang, “Monolithically integrated, ultrahigh-frequency cavity nano-optoelectromechanical system with on-chip germanium waveguide photodetector,” Opt. Lett. 39(8), 2514–2517 (2014). [CrossRef]
13. C. Reinhardt, T. Müller, A. Bourassa, and J. C. Sankey, “Ultralow-noise SiN trampoline resonators for sensing and optomechanics,” Phys. Rev. X 6(2), 021001 (2016). [CrossRef]
14. B. Khanaliloo, H. Jayakumar, A. C. Hryciw, D. P. Lake, H. Kaviani, and P. E. Barclay, “Single-crystal diamond nanobeam waveguide optomechanics,” Phys. Rev. X 5(4), 041051 (2015). [CrossRef]
15. C. Lee, X. Wei, J. W. Kysar, and J. Hone, “Measurement of the elastic properties and intrinsic strength of monolayer graphene,” Science 321(5887), 385–388 (2008). [CrossRef]
16. A. K. Geim, “Graphene: status and prospects,” Science 324(5934), 1530–1534 (2009). [CrossRef]
17. F. Zangeneh-Nejad and R. Safian, “Hybrid graphene–molybdenum disulphide based ring resonator for label-free sensing,” Opt. Commun. 371, 9–14 (2016). [CrossRef]
18. X. Wang, Z. Cheng, K. Xu, H. K. Tsang, and J.-B. Xu, “High-responsivity graphene/silicon-heterostructure waveguide photodetectors,” Nat. Photonics 7(11), 888–891 (2013). [CrossRef]
19. F. Zangeneh-Nejad and R. Safian, “Significant enhancement in the efficiency of photoconductive antennas using a hybrid graphene molybdenum disulphide structure,” J. Nanophotonics 10(3), 036005 (2016). [CrossRef]
20. C. Chen and J. Hone, “Graphene nanoelectromechanical systems,” Proc. IEEE 101(7), 1766–1779 (2013). [CrossRef]
21. S. Lee, V. P. Adiga, R. A. Barton, A. M. van der Zande, G.-H. Lee, B. R. Ilic, A. Gondarenko, J. M. Parpia, H. G. Craighead, and J. Hone, “Graphene metallization of high-stress silicon nitride resonators for electrical integration,” Nano Lett. 13(9), 4275–4279 (2013). [CrossRef]
22. S. Schmid, T. Bagci, E. Zeuthen, J. M. Taylor, P. K. Herring, M. C. Cassidy, C. M. Marcus, L. Guillermo Villanueva, B. Amato, A. Boisen, Y. Cheol Shin, J. Kong, A. S. Sørensen, K. Usami, and E. S. Polzik, “Single-layer graphene on silicon nitride micromembrane resonators,” J. Appl. Phys. 115(5), 054513 (2014). [CrossRef]
23. W. C. Tian, J. W. Weigold, and S. W. Pang, “Comparison of Cl2 and F-based dry etching for high aspect ratio Si microstructures etched with an inductively coupled plasma source,” J. Vac. Sci. Technol. B 18(4), 1890–1896 (2000). [CrossRef]
24. M. Soltani, S. Yegnanarayanan, and A. Adibi, “Ultra-high Q planar silicon microdisk resonators for chip-scale silicon photonics,” Opt. Express 15(8), 4694–4704 (2007). [CrossRef]
25. S. Yegnanarayanan, W. Roman, M. Soltani, G. Cremona, H. Lu, and A. Adibi, “On-chip integration of microfluidic channels with ultra-high Q silicon microdisk resonators for lab-on-a-chip sensing applications,” in LEOS 2007 - IEEE Lasers and Electro-Optics Society Annual Meeting Conference Proceedings (2007), 50–51.
26. W. C. Jiang, J. Zhang, and Q. Lin, “Compact suspended silicon microring resonators with ultrahigh quality,” Opt. Express 22(1), 1187–1192 (2014). [CrossRef]
27. M. Borselli, T. J. Johnson, and O. Painter, “Beyond the Rayleigh scattering limit in high-Q silicon microdisks: theory and experiment,” Opt. Express 13(5), 1515–1530 (2005). [CrossRef]
28. A. Naiman, B. Desiatov, L. Stern, N. Mazurski, J. Shappir, and U. Levy, “Ultrahigh-Q silicon resonators in a planarized local oxidation of silicon platform,” Opt. Lett. 40(9), 1892–1895 (2015). [CrossRef]
29. A. Griffith, J. Cardenas, C. B. Poitras, and M. Lipson, “High quality factor and high confinement silicon resonators using etchless process,” Opt. Express 20(19), 21341–21345 (2012). [CrossRef]
30. A. Biberman, M. J. Shaw, E. Timurdogan, J. B. Wright, and M. R. Watts, “Ultralow-loss silicon ring resonators,” Opt. Lett. 37(20), 4236–4238 (2012). [CrossRef]
31. Z. Shen, X. Han, C.-L. Zou, and H. X. Tang, “Phase sensitive imaging of 10 GHz vibrations in an AlN microdisk resonator,” Rev. Sci. Instrum. 88(12), 123709 (2017). [CrossRef]
32. P. Weber, J. Güttinger, A. Noury, J. Vergara-Cruz, and A. Bachtold, “Force sensitivity of multilayer graphene optomechanical devices,” Nat. Commun. 7, 12496 (2016). [CrossRef]
33. F. Zangeneh-Nejad and R. Safian, “A graphene-based THz ring resonator for label-free sensing,” IEEE Sens. J. 16(11), 4338–4344 (2016). [CrossRef]
34. H. J. Kimble, “The quantum internet,” Nature 453(7198), 1023–1030 (2008). [CrossRef]
35. S. Tallur and S. A. Bhave, “A silicon electromechanical photodetector,” Nano Lett. 13(6), 2760–2765 (2013). [CrossRef]
36. F. Zangeneh-Nejad and R. Fleury, “Topological analog signal processing,” Nat. Commun. 10, 2058 (2019). [CrossRef]
