Study of Broadband Cryogenic DC-Contact RF MEMS ... - IEEE Xplore

3442

IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES, VOL. 57, NO. 12, DECEMBER 2009

Study of Broadband Cryogenic DC-Contact RF MEMS Switches Songbin Gong, Student Member, IEEE, Hui Shen, Member, IEEE, and N. Scott Barker, Member, IEEE

Abstract—A dielectric-free DC-contact RF microelectromechanical system (MEMS) switch is designed and tested at room temperature and cryogenic temperatures. The switch demonstrates a 1contact resistance and 2 fF up-state capacitance at room temperature, with an insertion-loss less than 0.4 dB up to 50 GHz and less than 0.9 dB up to 75 GHz. The isolation is better than 24 dB up to 50 GHz and 18 dB up to 75 GHz at room temperature. At a cryogenic temperature of 1.6 K, the switch has an insertion loss less than 0.6 dB with isolation better than 24 dB up to 50 GHz. The effects of cryogenic temperatures on deformation of the cantilever beam, actuation voltage, and RF performance have been noted. The theoretical and experimental results of the switch performance are presented and compared.

Index Terms—Broadband, cantilever, contact resistance, cryogenic, DC-contact, low loss, MEMS, series switch.

I. INTRODUCTION

R

ADIO frequency (RF) microelectromechanical systems (MEMS) have been vastly researched for the last decade due to their small size, superior RF performance and low power consumption over a broad band of frequencies. Numerous applications, including phase shifters [1], [2], tunable filters [3], and reconfigurable matching networks [4], [5], have all been demonstrated using RF MEMS devices. Recently, research interests in RF MEMS have been extended to their performance under cryogenic temperatures owing to their potential integration with superconducting materials [6]. Superconductors exhibit extremely low-loss that enables the fabrication of a variety of passive microwave components in a far more compact structure when compared with conventional materials [7]. DC-contact cryogenic MEMS with superconductors can provide low insertion-loss signal routing for switching networks while capacitive MEMS with superconductors can enable tunable high- resonant structures. MEMS integrated with a high temperature superconductor (HTS) microstrip resonator and a tunable HTS filter using MEMS have been demonstrated [8], [9]. Other cryogenic MEMS applications, including phase shifters for radio astronomy instrumentation, are currently being developed. This paper explores the design Manuscript received April 15, 2009; revised September 15, 2009. First published November 17, 2009; current version published December 09, 2009. This work was supported by NASA under contract NNG05GJ82G and NSF under contract ECCS-0238967. This work was also supported in part by the U.S. Army National Ground Intelligence Center under Grant DASC01–01–C-0009. S. Gong and N. S. Barker are with the Department of Electrical and Computer Engineering, University of Virginia, Charlottesville, VA 22904 USA (e-mail: [email protected]; [email protected]). H. Shen is now with Steptoe & Johnson LLP, Washington, DC 20036 USA. Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/TMTT.2009.2033872

and fabrication of DC-contact RF MEMS to address several challenges imposed by working at cryogenic temperatures. In the following sections, simulation and experimental results of the cryogenic DC-contact RF MEMS switch are presented. II. CRYOGENIC DC-CONTACT RF-MEMS SWITCH DESIGN It has been reported that fixed-fixed beam RF-MEMS experience a drastic increase in the actuation voltage, at a rate of 0.3–0.5 V C, as the operating temperature drops [10]. According to the theoretical model, the cause of this effect over temperature is primarily due to the difference in the thermal expansion coefficient between the mechanical beam and the substrate [11]. Over a broad range of temperatures, this can cause the actuation voltage to increase by an order of magnitude [12]. This is problematic for multiple reasons including greatly reduced device lifetime due to high voltage actuation [13]. Cantilever beams, on the other hand, have a free end and thus do not suffer from the stress due to expansion mismatch between the beam and substrate. Therefore, it is expected that the cantilever configuration will demonstrate less actuation voltage variation between room temperature and cryogenic temperatures. However, there are other challenges associated with cryogenic cantilever MEMS switches. It is reported that under low temperatures, the dielectric discharging time-constant increases as temperature decreases [14]. The prolonged discharging time can lead to reduced lifetime if the switch incorporates a dielectric layer over the actuation pad. Moreover, the dielectric layers would introduce more defects when exposed to alpha rays [15] and thus render a much more serious dielectric charging problem. This could rule out space-instrumentation applications which is a promising field for cryogenic RF-MEMS. Therefore, a dielectric-free RF-MEMS cantilever device, as shown in Fig. 1, is developed to address these issues. The switch designed for this work uses the traditional DC-contact cantilever configuration with no dielectric layer on top of the actuation pad. The dimensions of the switch, as shown in Fig. 1, are carefully designed using a multi-physics finite-element solver, Coventor,1 such that upon actuation the cantilever tip will touch the transmission line through dimples, without collapsing the middle section of the cantilever onto the actuation pad. The down-state of the switch is shown in Fig. 2(a) with the free end of the cantilever simply supported by dimples. Due to the nature of this design, the operational actuation voltage of the switch is constricted by the secondary pull-down voltage as shown in Fig. 2(b). The pull-down voltage and the secondary pull-down voltage of the structure shown 1Coventor

is a registered trademark of Coventor Corporation, Cary, NC.

0018-9480/$26.00 © 2009 IEEE

GONG et al.: STUDY OF BROADBAND CRYOGENIC DC-CONTACT RF MEMS SWITCHES

3443

Fig. 3. SEM of fabricated DC-contact cantilever RF MEMS switch. Fig. 1. (a) cross-section view and (b) top view of the DC-contact cantilever MEMS switch fabricated in this work.

Fig. 2. Different actuation phases of the dielectric free DC-contact switch. (a) First pull-down; (b) secondary pull-down.

in Fig. 1, are simulated to be 72 and 102 V, respectively. The spring constant k of the cantilever is calculated to be 71 N/m. This high spring constant is desired for DC-contact cantilever switches to ensure good reliability [16]. The switch is designed for integration into a 50 CPW line m on top of a 500- mwith dimensions of thick fused-quartz substrate . The anchor of the switch rests on one side of the center conductor and the cantilever beam reaches out to connect the other side of the center conductor to form a series DC-contact switch. The bias line of the switch is made of gold rather than high resistivity material to avoid heat generation in large arrays operating at cryogenic temperatures. The switch is designed to have an up-state capacitance of 2 fF to ensure good isolation up to -band and contact force of 36 N at the pull-down voltage to ensure a 1- or lower contact resistance for low insertion loss [17]. III. LOW STRESS FABRICATION PROCESS Although cantilever beams may suffer less voltage change due to thermal expansion mismatch with the substrate, it is important to develop a low-stress fabrication process to create flat cantilevers that match the design profile (i.e., remain flat). Most cantilever fabrication processes based on polymer sacrificial layers have problems with out-of-plane deformation due

to stress gradients within the beam. Recent work indicates that this stress gradient is largely due to the mismatch in the coefficient of thermal expansion (CTE) between the sacrificial layer and the beam [18]. Therefore, an aluminum based sacrificial layer process is developed to reduce this effect. Aluminum has a CTE (21 ppm/K) much closer to the beam material, Au (14 ppm/K), than typical polymer materials such as photoresist ( 50 ppm/K) [19]. The fabrication process begins with the deposition of the circuit layer using a lift-off technique and is followed by planarization. The aluminum sacrificial layer of 1.2 m is then deposited with dimple features patterned in the top layer. The anchor for the cantilever structure is defined using a reactive ion etch (RIE). The anchor area is then planarized and the cantilever beam is defined with 2.2- m plated gold. The Al sacrificial layer is removed with a wet etchant and the RF MEMS switches are released using critical point drying. The scanning electron microscope (SEM) image of the fabricated switch shown in Fig. 3 demonstrates the ability of this process to yield flat beams. IV. MEASUREMENTS AND DISCUSSION A. Room Temperature Measurements Multiple switches were tested at room temperature and cryogenic temperatures. For consistency, switches are labeled as device A, B, C, and D corresponding to the measurement data. Room temperature S-parameter measurements were taken from 2–75 GHz with an HP 8510 network analyzer and wafer probe station. Calibration was done using NIST’s MultiCal2 TRL calibration routine and an on-wafer thru-refl-lines (TRL) kit to fix the reference planes 100 m away from the switch as indicated in Fig. 3. The CPW line loss measured from the TRL calibration is 3 dB/cm at 40 GHz. Device A is actuated by applying a step voltage of 88 V and the measured room temperature S-parameter performance is shown in Fig. 4(a) and (b) from 2–75 GHz. The switch demonstrates an insertion loss less than 0.4 dB and isolation better than 24 dB up to 50 GHz. An RLC circuit model, 2Multical

is a registered trademark of NIST, Boulder, CO.

3444


Fig. 5. Equivalent circuit model for the DC-contact RF-MEMS cantilever switch. (a) Up-state; (b) down-state.

Fig. 4. Room temperature S-parameter performance of the DC-contact RF-MEMS cantilever switch. (a) 2–50 GHz; (b) 50–75 GHz.

Fig. 6. Setup of cryogenic S-parameter measurements.

shown in Fig. 5, is constructed to fit the experimental results. The up-state capacitance of the switch is 2 fF which agrees with the design parameters, and the contact resistance is 1 . At frequencies above 50 GHz, the return loss of the switch degrades due to coupling between the bias pad and beam in the on-state. This coupling is problematic because the impedance of the bias lines is 85 . The switch performance can be improved through the use of a bias tee designed for higher frequencies where the coupling to the bias line is significant. B. Actuation Voltage Versus Temperature Cryogenic measurements were taken using a Lakeshore CPX-HF cryogenic probe station with an HP 8510 network analyzer as shown in Fig. 6. The cryogenic measurements were carried out over two different controlled temperature ranges: 77–300 K using liquid nitrogen and 1.6–45 K using liquid helium. The chamber pressure of the CPX-HF probe station was reduced to before cooling. The cooling process was carefully controlled to ensure that the radiation shield reached the target temperature before the cold stage. Thus, there was little condensation on the sample after the cooling process. The pull-down voltages of the cryogenic DC-contact

Fig. 7. The measured pull-down and hold-down voltages versus temperature.

switch at different temperatures were measured by applying a gradually increasing voltage until the switch was actuated; the results are plotted in Fig. 7. The hold-down voltages were measured by decreasing the actuation voltage, after turning on the


3445

Fig. 8. SEM of the fabricated deformation sensor cantilever.

switch, until the switch turned off. A consistent increase in both the pull-down voltage and hold-down voltage was observed for device A in [20], and device B, C, and D as shown in Fig. 7. Device B and C failed after trying to find the secondary pull-down voltage at 77.2 and 220 K, respectively. An increasing voltage gap between the pull-down and hold-down voltages was also noted. The increase in the pull-down voltage was roughly 60% from 300 to 4.2 K, varying slightly for different devices. This variation agrees with the results reported by Rantakari [21] and is explored further in the following section.

Fig. 9. Measured Isolation of the fabricated deformation sensor at various temperatures.

C. Deformation of the Cantilever at Cryogenic Temperatures In order to analyze the increase in pull-down voltage as the temperature is decreased, a deformation sensor, as shown in Fig. 6, was fabricated with the switches to acquire more measurements of the cantilever deformation under cryogenic conditions. The cantilever deformation sensor, with the same dimensions as the switch cantilever, has the transmission line extended below it resulting in a sensing capacitance of 27 fF. The value of the sensing capacitance is simulated using Maxwell 3D.3 The sensor’s S-parameter performance is measured at cryogenic temperatures from 77.2 to 300 K to determine the de-embeded capacitance value of the sensor, which is dictated by the deformation of the cantilever. The measured S21 versus temperature is plotted in Fig. 9. As can be seen, there is almost no difference as the temperature is varied. A simple LC circuit model is used to fit the measurement results at 77.2 and 300 K. As the fitted curves show in Fig. 10, the de-embedded capacitance value of the sensor decreases to 25 fF as the temperature drops to 77.2 K. However, the isolation measurement of the switch, as shown in Fig. 14, shows virtually no change with temperature. A thermal finite-element-method (FEM) simulation was done using Ansys4 to investigate the possible deformation mechanism. As shown in Fig. 11, the cantilever shows a small out-of-plane kink at the anchor with the cantilever section being straight and flat. This is believed to be caused by the stress gradient in the anchor section induced by the CTE mismatch between gold (14 ppm/K) and fused quartz (0.5 ppm/K). Fig. 11 shows an average increase of 0.35 m in beam height over the actuation pad section, which is a 29% increase over the original gap of 1.2 m. Considering the actuation voltage is related to , this deformation causes a 47% increase in actuation 3Maxwell 4Ansys

is a registered trademark of Ansoft Corporation, Pittsburgh, PA.

is a registered trademark of Ansys Inc., Canonsburg, PA.

Fig. 10. Fitted model demonstrating deformation sensor capacitance changes.

Fig. 11. FEM simulation results of the cantilever deformation at 4.2 K.

voltage. The reason the increase in beam height does not significantly impact the isolation of the original switch ( 1.6 dB at 50 GHz) is due to a high degree of fringing capacitance around the tip of the beam which is confirmed by Maxwell 3D simulations. Another factor that should not be neglected is the increase in Young’s modulus for gold as the temperature decreases. Research on nanocrystalline gold films has shown a 20% increase in Young’s modulus from room temperature to 4 K, although this property may vary for different film deposition methods [22]. This change accounts for a 9% increase in pull-down voltage due to the actuation voltage varying as the square root of Young’s modulus. Therefore, it is believed the increase in actuation voltage under cryogenic temperatures for

3446


D. RF Performance Versus Temperature

Fig. 12. Measured insertion loss of cryogenic broadband switch at various temperatures.

Fig. 13. Measured return loss of cryogenic broadband switch at various temperatures.

1) 77.2 to 300 K Measurements: In order to study temperature’s effect on the RF performance of the switch, the insertion loss of the switch was measured at different temperatures ranging from 77.2 to 300 K (TRL calibrations were done at each measurement temperature) with the measurement actuation voltage having the same offset (16 V) over the hold-down voltage as shown in Fig. 7. The switch was first measured at 300 K after 50 switching cycles. Then, the switch was cooled to 77.2 K and measured after 30 switching cycles. The temperature was then increased to 100, 120, 160, and 220 K, in sequence, and measured after 30 switching cycles at each temperature. The measured insertion loss, return loss, and isolation are shown in Figs. 12–14. A steady increase in insertion loss and decrease in return loss are observed as temperature decreases. This is most likely caused by multiple factors affecting the contact resistance, including variation in the contact force and plastic deformation of the contact spots. Considering the deformation of the switch and the uncertainty in Young’s modulus at cryogenic temperatures, it is difficult to precisely calculate the contact force. A simple analytical model is used to bound the contact forces between 36 and 45 N at cryogenic testing temperatures. Moreover, the hardening of the gold metal at lower temperature [23] could potentially affect the contact area since the gold cantilever shrinks by 0.23 m from room temperature to 77 K. This could lead to larger contact resistance [17], [24]. 2) 1.6 to 77 K Measurements: Below 77 K cryogenic measurements were done with device D in a similar way with liquid helium as the cryogen. After the cold stage reaches 4.2 K, an extra pump is connected to reduce the pressure in the helium flow lines to cool the stage down to 1.6 K. The measured S-parameter performance of the switch at 1.6 and 4.2 K are shown in Figs. 15 and 16, respectively. The switch demonstrates an insertion loss less than 0.6 dB and an isolation better than 25 dB up to 50 GHz. The actuation voltage increases to 131 V at 1.6 K, which follows the same trend observed at temperatures above 77 K. Due to the increasing gap between the pull-down voltage and hold-down voltage, the pull-down voltage is higher than the hold-down voltage plus the 16 V offset. The S-parameter performance shown in Figs. 15 and 16 are measured with a pull-down voltage of 131 V, which has an offset of 19 V from the hold-down voltage. The insertion loss performance at 1.6 K is better compared to its performance at 77 K but shows little difference in comparison with the 4.2 K measurement. This is most likely due to the significant increase in the conductivity of gold (2 orders of magnitude) from 60 to 10 K with little change below 10 K [25]. E. RF Performance Versus Biasing Current

Fig. 14. Measured isolation of cryogenic broadband switch at various temperatures.

the DC-contact cantilever switch is caused by multiple factors, including the deformation-induced beam height increase, and Young’s modulus increase as the temperature is decreased.

Different DC biasing currents were injected into the switch in the down-state to investigate the effect of increased power on contact quality. At 77.5 K, the insertion loss was measured with device A while the switch was subjected to four different DC biasing currents as shown in Fig. 17. The device burned out after biasing it with a current higher than 150 mA. The insertion loss improved as more current was passed through the


Fig. 15. Measured RF performance of cryogenic broadband switch at 1.6 K.

Fig. 16. Measured RF performance of cryogenic broadband switch at 4.2 K.

3447

Fig. 18. Measured insertion loss of cryogenic broadband switch as a function of actuation voltage at 4.2 K.

Fig. 19. Measured insertion loss of cryogenic broadband switch as a function of actuation voltage at 45 K.

Fig. 17. Measured insertion loss of the cryogenic DC-contact RF MEMS switch at 77.5 K with different DC bias current.

Fig. 20. Measured insertion loss of cryogenic broadband switch as a function of actuation voltage at 220 K.

switch. It is believed the current heats up the contact region locally and causes the metal to soften to enable smaller resistance [26]. Hence, better insertion loss is observed at larger biasing current values.

F. RF Performance Versus Applied Voltage Additional contact force can be obtained by increasing the applied voltage. However, the amount of voltage increase is limited by the secondary pull-down voltage. In order to investi-

3448


gate how increased actuation voltage translates into enhanced RF performance within the operational voltage range, device C and D were tested with increasing voltages applied at 220, 45, and 4.2 K, respectively. As shown in Fig. 18, the insertion loss improves 0.2 dB as the actuation voltage increases from 131 to 138 V at 4.2 K, and similar increases are observed in Figs. 19 and 20 at 45 and 220 K for device D and device C, respectively. V. CONCLUSION A dielectric-free cryogenic DC-contact switch has been designed, fabricated and demonstrated. The switch shows an insertion loss less than 0.9 dB with an isolation better than 18 dB up to 75 GHz at room temperature, and an insertion loss less than 0.6 dB with an isolation better than 24 dB up to 50 GHz at cryogenic temperature (1.6 K). The effect of temperature on deformation of the cantilever, actuation voltage, and RF performance of the switch has been studied and interpretations of the observed effects have been offered. The effect of other factors, including power handling and actuation voltage, on RF performance of the switch have also been reported. ACKNOWLEDGMENT The authors would like to thank Dr. A. Kogut and Dr. E. Wollack of NASA Goddard Space Flight Center for useful discussions. REFERENCES [1] N. S. Barker and G. M. Rebeiz, “Distributed MEMS true-time delay phase shifters and wide-bandswitches,” IEEE Trans. Microw. Theory Tech., vol. 46, no. 11, pp. 1881–1890, Nov. 1998. [2] J.-J. Hung, L. Dussopt, and G. Rebeiz, “Distributed 2- and 3-bit W-band MEMS phase shifters on glass substrates,” IEEE Trans. Microw. Theory Tech., vol. 52, no. 2, pp. 600–606, Feb. 2004. [3] K. Entesari and G. M. Rebeiz, “A 12–18 GHz three-pole RF MEMS tunable filter,” IEEE Trans. Microw. Theory Tech., vol. 53, no. 8, pp. 2566–2571, Nov. 2005. [4] Q. Shen and N. S. Barker, “Distributed MEMS tunable matching network using minimal-contact RF-MEMS varactors,” IEEE Trans. Microw. Theory Tech., vol. 54, no. 6, pp. 2646–2659, 2006. [5] T. Vaha-Heikkila and G. Rebeiz, “A 20–50 GHz reconfigurable matching network for power amplifier applications,” in 2004 IEEE MTT-S Int. Microwave Symp. Dig., Jun. 2004, vol. 2, pp. 717–720. [6] C. Brown, A. Morris, A. Kingon, and J. Krim, “Cryogenic performance of RF MEMS switch contacts,” J. Microelectromech. Syst., vol. 17, no. 6, pp. 1460–1467, Dec. 2008. [7] R. W. Simon, R. B. Hammond, S. J. Berkowitz, and B. A. Willemsen, “Superconducting microwave filter systems for cellular telephone base stations,” in Proc. IEEE, Oct. 2004, vol. 92, pp. 1585–1595. [8] J. Noel, Y. Hijazi, J. Matinez, Y. A. Vlasov, and J. G. L. Larkins, “A switched high-Tc superconductor microstrip resonator using a MEMS switch,” Supercond. Sci. Technol., vol. 16, pp. 1438–1441, Nov. 2003. [9] E. M. Prophet, J. Musolf, B. F. Zuck, S. Jimenez, K. E. Kihlstrom, and B. A. Willemsen, “Highly-selective electronically-tunable cryogenic filters using monolithic, discretely-switchable MEMS capacitor arrays,” IEEE Trans. Appl. Supercond., vol. 15, no. 2, pp. 956–959, 2005. [10] B. Schauwecker, J. Mehner, K. Strohm, H. Haspeklo, and J. F. Luy, “Investigation of RF shunt airbridges among different environmental conditions,” Sens. Actuators A, Phys., vol. 114, pp. 49–58, May 2004. [11] C. L. Goldsmith and D. I. Forehand, “Temperature variation of actuation voltage in capacitive MEMS switches,” IEEE Microw. Wireless Compon. Lett., vol. 5, no. 10, pp. 718–720, Oct. 2005. [12] J. G. Noel, A. Bogozi, Y. A. Vlasov, and J. G. L. Larkins, “Cryogenic pull-down voltage of microelectromechanical switches,” J. Microelectromech. Syst., vol. 17, no. 2, pp. 351–355, 2008.

[13] C. Goldsmith, J. Ehmke, A. Malczewski, B. Pillans, S. Eshelman, Z. Yao, J. Brank, and M. Eberly, “Lifetime characterization of capacitive RF MEMS switches,” in 2001 IEEE MTT-S Int. Microwave Symp. Dig., May 2001, vol. 1, pp. 227–230. [14] X. Yuan, Z. Peng, J. C. M. Hwang, D. Forehand, and C. L. Goldsmith, “Acceleration of dielectric charging in RF MEMS capacitive switches,” IEEE Trans. Device Mater. Rel., vol. 6, no. 4, pp. 556–564, Dec. 2006. [15] M. Exarchos, E. Papandreou, P. Pons, M. Lamhamdi, G. J. Papaioannou, and R. Plana, “Charging of radiation induced defects in RF MEMS dielectric films,” Microelectron. Reliab., vol. 46, pp. 1695–1699, 2006. [16] G. M. Rebeiz, RF MEMS: Theory, Design and Technology. New York: Wiley, 2003. [17] S. Majumder, N. E. McGruer, G. G. Adams, P. M. Zavarcky, R. H. Morrison, and J. Krim, “Study of contacts in an electrostatically actuated microswitch,” Sens. Actuators A, Phys., vol. 93, pp. 19–26, 2001. [18] J. R. Stanec, C. H. Smith, III, I. Chasiotis, and N. S. Barker, “Realization of low-stress Au cantilever beams,” J. Micromech. Microeng., vol. 17, pp. N7–N10, Jan. 2007. [19] J. R. Stanec, M. R. Begley, and N. S. Barker, “Mechanical properties of sacrificial polymers used in RF-MEMS applications,” J. Micromech. Microeng., vol. 16, pp. 2086–2091, Aug. 2006. [20] S. Gong, H. Shen, and N. S. Barker, “A cryogenic broadband DC contact RF MEMS switch,” in 2009 IEEE MTT-S Int. Microwave Symp. Dig., 2009, vol. 1, pp. 227–230. [21] P. Rantakari and T. Vaha-Heikkila, “Characterization of CMOS compatible RF MEMS switch at cryogenic temperatures,” in Proc. Int. Solid-State Sensors, Actuators and Microsystems Conf. (TRANSDUCERS ), Jun. 2007, pp. 639–642. [22] S. Sakai, H. Tanimoto, K. Otsuka, T. Yamada, Y. Koda, E. Kita, and H. Mizubayashi, “Elastic behaviors of high density nanocrystalline gold prepared by gas deposition method,” Scripta Materialia, vol. 45, pp. 1313–1319, 2001. [23] H. D. Merchant, G. S. Murty, S. N. Bauadur, L. Dwivedi, and Y. Mehrotra, “Hardness-temperature relationships in metals,” J. Mater. Sci., vol. 8, no. 3, pp. 437–442, Dec. 1973. [24] K. L. Johnson, K. Kendall, and A. D. Roberts, “Surface energy and the contact of elastic solid,” Proc. R. Soc. Lond. A, Math. Phys. Sci., vol. 324, pp. 301–313, 1971. [25] P. V. Attekum, P. H. Woerlee, G. C. Verkade, and A. A. M. Hoeben, “Influence of grain boundaries and surface Debye temperature on the electrical resistance of thin gold films,” Phys. Rev. B, Condens. Matter., vol. 29, no. 2, pp. 640–650, Jan. 1984. [26] B. Jensen, L. Chow, K. Huang, K. Saitou, J. Volakis, and K. Kurabayashi, “Effect of nanoscale heating on electrical transport in RF MEMS switch contacts,” J. Microelectromech. Syst. , vol. 14, no. 5, pp. 935–946, Oct. 2005.

Songbin Gong (S’05) received the B.S. degree in electrical engineering from Huazhong University of Science and Technology, Wuhan, China, in 2004 and the M.E. degree in electrical engineering from University of Virginia, Charlottesville, in 2006, where he is currently working towards the Ph.D. degree. His research interests include design and development RF-MEMS technology, phase shifters, and millimeter wave integrated circuits.

Hui Shen (S’03–M’08) received the B.S. and M.S. degrees in electrical engineering from the Huazhong University of Science and Technology, Wuhan, China, in 1998 and 2001, respectively, and the Ph.D. degree in electrical engineering from the University of Virginia, Charlottesville, in 2008. His Ph.D. work involved the applications of radio frequency microelectromechanical system (RF MEMS) switches and phase shifters, and microfabrication techniques for multimetal layer planarization, and low stress microstructure.


N. Scott Barker (S’94–M’99) received the B.S.E.E. degree from the University of Virginia, Charlottesville, in 1994 and the M.S.E.E. and Ph.D. degrees in electrical engineering from The University of Michigan at Ann Arbor in 1996 and 1999, respectively. From 1999 to 2000, he was a Staff Scientist with the Naval Research Laboratory. In January 2001, he joined the Charles L. Brown Department of Electrical and Computer Engineering, University of Virginia, Charlottesville, where he is currently an Associate

3449

Professor. His research interests include applying MEMS to the development of microwave and millimeter-wave circuits and components. He is also interested in micromachining techniques for submm-wave circuits and RF system and circuit design. Prof. Barker was the recipient of the 2003 National Science Foundation CAREER Award, the 2000 IEEE Microwave Prize, and First and Second Place, respectively, in the 1999 and 1997 Student Paper Competition of the IEEE Microwave Theory and Techniques Society (IEEE MTT-S) International Microwave Symposium (IMS).