Fabrication of SiO2 Microcantilever Using Isotropic Etching With ICP

1632

IEEE SENSORS JOURNAL, VOL. 7, NO. 12, DECEMBER 2007

Fabrication of SiO2 Microcantilever Using Isotropic Etching With ICP Qi Chen, Ji Fang, Hai-Feng Ji, and Kody Varahramyan

Abstract—This paper reports a new design and microfabrication process for high sensing guard-armed silicon dioxide (SiO2 ) microcantilever sensor, which can be widely used in chemical, environmental and biomedical applications. One sensor platform consists of two SiO2 cantilever beams as the sensing and reference elements, two connecting wings, and three guard arms. The guard arms prevent damage to the cantilever beam from collision. To efficiently release the SiO2 cantilevers from the silicon substrate, an isotropic etch method using inductively coupled plasma (ICP) was developed. The isotropic etching with ICP system provides an advantage in good profile control and high etching rate than wet isotropic etching or conventional RIE (reactive ion etching); however, it has not been gained many attentions. In this work, the effects of chamber pressure, plasma source power, substrate power, SF6 flow rate relating with Si etching rate, undercutting rate, and isotropic ratio were investigated and discussed. The optimum isotropic etching process achieved a 9.1 m/min etch rate, 70% isotropic ratio, and 92% etching uniformity. The SiO2 cantilever sensor was fabricated and the cantilever beam was successfully released using a patterned photoresist layer as an etching mask. This plasma isotropic etching release processing can be also applied to release other SiO2 or metal suspended MEMS structures, such as bridges and membranes, with a fast etch rate and reasonable isotropic ratio. Index Terms—Isotropic etching with inductively coupled plasma (ICP) and cantilever design, silicon dioxide (SiO2 ) cantilever.

I. INTRODUCTION

M

ICROCANTILEVERS have recently attracted considerable interests in the development of physical, chemical, and biological sensors [1]–[9]. Microcantilever can undergo bending due to molecular adsorption or absorption by confining the adsorption and absorption to one side of the cantilever where the sensing material is coated. Adsorption or interaction of the analyte will change the surface characteristics of the microcantilever or the film volume on the cantilever, and result in the bending of the microcantilever. This concept has already been used to demonstrate the feasibility of chemical detection of a number of vapor phase analytes, as well as highly sensitive detection of chemical and biological species in solutions. Most of these microcantilever sensors were made of silicon. Due Manuscript received January 26, 2007; revised April 15, 2007; accepted April 30, 2007. This work was supported in part by the National Science Foundation Sensor and Sensor Network under ECS-0428263 and in part by the Board of Regent Industrial Ties and Research Subprogram under Contract LEQSF(200504)-RD-B-19. The associate editor coordinating the review of this paper and approving it for publication was Prof. Michael Pishko. The authors are with the Institute for Micromanufacturing, Louisiana Tech University, Ruston, LA 71270 USA (e-mail: [email protected]; jfang@latech. edu; [email protected]; [email protected]). 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/JSEN.2007.908922

to the relatively large Young’s modulus of silicon material (169 GPa), the bending response of the silicon microcantilever is too weak to be measured when surface stress change is rather small. In the previous works [10], we developed SiO cantilever sensor for better sensitivity [10]–[13] by taking advantage of the smaller Young’s modulus of SiO (76.5–97.2 GPa) in comparison with silicon or silicon nitride microcantilever. We used a plasma anisotropic dry etching process to release the SiO microcantilever beams. To completely release the cantilever beam, the anisotropic dry etching must go through the 500- m-thick sacrificial silicon from the backside of wafer. Common photoresist can not be used as an etching mask. The etching mask material must be carefully selected. In this work, we developed a SiO cantilever sensor fabrication process with a very high etching rate by using isotropic dry etching to completely release the cantilever beam from both sides of substrate. For a better understanding the isotropic etching, the effects of reaction chamber pressure, inductively coupled plasma (ICP) source power, substrate power, and flow rate on the etch rate, undercut rate, and isotropic ratio have been investigated. II. EXPERIMENTS FOR INVESTIGATING ISOTROPIC ETCH PROPERTIES A. Isotropic Plasma Etching Dry etching with ICP has been extensively studied for the fabrication of high aspect ratio micromechanical structures. Among the available process schemes and recipes, the Bosch process [14], [15] has made high aspect ratio silicon structures with vertical sidewalls possible [16]. The standard Bosch process is an anisotropic etching procedure and works as follows. It starts with a shallow etching step under a low pressure SF plasma generated with an RF source, fluorine radicals generated in SF plasma then react with silicon to form volatile SiF , and continues with a passivation step enhanced via an ICP source, where C F gas molecules are dissociated to form a polymeric layer over the exposed silicon surface. During the subsequent etch step, this layer is preferentially withdrawn on the horizontal surface rather than on the sidewalls because of ionic physical etching. Both steps repeat until complete etching of the silicon layer is accomplished. Experiments and optimization settings for anisotropic plasma etching with ICP system have been reported [17], [18]. It was recently found that the isotropic etching capability of ICP has certain advantages over anisotropic etching, including good controllability [19], high silicon etching rate, and high SiO :Si etching selectivity [10], etc. The silicon isotropic etching procedure is modified from the standard Bosch process [14], [15]. In this modified process, the passivation step was

1530-437X/$25.00 © 2007 IEEE

CHEN et al.: FABRICATION OF SIO MICROCANTILEVER USING ISOTROPIC ETCHING WITH ICP

1633

Fig. 1. Pattern of the PR (photoresist) as a mask for cantilever beam release.

dismissed and SF was the only process gas. Since no passivation layer was formed along the sidewall of the trench, the ions reacted freely with the silicon atoms under the mask horizontally. The isotropic etching was realized largely by was dismissed in ion-enhanced chemical etching after the etching process. B. Experimental Setup and Procedure A high-density plasma ICP system (Alcatel 601E) was used to conduct the experiment. The plasma was generated by a 13.56 MHz RF power supply and diffused down to a low pressure chamber where the wafer was processed. The ion energy was independently controlled from ion flux by a radio frequency (RF) power applied to the chuck. The wafer was mechanically clamped on the chuck, which was thermally regulated with liquid nitrogen and a heating element. The C to 50 C. A temperature could be adjusted from helium pressure-controlled film between the chuck and the wafer ensured better heat transfer. A load chamber optimized the load/unload operations without breaking vacuum in the process chamber. The starting material was a commercially available [Silicon silicon Inc.] double side polished four-inch (100-mm), wafer with 500 m thick, and a 1- m-thick SiO layer on both surfaces. To achieve the structure of microcantilever described previously, four masks were designed and applied. Among these, mask 2 was specifically designed for cantilever beam release by isotropic plasma dry etching combined with an anisotropic etching process in Fig. 1. Here, Fig. 1(b) depicts an enlarged part of the mask pattern in Fig. 1(a) and the related dimension of mask 1 and mask 2. The photoresist pattern of mask 2 served as an etching mask in the ICP etching process to release the microcantilever beam from bulk silicon. In order to protect the edges of the cantilever beam, the photoresist pattern covering on the cantilever beam was about 5 m longer along the three edges than the SiO beam. This also compensated for any slightly alignment errors. To ensure the cantilever beam released completely and the foot of the beam remained at the expected position after release, the photoresist pattern at the foot of beam was designed at 60 m longer than mask 1. The experimental processes for investigating the isotropic etch to release the SiO microcantilever beam is shown in Fig. 2. First, after baking for 30 min at 250 C, the wafer was spin-coated with a 1- m-thick layer of Shipley 1813 positive tone photoresist. Then, the photoresist was patterned by the lithography process using mask 1 (Fig. 2(a)). Next, the SiO cantilever beams were formed by wet etching with buffered in volume). Then, oxidation etchant (

Fig. 2. The process for investigation of isotropic plasma etching. (a) Pattern resist. (b) Etch SiO . (c) Pattern resist for cantilever beam release. (d) Isotropic etching for 10 min.

the photoresist was removed from the surface of the wafer [Fig. 2(b)]. A 3 m-thick layer of photoresist 1813 was spun onto the wafer and patterned with mask 2 [Fig. 2(c)]. Finally, the isotropic etching processing was applied by the ICP system to release the cantilever [Fig. 2(d)]. For the best cantilever beam release results, the isotropic etching process was compared using 16 distinct recipes. Each process was initiated with an oxygen cleaning process to remove impurities in the chamber, which would influence the isotropic etching. This cleaning step was applied on a dummy wafer for 10 min. The chamber preconditioning process was followed by the designated recipe within 10 min. After this step, a processing substrate was loaded into the system for isotropic etching. Pure SF is the only process gas used in the silicon isotropic etching. A 10-min etching time was conducted in all testing processes in these experiments in order to compare the process parameters for isotropic etching. The process parameters varied in the following range: chamber pressure (14–66 mTorr), ICP source power (800–1800 W), substrate power (0–30 W), and SF flow rate (150–300 sccm). The profiles of released cantilevers were observed by scanning electron microscopy (SEM). The resulting SEM pictures were analyzed using Matlab software. The etching depths were measured by a KLA Tencor profilometer (TENCOR Alpha step m. 500 profiler; TENCOR Inc.) with a precision of Table I lists the detail of the processing recipes tested and the best result obtained, as shown in Fig. 3. C. Results and Discussion Fig. 3 shows SEM images of the cross section of the etched trenches at the wafer center and the edge (about 7 mm from the wafer edge) separately after isotropic etching. The parameters of the etching process were as the following: mTorr, source W, substrate W, standard cubic centimeters per minute (sccm). The SF microcantilever beams were successfully released (Fig. 4). Table I shows that the chamber pressure, ICP source power, ICP substrate power, and SF flow rate affect the etching profiles. 1) Chamber Pressure Effect: The chamber pressure exerts a great impact on the plasma isotropic etching, as shown in Fig. 5. The source power and substrate power were set to 1800 and 30 W, respectively. The SF flow rate was set to 300 sccm. Both etch rate [Fig. 5(a) black solid line) and undercut rate (Fig. 5(a) red dash line] increased as the pressure increased from 14 to

1634

IEEE SENSORS JOURNAL, VOL. 7, NO. 12, DECEMBER 2007

Fig. 3. SEM images of the cross section of etched trenches. The mask opening was 200 m in width, and the etching time was 600 s. (a) A trench near the center of wafer. (b) A trench at the edge of wafer. TABLE I PROCESS PARAMETERS AND ETCHING PROFILES

higher number of fluorine radicals) in the chemical reaction at higher pressure [20], [21]. When the pressure was higher than 46 mTorr, the chemical reaction tended to saturate and higher pressure would increase the collision frequency, which resulted in the reduction of ion energy [20]. The isotropic ratio in Table I was defined as

Fig. 4. SEM image of the released cantilever beams.

46 mTorr. When the pressure higher than 46 mTorr, the etch rate was the same but undercut rate slightly dropped. Under the pressure range from 14 to 46, the increases in etch rate and undercut rate were due to the higher plasma density (especially the

where is the horizontal dimension extent of the foot edge from the mask edge or the undercut rate, and is the vertical etched depth or the etch rate. The ratio of undercut rate over etch rate gave an isotropic ratio, as shown in Fig. 5(b). The highest isotropic ratio appears at 23 mTorr. In the 14–23 mTorr range, the isotropic ratio increased when the pressure increased, indicating the enhanced chemical reaction rate which exerted more impact on undercut rate than on etch rate. Above 23 mTorr, the isotropic ratio decreased as the pressure increased, suggesting differences between undercut and etch rates. Interactions of reactive

CHEN et al.: FABRICATION OF SIO MICROCANTILEVER USING ISOTROPIC ETCHING WITH ICP

1635

Fig. 5. Silicon isotropic etching profiles as a function of pressure. (a) Etch rate and undercut rate. (b) Isotropic ratio.

Fig. 6. Silicon isotropic etching profiles as a function of plasma source power. (a) Etch rate and undercut rate. (b) Isotropic ratio.

Fig. 7. Silicon isotropic etching profiles as a function of substrate power. (a) Etching rate and undercut rate. (b) Isotropic ratio.

plasmas with surfaces often involve two components: physical and chemical interactions. The lateral etch rate (undercut rate) mainly depends on the chemical plasma reaction with the silicon substrate horizontally, as reported by Jae-Ho Min et al. [22]. Both the physical and chemical etching contribute to the overall etch rates especially in the vertical direction [22]. These results suggest that the physical etching rate in the vertical direction increased faster than in horizontal direction when the pressure was higher than 23 mTorr. 2) ICP Source Power Effect: Fig. 6 shows the plasma isotropic etching profiles as a function of ICP plasma source power. The pressure, substrate power, and SF flow rate were set to 46 mTorr, 30 W, and 300 sccm, respectively. It was found that an increase in the ICP source power increased the etch rate [Fig. 6(a) black solid line] and the undercut rate [Fig. 6(a) red dash line]. It has been reported that an increase in the ICP source

power would produce a large number of electrons with higher energy for dissociation and ionization [20], which create higher concentrations of reactive species and higher ion flux. These more reactive species would increase the bond breaking and desorption efficiency. The etch rate increased faster than the undercut rate when the plasma source power increased. This resulted in a decrease in the isotropic ratio as the source power increased [Fig. 6(b)]. Fig. 6(b) can also be explained by a difference in physical etching influence between the horizontal direction and the vertical direction with the plasma density rising. 3) Substrate Power Effect: Fig. 7 shows the plasma isotropic etching properties as a function of substrate power. The chamber pressure, source power, and SF flow rate were set to 46 mTorr, 1800 W, and 300 sccm, respectively. An increase in the substrate power applied to the wafer had no significant impact on the etch rate [Fig. 7(a) black solid line] and undercut rate [Fig. 7(a) red

1636

IEEE SENSORS JOURNAL, VOL. 7, NO. 12, DECEMBER 2007

Fig. 8. Silicon isotropic etching profiles as a function of SF flow rate. (a) Etch rate and undercut rate. (b) Isotropic ratio.

Fig. 9. Schematic diagram of the SiO microcantilever design.

dash line] because the fluorine ion density in the plasma area surrounding the wafer was almost independent of substrate power [21], [23]. A slight increase in etch rate and undercut rate from 10 to 20 W [Fig. 7(a) and (b)] may be due to an increase in the ion bombardment energy [21], which allows the fluorine radicals to react with silicon more efficiently. The isotropic ratio also increased slightly when the substrate power increased from 10 to 20 W, as shown in Fig. 7(b). 4) SF Flow Rate Effect: Fig. 8 shows the etch rate, undercut rate, and isotropic ratio as a function of SF flow rate. The chamber pressure, source power, and SF flow rate were held constantly at 46 mTorr, 1800 W, and 300 sccm, respectively. An increase in the etch rate [Fig. 8(a) black solid line] and undercut rate [Fig. 8(a) red dash line] with increasing SF flow rate from 150 to 250 sccm was observed. This indicates that an increase in the SF flow rate contribute to an increase of the number of SF molecules that could produce more fluorine radicals and resulted in an increase of the etch rate and the undercut rate. At 300 sccm flow rate, the etch rate and undercut rate were saturated because of: a) insufficient source power to promote the dissociation of SF and produce fluorine radicals and/or b) the recombination of radicals before reaching the sample surface [23]. The isotropic ratio decreased with increasing SF flow rate [Fig. 8(b)] since the speed of increase of the undercut rate was slower than that of the etch rate, which was similar to the effect of source power, (i.e., the plasma density increase resulted in a physical etching reaction in the vertical direction faster than in the horizontal direction). III. DEVICE FABRICATION A schematic diagram of the SiO microcantilever structure design is shown in Fig. 9. The device consists of two adjacent SiO cantilever beams, connecting wings on both sides, and

Fig. 10. Process flow for SiO microcantilever fabrication. (a) Pattern SiO microcantilever beam by BOE etching. (b) Pattern the guide arm by ICP etching. (c) Pattern the connecting wing by ICP etching. (d) Microcantilever beam released by ICP isotropic etching.

three guard arms. The dimensions of the designed SiO microcantilever beams are 250 m in length, 100 m in width, and 1 m in thickness. The connecting wings were connected with the adjoining cantilever bodies after the beams were released, so that all the cantilevers were held on the substrate for further processing. The guard arm was designed to protect the microcantilever beam from damage by collision during separation and handling procedures. Fig. 10 illustrates the main steps of the fabrication process. The fabrication process is described as follows. A. Pattern the SiO Cantilever Beam Shipley 1813 positive tone photoresist was spun on one surface of a silicon wafer. A microcantilever beam pattern was transferred with mask 1 to the photoresist layer on the front side of the wafer by a standard photolithography process, and then the SiO cantilever shapes were defined by wet etching with Buffered Oxidation Etchant (BOE HF:NH F=1:6 in volume). In the mean time, the entire SiO layer on the backside was etched off. The photoresist was then cleaned by acetone and DI water [Fig. 10(a)].

CHEN et al.: FABRICATION OF SIO MICROCANTILEVER USING ISOTROPIC ETCHING WITH ICP

1637

B. Wafer Backside Etching 1: Pattern Cantilever Base With Guard Arm A 3- m-thick 1813 photoresist layer was spun on the backside of the wafer, and then patterned with a photolithography process to form the base of the cantilever sensor. The photoresist pattern served as a mask for ICP plasma etching to create the guard arm in the next fabrication step. The wafer was etched by about 70 m by the ICP anisotropic etching process [Fig. 10(b)]. The etching depth of about 70 m determined the thickness of the connecting arm. This etched area will be completely opened through the wafer after releasing the cantilever beam.

Fig. 11. SEM image of rectangular cantilever beam.

C. Wafer Backside Etching 2: Pattern Cantilever Connecting Wing and Guard Arm A 3 m layer of photoresist 1813 was spun on the backside of the wafer again and then plasma anisotropic etching applied to etch about 260 m. This etching step is also applied to etch the connecting wings, guard arms, and open window from the backside of the substrate [Fig. 10(c)]. D. Cantilever Beam Release A new process, which involves two steps was developed to release the SiO cantilever beam. The cantilever beam was first patterned by anisotropic dry etching and followed by the isotropic plasma dry etching to completely release the cantilever beam. In the first step, anisotropic etching was applied to open a window and to ensure the fluorine radicals can react efficiently with the silicon underneath the SiO beam during the isotropic plasma etching. Then, the isotropic plasma etching was applied to release the SiO beam from the bulk silicon substrate. A 3 m thick film of photoresist 1813 was spun on the front side of the wafer, and then the front side of the wafer was patterned with mask 2 (Fig. 1). The photoresist pattern served as an etching mask for ICP etching process to release the microcantilever beam from the bulk silicon. The cantilever beam was released by two plasma dry etching steps: 90- m-thick anisotropic etching, and then isotropic etching until all microcantilever beams were released [Fig. 10(d)]. The processing time for releasing the cantilever was 20 min. The optimal recipe for isotropic etching in this step was chosen by the following investigation and discussions. The etch rate, undercut rate, and isotropic ratio have been taken into consideration to obtain an optimal set of plasma etching parameters. In fact, for the isotropic etch, we hope to achieve higher etch rate, as well as better isotropic ratio at the same time. Therefore, recipe 13 in Table I (46 mTorr pressure, 1800 W source power, substrate power 20 W, 300 sccm SF flow rate, and 20 C chuck temperature) was adopted as the best one. For this recipe, the etching rate was about 9.1 m/min, uniformity was 92% and isotropic ratio was 66%. Using this process, the cantilever beam was successfully released from the substrate. Fig. 11 shows the SEM pictures of one cantilever. The configuration of the several devices is shown in Fig. 12. The connecting wing between two adjacent microcantilever bodies should be about 50–70 m thick, so that the thickness is not only sufficient to connect them together, but also

Fig. 12. SEM image of SiO microcantilevers.

easily broken just by applying a little force. The cantilevers can be easily coated with a sensing polymer or films for different chemical or biological sensing applications. Another advantage of this process is that it is easy for us to handle the problem associated with the photoresist mask, which occurred in our previous experiment. During an anisotropic dry etching go through an entire wafer with ICP, the photoresist mask was damaged or etched off because the ions bombard the surface of the photoresist. The fabrication process we developed was to divide the release process into three dry etching steps. Meanwhile, the release time is greatly reduced at the higher isotropic etching rate (9.1 m/min). Therefore, the 3- m-thick photoresist 1813 can be an appropriate mask layer for releasing the cantilever beam. IV. CONCLUSION New SiO microcantilever sensors were designed and fabricated using the isotropic dry etching process herein developed. The release of SiO microcantilever beams and the effects of plasma etching properties on etching rates and undercut rates were investigated and discussed. The results showed that the plasma density played a major role in the isotropic etching results; higher plasma density would result in a higher etch rate and higher undercut rate. The higher plasma density could be achieved by increasing pressure, source power or SF flow rate. One optimal recipe with a balance between etch rate and isotropic ratio for releasing the designed microcantilever beams using isotropic dry etching process was shown in Table I (recipe 13 at 46 mTorr pressure, 1800 W source power, 20 W substrate power, 300 sccm SF flow rate, and 20 C chuck temperature). Using this recipe, a 9.1 m/min etch rate and a 6.0 m/min

1638

undercut rate has been achieved. The cantilever beam was completely released with a 3 m thickness photoresist mask and 20 min etching time by ICP plasma dry etching. This process was valuable for fabricating SiO cantilever beams and releasing other suspending parts from the silicon substrate with high etch rate and good processing controls. The developed SiO cantilever sensor can provide a higher sensitivity with easy fabrication for many chemical, biochemical, and environmental applications. REFERENCES [1] M. D. Chabot and J. Moreland, “Micrometer-scale magnetometry of thin Ni-Fe films using ultra-sensitive microcantilevers,” J. Appl. Phys., vol. 93, pp. 7897–7899, 2003. [2] T. Thundat, P. I. Oden, and R. J. Warmack, “Microcantilever sensors,” Microscale Thermophys. Eng., vol. 1, pp. 185–199, 1997. [3] X. Xu, T. G. Thundat, G. M. Brown, and H.-F. Ji, “Detection of Hg using microcantilever sensors,” Anal. Chem., vol. 74, pp. 3611–3615, 2002. [4] H.-F. Ji, T. Thundat, R. Dabestani, G. M. Brown, P. F. Britt, and P. V. Bonnesen, “Ultrasensstivite detection of CrO using a microcantilever sensor,” Anal. Chem., vol. 73, pp. 1572–1576, 2001. [5] K. M. Hansen, H.-F. Ji, G. Wu, R. Datar, R. Cote, A. Majumadar, and T. Thundat, “Cantilever-based optical deflection assay for discrimination of DNA single nucleotide mismatches,” Anal. Chem., vol. 73, pp. 1567–1571, 2001. [6] G. Wu, H.-F. Ji, K. Hansen, T. Thundat, R. Datar, R. Cote, M. F. Hagan, A. K. Chakraborty, and A. Majumdar, “Nanomechanical signatures of biomolecular recognition and interactions,” in Proc. Natl. Acad. Sci., 2001, vol. 98, pp. 1560–1564. [7] H.-F. Ji and T. G. Thundat, “Trace Ca microcantilever sensor,” Biosens. Bioelectron., vol. 17, pp. 337–343, 2002. [8] P. Belaubre, M. Guirardel, and G. Garcia, “Fabrication of biological microarrays using microcantilevers,” Appl. Phys. Lett., vol. 82, pp. 3122–3124, 2003. [9] J. Fritz et al., “Translating biomolecular recognition into nanomechanics,” Science, vol. 288, pp. 316–318, 2000. [10] Y. Tang, J. Fang, X. Yan, and H.-F. Ji, “Fabrication and characterization of SiO microcantilever for microsensor application,” Sens. Actuators, B: Chem., vol. 97, pp. 109–113, 2004. [11] P. Li and X. Li, “A single-sided micromachined piezoresistive SiO cantilever sensor for ultra-sensitive detection of gaseous chemicals,” J. Micromech. Microeng., vol. 16, pp. 2539–2546, 2006. [12] P. Li, X. Li, G. Zuo, J. Liu, Y. Wang, M. Liu, and D. Jin, “Silicon dioxide microcantilever with piezoresistive element integrated for portable ultraresoluble gaseous detection,” Appl. Phys. Lett., vol. 89, no. 7, p. 074104, 2006. [13] M. H. Bao, , S. Middelhoek, Ed., Micro mechanical transducers (Handbook of Sensors and Actuators). Amsterdam, The Netherlands: Elsevier, 2000. [14] F. Laermer and A. Schilp, “Method of anisotropically etching silicon,” in U.S. Patent 5501893 1996. [15] F. Laermer, A. Schilp, K. Funk, and M. Offenberg, “Bosch deep silicon etching: Improving uniformity and etch rate for advanced MEMS applications,” in Proc. Tech. Dig. MEMS’99, Florida, 1999, pp. 211–216. [16] E. Quevy, B. Parvais, J. P. Raskin, L. Buchaillot, D. Flander, and D. Collard, “A modified Bosch-type process for precise surface micromachining of polysilicon,” J. Micromech. Microeng., vol. 12, pp. 328–333, 2002. [17] S. Aachboun and P. Ranson, “Deep anisotropic etching of silicon,” J. Vac. Sci. Technol. A, vol. 17, pp. 2270–2273, 1999. [18] S. Aachboun, P. Ranson, C. Hibert, and M. Boufnichel, “Cryogenic etching of deep narrow trenches in silicon,” J. Vac. Sci. Technol. A, vol. 18, pp. 1848–1852, 2000. [19] K. P. Larsen, J. T. Ravnkilde, and O. Hansen, “Investigations of the isotropic etch of an ICP source for silicon microlens mold fabrication,” J. Micromech. Microeng., vol. 15, pp. 873–882, 2005. [20] S. Frederico and C. Hibert, “Silicon sacrificial layer dry etching (SSLDE) for free-standing RF MEMS architectures,” Proc. IEEE Micro Electro Mech. Syst. (MEMS), pp. 570–573, 2003. [21] B. Kim, S.-M. Kong, and B.-T. Lee, “Modeling SiC etching in C F =O inductively coupled plasma using neural networks,” J. Vac. Sci. Technol. A, vol. 20, pp. 46–52, 2001.

IEEE SENSORS JOURNAL, VOL. 7, NO. 12, DECEMBER 2007

[22] J.-H. Min, G.-R. Lee, J.-K. Lee, and S. H. Moon, “Angular dependence of etch rates in the etching of poly-Si and fluorocarbon polymer using SF ; C F , and O plasmas,” J. Vac. Sci. Technol. A, vol. 22, pp. 661–669, 2004. [23] F. Gaboriau, M.-C. Peignon, G. Cartry, L. Rolland, D. Eon, C. Cardinaud, and G. Turban, “Langmuir probe measurements in an inductively coupled plasma: Electron energy distribution functions in polymerizing fluorocarbon gases used for selective etching of SiO ,” J. Vac. Sci. Technol. A, vol. 20, pp. 919–927, 2002.

Qi Chen received the B.S. and M.S. degrees from the Department of Electrical Engineering, Nanchang University (NCU), Nanchang, China, in 1996 and 2001, respectively. Currently, he is working towards the Ph.D. degree in engineering at the Department of Electrical Engineering, Institute of Micromanufacturing, Louisiana Tech University, Ruston. His research interests are in the simulation, design, fabrication, and testing of MEMS devices and sensors.

Ji Fang received the B.S. degree in electrical engineering from Tianjing University, Tianjing, China, in 1965. From 1965 to 1991, he was a Research Engineer, Senior Engineer, Department Director, and member of the Academic Board at the Research Institute for Petroleum Processing, SINOPEC, Beijing, China. From 1991 to 1992, he was a Senior Visiting Scholar at the University of Cincinnati, Cincinnati, OH. Since 1992, he has been a Senior Research Engineer and Graduate Faculty Member at the Institute for Micromanufacturing, Louisiana Tech University, Ruston. His research interests include microfluidic devices and systems, sensors, microreactors, total analysis system-on-chip, optical lens systems, and micro/nano fabrication technologies.

Hai-Feng Ji received the Ph.D. degree in chemistry from the Chinese Academy of Science, Beijing, China, in 1996. Currently, he is an Associate Professor at the Institute for Micromanufacturing, Louisiana Tech University, Ruston. He is currently a coauthor of 75 peer-viewed journal articles and book chapters and is one of the most-active researchers in the microcantilever sensors field. His research interests focus on MEMS devices, surface modification, and self-assembled nanostructures.

Kody Varahramyan received the Ph.D. degree in electrical engineering from the Rensselaer Polytechnic Institute, Troy, NY, in 1983. From 1982 to 1992, he was with IBM Microelectronics, conducting research and development in the realization of advanced semiconductor technologies. Since 1992, he has been with Louisiana Tech University, Ruston, where he is the Entergy/LP&L/NOPSI Professor of Electrical Engineering, in recognition of his teaching and research contributions in the microsystems and nanotechnology areas. Since September 2000, he has been the Director of the Institute for Micromanufacturing, where, since 1992, he has contributed to the growth and development of the Institute, including through planning and setting up of laboratory resources and facilities, development and implementation of major sponsored research efforts, and realization of academic courses and curricula, on the science and engineering of materials, processes, and devices for the realization of micro/nanoscale systems.