Spin-coatable, photopatternable magnetic nanocomposite thin films for MEMS device applications

Abstract

Magnetic nanomaterials' (especially metals) air stability and compatibility with standard micro-fabrication technologies are often a concern for development of MEMS-based magnetic devices. In this paper, we report an air-stable, photo-patternable and spin-coatable magnetic thin film preparation process for MEMS applications. This magnetic nanocomposite thin film was prepared by incorporating carbon capped ferromagnetic cobalt nanoparticles of dimension 20–80 nm into the SU-8 matrix. TEM, XRD and EDAX analyses were done, to investigate the crystal structure, dispersion and phase stability of the films. The SQUID magnetometry and MFM measurements of the film confirmed its magnetic response at room temperature and the retention of its magnetic properties over a period of time. The material compatibility for MEMS device applications was demonstrated through fabrication of a suspended circular membrane of radius ∼250 μm, having four U-shaped beams, of dimension ∼270 × 50 μm each. Three conventional lithography steps and a sacrificial release layer of ∼1 μm thick oxide was used for the fabrication. The membrane was characterized by evaluating its spring constant and resonant frequency. The spring constant and resonant frequencies were estimated to be ∼4.2 N m⁻¹ and ∼29 kHz respectively. Finally, we demonstrated the actuation of the magnetic membrane by an off-chip generated magnetic field, for its possible use as a MEMS device.

---

1. Introduction

Nanocomposite materials with their well-defined physical and chemical properties have attracted considerable attention in the recent past, for applications in micro-fabrication technologies. The challenge, however, has been the tailoring of the materials' piezoresistive, piezoelectric, magnetic, and chemical properties with high biocompatibility.^1–5 Magnetic nanocomposites with different magnetic fillers and a combination of various polymer matrices are areas of current research interest because of their numerous applications in the area of nanoelectronics and biomedical devices.^6–11 The biggest challenge with using magnetic materials for device applications arises from their traditional method of deposition; etch chemistry and high affinity towards oxygen. The formation of an oxide layer over a core magnetic material is always undesirable, as it deteriorates the magnetic properties over a period of time.^12–14 Therefore, there is a need to develop new types of magnetic materials, which are amenable for micro-fabrication applications. Because of the recent progress in the synthesis of magnetic-capped nanoparticles and significant innovations in nanocomposite thin films, there is an opportunity to develop low cost photopatternable magnetic nanocomposite materials. During the last decade, a significant progress has been made in the development of various magnetic nanoparticle synthesis for applications in novel imaging, specific drug delivery and recovery of catalysts. Previously, polymer based magnetic nanocomposite materials have been explored for electromagnetic shielding applications by various research groups.^15–17 Also, in the past researchers have explored these materials for their application in micro robotics and energy scavenging. Ergeneman et al. uses nickel–cobalt as a material for realization of micro resonator which utilizes the magnetic actuation and readout system for detection of vibration.¹⁸ Sutter et al. also demonstrated a cantilever based MEMS actuator, by incorporation of the super paramagnetic Fe₃O₄ nanoparticles into SU-8 matrix.^19,20 Recently, Alfadhel et al. reported a magnetic flow sensor, which was realized with integration of iron nanowires into polydimethylsiloxane (PDMS).²¹ Similarly, Singh et al. also reported a cobalt based PDMS nanocomposite for actuation applications.²² PDMS based nanocomposites have an issue with patterning, therefore, it is tough to pattern them using a standard micro fabrication technology. In this work, we report development of air stable, low cost, spin-coatable, magnetic nanocomposite, which can be used for fabrication of MEMS structures. For this work, carbon coated cobalt nanoparticles were selected as nano fillers, for realization of a photo-patternable magnetic nanocomposite (MNC) thin film, because of their good magnetic properties. SU-8 (ref. 23) was used as matrix for this work, as the integration of these thin films for device applications was the main objective.

In the present study, cobalt nano fillers and nanocomposite thin films were characterized by different methods such as X-ray diffraction (XRD), transmission electron microscopy (TEM) and superconducting quantum interference devices (SQUID) with the aim to investigate the morphology, size, phase, dispersion and magnetization. Thereafter, a MEMS device was demonstrated with the fabrication of a magnetic micro membrane. Dynamic characterization of the fabricated magnetic membrane was carried out using a Laser Doppler Vibrometer. In order to study the stiffness of fabricated membrane, a static evaluation of membrane was done with the help of a calibrated nanoindentaion technique. Finally, the magnetic response of the fabricated membrane was analyzed qualitatively using an off-chip generated magnetic field.

2. Result and discussion

2.1. Materials & methods for the synthesis of magnetic nanocomposite (MNC)

The SU-8/cobalt nanocomposite was prepared by incorporating carbon coated cobalt nanoparticles of size ∼50 nm (Sigma-Aldrich) into a photosensitive polymer SU-8 (Microchem). The illustrations of the used process are as shown in Fig. 1(a and b). Briefly, 25 mg of cobalt nanoparticles were mixed into 1 ml of cyclopentanone (SU-8 Thinner, Microchem). The prepared solution was sonicated for 6 min for proper dispersion of nanoparticles into cyclopentonone. Then a solution of 1 ml SU-8-2002 was added and the mixer was again sonicated for 20 min with an interval of 1 min. In order to compensate the excessive heat generation during sonication process, the whole set was kept in an ice bath. After that, the nanocomposite suspension was transferred over 2′′ silicon wafer using the spin coating process. The spin speed of 2200 rpm for 30 s was used and thickness of 1.5 μm was achieved for MNC film. Then, the spun MNC films were prebaked at 70 °C for 3 min and 90 °C for 4 min for evaporation of the solvent. For patterning these thin films, UV exposure of dose 200 mJ cm⁻² was used in two cycles with a 1 min gap time using standard Karl-Suss MJB4 mask aligner. Subsequently, the films were cross-linked by a post-baking process using similar parameters as used for the pre-baking process.

Fig. 1 Schematic illustration of the process for fabrication of magnetic nanocomposite thin films: (a) the nanocomposite materials; carbon capped cobalt nanoparticles (shown in red circles) and SU-8 monomer are mixed and sonicated (b) representation of nanocomposite thin film spin coating process and patterning with UV light.

2.2. Material characterizations for MNC thin films

2.2.1. Micropatterning study. In order to evaluate the minimum resolution and patterning quality of magnetic nanocomposite thin films, simple structure with different physical dimensions and shapes were chosen and patterned. The results for patterned film are shown in Fig. 2(a–c). It was observed that the nanocomposite thin film could be patterned to various complex shapes such as square and curved geometries with a resolution of 50–60 μm. Further, we have also extended our study to characterize the uniformity in patterned structure by selecting a comb structure in which the fingers of the beam were spaced apart by ∼120 μm. It could be seen from Fig. 2(c) that the patterned structures attained good uniform width.

Fig. 2 Patterned micro-structures with magnetic nanocomposite thin film (a) simple squares and rectangles patterns (b) L and + patterns (c) comb micro structures.

2.2.2. XRD, TEM and SEM analysis. The crystal structure, purity and morphology of used Co nanofillers were investigated by using XRD and high-resolution transmission electron microscopy (HRTEM) characterization. The XRD patterns for cobalt nanoparticles are shown in Fig. 3(a). The peaks at 2θ positions; 44.4, 51.7 and 76 correspond to crystal planes (111), (200), (220) and show face cubic crystal structure of cobalt (JCPDS no-00-015-0806).^13,14 As no other peaks were observed, it confirms the purity of used nanoparticles. HRTEM analysis was carried out to investigate the morphology and size distribution of Co nanoparticles and the results of the same are as shown in Fig. 3(b, c and f). It can be seen from Fig. 3(b–d) that these nanoparticles possess spherical morphology with an average particle size distribution that lies within the range of 20–80 nm. Fig. 3(f) shows nanoparticles consisting of a stack of carbon layers of thickness 3.5 nm around the cobalt core. This microstructure arrangement helps in maintaining the magnetic properties over a period of time by providing stability towards corrosion and air.^12–14 Thereafter, we have looked at the dispersion quality and composition of nanocomposite thin films. Fig. 3(e) and its inset show the scanning electron microscopy images of SU-8/Co nanocomposite film and EDAX respectively. It could be seen from the Fig. 3(e) that the nanoparticles were entirely distributed inside the SU-8 matrix, but agglomeration was also observed. However, this can be minimized by using proper surface functionalization of carbon layer of cobalt nanoparticles with TETA (triethylenetetramine),²⁴ addition of surfactant such as sodium dodecyl sulfate (SDS)²⁵ and/or addition of glycidol to nanocomposite²⁶ as reported recently. This can help to achieve a better dispersion quality which in turn can minimize the variation across different magnetic fields. The compositional analysis results are as shown in the inset of Fig. 3(e). The measured composition of carbon and cobalt in nanocomposite thin film was 39% and 49% respectively.

Fig. 3 (a) XRD of cobalt nanoparticles, (b, c & f) are TEM micrographs of cobalt nanoparticles (d) histogram for size distribution of nanoparticles; (e) and the inset figure shows the scanning electron microscopy and EDAX images of nanocomposite thin film.

2.2.3. The M–H loop analysis. In order to investigate the magnetic properties of nanoparticles and nanocomposite thin films at room temperature, magnetization against the magnetic field (M–H curve) measurements were studied by using SQUID magnetometry. Fig. 4(a) shows M–H curve for Co nanoparticles. The maximum magnetization was found to be ∼164 emu g⁻¹ at 10 000 Oe applied magnetic field. The inset of figure shows magnified view of M–H curve, with coercivity and remnant magnetization of ∼200 Oe and ∼4 emu g⁻¹ respectively. Fig. 4(b) shows the M–H curve for nanocomposite film and the magnetic moment was found to be 3.9 × 10⁻⁴ emu at 10 000 Oe magnetic fields. The coercivity and remnant magnetic moment was found to be 237 Oe and 6.7 × 10⁻⁴ emu respectively. The performance of the device depends on the stability of its magnetic properties over a period of time. The use of carbon coating over the cobalt nanoparticles helps to maintain the magnetic properties, by protecting the cobalt nanoparticles from oxidation, which is responsible for the stability observed in this work. Further, to investigate the magnetic stability of nanocomposite thin film, in ambient condition, the measurement was repeated after ∼180 days and comparison graphs are as shown in Fig. 4(c). This measurement confirms that, the magnetic film retains its magnetic properties over the long span of time, because of the carbon capping of the nanoparticles.

Fig. 4 M–H loop curve at room temperature (a) Co nanoparticles (b) thin film (c) comparison results of magnetic measurement of thin film sample after 180 days.

2.2.4. MFM analysis. MFM²⁷ was used to study the magnetic force gradient of the SU8–cobalt nanocomposite, by using a magnetic probe. A Multimode™ scanning probe microscope, equipped with nanoscope IV controller (Veeco Instruments, Inc., Santa Barbara, CA) was used for these measurements. The MFM probe; Multi-75M G has magnetic coating on the tip and aluminum reflex coating on the back of the cantilever. The nominal spring constant of cantilever was 3 N m⁻¹ and tip radius was less than 60 nm. The tip was magnetized by strong magnetic field prior to experiment. The resonance frequency of cantilever was experimentally found to be 68.6 kHz. The data acquisitions were done by two-pass technique (oscillating mode). The surface topography (height image) was obtained in the first pass, while the magnetic information (phase image) was obtained in the second pass by setting the lift height (30–100 nm) of cantilever above the sample surface.

The MFM measurements taken at different tip-surface distance are as shown in Fig. 5. No significant change occurred in the topography information as seen from the height images (a, c and e). Whereas, their corresponding magnetic phase images (b, d and e) showed significant contrast in topography of surface with respect to the decrease in the cantilever tip height from the surface. Further, to explore the magnetic gradient force on the nanoparticles, we have analyzed the magnitude of the phase changes by studying the line profile taken at different tip-surface distance, as shown in figure (g and h). The maximum phase shift was observed when the tip-surface distance was 30 nm, leading to the increased magnetic interaction between the magnetic probe and nanocomposite surface and depicting an on/off behavior of embedded cobalt nanoparticles inside the SU8 matrix. It can be seen from the MFM images (Fig. 5) that the nanoparticles morphology was different, as compared to the morphology observed in TEM images (Fig. 3(b)). This difference in the images is due to the agglomeration of nanoparticles inside SU-8 matrix and stray magnetization field²⁸ of MFM tip.

Fig. 5 The images in the left column show topography (height) information (a, c and e) and the images in right column show their corresponding magnetic phase image taken at the varying tip-sample distances of (b) 30 nm (top row), (d) 50 nm (second row) and (f) 100 nm (third row), respectively. The red arrows indicate the position of the nanoparticles on which profiles have been plotted (g and h).

3. Fabrication of a micro membrane

A flip chip technique and three levels of lithography process were used for fabrication of a test membrane structure. The schematic representation of process flow is shown in Fig. 6.

Fig. 6 Process flow for fabrication of magnetic membrane (a) substrate (Si, p-type 4–7 ohm cm) (b) grown oxide sacrificial layer (1 μm) (c) SU-8 2002 device layer (∼1.8 μm) (d) SU8–Co nanocomposite layer (∼1.5 μm) (e) SU8-100 anchor layer (∼150 μm) (f) released membrane schematic and microscopy picture.

Briefly, it starts with the RCA cleaned²⁹ p-type silicon wafer of resistivity 4–7 ohm cm, followed by a 1 μm sacrificial oxide layer grown over it. A thin layer of thickness 1.8 μm of SU-8 2002, at a spin speed of 3000 rpm was spun over the oxide layer. This layer was patterned to define the membrane structure. Thereafter, a nanocomposite suspension of 25 mg ml⁻¹ Co concentration was prepared as discussed previously and spin coated over the patterned SU-8 layer. Spin speed of 2200 rpm, prebake temperature of 70 °C for 3 minute and 90 °C for 4 minute were used for evaporation of solvent. Subsequently, a nanocomposite film was exposed to 200 mJ cm⁻² in two cycles, with 1 minute interval. For proper crosslinking of exposed films a post-bake process was carried out by setting a similar temperature as that of the pre-bake process. Then an anchor of thickness 250 μm was defined and patterned with a thick SU8-100 resist layer. Structures were released from oxide surface by complete etching of oxide layer using buffer HF concentration of 5 : 1.

4. Device characterizations

Resonant frequency & stiffness measurements for the freely suspended micro membranes were carried out to study the nanomechanical behavior of the fabricated magnetic MEMS devices under dynamic and static conditions.

4.1. Static analysis for the evaluation of spring constant

The microscopic image of fabricated membrane is shown in Fig. 6(f). It consists of a membrane which has a circular shape of radius 250 μm supported by four U-shaped cantilever beams of dimension 270 × 50 μm. Beam bending approach using AFM and nanoindentation study are common techniques used for measuring the stiffness of MEMS structures.³⁰ In order to estimate the spring constant of the fabricated MEMS structures, a static analysis was performed using the nanoindentation study. For our study, a displacement controlled-mode was chosen for the analysis. The tip of the nanoindentation apex was placed at the center of the membrane as shown in the inset of Fig. 7(b) so that the load can be uniformly distributed over the beams. A calibrated nanoindentation tip was used for experiments. A constant load of 1.5 μN at a loading rate 10 nm s⁻¹ was used for nanoindentation experiments. Thereafter, the displacement of tip was varied from 0 nm to 300 nm and data for load vs. displacement was analyzed. Membranes showed a linear response on application of a load. Loading and unloading characteristics are as shown in Fig. 7(a). The spring constant of 4.9 N m⁻¹ was estimated from the slope of force–displacement curve as shown in Fig. 7(b). The effect of cobalt nanoparticles on mechanical properties of SU-8 were studied and compared and the results are as shown in Fig. 7(c). These results illustrate that the penetration depth shifted to the lower value of 69 nm, as compared to the earlier value of 139 nm for 400 μN load. Also, Young's modulus of SU-8 increased from 7 to 12 GPa, which is due to higher material density of cobalt nanoparticles.

Fig. 7 Nanoindentation results for evaluation of spring constant (a) loading and unloading characteristics of membrane with displacement controlled mode (b) force vs. deflection characteristic of membrane to evaluate the spring constant. (c) Load displacement curve for SU-8 and SU-8/Co nanocomposite.

4.2. Resonant frequency measurement

Resonant frequency measurement for freely suspended micro membranes was estimated under dynamic condition of vibration, using a calibrated piezoactuator. A Laser Doppler Vibrometer (LDV) was used for observation of in-plane motion of membrane.³¹ The membrane's dynamic motion characteristic was analyzed in the range 0–100 kHz. The laser beam of LDV was scanned over the defined grid points on the surface of the membrane and ‘Doppler shift’ of reflected laser beam due to the motion of the membrane was recorded. The recorded mode shapes for the membrane are as shown in Fig. 8(a–c). Three mode shapes for actuation of magnetic membrane were recorded. The mode shape corresponding to 29 kHz (Fig. 8(a)) was identified as its fundamental mode of resonance. In this case, the motion of the suspended membrane was confined to Z direction only which is the desired mode of operation. The other two modes as shown in Fig. 8(b and c) were observed to have the frequencies of 35 and 50 kHz respectively. These mode shapes correspond to torsion motion of membrane.

Fig. 8 (a) Resonant frequency of fabricated membrane and (b & c), represent the torsion mode actuation characteristic of membrane.

One of the possible applications for this membrane is its use as self-calibrated active mass sensors, so the quality factor needs to be optimized. The quality factor was estimated experimentally from the resonant frequency measurements using the half power point method.³² It can be estimated using the following equation.³³

Where, f₀ and Δf are the resonance frequency and bandwidth respectively. As shown in Fig. 8(a), the bandwidth at the midgap was observed to vary from 29.04 kHz to 30.27 kHz. Using the calculation using eqn (1), a quality factor of 23 was observed for the desired fundamental mode.

4.3. Qualitative analysis of magnetic actuation

Sensitivity and accuracy of the fabricated magnetic sensors, depends on the precise control of the deflection of sensors, which is generally in the range of a few nm to a few μm. The micro actuation analysis was done qualitatively with white light interferometry technique.³⁴ The line profiles for membrane surface at various fixed points was first analyzed without the application of magnetic field and thereafter the same pattern was analyzed with the application of magnetic field of 8.7 × 10³ A m⁻¹ in Z-direction. For these measurements the vibration was avoided, to exclude the other disturbances. The whole analysis was restricted to a single line, located at 420 μm at X-axis. The recorded topography images are as shown in insets of Fig. 9(a and b) where a significant change in the surface topography was observed, with the change in color contrast of images. The line profile for membrane located at 420 μm was plotted as shown in Fig. 9.

Fig. 9 Line profiles of membrane with (red) and without (black) the magnetic field. Inset picture (a and b) shows the topography images of the membrane.

A height variation of 10 μm was observed, in the line profiles of membrane about Z-axis on the application of magnetic field. This change is due to the magnetic nature of the embedded Co nanoparticles. When the magnetic field is switched on, a gradient force on the embedded magnetic nanoparticles gets exerted, which changes the position of membrane and hence a change in the line profiles is observed.

5. Conclusions

In summary, a spin coatable, photopatternable nanocomposite with magnetic property was developed for MEMS based magnetic structures. To ensure stability in ambient conditions and better magnetic properties, carbon capped ferromagnetic particles were chosen as fillers for SU-8 based nanocomposite. Thin films of the nanocomposite were evaluated for their phase, elemental analysis and magnetic properties with XRD, TEM, EDAX and MFM techniques. The magnetic nanocomposite thin film showed good magnetic properties and better stability in ambient conditions. Further, for a demonstration of these nanocomposite thin films for MEMS applications, a membrane test structure was fabricated using a flip-chip micro-fabrication technique. Dynamic and static behavior of the fabricated MEMS structures were also studied for their practical applications. A resonant peak at 29 kHz was observed as a fundamental mode of operation. The static evaluation of fabricated membranes showed a spring constant of magnitude 4.3 N m⁻¹. Finally, the fabricated micro membrane was actuated successfully by off-chip generated external magnetic field for its magnetic actuation applications.

6. Experimental details

6.1. XRD, SEM, EDAX and TEM measurements

The phase investigation of cobalt nanoparticles was performed with X-ray diffraction (XRD) techniques using Panalytical X'Pert Pro X-ray diffractometer with the copper K_α radiation (λ = 1.5418 Å). The scanning range was set to 10–80° for recording the XRD diffraction patterns. Morphology, dispersion and elemental analysis of cobalt nanoparticles and nanocomposite thin films were investigated by using scanning electron microscopy (model-S-3400N) and transmission electron microscopy, PHILIPS (model no-CM200).

6.2. SQUID measurements

The room temperature field dependent magnetic measurements (M vs. H) were carried out on nanoparticles, thin films and devices after fabrication, by tightly fixing them in the sample holder and varying the magnetic field. Vibrating sample magnetometer (LakeShore, VSM – 7410) was used for these experiments.

6.3. Nanoindentaion and laser Doppler vibrometer experiments

Laser Doppler instrument, Polytec and nanoindentation Hysitron Inc Minneapolis USA (model TI-900) was used for estimation of spring constant and dynamic analysis of fabricated magnetic membrane.

Acknowledgements

The authors wish to acknowledge the support provided by the Department of Electronics and Information Technology, MCIT and DST, Government of India for their financial support and SAIF for providing the SEM and TEM facility. We also extend our acknowledgements to the institute nanoindentation facility, Suman Mashruwala Advanced Micro Engineering Laboratory for the Laser Doppler Vibrometer characterization facility and the IIT Bombay Nano-fabrication facility for the fabrication of MEMS devices.

References

1. V. Seena, A. Fernandes, P. Pant, S. Mukherji and V. R. Rao, Nanotechnology, 2011, 22, 295501.
2. H. C. Chiamori, J. W. Brwon, E. V. Adhiprakasha, E. T. Hantsoo, J. B. Straalsund, N. A. Melosh and B. L. Pruitt, Microelectron. J., 2008, 39, 228.
3. M. Nordström, S. Keller, M. Lillemose, A. Johansson, S. Dohn, D. Haefliger, G. Blagoi, M. H. Jakobsen and A. Boisen, Sensors, 2008, 8, 1595–1612.
4. K. Prashanthi, M. Naresh, V. Seena, T. Thundat and V. R. Rao, J. Microelectromech. Syst., 2012, 21, 259–261.
5. M. Kandpal, C. Sharan, P. Poddar, K. Prashanthi, P. R. Apte and V. R. Rao, Appl. Phys. Lett., 2012, 101, 104102–104105.
6. A. H. Lu, E. L. Salabas and F. Schuth, Angew. Chem., Int. Ed. Engl., 2007, 46, 1222–1244.
7. J. Gallo, N. J. Long and E. O. Aboagye, Chem. Soc. Rev., 2013, 42, 7816.
8. J. Liu, S. Z. Qiao, Q. H. Hu and G. Q. Lu, Small, 2011, 7, 425–443.
9. S. Behrens, Nanoscale, 2011, 3, 877–892.
10. T. Wen and K. M. Krishnan, J. Phys. D: Appl. Phys., 2011, 44, 393001.
11. L. Y-Wei, Z. Q. Feng and L. R. Wei, Chin. Phys. B, 2013, 22(12), 127502.
12. Y. Xu, M. Mahmood, Z. Li, E. Dervishi, S. Trigwell, V. P. Zharov, N. Ali, V. Saini, A. R. Biris, D. Lupu, D. Boldor and A. S. Biris, Nanotechnology, 2008, 19, 435102.
13. H. Li, N. Zhao, C. He, C. Shi, X. Du and J. Li, J. Alloys Compd., 2008, 458, 130–133.
14. N. A. M. Barakat, B. Kim, S. J. Park, Y. Jo, M. H. Jung and H. Y. Kim, J. Mater. Chem., 2009, 19, 7371–7378.
15. J. Gass, P. Poddar, J. Almand, S. Srinath and H. Srikanth, Adv. Funct. Mater., 2006, 16, 71–75.
16. P. Poddar, J. L. Wilson, H. Srikanth, S. A. Morrison and E. E. Carpenter, Nanotechnology, 2004, 15, 570–574.
17. O. Yavuz, M. K. Ram, M. Aldissi, P. Poddar and H. Srikanth, Synth. Met., 2005, 151, 211–217.
18. O. Ergeneman, P. Eberle, M. Suter, G. Chatzipirpiridis, K. M. Sivaraman, S. Pané, C. Hierold and B. J. Nelson, Sens. Actuators, A, 2012, 188, 120–126.
19. M. Sutter, O. Ergeneman, J. Zurcher, C. Moitzi, S. Pane, T. Rudin, S. E. Pratsinis, B. J. Nelson and C. Hierold, Sens. Actuators, B, 2011, 156, 433.
20. M. Sutter, O. Ergeneman, J. Zurcher, S. Schimd, A. Camenzind, B. J. Nelson and C. Hierold, J. Micromech. Microeng., 2011, 21, 025023.
21. A. Alfadhel, B. Li, A. Zaher, O. Yassine and J. Kosel, Lab Chip, 2014, 14, 4362–4369.
22. A. Singh, L. Hirsinger, P. Delobelle and C. K. Malek, Microsyst. Technol., 2014, 20, 427–436.
23. W. H. Teh, U. Durig, U. Drechsler, C. G. Smith and H. J. Guntherodt, J. Appl. Phys., 2005, 97, 054907.
24. K. Yang, M. Gu, Y. Guo, X. Pan and G. Mu, Carbon, 2009, 47, 1723–1737.
25. A. G. Kahrizsangi, J. Neshati, H. Shariatpanahi and E. Akbarinezhad, Polym. Bull., 2015, 72, 2297–2310.
26. M. Vinchurkar, A. Joshi, S. Pandey and V. R. Rao, J. Microelectromech. Syst., 2015, 24, 1111–1116.
27. J. A. Sidles, J. L. Garbini, K. J. Bruland, D. Rugar, O. Zuger, S. Hoen and C. S. Yannoni, Rev. Mod. Phys., 1995, 67, 249.
28. M. Kleiber, F. Kummerlen, M. Lohndorf, A. Wadas, D. Weiss and R. Wiesendanger, Phys. Rev. B: Condens. Matter Mater. Phys., 1998, 58, 5563–5567.
29. K. Werner, J. Electrochem. Soc., 1990, 137(6), 1887–1892.
30. E. Finot, A. Passian and T. Thundat, Sensors, 2008, 8, 3497–3541.
31. P. Castellini, B. Marchetti and E. P. Tomasini, Proc. IMAC-XXII, Conf., Exposit. Struct.,Dyn., 2004, pp. 85–94.
32. C. F. Beards, Structural vibration analysis, Wiley, New York, 1983, vol. 74.
33. L. Kiesewetter, J. M. Zhang, D. Houdeau and A. Steckenborn, Sens. Actuators, A, 1992, 35, 153–159.
34. J. C. Wyant, White light interferometry, Aero Sense International Society for Optics and Photonics, 2002.

---