Ultramicroscopy 91 (2002) 103–110
Read/write mechanisms and data storage system using atomic force microscopy and MEMS technology Hyunjung Shina,*, Seungbum Honga, Jooho Moonb, Jong Up Jeonc a
Storage Laboratory, Samsung Advanced Institute of Technology, P.O. Box 111, 440-600 Suwon, South Korea b Department of Ceramic Engineering, YonSei University, Seoul, South Korea c School of Mechanical and Automotive Engineering, University of Ulsan, Ulsan, South Korea Received 29 May 2001; received in revised form 10 January 2002
Abstract Information storage system that has a potentially ultrahigh storage density based on the principles of atomic force microscopy (AFM) has been developed. Micro-electro-mechanical systems (MEMS) technology plays a major role in integration and miniaturization of the standard AFM. Its potential application for ultrahigh storage density has been demonstrated by AFM with a piezoresponse mode to write and read information bits in ferroelectric Pb(ZrxTi1 x)O3 films. With this technique, bits as small as 40 nm in diameter have been achieved, resulting in a data storage density of simply more than 200 Gb/in2. Retention loss phenomenon has also been observed and investigated by AFM in the piezoresponse mode. Finally, local piezoelectric measurements of PZT films by different processing technologies are discussed in detail. r 2002 Published by Elsevier Science B.V. Keywords: Atomic force microscopy; Piezoresponse imaging; Ferroelectric thin film; R=W mechanism
1. Introduction Since IBM introduced the first disk drive in 1957, the data storage systems based on magnetic recording technology have shown remarkable advancements. Currently, area density of manufactured hard disk drive (HDD) reaches almost 40 Gbits/in2 and has increased at a rate of almost 100% per year in the 1990s. No technological limitation seemed to stop the need for greater storage capacity in less space. However, it is known that the current magnetic recording tech*Corresponding author. Tel.: +82-31-280-6907; fax: +8231-280-6955. E-mail address:
[email protected] (H. Shin).
nology used in HDD has a physical—superparamagnetic—limit that is due to the thermal energy causing the de-magnetization of the ‘‘written’’ information bits in nano-meter size. Next generation data storage systems need much higher density beyond the incremental advances of an existing mass-information storage technology. Therefore, new reading and writing technology in nano-meter scale needs to be explored. Until recently, scanning probe microscopy (SPM) has been used for the characterization of the materials’ surfaces and also for probing various surface properties in local area, i.e. electrical, magnetic, mechanical as well as optical properties. Using the advancement of the local probing technique, many researchers have
0304-3991/02/$ - see front matter r 2002 Published by Elsevier Science B.V. PII: S 0 3 0 4 - 3 9 9 1 ( 0 2 ) 0 0 0 8 8 - 8
104
H. Shin et al. / Ultramicroscopy 91 (2002) 103–110
proposed many (scanning tunneling microscope (STM), atomic force microscope (AFM), electrostatic force microscope (EFM), scanning capacitance microscope (SCM), magnetic force microscope (MFM), and scanning near-field optical microscope (SNOM), etc.) SPMs—based storage techniques as future ultra-high density data storage systems [1]. However, most of the works are addressing the fundamental understanding of read/write mechanisms. In the case of AFM-based storage, works toward storage devices in our research group as well as in the other group [2] have progressed beyond simply demonstrating the ability to make small information bits and sense them back in nano-meter scale. Such an information storage system based on the principles of AFM will soon become an evolutionary one performing nano-scaled recording [3]. This is simpler because the storage devices using AFM can be achieved higher information storage density (more than terabits per inch square) than any other methods being investigated currently. Most of the AFM-based storage, which store information with spatial resolution in tens of nano-meter or below, demonstrate successfully write-once read-many (WORM) type storage. However, utilizing ferroelectric thin film as the recording medium in our present study provides a unique way to rewrite information bits as many times even in the AFM-based storage system. Recently, AFM has opened up a new vista of research into the nanoscopic responses of ferroic materials [4–6]. Particularly, AFM in the piezoresponse detection mode has been employed as an effective tool for visualization of domain structure in ferroelectric thin films together with their surface morphology [7,8]. This technique allows non-destructive and high-resolution domain imaging as well as local piezoelectric measurements [9–12] Several groups [13,14] including us in Samsung Advanced Institute of Technology, have demonstrated the ability to explore and exploit the inverse piezoelectric effect that is available through the interaction of ferroelectric materials and the AC signal imposed on the AFM tip. Using local piezoelectric measurement technique, Roelof et al. differentiated 1801 and 901 domains switching in ferroelectric polycrystalline PT (PbTiO3) films [15].
Monitoring the evolution of the domain structure with time intervals typical of AFM (a few minutes per scan) also provides another unique research opportunity. Real time observation allows the investigation of relaxation dynamics, mainly domain branching, coarsening, pinning, even during heat treatments. More generally it also allows observation on phase transformations at the submicrometer scale [16,17]. Fatigue studies using this technique are notable examples [18]. Colla et al. recently observed ‘‘ferroelectrically dead’’ regions in fatigued PZT films [19]. The ability to probe the microscopic details and to write a sub-100 nm ‘‘artificial domain’’ has already been demonstrated in continuous thin films [20–22]. As described, many observations using the piezo-response imaging of ferroelectric domains help to understand their behaviors and properties [23–26] in particular, spontaneous polarization reversal [8,27] and fatigue [18,28]. However, works are reported on the formation of the ferroelectric domains using AFM tip in nano-meter scale as well as retention loss phenomena of the formed ones [29]. In this study, the formation of ferroelectric domains and the retention loss phenomena of the formed ones are investigated in detail with future massive information storage devices in mind. In addition, local piezoelectric measurements are also performed. This particular measurement helps us to understand the effect of processing methods on the local piezoelectric response. With this understanding one can increase signal-to-noise ratio choosing adequate processing for the recording media. Miniaturization and integration of the standard AFM by MEMS technology is also briefly described.
2. Intellectual AFM probe array and x- and y-axis micro-stage Besides the investigation of recording media and their physical mechanisms for reading and writing (R=W ) of the bit, the micro-cantilever with an AFM tip as an R=W head and micro-actuator for ultraprecise positioning (less than 10 nm) and large movements (750 mm) in x- and y-axis are required to be fabricated. Micro-electro-mechanical system
H. Shin et al. / Ultramicroscopy 91 (2002) 103–110
(MEMS) technology is being used for the fabrication of integrated and miniaturized nano-storage devices. Current, conventional AFM uses an optical detection system and in this a focused laser beam is used to measure the deflection of the cantilever, and through this atomic force between the cantilever and the sample’s surfaces is created. To apply the focused laser beam to the cantilever end, the irradiated system must be aligned. Position-sensitive photo-detector must also be properly located both for measuring and amplifying the deflection of the cantilever. These cause complications in the design, alignment, and operation of the AFM and make the system bulky. Therefore it is required to develop an intelligent micro-cantilever for integration and miniaturization. Firstly, as for a force-sensing microcantilever, piezoelectric [30–32], piezoresistive [33,34], and capacitive sensors [35] have been developed. Several research groups [36–38] introduced the downsized AFM with a piezoelectric cantilever. Their advantage is that they have deflection sensing and direct oscillating functions in themselves. More importantly, a scanning multi-probe microscopic system should be constructed into a single chip. Even though AFM technique has shown considerable promise for, aforementioned, high information storage density, current single AFM cantilever will not achieve data transfer rate that is useful for the storage devices. Therefore, only multi-probe (an array of 1000 or more) system could realize the rates comparable to those of current magnetic disk drive [39]. Integration deflection, or force, sensors onto the microcantilever simplify the operation of the parallel system. The piezoresistive AFM sensor has been incorporated into parallel arrays and has been used for imaging and lithography [40] Lutwyche et al. has fabricated and used twodimensional arrays of piezoresistive cantilever [41]. Parallel piezoresistive cantilevers for data storage applications have been fabricated by Chui et al. [42] and Ried et al. [43]. Itoh et al. have also developed an array of individually controlled piezoelectric microcantilevers for a multi-probe AFM [44].
105
More advanced, a 32 32 (1024) 2D cantilever array chip has been fabricated for ultrahighdensity and high-speed data storage applications by IBM research group [2]. In our group, also, researchers are being carried out to develop intelligent multi-probe array equipped with PZT films. Electrostatic and electromagnetic micro-actuators as planar scanners with wide range (more than 750 mm) have already been fabricated in our group. Therefore, in the near future, new data storage devices fabricated by MEMS technology will come out in the market as a commercial product.
3. Experimental procedure A commercial AFM (M5, Park Scientific Instruments, PSI, USA) equipped with a conductive tip (Ultralever) was used to polarize and image the ferroelectric domains. AFM tips are made of boron doped p-type Si, and their height and radius are about 3 mm and 20 nm, respectively, estimated using scanning electron microscopy. The width, length and thickness of the cantilever are 25, 180 and 1 mm, respectively. The force constant and the resonance frequency of the cantilever are 0.26 N/m and 40 kHz, respectively. The AFM tip was scanned over the surface of a Pb(ZrxTiy)O3 (hereafter PZT) film deposited on a Pt-passivated Si substrate while a voltage was applied between the tip and the Pt electrode. All scanning processes were performed in a contact mode at room temperature and ambient pressure. For piezoresponse imaging, an AC modulation signal (Vac ) is applied to the conductive tip while the bottom Pt electrode is grounded. The modulation voltage was 1Vpp (peak to peak) and modulation frequency was 17 kHz. The tip vibration signal induced by Vac on the tip was detected by a position-sensitive photo diode (PSPD). The lock-in amplifier detects the first harmonic signal while the low pass filter sends only the topographic information to the z-control feedback circuit. Signals from the lock-in amplifier (A cos f; where A is the amplitude of the first harmonic signal and f is the phase difference between the AC
106
H. Shin et al. / Ultramicroscopy 91 (2002) 103–110
modulation signal and the first harmonic signal) are recorded. The measured phase difference between the tip vibration signal and the AC modulation voltage indicates the domain polarization orientation whereas the tip vibration amplitude is proportional to the piezoelectric coefficient. Finally, most of this study was carried out at or near the sample edge to minimize unfavorable cantilever–sample capacitive force interaction. Details of this interaction and contribution to the domain imaging are published elsewhere [45]. Ferroelectric thin films used in the present work were lead zirconate titanate (Pb(Zr0.52Ti0.48)O3) films fabricated by various methods, such as sol– gel processing, sputtering, laser ablation and MOCVD. They were polycrystalline deposited on Pt electrode as well as epitaxial films on oxide electrode as a buffer layer on Si or single crystal substrates.
4. Results and discussion Figs. 1(a) and (b) show piezoresponse images in amplitude (a) and phase (b) of as-prepared ferroelectric domains in a laser-ablated PZT films with a thickness of 100 nm. As described earlier, amplitude provides information about the displacement (d33) of the piezoelectric films and phase the direction of the ferroelectric polarization. Areas with bright contrast in Fig. 1(a) indicate the region of the maximum displacement in the
films, while bright and dark contrasts in Fig. 1(b) represent opposite directions of the polarization. Piezoelectric domain imaging was conducted on the edge of a sample to exclude the undesirable cantilever–sample capacitive force. We clarified the piezoresponse imaging mechanism and, thus, improved the imaging resolution by excluding the interaction [45,46]. Piezoresponse image with spatial resolution of about 10 nm could be obtained. As shown in the image, the observed domain wall width was about 17 nm. In the storage-application point of view, domain boundaries may impose physical limitation on minimum spacing between two adjacent information bits. Applying DC pulses through a conductive tip, the electric polarization—ferroelectric domain—in nano-meter size can be purposely formed in any two different directions. This domain can be labeled 0 or 1 depending on its polarization direction, thereby acting as a bit in a memory device shown in Fig. 2 with amplitude (a) and phase (b), respectively. The size of the formed ‘‘written’’ information bits was about 60 nm in diameter. In the storage application, the lateral size of the information bits reaches few tens of nano-meters and have a homogeneity in piezoelectric properties of the recording media is strongly required throughout the whole area. Detailed experimental works have been performed focusing on the dependencies of bit (i.e., induced artificial ferroelectric domains) size formed on the applied pulse width and amplitude
Fig. 1. As-received ferroelectric domain images of the laser-ablated PZT. Domain map is represented in amplitude (a) and phase (b) ( in its width. images of the first harmonic signal from lock-in amplifier. The arrowed boundary is 166 A
H. Shin et al. / Ultramicroscopy 91 (2002) 103–110
107
Fig. 2. Array of induced ‘‘written’’ information bits which average size was 60 nm in diameter imaging in amplitude (a) and in phase (b).
Fig. 3. The procedure of the retention loss phenomenon occurring in the ‘‘written’’ bit as a function of time right after the writing process: (a) 2 min, (b) 22 min, (c) 1 h, (d) 2 h, (e) 4 h 29 min and (f) 7 h 21 min.
as well as the films’ thickness. They are published elsewhere [47]. Dynamic behaviors of the ferroelectric domains formed were also investigated using the same technology. Retention loss phenomena, which is probably the spontaneous polarization reversal of the ‘‘written’’ induced ferroelectric domains without any external applied voltage, were observed as a function of time and characterized with their characteristic retention time dependent upon the size of ‘‘written’’ domains and the condition of the
writing procedures. Retention loss is a critical issue on the reliability of the stored information. Smaller ‘‘written’’ domains with longer interval of the applied pulse for inducing the domains are more stable [48]. The procedure of the retention loss phenomenon occurring in the ‘‘written’’ bit as a function of time right after the writing process is shown in Figs. 3(a)–(f): (a) 2 min, (b) 22 min, (c) 1 h, (d) 2 h, (e) 4 h 29 min and (f) 7 h 21 min. It is found that the retention loss occurs ‘‘region by region’’ showing local variation of the rate of the
H. Shin et al. / Ultramicroscopy 91 (2002) 103–110
108
350
250
300 200
Amplitude ( µ V)
Amplitude ( µ V)
250 150
100
50
200 150 100 50 0
100
100
50
50
Phase (deg.)
Phase (deg.)
0
0
-50
0
-50
-100
-100 -10
(a)
-5
0
5
-35 -30 -25 -20 -15 -10 -5 0
10
(b)
Applied Voltage (V)
5 10 15 20 25 30 35
Applied Voltage (V)
150 350 300
Amplitude ( µ V)
Amplitude ( µ V)
250 200 150 100 50
100
50
0 100
50
50
Phase (deg.)
Phase (deg.)
0 100
0
-50
-50
-100
-100 -10
(c)
0
-5
0
5
-10
10
Applied Voltage (V)
(d)
-5
0
5
10
Applied Voltage (V)
Fig. 4. Piezoelectric hysteresis loops that are deconvoluted into the amplitude and the phase. Four PZT thin films are deposited onto the bottom electrodes with different deposition techniques: (a) sol–gel (250 nm), (b) laser ablation (500 nm), (c) sputter (100 nm), (d) reactive evaporation (60 nm).
loss. A method to improve the retention characteristics is also suggested. Making the artificial domains in the oppositely pre-poled region was stable enough from the practical point of view. Works are in progress to quantify the retention characteristic time by observing relaxation at elevated temperatures. The depolarization field
may come from only the bit when the matrix is to be randomly polarized. By poling the matrix in a direction anti-parallel to the bit, one can minimize the depolarization field, which is the main driving force for the retention loss. As a result, relaxation time constant was increased by a factor of nearly 20 by making the ‘‘written’’ bits in pre-poled
H. Shin et al. / Ultramicroscopy 91 (2002) 103–110
109
matrix. Furthermore, the total retention loss can be successfully described by an extended exponential decay that implies a narrow distribution of the relaxation times of the domains. However, further research is needed to understand the mechanism clearly. Local piezoelectric measurements were also performed on the PZT films produced by different methods. The amplitude gives information about displacement within an applied DC pulse range, while phase about the coercive voltage at which the polarization switching has occurred. To accurately measure the magnitude of the coefficient, these measurements could be directly related to the longitudinal piezoelectric coefficient (d33) of the films with very high resolutions perpendicular to the film. PZT films with larger coefficient could provide larger recording signal so as to increase the signal-to-noise ratio. Figs. 4(a)–(d) show local piezoelectric response loops deconvoluted into amplitude and phase from sol–gel processed (a), laser ablated (b), sputtered (c) and reactive evaporated (d) PZT films. Piezoelectric hysteresis loops de-convoluted into phase as well as amplitude indicated that the sol–gel processed polycrystalline PZT (250 nm in thickness, in the right) showed broad range of coercive voltages, whereas sharp and well-defined coercive voltages were shown in epitaxially grown PZT films (Fig. 4(d)). This comes mainly from the poly crystalline nature of sol–gel derived films. Though the sharp AFM tip can probe highly localized piezoelectric properties in lateral (B20 nm), it actually interacts with the effective volume which contains vertical direction of the film below the tip. Hence, two or more grains, which have different spontaneous polarization directions, may contribute to the piezoresponse in the films. Local piezoresponse measurement is an important characterization tool for the evaluation of the piezoelectric recording media in the future storage devices.
with size of 60 nm diameter. The formed (written) information bits were recorded back using ACmodulation and lock-in amplification with very high spatial resolution (B10 nm). Homogeneity of the ferroelectric thin films, PZT films in this study, is the most critical issues on R=W mechanism. Retention loss phenomena, i.e. spontaneous polarization reversal of the ‘‘written’’ induced ferroelectric domains without any externally applied voltage, were observed as a function of time and characterized with their characteristic relaxation time dependent on the size of domains and the condition of the writing procedures. Retention loss is a critical issue for reliability of the stored information in practical applications. By poling the matrix in a direction anti-parallel to the bit, one can minimize the depolarization field that is the main driving force for the retention loss, so as to increase the stability of the bits. Relaxation time constant was increased by a factor of nearly 20. Local piezoelectric measurements were also performed on the PZT films produced by different methods. This measurement technique is an essential characterization tool for assessing the required properties in nano-meter scale as a recording medium. In the future storage devices using the principles of AFM and ferroelectric recording medium will replace current storage devices and will be fabricated by MEMS technology.
5. Conclusion
References
Induced ferroelectric domains as applying through the AFM tip formed information bits
Acknowledgements This work was supported partially by the Korean Ministry of Science and Technology through Creative Research Initiative (CRI) Program and National Research Laboratory (NRL). The authors would like to thank Prof. No’s research group for preparation of the sputtered PZT samples and helpful discussion.
[1] H. Shin, J.-H. Lee, K. Lee, W.K. Moon, J.U. Jeon, G. Lim, Y. Pak, J.H. Park, K.H. Yoon, IEEE Trans.
110
[2]
[3] [4] [5] [6] [7] [8]
[9] [10]
[11] [12] [13] [14] [15]
[16] [17]
[18] [19] [20]
[21] [22] [23] [24]
H. Shin et al. / Ultramicroscopy 91 (2002) 103–110 Ultrasonics, Ferroelectrics Frequency Control 47 (4) (2000) 801–807. M. Despont, J. Brugger, U. Drescher, U. Durig, W. Haberle, M. Lutwyche, H. Roithuisen, R. Widmer, H. Rohrer, G.K. Binnig, P. Vettiger, IEEE Int. Micro Electro Mech. Syst. Tech. Dig. 564 (1999) H.J. Mamin, B.D. Terris, L.S. Fan, S. Hoen, R.C. Barrett, D. Rugar, IBM J. Res. Develop. 39 (6) (1995) 681–700. F. Saurenbach, B.D. Terris, Appl. Phys. Lett. 56 (1990) 1703. O. Kosolov, A. Gruvermann, J. Hatano, K. Takahashi, H. Tokumoto, Phys. Rev. Lett. 74 (1995) 4309. G. Zavala, J.H. Fendler, S. Troiler-McKinstry, J. Appl. Phys. 81 (1997) 7480. O. Auciello, A. Gruverman, H. Tokumoto, S.A. Prakash, S. Aggarwal, R. Ramesh, MRS Bull. 23 (1) (1998) 33. A. Gruverman, H. Tokumoto, A.S. Prakash, S. Aggarwal, B. Yang, R. Ramesh, O. Auciello, T. Venkatesan, Appl. Phys. Lett. 71 (1997) 3492–3494. P. Guethner, K. Dransfield, Appl. Phys. Lett. 61 (1992) 1137. T. Hidaka, T. Maruyama, M. Saitoh, N. Mikoshiba, M. Shimizu, T. Shiosaki, L.A. Wills, R. Hiskes, S.A. Dicarolis, J. Amano, Appl. Phys. Lett. 68 (1996) 2358. J.A. Christman, R.R. Woolcott Jr., A.I. Kingon, R.J. Nemanich, Appl. Phys. Lett. 73 (260) (1998) 3851. G.D. Hu, J.B. Xu, I.H. Wilson, Appl. Phys. Lett. 75 (11) (1999) 1610–1612. L.M. Eng, Nanotechnology 10 (1999) 405–411. C. Durkan, M.E. Welland, D.P. Chu, P. Migliorato, Phys. Rev. B 60 (23) (1999) 16198–16204. A. Roelofs, U. Bottger, R. Waser, F. Schlaphof, S. Trogisch, L.M. Eng, Appl. Phys. Lett. 77 (21) (2000) 3444–3446. V. Likodimos, X.K. Orlik, L. Pardi, M. Labardi, M. Allegrini, J. Appl. Phys. 87 (1) (2000) 443–451. C.S. Ganpule, V. Nagarajan, S.B. Ogale, A.L. Roytburd, E.D. Williams, R. Ramesh, Appl. Phys. Lett. 77 (20) (2000) 3275–3277. A. Gruverman, O. Auciello, H. Tokumoto, Appl. Phys. Lett. 69 (1996) 3191. E.L. Colla, S. Hong, D.V. Taylor, A.K. Tangantsev, N. Setter, Appl. Phys. Lett. 72 (1997) 2478. C.H. Ahn, T. Tybell, L. Antognazza, K. Char, R.H. Hammond, M.R. Beasley, O. Fischer, J.-M. Triscone, Science 276 (1997) 1100. A. Gruverman, O. Auciello, H. Tokumoto, J. Vac. Sci. Technol. B 14 (2) (1996) 602–605. T. Maruyama, M. Saitoh, I. Sakai, T. Hidaka, Y. Yano, T. Noguchi, Appl. Phys. Lett. 73 (1998) 3524–3526. O. Kolosov, A. Gruverman, J. Hatano, K. Takahashi, H. Tokumoto, Phys. Rev. Lett. 4 (21) (1995) 4309. O. Auciello, A. Gruverman, H. Tokumoto, S.A. Prakash, S. Aggarwal, R. Ramesh, MRS Bull. 23 (1) (1998) 33.
[25] G. Zavala, H. Janos, S. Fendler, Trolier-McKinstry, J. Appl. Phys. 81 (11) (1997) 7480. [26] S. Hong, E.L. Colla, E. Kim, K. No, D.V. Taylor, A.K. Tagantsev, N. Setter, J. Appl. Phys. 86 (1998) 607. [27] H. Shin, K. Lee, G. Lim, J.U. Jeon, Y. Pak, S. Hong, K. No, SPIE Smart Struct. Mater. 3675 (1999) 94–103. [28] E.L. Colla, S. Hong, D.V. Taylor, A.K. Tangantsev, N. Setter, Appl. Phys. Lett. 72 (1997) 2478. [29] T. Hidaka, T. Maruyama, M. Saitoh, N. Mikoshiba, M. Shimizu, T. Shiosaki, L.A. Wills, R. Hiskes, S.A. Dicarolis, J. Amano, Appl. Phys. Lett. 68 (1996) 2358–2359. [30] T. Itoh, T. Suga, J. Vac. Sci. Tech. B12 (3) (1994) 1581–1585. [31] T. Itho, T. Suga, Nanotechnology 4 (1993) 218–224. [32] M. Tortonese, R.C. Barrett, C.F. Quate, Appl. Phys. Lett. 62 (1993) 834. [33] Toru Fujii, Shunji Watanabe, Masatoshi Suzuki, Takamitsu Fujiu, J. Vac. Sci. Tech. B 13 (1995) 1119. [34] R. Linnemann, T. Gotszalk, I.W. Rangelow, P. Dumania, E. Oesterschulze, J. Vac. Sci. Tech. B 14 (2) (1996) 856–860. [35] N. Blanc, J. Brugger, N.F. de Rooji, U. Durig, J. Vac. Sci. Tech. B 14 (1996) 901. [36] T. Itoh, C. Lee, T. Suga, Appl. Phys. Lett. 69 (14) (1996) 2036–2038. [37] S.C. Minne, S.R. Manalis, A. Atalar, C.F. Quate, Appl. Phys. Lett. 68 (10) (1996) 1427–1429. [38] J. Chu, T. Itoh, C. Lee, T. Suga, K. Watanabe, J. Vac. Sci. Tech. B 14 (1996) 901. [39] E. Grochowski, R.F. Hoyt, IEEE Trans. Magnetics 32 (3) (1996) 1850–1854. [40] S.C. Minne, Ph. Flueckiger, H.T. Soh, C.F. Quate, J. Vac. Sci. Tech. B 13 (3) (1995) 1380. [41] M. Lutwyche, C. Andreoli, G. Binnig, J. Brugger, U. Drechsler, W. Haeberle, H. Rohrer, H. Rothuizen, P. Vettiger, Proceedings IEEE International Workshop on Microelectro Mechanical Systems (MEMS 98), Heidelberg, Germany, January 25–29, 1998. [42] B.W. Chui, T.D. Stowe, T.W. Kenny, H.J. Mamin, B.D. Terris, D. Rugar, Appl. Phys. Lett. 69 (18) (1996) 2767–2769. [43] R.P. Ried, H.J. Mamin, B.D. Terris, L.S. Fan, D. Rugar, J. Microelectromech. Syst. 6 (1997) 294. [44] T. Itoh, C. Lee, J. Chu,T. Suga, 1997, to be published. [45] S. Hong, E. Kim, J. Woo, H. Shin, J.-U. Jeon, E.Y. Pak, E.L. Colla, N. Setter, K. No, J. Appl. Phys. 89 (2) (2001) 1377–1386. [46] S. Hong, J. Woo, H. Shin, E. Kim, K.-H. Kim, J.U. Jeon, Y. Pak, K. No, J. Vac. Sci. Tech. B 18 (6) (2000) 2688–2691. [47] H. Shin, J. Woo, S. Hong, J.U. Jeon, Y. Pak, K. No, Integrat. Ferroelectrics 31 (2000) 163–171. [48] K. No, H.W. Song, J. Hong, J. Woo, H. Shin, S. Hong, Ferroelectrics, 2001, accepted for the publication.
