- Email: [email protected]
Self-sensing piezoresistive cantilever and its magnetic force microscopy applications
Ultramicroscopy 91 (2002) 63–72
Self-sensing piezoresistive cantilever and its magnetic force microscopy applications Hiroshi Takahashi*, Kazunori Ando, Yoshiharu Shirakawabe Scientific Instruments Division, Seiko Instruments Inc., 563, Takatsukashin-den, Matsudo-shi, 270-2222 Chiba, Japan Received 30 May 2001; received in revised form 17, December 2001
Abstract A newly developed Si self-sensing piezoresistive cantilever is presented. Si piezoresistive cantilevers for scanning microscopy are fabricated by Si micro-machining technique. The sensitivity of the piezoresistive cantilever is comparable to the current laser detecting system. Topographic images are successfully obtained with the piezoresistive cantilever and some comparisons are made with the laser detecting system. Furthermore, the magnetic film (Co–Cr–Pt) is coated on the tip of the piezoresistive cantilever for magnetic force microscopy (MFM) application. The magnetic images are successfully obtained with the self-sensing MFM piezoresistive cantilever. The self-sensing piezoresistive cantilevers have been successfully applied in scanning probe microscopy and MFM. r 2002 Elsevier Science B.V. All rights reserved. PACS: 07.79; 07.79.P; 72.20.F Keywords: Si micro-cantilever; Piezoresistor; Scanning probe microscopy; Magnetic force microscopy
1. Introduction Scanning probe microscopy (SPM) [1–4] is widely used technique to analyze very small size material surface down to sub-micron order for material research, quality control and metrology. Recently, SPM technology is highly attractive for researching the high-density memory devices such as digital versatile disk (DVD) or other memory devices. In these days, studying physical properties, such as mechanical, electrical, and magnetic *Corresponding author. Tel.: +81-47-391-2004; fax: +8147-391-0960. E-mail address: [email protected] (H. Takahashi).
properties, of very small area surface is important for material research. SPM is expected to be powerful technique to measure such physical properties in addition to its topographic imaging [5–10]. Optical laser detecting system is commonly used technique detecting cantilever deformation caused by tip–sample force to control z-distance between tip and sample surface. Although the sensitivity of the optical laser detecting system is very high, its equipment is complicated and needs much space to install on instruments. Another problem of the optical laser detecting system is that it is not desirable for some applications such as high vacuum measurement, in situ solution
0304-3991/02/$ - see front matter r 2002 Elsevier Science B.V. All rights reserved. PII: S 0 3 0 4 - 3 9 9 1 ( 0 2 ) 0 0 0 8 3 - 9
64
H. Takahashi et al. / Ultramicroscopy 91 (2002) 63–72
measurement, and optical functioning material measurement. To solve above problems, several self-sensing cantilever techniques had been reported such as capacitive cantilevers [11], piezoelectric (ZnO, PZT) cantilevers [12], and piezoresistive cantilevers [13,14]. We present newly developed Si piezoresistive cantilever having constricted part on the cantilever to concentrate the deformation strain caused by tip–sample force to achieve much higher sensitivity. The sensitivity of the piezoresistive cantilever is comparable to the optical laser detecting system and we successfully obtained topography images. We have made some comparison with the optical laser detecting system. Furthermore, we coated the ferromagnetic film (Co–Cr–Pt) on a tip of the piezoresistive cantilever for magnetic force microscopy (MFM) application. We successfully obtained magnetic images with the piezoresistive MFM cantilever.
2. Experiment 2.1. Fabrication Fig. 1 shows the schematic diagram of the piezoresistive cantilever. We fabricated two different length cantilever, PRC120 for 120 mm long, and PRC400 for 400 mm long. The width of both cantilevers is 50 mm. The constriction size is 5 mm in length (l) and 4 mm in width (w). Fig. 2 shows the fabrication steps of the piezoresistive cantilevers. The substrate is 4-in siliconon-insulator (SOI) wafer with 16 mm thick n-type
Fig. 2. Process flow diagram of self-sensing piezoresistive cantilever.
Fig. 1. Schematic diagram of the self-sensing piezoresistive cantilever and basic design.
H. Takahashi et al. / Ultramicroscopy 91 (2002) 63–72
silicon film on 1 mm buried oxide film. The oxide film is thermally grown on silicon film and patterned with buffered HF (BHF) to form the oxide mask for tip etching (Fig. 2(a)). The tips are formed by isotropic reactive ion etching (RIE) with SF6 gas system (Fig. 2(b)). The tips are sharpened by subsequent thermal oxidation. (Fig. 2(c)) The cantilevers are formed by anisotropic RIE with SF6/O2 gas system (Fig. 2(d)). The piezoresistor is formed by boron ion implantation and subsequent thermally dopant activation (Fig. 2(e)). The contact openings are etched with BHF for connecting piezoresistor and Al wiring. The Al film is deposited by metal sputtering method for 1 mm thick. The Al film is patterned for wiring of the piezoresistor by chemical wet etchant and annealed at 4301C for 1 h to alloy Si–Al contact (Fig. 2(f)). The backside oxide layer is patterned with BHF and subsequently supporting Si layer is etched with 40% KOH aqueous solution at 601C (Fig. 2(g)). The last step is removing the buried oxide film by BHF/glycerol to release the cantilever (Fig. 2(h)). Fig. 3 shows photograph and scanning electron microscope images of the fabricated piezoresistive cantilever. The thickness of the cantilever is 5–6 mm and the tip height is 6–8 mm and typical apex radius is about 20 nm measured from SEM images. The piezoresistive cantilever characteristics are listed in Table 1. Fabricating the piezoresistive cantilever for magnetic force measurement, thin titanium film (5 nm thick for adhesion promoter of silicon tip and magnetic film) and the magnetic film (Co–Cr– Pt 70.9:17.2:11.9 mol%, 50 and 24 nm in thickness) are deposited by sputtering just before KOH substrate etching. The titanium and the magnetic film are patterned by lift-off technique or etched by aqua regia at 501C (for magnetic film) and diluted HF aqueous solution (for titanium film) to leave the film only on the tip. Subsequently, the cantilever release steps are performed to finish the
65
fabrication of the piezoresistive magnetic force microscope cantilevers. Fig. 4 shows the SEM image of the magnetic film coated tip.
2.2. Sensitivity and topography measurement We measured the sensitivity of piezoresistive cantilever with the SPA300 system setup (with SPI3800N Controller, Seiko Instruments Inc.). The piezoresistive cantilever and its reference are connected to Wheatstone bridge and the bridge is connected to the differential amplifier to obtain the differential voltage signal. The signal is amplified 2000 times by the wide-range amplifier (Fig. 5). For the sensitivity measurement, the amplified signal (cantilever deformation signal) is monitored with oscilloscope. The Wheatstone bridge circuit setup and laser detection system for feedback control are both installed in SPI300 system. The cantilever is approached to a sample until the tip is contact to the sample surface. Then the cantilever is further pushed 10 nm down to z-direction and voltage change is read directly from oscilloscope. The measured sensitivity of each cantilever is listed in Table 1. The sensitivity is measured at various bridge bias voltages to study sensitivity dependence on bridge bias voltage. The sensitivity is increased with bridge bias voltage increases and the result is shown in Fig. 6. The topographic image of titanium film (100 nm thick) on a Si wafer is measured with the piezoresistive cantilever (PRC120) in dynamic mode at 1.0 V bridge bias voltage. The piezoresistive signal is connected to the SPI system for feedback control and imaging. The laser detecting system is also installed in SPI system so that self-sensing image and laser detecting image are obtained at the same time. Fig. 7 shows both self-sensing image and laser detecting image of the titanium film. We also measured the topographic image of the same titanium film with commercially available cantilever (Nanosensors NCHR k ¼ 40 N/m) and laser detecting system for comparison (Fig. 8). Both topographic image data are listed in Table 2.
66
H. Takahashi et al. / Ultramicroscopy 91 (2002) 63–72
Fig. 3. Photograph and SEM images of the self-sensing piezoresistive cantilever. (a) Photograph of the cantilever (PRC120) and its reference. (b) SEM image of the cantilever beam free end and the tip. (c) SEM image of the tip.
2.3. MFM measurement MFM image of magnetic recording tape is measured with the magnetic film coated piezoresistive cantilevers (50 nm thick magnetic film coated on tip) and SPA300 setup (the same instruments used in topography measurement). Also MFM image of the same sample is measured with commercially available cantilever (MFMR Nanosensors) cantilever and laser detection system. Fig. 9 shows the MFM images of magnetic recording tape.
MFM image of the same sample is measured with the 24 nm thick magnetic film coated piezoresistive cantilevers under atmospheric air and high vacuum at 1.0 E4 Pa to study the lower magnetic moment probe measurement (see Fig. 10). The very low contrasted data recorded lines are observed at upper half of the MFM image measured in the air with PRC400 magnetic coated on the tip because of its low magnetic sensitivity. The strong and clear contrasted data recorded lines are observed at upper one-third of MFM image measured in vacuum (1.0 E–4 Pa).
H. Takahashi et al. / Ultramicroscopy 91 (2002) 63–72 Table 1 The properties of fabricated piezoresistive cantilevers
Lever length (mm)a Lever width (mm)a Lever thickness (mm)a Constriction size l w (mm)a Resonant frequency (kHz) Quality factorb Tip radius (nm) Tip height (mm) Force constant (N/m)c Resistance (O) Sensitivity DR=R (1/nm) Sensitivity (mV/nm)d
67
3. Discussion
PRC120
PRC400
3.1. Sensitivity and topography measurement
120 50 4–5 54 250–300 350–500 20–50 6–8 30–40 400–700 3.5E5 18.5 at Vb ¼ 1 V
400 50 4–5 54 35–40 100–160 20–50 6–8 2–4 400–700 0.5E5 6.5 at Vb ¼ 2 V
The sensitivity of PRC120 (120 mm long, k ¼ 30240 N/m) is comparable to the laser detection system. The sensitivity of PRC400 (400 mm long, k ¼ 224 N/m) is low compared to the sensitivity of PRC120. It is explained by a simple rectangular cantilever beam modeling. The piezoresistive sensitivity of rectangular cantilever beam is
a
Designed size (Mask size). Measured in atmospheric condition. c Calculated. d Measured with 2000 amplified. The Sensitivity of SPA300 laser detection is about 26.0 mV/ nm.
DR ¼ ps; R
b
ð1Þ
6lk DZ; wt2
ð2Þ
DR 6l ¼ p 2 kDZ; R wt
ð3Þ
s¼
where R is the resistivity of piezoresistor, p is the piezoresistive constant of material, s is the maximum strain stress, l is the length of cantilever beam, w is the width of cantilever beam, t is the thickness of cantilever beam, k is the spring constant of cantilever beam, and Z is the displacement of cantilever beam end. The sensitivity of piezoresistive cantilever is dependent on its spring constant (force constant), the higher the spring constant, the higher the sensitivity. In the present case, the cantilever beam has constricted part for concentrating the deformation strain. Formula (3) turns to DR 6l w=2wc ; ¼ p 2 kDZ R wt 1 þ ðw=2wc 1Þ 1 ð1 2lc =lÞ3 ð4Þ Fig. 4. SEM image of the magnetic film (Co–Cr–Pt) coated tip.
It is because that the air dumping effect on cantilever beam is removed and magnetic sensitivity is improved in the same way as quality factor. The magnetic sensitivity of each measurement is listed in Table 3.
where wc is the constriction width, lc is the constriction length. (This formula is calculated in single beam constriction structure approximation.) The sensitivity dependence on bridge bias is explained that the net sensitivity is ratio of DR and R and dose not change but the deformation signal is measured by voltage differential through Wheatstone bridge. The more electrical current flow to the piezoresistor, the higher the output signal voltage.
H. Takahashi et al. / Ultramicroscopy 91 (2002) 63–72
68
Fig. 5. Circuit diagram of the self-sensing piezoresistive cantilever setup connected to Wheatstone bridge.
Fig. 6. Sensitivity vs. bridge bias voltage: (a) for PRC120 (length=120 mm, k ¼ 30240 N/m); and (b) for PRC400 (length=400 mm, k ¼ 224 N/m).
The sensitivity is higher with applied bridge bias but the noise is also increases. It is necessary to choose the bridge bias to gain high enough sensitivity and keep low enough noise level for imaging. 3.2. MFM measurement MFM piezoresistive cantilever with 50 nm magnetic film on tip has as the same ability for imaging as laser detection MFM measurement. The magnetic sensitivity is higher with PRC400 (400 mm long, k ¼ 224 N/m) compared to PRC120
(120 mm long, 30–40 N/m). It is because the tip is lifted certain distance from the sample during magnetic imaging and the higher spring constant cantilever is mechanically insensitive to the magnetic force gradient compared to the lower spring constant cantilever. So piezoresistive sensitivity and magnetic sensitivity are in trade-off relation, cantilever dimensions should be carefully designed to achieve appropriate performance. MFM measurement of lower magnetic moment probing is successfully performed under high vacuum condition with 24 nm thick magnetic film
H. Takahashi et al. / Ultramicroscopy 91 (2002) 63–72
69
Fig. 8. Topographic image of Ti film (100 nm thick) on Si wafer imaged with nanosensors cantilever NCHR (k ¼ 40 N/m) by optical detection.
Table 2 Ti film (100 nm thick) on Si wafer topographic data
Fig. 7. Topographic images of Ti film (100 nm thick) on Si wafer. (a) Imaged with self-sensing piezoresistive cantilever. (b) Imaged with laser detection signal.
coated MFM piezoresistive cantilever. It is in general very difficult to observe magnetic image with the low magnetic moment probe (tip with thinner magnetic film coat) in atmospheric air because of its low magnetic sensitivity and the air dumping effect on cantilever (Fig. 10(b)). Nevertheless, magnetic image of the sample is easily
Measure mode
Piezoresistive self-sensing (nm)
Piezoresistive laser detection (nm)
Nanosensors laser detection (nm)
Ra RMS Max P2V
1.569 8.020 1.734
1.245 8.452 1.628
1.167 15.35 1.574
observed in high vacuum condition with the newly developed thinner magnetic film MFM piezoresistive cantilever. Under high vacuum condition, the air dumping effect on cantilever beam is removed and it is possible to gain high enough magnetic sensitivity for imaging with lower magnetic moment probe. It is applicable for measuring the sample of which magnetic field is easily interfered by outer magnetic field even by probing. The piezoresistive self-sensing cantilever is versatile and promising technology for SPM application and will be powerful tool for complicated and advanced material research.
70
H. Takahashi et al. / Ultramicroscopy 91 (2002) 63–72
Fig. 9. MFM images of magnetic recording tape. (a) Imaged with PRC120 cantilever magnetic film coated on the tip (magnetic film thickness=50 nm). (b) Imaged with PRC400 cantilever magnetic film coated on the tip (magnetic film thickness=50 nm). (c) Imaged with MFM cantilever MFMR (length=225, k ¼ 2:8 N/m, magnetic film thickness=40 nm) by optical detection.
4. Conclusion The self-sensing piezoresistive cantilever with higher sensitivity is successfully fabricated. The
sensitivity of the piezoresistive cantilever is comparable to the current laser detecting system and topographic images are successfully demonstrated with the piezoresistive cantilever.
H. Takahashi et al. / Ultramicroscopy 91 (2002) 63–72
71
Fig. 10. Topographic and MFM images of magnetic recording tape imaged with PRC400 cantilever magnetic film coated on the tip (magnetic film thickness=24 nm). (a) Topographic image taken in the atmospheric pressure. (b) MFM image taken in the atmospheric pressure. (c) Topographic image taken in vacuum (at 1.0 E–4 Pa). (d) MFM image taken in vacuum (at 1.0 E–4 Pa).
Furthermore, MFM piezoresistive cantilever with the magnetic film (Co–Cr–Pt) on the tip is also fabricated and tested. We successfully obtain magnetic images with the piezoresistive MFM cantilever. Piezoresistive cantilevers with lower magnetic moment probe are also fabricated and the MFM image is observed under high vacuum condition.
Acknowledgements The authors wish to thank M. Despont, P. Vettiger, and G. Binnig from IBM Zurich Research Laboratory for their assistance of cantilever design and fabrication. The authors also thank T. Yamaoka for assistance and useful discussion on MFM measurement.
H. Takahashi et al. / Ultramicroscopy 91 (2002) 63–72
72
Table 3 Magnetic sensitivity measured with various cantilevers Magnetic sensitivity (deg.) PRC120 with 50 nm magnetic film PRC400 with 50 nm magnetic film PRC400 with 24 nm magnetic film MFM cantilever MFMR
0.22 3.05 2.1 (atmospheric pressure) 47.5 (in vacuum at 1.0E4 Pa) 3.65
References [1] G. Binnig, H. Rohrer, Ch. Gerber, E. Weibel, Phys. Rev. Lett. 49 (1982) 57. [2] H.K. Wickramasinghe, J. Vac. Sci. Technol. A 8 (1990) 363. [3] G. Binnig, C.F. Quate, Ch. Gerber, Phys. Rev. Lett. 56 (1986) 930.
[4] O. Wolter, Th. Bayer, J. Grescher, J. Vac. Sci. Technol. B 9 (1991) 1353. [5] C. Mate, G. McClelland, R. Erlandsson, S. Chiang, Phy. Rev. Lett. 59 (1987) 1942. [6] K. Yamanaka, E. Tomita, Jpn. J. Appl. Phys. 34 (1995) 3350. [7] M. Tanimoto, O. Vatel, J. Vac. Sci. Technol. B 94 (1996) 1547. [8] Y. Martin, H. Wickramasinghe, Appl. Phys. Lett. 50 (1987) 1455. [9] U. Hartmann, J. Vac. Sci. Technol. A 8 (1990) 411. [10] A. Wadas, P. Gruetter, H. Guentherodt, J. Vac. Sci. Technol. A 8 (1990) 416. [11] N. Blanc, J. Brugger, N.F. de Rooij, J. Vac. Sci. Technol. B 14 (1996) 903. [12] S. Akamine, T. Alabrecht, M. Zdeblick, C.F. Quate, IEEE Elec. Dev. Lett. EDL-10 (1989) 490. [13] F. Giessibl, B. Trafas, Rev. Sci. Instrum. 65 (1994) 1923. [14] M. Despont, H. Takahashi, S. Ichihara, Y. Shirakawabe, N. Shimizu, A. Inoue, W. Haeberle, G. Binnig, P. Vettiger, Proceedings on the IEEE Micro Electro Mechanical Systems 2000, MEMS2000, 2000, p. 126.
Self-sensing piezoresistive cantilever and its magnetic force microscopy applications Hiroshi Takahashi*, Kazunori Ando, Yoshiharu Shirakawabe Scientific Instruments Division, Seiko Instruments Inc., 563, Takatsukashin-den, Matsudo-shi, 270-2222 Chiba, Japan Received 30 May 2001; received in revised form 17, December 2001
Abstract A newly developed Si self-sensing piezoresistive cantilever is presented. Si piezoresistive cantilevers for scanning microscopy are fabricated by Si micro-machining technique. The sensitivity of the piezoresistive cantilever is comparable to the current laser detecting system. Topographic images are successfully obtained with the piezoresistive cantilever and some comparisons are made with the laser detecting system. Furthermore, the magnetic film (Co–Cr–Pt) is coated on the tip of the piezoresistive cantilever for magnetic force microscopy (MFM) application. The magnetic images are successfully obtained with the self-sensing MFM piezoresistive cantilever. The self-sensing piezoresistive cantilevers have been successfully applied in scanning probe microscopy and MFM. r 2002 Elsevier Science B.V. All rights reserved. PACS: 07.79; 07.79.P; 72.20.F Keywords: Si micro-cantilever; Piezoresistor; Scanning probe microscopy; Magnetic force microscopy
1. Introduction Scanning probe microscopy (SPM) [1–4] is widely used technique to analyze very small size material surface down to sub-micron order for material research, quality control and metrology. Recently, SPM technology is highly attractive for researching the high-density memory devices such as digital versatile disk (DVD) or other memory devices. In these days, studying physical properties, such as mechanical, electrical, and magnetic *Corresponding author. Tel.: +81-47-391-2004; fax: +8147-391-0960. E-mail address: [email protected] (H. Takahashi).
properties, of very small area surface is important for material research. SPM is expected to be powerful technique to measure such physical properties in addition to its topographic imaging [5–10]. Optical laser detecting system is commonly used technique detecting cantilever deformation caused by tip–sample force to control z-distance between tip and sample surface. Although the sensitivity of the optical laser detecting system is very high, its equipment is complicated and needs much space to install on instruments. Another problem of the optical laser detecting system is that it is not desirable for some applications such as high vacuum measurement, in situ solution
0304-3991/02/$ - see front matter r 2002 Elsevier Science B.V. All rights reserved. PII: S 0 3 0 4 - 3 9 9 1 ( 0 2 ) 0 0 0 8 3 - 9
64
H. Takahashi et al. / Ultramicroscopy 91 (2002) 63–72
measurement, and optical functioning material measurement. To solve above problems, several self-sensing cantilever techniques had been reported such as capacitive cantilevers [11], piezoelectric (ZnO, PZT) cantilevers [12], and piezoresistive cantilevers [13,14]. We present newly developed Si piezoresistive cantilever having constricted part on the cantilever to concentrate the deformation strain caused by tip–sample force to achieve much higher sensitivity. The sensitivity of the piezoresistive cantilever is comparable to the optical laser detecting system and we successfully obtained topography images. We have made some comparison with the optical laser detecting system. Furthermore, we coated the ferromagnetic film (Co–Cr–Pt) on a tip of the piezoresistive cantilever for magnetic force microscopy (MFM) application. We successfully obtained magnetic images with the piezoresistive MFM cantilever.
2. Experiment 2.1. Fabrication Fig. 1 shows the schematic diagram of the piezoresistive cantilever. We fabricated two different length cantilever, PRC120 for 120 mm long, and PRC400 for 400 mm long. The width of both cantilevers is 50 mm. The constriction size is 5 mm in length (l) and 4 mm in width (w). Fig. 2 shows the fabrication steps of the piezoresistive cantilevers. The substrate is 4-in siliconon-insulator (SOI) wafer with 16 mm thick n-type
Fig. 2. Process flow diagram of self-sensing piezoresistive cantilever.
Fig. 1. Schematic diagram of the self-sensing piezoresistive cantilever and basic design.
H. Takahashi et al. / Ultramicroscopy 91 (2002) 63–72
silicon film on 1 mm buried oxide film. The oxide film is thermally grown on silicon film and patterned with buffered HF (BHF) to form the oxide mask for tip etching (Fig. 2(a)). The tips are formed by isotropic reactive ion etching (RIE) with SF6 gas system (Fig. 2(b)). The tips are sharpened by subsequent thermal oxidation. (Fig. 2(c)) The cantilevers are formed by anisotropic RIE with SF6/O2 gas system (Fig. 2(d)). The piezoresistor is formed by boron ion implantation and subsequent thermally dopant activation (Fig. 2(e)). The contact openings are etched with BHF for connecting piezoresistor and Al wiring. The Al film is deposited by metal sputtering method for 1 mm thick. The Al film is patterned for wiring of the piezoresistor by chemical wet etchant and annealed at 4301C for 1 h to alloy Si–Al contact (Fig. 2(f)). The backside oxide layer is patterned with BHF and subsequently supporting Si layer is etched with 40% KOH aqueous solution at 601C (Fig. 2(g)). The last step is removing the buried oxide film by BHF/glycerol to release the cantilever (Fig. 2(h)). Fig. 3 shows photograph and scanning electron microscope images of the fabricated piezoresistive cantilever. The thickness of the cantilever is 5–6 mm and the tip height is 6–8 mm and typical apex radius is about 20 nm measured from SEM images. The piezoresistive cantilever characteristics are listed in Table 1. Fabricating the piezoresistive cantilever for magnetic force measurement, thin titanium film (5 nm thick for adhesion promoter of silicon tip and magnetic film) and the magnetic film (Co–Cr– Pt 70.9:17.2:11.9 mol%, 50 and 24 nm in thickness) are deposited by sputtering just before KOH substrate etching. The titanium and the magnetic film are patterned by lift-off technique or etched by aqua regia at 501C (for magnetic film) and diluted HF aqueous solution (for titanium film) to leave the film only on the tip. Subsequently, the cantilever release steps are performed to finish the
65
fabrication of the piezoresistive magnetic force microscope cantilevers. Fig. 4 shows the SEM image of the magnetic film coated tip.
2.2. Sensitivity and topography measurement We measured the sensitivity of piezoresistive cantilever with the SPA300 system setup (with SPI3800N Controller, Seiko Instruments Inc.). The piezoresistive cantilever and its reference are connected to Wheatstone bridge and the bridge is connected to the differential amplifier to obtain the differential voltage signal. The signal is amplified 2000 times by the wide-range amplifier (Fig. 5). For the sensitivity measurement, the amplified signal (cantilever deformation signal) is monitored with oscilloscope. The Wheatstone bridge circuit setup and laser detection system for feedback control are both installed in SPI300 system. The cantilever is approached to a sample until the tip is contact to the sample surface. Then the cantilever is further pushed 10 nm down to z-direction and voltage change is read directly from oscilloscope. The measured sensitivity of each cantilever is listed in Table 1. The sensitivity is measured at various bridge bias voltages to study sensitivity dependence on bridge bias voltage. The sensitivity is increased with bridge bias voltage increases and the result is shown in Fig. 6. The topographic image of titanium film (100 nm thick) on a Si wafer is measured with the piezoresistive cantilever (PRC120) in dynamic mode at 1.0 V bridge bias voltage. The piezoresistive signal is connected to the SPI system for feedback control and imaging. The laser detecting system is also installed in SPI system so that self-sensing image and laser detecting image are obtained at the same time. Fig. 7 shows both self-sensing image and laser detecting image of the titanium film. We also measured the topographic image of the same titanium film with commercially available cantilever (Nanosensors NCHR k ¼ 40 N/m) and laser detecting system for comparison (Fig. 8). Both topographic image data are listed in Table 2.
66
H. Takahashi et al. / Ultramicroscopy 91 (2002) 63–72
Fig. 3. Photograph and SEM images of the self-sensing piezoresistive cantilever. (a) Photograph of the cantilever (PRC120) and its reference. (b) SEM image of the cantilever beam free end and the tip. (c) SEM image of the tip.
2.3. MFM measurement MFM image of magnetic recording tape is measured with the magnetic film coated piezoresistive cantilevers (50 nm thick magnetic film coated on tip) and SPA300 setup (the same instruments used in topography measurement). Also MFM image of the same sample is measured with commercially available cantilever (MFMR Nanosensors) cantilever and laser detection system. Fig. 9 shows the MFM images of magnetic recording tape.
MFM image of the same sample is measured with the 24 nm thick magnetic film coated piezoresistive cantilevers under atmospheric air and high vacuum at 1.0 E4 Pa to study the lower magnetic moment probe measurement (see Fig. 10). The very low contrasted data recorded lines are observed at upper half of the MFM image measured in the air with PRC400 magnetic coated on the tip because of its low magnetic sensitivity. The strong and clear contrasted data recorded lines are observed at upper one-third of MFM image measured in vacuum (1.0 E–4 Pa).
H. Takahashi et al. / Ultramicroscopy 91 (2002) 63–72 Table 1 The properties of fabricated piezoresistive cantilevers
Lever length (mm)a Lever width (mm)a Lever thickness (mm)a Constriction size l w (mm)a Resonant frequency (kHz) Quality factorb Tip radius (nm) Tip height (mm) Force constant (N/m)c Resistance (O) Sensitivity DR=R (1/nm) Sensitivity (mV/nm)d
67
3. Discussion
PRC120
PRC400
3.1. Sensitivity and topography measurement
120 50 4–5 54 250–300 350–500 20–50 6–8 30–40 400–700 3.5E5 18.5 at Vb ¼ 1 V
400 50 4–5 54 35–40 100–160 20–50 6–8 2–4 400–700 0.5E5 6.5 at Vb ¼ 2 V
The sensitivity of PRC120 (120 mm long, k ¼ 30240 N/m) is comparable to the laser detection system. The sensitivity of PRC400 (400 mm long, k ¼ 224 N/m) is low compared to the sensitivity of PRC120. It is explained by a simple rectangular cantilever beam modeling. The piezoresistive sensitivity of rectangular cantilever beam is
a
Designed size (Mask size). Measured in atmospheric condition. c Calculated. d Measured with 2000 amplified. The Sensitivity of SPA300 laser detection is about 26.0 mV/ nm.
DR ¼ ps; R
b
ð1Þ
6lk DZ; wt2
ð2Þ
DR 6l ¼ p 2 kDZ; R wt
ð3Þ
s¼
where R is the resistivity of piezoresistor, p is the piezoresistive constant of material, s is the maximum strain stress, l is the length of cantilever beam, w is the width of cantilever beam, t is the thickness of cantilever beam, k is the spring constant of cantilever beam, and Z is the displacement of cantilever beam end. The sensitivity of piezoresistive cantilever is dependent on its spring constant (force constant), the higher the spring constant, the higher the sensitivity. In the present case, the cantilever beam has constricted part for concentrating the deformation strain. Formula (3) turns to DR 6l w=2wc ; ¼ p 2 kDZ R wt 1 þ ðw=2wc 1Þ 1 ð1 2lc =lÞ3 ð4Þ Fig. 4. SEM image of the magnetic film (Co–Cr–Pt) coated tip.
It is because that the air dumping effect on cantilever beam is removed and magnetic sensitivity is improved in the same way as quality factor. The magnetic sensitivity of each measurement is listed in Table 3.
where wc is the constriction width, lc is the constriction length. (This formula is calculated in single beam constriction structure approximation.) The sensitivity dependence on bridge bias is explained that the net sensitivity is ratio of DR and R and dose not change but the deformation signal is measured by voltage differential through Wheatstone bridge. The more electrical current flow to the piezoresistor, the higher the output signal voltage.
H. Takahashi et al. / Ultramicroscopy 91 (2002) 63–72
68
Fig. 5. Circuit diagram of the self-sensing piezoresistive cantilever setup connected to Wheatstone bridge.
Fig. 6. Sensitivity vs. bridge bias voltage: (a) for PRC120 (length=120 mm, k ¼ 30240 N/m); and (b) for PRC400 (length=400 mm, k ¼ 224 N/m).
The sensitivity is higher with applied bridge bias but the noise is also increases. It is necessary to choose the bridge bias to gain high enough sensitivity and keep low enough noise level for imaging. 3.2. MFM measurement MFM piezoresistive cantilever with 50 nm magnetic film on tip has as the same ability for imaging as laser detection MFM measurement. The magnetic sensitivity is higher with PRC400 (400 mm long, k ¼ 224 N/m) compared to PRC120
(120 mm long, 30–40 N/m). It is because the tip is lifted certain distance from the sample during magnetic imaging and the higher spring constant cantilever is mechanically insensitive to the magnetic force gradient compared to the lower spring constant cantilever. So piezoresistive sensitivity and magnetic sensitivity are in trade-off relation, cantilever dimensions should be carefully designed to achieve appropriate performance. MFM measurement of lower magnetic moment probing is successfully performed under high vacuum condition with 24 nm thick magnetic film
H. Takahashi et al. / Ultramicroscopy 91 (2002) 63–72
69
Fig. 8. Topographic image of Ti film (100 nm thick) on Si wafer imaged with nanosensors cantilever NCHR (k ¼ 40 N/m) by optical detection.
Table 2 Ti film (100 nm thick) on Si wafer topographic data
Fig. 7. Topographic images of Ti film (100 nm thick) on Si wafer. (a) Imaged with self-sensing piezoresistive cantilever. (b) Imaged with laser detection signal.
coated MFM piezoresistive cantilever. It is in general very difficult to observe magnetic image with the low magnetic moment probe (tip with thinner magnetic film coat) in atmospheric air because of its low magnetic sensitivity and the air dumping effect on cantilever (Fig. 10(b)). Nevertheless, magnetic image of the sample is easily
Measure mode
Piezoresistive self-sensing (nm)
Piezoresistive laser detection (nm)
Nanosensors laser detection (nm)
Ra RMS Max P2V
1.569 8.020 1.734
1.245 8.452 1.628
1.167 15.35 1.574
observed in high vacuum condition with the newly developed thinner magnetic film MFM piezoresistive cantilever. Under high vacuum condition, the air dumping effect on cantilever beam is removed and it is possible to gain high enough magnetic sensitivity for imaging with lower magnetic moment probe. It is applicable for measuring the sample of which magnetic field is easily interfered by outer magnetic field even by probing. The piezoresistive self-sensing cantilever is versatile and promising technology for SPM application and will be powerful tool for complicated and advanced material research.
70
H. Takahashi et al. / Ultramicroscopy 91 (2002) 63–72
Fig. 9. MFM images of magnetic recording tape. (a) Imaged with PRC120 cantilever magnetic film coated on the tip (magnetic film thickness=50 nm). (b) Imaged with PRC400 cantilever magnetic film coated on the tip (magnetic film thickness=50 nm). (c) Imaged with MFM cantilever MFMR (length=225, k ¼ 2:8 N/m, magnetic film thickness=40 nm) by optical detection.
4. Conclusion The self-sensing piezoresistive cantilever with higher sensitivity is successfully fabricated. The
sensitivity of the piezoresistive cantilever is comparable to the current laser detecting system and topographic images are successfully demonstrated with the piezoresistive cantilever.
H. Takahashi et al. / Ultramicroscopy 91 (2002) 63–72
71
Fig. 10. Topographic and MFM images of magnetic recording tape imaged with PRC400 cantilever magnetic film coated on the tip (magnetic film thickness=24 nm). (a) Topographic image taken in the atmospheric pressure. (b) MFM image taken in the atmospheric pressure. (c) Topographic image taken in vacuum (at 1.0 E–4 Pa). (d) MFM image taken in vacuum (at 1.0 E–4 Pa).
Furthermore, MFM piezoresistive cantilever with the magnetic film (Co–Cr–Pt) on the tip is also fabricated and tested. We successfully obtain magnetic images with the piezoresistive MFM cantilever. Piezoresistive cantilevers with lower magnetic moment probe are also fabricated and the MFM image is observed under high vacuum condition.
Acknowledgements The authors wish to thank M. Despont, P. Vettiger, and G. Binnig from IBM Zurich Research Laboratory for their assistance of cantilever design and fabrication. The authors also thank T. Yamaoka for assistance and useful discussion on MFM measurement.
H. Takahashi et al. / Ultramicroscopy 91 (2002) 63–72
72
Table 3 Magnetic sensitivity measured with various cantilevers Magnetic sensitivity (deg.) PRC120 with 50 nm magnetic film PRC400 with 50 nm magnetic film PRC400 with 24 nm magnetic film MFM cantilever MFMR
0.22 3.05 2.1 (atmospheric pressure) 47.5 (in vacuum at 1.0E4 Pa) 3.65
References [1] G. Binnig, H. Rohrer, Ch. Gerber, E. Weibel, Phys. Rev. Lett. 49 (1982) 57. [2] H.K. Wickramasinghe, J. Vac. Sci. Technol. A 8 (1990) 363. [3] G. Binnig, C.F. Quate, Ch. Gerber, Phys. Rev. Lett. 56 (1986) 930.
[4] O. Wolter, Th. Bayer, J. Grescher, J. Vac. Sci. Technol. B 9 (1991) 1353. [5] C. Mate, G. McClelland, R. Erlandsson, S. Chiang, Phy. Rev. Lett. 59 (1987) 1942. [6] K. Yamanaka, E. Tomita, Jpn. J. Appl. Phys. 34 (1995) 3350. [7] M. Tanimoto, O. Vatel, J. Vac. Sci. Technol. B 94 (1996) 1547. [8] Y. Martin, H. Wickramasinghe, Appl. Phys. Lett. 50 (1987) 1455. [9] U. Hartmann, J. Vac. Sci. Technol. A 8 (1990) 411. [10] A. Wadas, P. Gruetter, H. Guentherodt, J. Vac. Sci. Technol. A 8 (1990) 416. [11] N. Blanc, J. Brugger, N.F. de Rooij, J. Vac. Sci. Technol. B 14 (1996) 903. [12] S. Akamine, T. Alabrecht, M. Zdeblick, C.F. Quate, IEEE Elec. Dev. Lett. EDL-10 (1989) 490. [13] F. Giessibl, B. Trafas, Rev. Sci. Instrum. 65 (1994) 1923. [14] M. Despont, H. Takahashi, S. Ichihara, Y. Shirakawabe, N. Shimizu, A. Inoue, W. Haeberle, G. Binnig, P. Vettiger, Proceedings on the IEEE Micro Electro Mechanical Systems 2000, MEMS2000, 2000, p. 126.