- Email: [email protected]
Characterization of magnetic materials using a scanning microwave microprobe
ARTICLE IN PRESS Ultramicroscopy 108 (2008) 1030– 1033
Contents lists available at ScienceDirect
Ultramicroscopy journal homepage: www.elsevier.com/locate/ultramic
Characterization of magnetic materials using a scanning microwave microprobe Harutyun Melikyan, Artur Hovsepyan, Tigran Sargsyan, Youngwoon Yoon, Hyungun Yoo, Arsen Babajanyan, Kiejin Lee Department of Physics and Interdisciplinary Program of Integrated Biotechnology, Sogang University, Seoul 121-742, Republic of Korea
article info PACS: 07.79.–v 78.20.Ci 75.50.Cc 75.70.–i 42.60.Da 73.61.–r Keywords: Scanning microwave microprobe Near-field Thin films Magnetic materials
a b s t r a c t We investigated the electromagnetic properties of metals of iron, nickel, cobalt, aluminum, gold, copper, silver, and permalloy thin films on SiO2 substrates using a near-field microwave microprobe. The electromagnetic properties of metal sheets were estimated by measuring the microwave reflection coefficient S11 and compared with the theoretical values. We observed the hysteresis behavior of permalloy thin films on SiO2 substrates using a near-field scanning microwave microprobes (NSMM) system. Experimental results are in good agreement with the theoretical model of transmission theory. In order to better characterize the electromagnetic properties of metals and magnetic metals instead of the usual method, we take advantage of the noncontact microprobing evaluation capabilities using a nearfield microwave. & 2008 Elsevier B.V. All rights reserved.
1. Introduction Scanning probes using a microwave for local and surface area characterization of conducting films have attracted considerable attention due to the possibility of applications in the electronic and display industries [1,2]. Conducting thin films with the thickness range of 0.01–1 mm are most widely used in semiconductor and display industry. For this purpose, various techniques have been developed enabling nano-through micro-scale microprobe and imaging of various materials. Recently, several nearfield scanning microwave microprobes (NSMM) have been developed for the microwave and millimeter-wave ranges [3,4]. An important ability of the NSMM is noncontact characterization of conducting materials. In order to better characterize the conductivity metal sheets and thin films instead of the usual contact method of a standard four-point probe, we take advantage of the noncontact evaluation capabilities of a NSMM [5,6]. The main advantages of NSMM compared with the usual contact method are that NSMM can directly image the electromagnetic properties of metals and metal thin films. Another advantage of microwave probing is its larger penetration depth, which makes the surface properties of the sample detectable as well as the surface properties. Note that, the Corresponding author. Tel.: +82 2 705 8429; fax: +82 2 715 8429.
E-mail address: [email protected] (K. Lee). 0304-3991/$ - see front matter & 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.ultramic.2008.04.010
skin depth in microwave range of 1–10 GHz is of the order of 0.5–5 mm. In this paper, we study the conductivity of metal sheets of iron (Fe), nickel (Ni), cobalt (Co), aluminum (Al), gold (Au), copper (Cu), silver (Ag) and permalloy thin films on SiO2 substrates with different thickness using a NSMM. We used a NSMM coupled to a high-quality resonator with a tuning fork distance regulation system at an operating frequency about f ¼ 4.1 GHz for TE011. The changes of conductivity of the metal samples were investigated by measuring the reflection coefficient S11 of the resonator. The change of the conductivity is directly related to the change of the reflection coefficient due to a near-field electromagnetic interaction between the probe tip and the metal samples. The conductivity of sheet metals and the permalloy thin films on SiO2 substrates with different thickness was investigated by measuring the reflection coefficient S11 and compared with microwave transmission line theory.
2. Theory The microwave reflection coefficient S11 depended on the conductivity of metals. A formula showing how the reflection coefficient S11 depended on conductivity of the metals can be derived by using standard transmission line theory [7,8] and is given by assuming impedance matching between the probe tip
ARTICLE IN PRESS H. Melikyan et al. / Ultramicroscopy 108 (2008) 1030–1033
3. Experiment
S11
ka t a / 106 51.
(1)
where Z0 is the impedance of the probe (Z0 ¼ 50 O), ka is the wave vector of air (ka ¼ 84 m1 at 4.1 GHz), and ta is the distance between probe tip and sample (ta ¼ 10 nm) and Zl is the modulus of the complex impedance of sample. Here and henceforth the subscripts, l ¼ m, ps indicated quantities related to metal sheet and the permalloy/SiO2 thin film substrate structure, respectively. The impedance of a metal sheet Zm is given by pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Z m ¼ ð1 þ jÞ pf mm m0 =sm ¼ 2pf mm m0 =sm , where sm is the conductivity of metal, m0 is the permeability of free-space (m0 ¼ 4p 107 H/m), mm is the relative permeability of metal and f is the operating frequency (f ¼ 4.1 GHz for TE011 mode). The relative permeability of diamagnetic and paramagnetic materials is mE1 and ferromagnetic materials is mb1 (Co m ¼ 250; Ni m ¼ 600; Fe m ¼ 2000). Thus, the reflection coefficient S11 due to the metal conductivity can be estimated as pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2pf m m =s Z m 0 m 0 S11 ¼ 20 logpffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi . 2pf mm m0 =sm þ Z 0
(2)
The calculation of the reflection coefficient S11 of permalloy thin film on SiO2 substrate structure depended on the thickness of the permalloy thin films. The modulus of the complex impedance of the permalloy thin film on SiO2 substrate structure Zps is given by [7,8] Z s þ jZ p tan ðkp t p Þ , Z ps ¼ Z p Z þ jZ tan ðk t Þ p
s
(3)
p p
where tp is the thickness of the permalloy thin film layer, Zs is the impedance of SiO2 substrate, Zp is the impedance of permalloy qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi layer and given by Z p ¼ ð1 þ jÞ pf mp m0 =sp, kp is the wave number qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi in the permalloy. Wave number given by kp ¼ ð1 þ jÞ pf mp m0 sp,
To characterize the conductivity of metals we measured the changes of reflection coefficient S11 at the resonance frequency of our NSMM system. The cavity and resonator stored electromagnetic energy at the resonant frequency depending on the conductivity. By measuring the reflection coefficient S11 of the microwaves input into the resonant cavity it is possible to determine the conductivity of the metal samples. Fig. 1 shows the experimental setup [10]. We applied the external field to samples using a magnet with controller. The magnetic field applied parallel to the sample surface. To obtain high sensitivity of the NSMM, a stainless-steel probe tip with a round apex with a diameter of 50 mm was used and glued onto one of the prongs of the tuning fork, which attached to the resonator. The probe tipsample distance was fixed by using a quartz tuning fork feedback control system at about 10 nm. Permalloy metal films with a thickness of 20, 120 and 200 nm were deposited on SiO2 substrates. The metal sheet samples were iron (Fe s ¼ 1 107 S/m), nickel (Ni s ¼ 1.5 107 S/m), cobalt (Co s ¼ 1.6 107 S/m), aluminum (Al s ¼ 3.8 107 S/m), gold (Au s ¼ 4.1 107 S/m), copper (Cu s ¼ 5.8 107 S/m) and silver (Ag s ¼ 6.2 107 S/m) with sizes of 50 mm 50 mm 1 mm. We measured the reflection coefficient S11 of the resonator of the NSMM at an operating frequency about 4.1 GHz for TE011 mode. All measurements were done at room temperature. At resonance, the modes we used were TE011 the unloaded Q factor was 24,000. A network analyzer (Agilent 8722 ES) was used in measuring the reflection coefficient S11.
4. Results and discussion We measured the reflection coefficient S11 using a network analyzer at the probe tip-sample separation of about 10 nm. Both the resonance frequency and the quality factor of the resonator
where mp is the relative permeability of permalloy (mp ¼ 100,000), sp is the conductivity of permalloy (sp ¼ 4.25 106 S/m at 4.1 GHz) [9], m0 is the permeability of free-space (m0 ¼ 4p 107 H/m).
-20
-25
Network Analyzer 8722ES
-30 S11 (dB)
Computer
Resonator
Feedback circuit
-35
-40 Lock-in amp.
Probe tip
(g) (f) (e)
-30
(d)
-35
(c)
S21
S11
(h)
S11 (dB)
and the microwave source Z Z 0 ; ¼ 20 log l Zl þ Z0
1031
-40
(b)
Sample
Tuning fork Magnet
20
-45
(a)
40
60
Conductivity (x106 S/m)
PZT
4.162 x-y-z stage Fig. 1. Experimental setup of NSMM system with external magnetic field. The microwave reflection coefficient S11 was measured using a network analyzer. A quartz tuning-fork used for distance control. The distance is about 10 nm between the tip and the sample.
4.164
4.166 4.168 Frequency (GHz)
4.17
4.172
Fig. 2. Measured reflection coefficient S11 of (a) air, (b) Fe, (c) Ni, (d) Co, (e) Al, (f) Au, (g) Cu, and (h) Ag sheets. The inset shows the reflection coefficient S11 of the metal sheets plotted against the conductivity of metal. The solid lines shows a fit to Eq. (2) with Z0 ¼ 50 O.
ARTICLE IN PRESS H. Melikyan et al. / Ultramicroscopy 108 (2008) 1030–1033
-20
(d) -30 S11 (dB)
(c)
-40
(b)
S11 (dB)
-30
-40
0
(a)
4.17
4.1725
500 250
-26
-27
0 -250 -500 -1000 -500
-28
0 500 Hext (Oe)
1000
-29
-30
-31
-32 -1000
-500
0
500
1000
Hext (Oe) Fig. 4. Measured reflection coefficient S11 20 nm permalloy thin film on SiO2 substrate as a function of the external magnetic field. The arrow indicates the sweep direction. Applied external magnetic field was parallel to the sample surface. The inset shows the magnetization for 20 nm permalloy thin film on SiO2 substrate as a function of the applied magnetic field.
magnetization amplitude increased [11]. The arrow indicates the sweep direction. Applied external magnetic field was parallel to the sample surface. The inset shows the measured magnetization for permalloy thin film with 20 nm on SiO2 substrate measured as a function of the applied magnetic field. By measuring the quantitative change of reflection coefficient S11 due to the magnetization changes, we could find the magnetic density changes of permalloy thin films due to the external magnetic fields. It will be published separately in detail. An important metric for future applications is sensitivity of measurements, which requires characterization of measurement noise. The rootmean-square statistical noise in S11 is about 5.4 107 in linear scale [12]. The measured signal-to-noise was about 61 dB. These results indicate NSMM could observe the high-contrast electromagnetic properties images.
5. Conclusion
-50
-50
-25
M (G)
are affected by the conductivity of metal samples. The resonant frequency is sensitive to the near-field interaction of the tip with the samples. The cavity and resonator store electromagnetic energy with the resonant frequency depending strongly on conductivity. By measuring the resonance frequency and the reflection coefficient of the microwaves input into the cavity it is possible to determine the conductivity of the samples. Fig. 2 shows the microwave reflection coefficient S11 profiles of sheet metals for (a) air, (b) Fe, (c) Ni, (d) Co, (e) Al, (f) Au, (g) Cu, and (h) Ag. The inset of Fig. 2 shows the microwave reflection coefficient as a function of conductivity of the metal sheets at the resonant frequency. As the conductivity of metal increased, the intensity of the reflection coefficient S11 increased as expected from Eq. (2). The intensity of reflection coefficient S11 of air was 63.4 dB, which are the reference levels of S11 of our measurements. The reflection coefficient of S11 of Fe shows a minimum value 40.35 dB. The maximum intensity of the reflection coefficient is observed for Ag with 27.33 dB. Clearly the change of the conductivity of the metals affected the microwave reflection coefficient S11. Fig. 3 shows the microwave reflection coefficient curves for (a) air and for the permalloy thin films on SiO2 substrate with film thicknesses of (b) 20 nm, (c) 120 nm, and (d) 200 nm. The reflection coefficient S11 changed dependent on the permalloy film thickness. The reflection coefficient S11 increased as the thickness of permalloy thin films increased. The inset of Fig. 3 shows the microwave reflection coefficient dependence on thicknesses of permalloy thin films. The solid line corresponds to theory (Eqs. (1) and (3)). We found a good agreement with theory between the variation of reflection coefficient and the impedance of pentacene. Fig. 4 shows the measured magnetic hysteresis behavior of permalloy thin films with the thickness of 20 nm on SiO2 substrate using a NSMM system. As the external magnetic filed increased, the measured reflection coefficient S11 increased due to the
S11 (dB)
1032
4.175
40
80 120 160 Thickness (nm)
4.1775
200
4.18
Frequency (GHz) Fig. 3. Measured reflection coefficient S11 for (a) air and for permalloy thin films on SiO2 substrates with film thickness of (b) 20 nm, (c) 120 nm, and (d) 200 nm. The inset shows the measured reflection coefficient S11 at the resonant frequency of permalloy thin films on SiO2 substrates for different thicknesses. The solid line shows a fit to Eqs. (1) and (3) with Z0 ¼ 50 O.
We demonstrate the measurement of electromagnetic properties of permalloy thin films on SiO2 surfaces and Fe, Ni, Co, Al, Au, Cu and Ag metal sheets samples using a NSMM. As the conductivity of metal increased, the intensity of the reflection coefficient S11 increased. The electromagnetic properties of metal sheets were estimated by measuring the microwave reflection coefficient S11 and compared with the theoretical values. Measured permalloy on SiO2 thin films magnetic hysteresis behavior using NSMM system. We developed a calculation model how the reflection coefficient S11 depended on the conductivity of the metal sheets by standard transmission line theory. These results clearly show the sensitivity and usefulness of this scanning microprobe microscope for investigating electromagnetic properties of metal thin films.
ARTICLE IN PRESS H. Melikyan et al. / Ultramicroscopy 108 (2008) 1030–1033
Acknowledgments This work was supported by Sogang University (2007), by the Korea Research Foundation (KRF-2005-042-C00058; KRF-2002005-CS0003), the Seoul Research and Business Development Program (10816) and by the Korea Science & Engineering Foundation (F01-2004-000-1082-0; R01-2006-000-11227-0). References [1] M. Abu-Teir, M. Golosovsky, D. Davidov, A. Frenkel, H. Goldberger, Rev. Sci. Instrum. 72 (2001) 2073. [2] T. Kasagia, T. Tsutaokab, K. Hatakeyamac, J. Magn. Magn. Mater. 272 (2004) 2224.
1033
[3] A. Babajanyan, K. Lee, E. Lim, T. Manaka, M. Iwamoto, B. Friedman, Appl. Phys. Lett. 90 (2007) 182104. [4] S. Lee, C. Vlahacos, B. Feenstra, A. Schwartz, D. Steinhauer, F. Wellstood, S. Anlage, Appl. Phys. Lett. 77 (2000) 4404. [5] M. Park, H. Yoo, H. Yoo, S. Na, S. Kim, K. Lee, B. Friedman, E. Lim, M. Iwamoto, Thin Solid Films 499 (2006) 318. [6] O. Acher, S. Queste, K. Barholz, R. Mattheis, J. Appl. Phys. 93 (2003) 6668. [7] D.M. Pozar, Microwave Engineering, Wesley, New York, 1990, pp. 224. [8] E. Silva, M. Lanacura, R. Marcon, Supercond. Sci. Technol. 9 (1996) 934. [9] C. Blaas, L. Szunyogh, P. Weinberger, C. Sommers, P.M. Levy, Phys. Rev. B. 63 (2001) 224408. [10] S. Yun, S. Na, A. Babajayan, H. Kim, B. Friedman, K. Lee, Thin Solid Films 515 (2006) 1354. [11] T. Kasagi, T. Tsutaoka, K. Hatakeyama, Appl. Phys. Lett. 88 (2006) 172502. [12] B. Friedman, M. Gaspar, S. Kalachikov, K. Lee, R. Levisky, G. Shen, H. Yoo, J. Am. Chem. Soc. 127 (2005) 9666.
Contents lists available at ScienceDirect
Ultramicroscopy journal homepage: www.elsevier.com/locate/ultramic
Characterization of magnetic materials using a scanning microwave microprobe Harutyun Melikyan, Artur Hovsepyan, Tigran Sargsyan, Youngwoon Yoon, Hyungun Yoo, Arsen Babajanyan, Kiejin Lee Department of Physics and Interdisciplinary Program of Integrated Biotechnology, Sogang University, Seoul 121-742, Republic of Korea
article info PACS: 07.79.–v 78.20.Ci 75.50.Cc 75.70.–i 42.60.Da 73.61.–r Keywords: Scanning microwave microprobe Near-field Thin films Magnetic materials
a b s t r a c t We investigated the electromagnetic properties of metals of iron, nickel, cobalt, aluminum, gold, copper, silver, and permalloy thin films on SiO2 substrates using a near-field microwave microprobe. The electromagnetic properties of metal sheets were estimated by measuring the microwave reflection coefficient S11 and compared with the theoretical values. We observed the hysteresis behavior of permalloy thin films on SiO2 substrates using a near-field scanning microwave microprobes (NSMM) system. Experimental results are in good agreement with the theoretical model of transmission theory. In order to better characterize the electromagnetic properties of metals and magnetic metals instead of the usual method, we take advantage of the noncontact microprobing evaluation capabilities using a nearfield microwave. & 2008 Elsevier B.V. All rights reserved.
1. Introduction Scanning probes using a microwave for local and surface area characterization of conducting films have attracted considerable attention due to the possibility of applications in the electronic and display industries [1,2]. Conducting thin films with the thickness range of 0.01–1 mm are most widely used in semiconductor and display industry. For this purpose, various techniques have been developed enabling nano-through micro-scale microprobe and imaging of various materials. Recently, several nearfield scanning microwave microprobes (NSMM) have been developed for the microwave and millimeter-wave ranges [3,4]. An important ability of the NSMM is noncontact characterization of conducting materials. In order to better characterize the conductivity metal sheets and thin films instead of the usual contact method of a standard four-point probe, we take advantage of the noncontact evaluation capabilities of a NSMM [5,6]. The main advantages of NSMM compared with the usual contact method are that NSMM can directly image the electromagnetic properties of metals and metal thin films. Another advantage of microwave probing is its larger penetration depth, which makes the surface properties of the sample detectable as well as the surface properties. Note that, the Corresponding author. Tel.: +82 2 705 8429; fax: +82 2 715 8429.
E-mail address: [email protected] (K. Lee). 0304-3991/$ - see front matter & 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.ultramic.2008.04.010
skin depth in microwave range of 1–10 GHz is of the order of 0.5–5 mm. In this paper, we study the conductivity of metal sheets of iron (Fe), nickel (Ni), cobalt (Co), aluminum (Al), gold (Au), copper (Cu), silver (Ag) and permalloy thin films on SiO2 substrates with different thickness using a NSMM. We used a NSMM coupled to a high-quality resonator with a tuning fork distance regulation system at an operating frequency about f ¼ 4.1 GHz for TE011. The changes of conductivity of the metal samples were investigated by measuring the reflection coefficient S11 of the resonator. The change of the conductivity is directly related to the change of the reflection coefficient due to a near-field electromagnetic interaction between the probe tip and the metal samples. The conductivity of sheet metals and the permalloy thin films on SiO2 substrates with different thickness was investigated by measuring the reflection coefficient S11 and compared with microwave transmission line theory.
2. Theory The microwave reflection coefficient S11 depended on the conductivity of metals. A formula showing how the reflection coefficient S11 depended on conductivity of the metals can be derived by using standard transmission line theory [7,8] and is given by assuming impedance matching between the probe tip
ARTICLE IN PRESS H. Melikyan et al. / Ultramicroscopy 108 (2008) 1030–1033
3. Experiment
S11
ka t a / 106 51.
(1)
where Z0 is the impedance of the probe (Z0 ¼ 50 O), ka is the wave vector of air (ka ¼ 84 m1 at 4.1 GHz), and ta is the distance between probe tip and sample (ta ¼ 10 nm) and Zl is the modulus of the complex impedance of sample. Here and henceforth the subscripts, l ¼ m, ps indicated quantities related to metal sheet and the permalloy/SiO2 thin film substrate structure, respectively. The impedance of a metal sheet Zm is given by pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Z m ¼ ð1 þ jÞ pf mm m0 =sm ¼ 2pf mm m0 =sm , where sm is the conductivity of metal, m0 is the permeability of free-space (m0 ¼ 4p 107 H/m), mm is the relative permeability of metal and f is the operating frequency (f ¼ 4.1 GHz for TE011 mode). The relative permeability of diamagnetic and paramagnetic materials is mE1 and ferromagnetic materials is mb1 (Co m ¼ 250; Ni m ¼ 600; Fe m ¼ 2000). Thus, the reflection coefficient S11 due to the metal conductivity can be estimated as pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2pf m m =s Z m 0 m 0 S11 ¼ 20 logpffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi . 2pf mm m0 =sm þ Z 0
(2)
The calculation of the reflection coefficient S11 of permalloy thin film on SiO2 substrate structure depended on the thickness of the permalloy thin films. The modulus of the complex impedance of the permalloy thin film on SiO2 substrate structure Zps is given by [7,8] Z s þ jZ p tan ðkp t p Þ , Z ps ¼ Z p Z þ jZ tan ðk t Þ p
s
(3)
p p
where tp is the thickness of the permalloy thin film layer, Zs is the impedance of SiO2 substrate, Zp is the impedance of permalloy qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi layer and given by Z p ¼ ð1 þ jÞ pf mp m0 =sp, kp is the wave number qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi in the permalloy. Wave number given by kp ¼ ð1 þ jÞ pf mp m0 sp,
To characterize the conductivity of metals we measured the changes of reflection coefficient S11 at the resonance frequency of our NSMM system. The cavity and resonator stored electromagnetic energy at the resonant frequency depending on the conductivity. By measuring the reflection coefficient S11 of the microwaves input into the resonant cavity it is possible to determine the conductivity of the metal samples. Fig. 1 shows the experimental setup [10]. We applied the external field to samples using a magnet with controller. The magnetic field applied parallel to the sample surface. To obtain high sensitivity of the NSMM, a stainless-steel probe tip with a round apex with a diameter of 50 mm was used and glued onto one of the prongs of the tuning fork, which attached to the resonator. The probe tipsample distance was fixed by using a quartz tuning fork feedback control system at about 10 nm. Permalloy metal films with a thickness of 20, 120 and 200 nm were deposited on SiO2 substrates. The metal sheet samples were iron (Fe s ¼ 1 107 S/m), nickel (Ni s ¼ 1.5 107 S/m), cobalt (Co s ¼ 1.6 107 S/m), aluminum (Al s ¼ 3.8 107 S/m), gold (Au s ¼ 4.1 107 S/m), copper (Cu s ¼ 5.8 107 S/m) and silver (Ag s ¼ 6.2 107 S/m) with sizes of 50 mm 50 mm 1 mm. We measured the reflection coefficient S11 of the resonator of the NSMM at an operating frequency about 4.1 GHz for TE011 mode. All measurements were done at room temperature. At resonance, the modes we used were TE011 the unloaded Q factor was 24,000. A network analyzer (Agilent 8722 ES) was used in measuring the reflection coefficient S11.
4. Results and discussion We measured the reflection coefficient S11 using a network analyzer at the probe tip-sample separation of about 10 nm. Both the resonance frequency and the quality factor of the resonator
where mp is the relative permeability of permalloy (mp ¼ 100,000), sp is the conductivity of permalloy (sp ¼ 4.25 106 S/m at 4.1 GHz) [9], m0 is the permeability of free-space (m0 ¼ 4p 107 H/m).
-20
-25
Network Analyzer 8722ES
-30 S11 (dB)
Computer
Resonator
Feedback circuit
-35
-40 Lock-in amp.
Probe tip
(g) (f) (e)
-30
(d)
-35
(c)
S21
S11
(h)
S11 (dB)
and the microwave source Z Z 0 ; ¼ 20 log l Zl þ Z0
1031
-40
(b)
Sample
Tuning fork Magnet
20
-45
(a)
40
60
Conductivity (x106 S/m)
PZT
4.162 x-y-z stage Fig. 1. Experimental setup of NSMM system with external magnetic field. The microwave reflection coefficient S11 was measured using a network analyzer. A quartz tuning-fork used for distance control. The distance is about 10 nm between the tip and the sample.
4.164
4.166 4.168 Frequency (GHz)
4.17
4.172
Fig. 2. Measured reflection coefficient S11 of (a) air, (b) Fe, (c) Ni, (d) Co, (e) Al, (f) Au, (g) Cu, and (h) Ag sheets. The inset shows the reflection coefficient S11 of the metal sheets plotted against the conductivity of metal. The solid lines shows a fit to Eq. (2) with Z0 ¼ 50 O.
ARTICLE IN PRESS H. Melikyan et al. / Ultramicroscopy 108 (2008) 1030–1033
-20
(d) -30 S11 (dB)
(c)
-40
(b)
S11 (dB)
-30
-40
0
(a)
4.17
4.1725
500 250
-26
-27
0 -250 -500 -1000 -500
-28
0 500 Hext (Oe)
1000
-29
-30
-31
-32 -1000
-500
0
500
1000
Hext (Oe) Fig. 4. Measured reflection coefficient S11 20 nm permalloy thin film on SiO2 substrate as a function of the external magnetic field. The arrow indicates the sweep direction. Applied external magnetic field was parallel to the sample surface. The inset shows the magnetization for 20 nm permalloy thin film on SiO2 substrate as a function of the applied magnetic field.
magnetization amplitude increased [11]. The arrow indicates the sweep direction. Applied external magnetic field was parallel to the sample surface. The inset shows the measured magnetization for permalloy thin film with 20 nm on SiO2 substrate measured as a function of the applied magnetic field. By measuring the quantitative change of reflection coefficient S11 due to the magnetization changes, we could find the magnetic density changes of permalloy thin films due to the external magnetic fields. It will be published separately in detail. An important metric for future applications is sensitivity of measurements, which requires characterization of measurement noise. The rootmean-square statistical noise in S11 is about 5.4 107 in linear scale [12]. The measured signal-to-noise was about 61 dB. These results indicate NSMM could observe the high-contrast electromagnetic properties images.
5. Conclusion
-50
-50
-25
M (G)
are affected by the conductivity of metal samples. The resonant frequency is sensitive to the near-field interaction of the tip with the samples. The cavity and resonator store electromagnetic energy with the resonant frequency depending strongly on conductivity. By measuring the resonance frequency and the reflection coefficient of the microwaves input into the cavity it is possible to determine the conductivity of the samples. Fig. 2 shows the microwave reflection coefficient S11 profiles of sheet metals for (a) air, (b) Fe, (c) Ni, (d) Co, (e) Al, (f) Au, (g) Cu, and (h) Ag. The inset of Fig. 2 shows the microwave reflection coefficient as a function of conductivity of the metal sheets at the resonant frequency. As the conductivity of metal increased, the intensity of the reflection coefficient S11 increased as expected from Eq. (2). The intensity of reflection coefficient S11 of air was 63.4 dB, which are the reference levels of S11 of our measurements. The reflection coefficient of S11 of Fe shows a minimum value 40.35 dB. The maximum intensity of the reflection coefficient is observed for Ag with 27.33 dB. Clearly the change of the conductivity of the metals affected the microwave reflection coefficient S11. Fig. 3 shows the microwave reflection coefficient curves for (a) air and for the permalloy thin films on SiO2 substrate with film thicknesses of (b) 20 nm, (c) 120 nm, and (d) 200 nm. The reflection coefficient S11 changed dependent on the permalloy film thickness. The reflection coefficient S11 increased as the thickness of permalloy thin films increased. The inset of Fig. 3 shows the microwave reflection coefficient dependence on thicknesses of permalloy thin films. The solid line corresponds to theory (Eqs. (1) and (3)). We found a good agreement with theory between the variation of reflection coefficient and the impedance of pentacene. Fig. 4 shows the measured magnetic hysteresis behavior of permalloy thin films with the thickness of 20 nm on SiO2 substrate using a NSMM system. As the external magnetic filed increased, the measured reflection coefficient S11 increased due to the
S11 (dB)
1032
4.175
40
80 120 160 Thickness (nm)
4.1775
200
4.18
Frequency (GHz) Fig. 3. Measured reflection coefficient S11 for (a) air and for permalloy thin films on SiO2 substrates with film thickness of (b) 20 nm, (c) 120 nm, and (d) 200 nm. The inset shows the measured reflection coefficient S11 at the resonant frequency of permalloy thin films on SiO2 substrates for different thicknesses. The solid line shows a fit to Eqs. (1) and (3) with Z0 ¼ 50 O.
We demonstrate the measurement of electromagnetic properties of permalloy thin films on SiO2 surfaces and Fe, Ni, Co, Al, Au, Cu and Ag metal sheets samples using a NSMM. As the conductivity of metal increased, the intensity of the reflection coefficient S11 increased. The electromagnetic properties of metal sheets were estimated by measuring the microwave reflection coefficient S11 and compared with the theoretical values. Measured permalloy on SiO2 thin films magnetic hysteresis behavior using NSMM system. We developed a calculation model how the reflection coefficient S11 depended on the conductivity of the metal sheets by standard transmission line theory. These results clearly show the sensitivity and usefulness of this scanning microprobe microscope for investigating electromagnetic properties of metal thin films.
ARTICLE IN PRESS H. Melikyan et al. / Ultramicroscopy 108 (2008) 1030–1033
Acknowledgments This work was supported by Sogang University (2007), by the Korea Research Foundation (KRF-2005-042-C00058; KRF-2002005-CS0003), the Seoul Research and Business Development Program (10816) and by the Korea Science & Engineering Foundation (F01-2004-000-1082-0; R01-2006-000-11227-0). References [1] M. Abu-Teir, M. Golosovsky, D. Davidov, A. Frenkel, H. Goldberger, Rev. Sci. Instrum. 72 (2001) 2073. [2] T. Kasagia, T. Tsutaokab, K. Hatakeyamac, J. Magn. Magn. Mater. 272 (2004) 2224.
1033
[3] A. Babajanyan, K. Lee, E. Lim, T. Manaka, M. Iwamoto, B. Friedman, Appl. Phys. Lett. 90 (2007) 182104. [4] S. Lee, C. Vlahacos, B. Feenstra, A. Schwartz, D. Steinhauer, F. Wellstood, S. Anlage, Appl. Phys. Lett. 77 (2000) 4404. [5] M. Park, H. Yoo, H. Yoo, S. Na, S. Kim, K. Lee, B. Friedman, E. Lim, M. Iwamoto, Thin Solid Films 499 (2006) 318. [6] O. Acher, S. Queste, K. Barholz, R. Mattheis, J. Appl. Phys. 93 (2003) 6668. [7] D.M. Pozar, Microwave Engineering, Wesley, New York, 1990, pp. 224. [8] E. Silva, M. Lanacura, R. Marcon, Supercond. Sci. Technol. 9 (1996) 934. [9] C. Blaas, L. Szunyogh, P. Weinberger, C. Sommers, P.M. Levy, Phys. Rev. B. 63 (2001) 224408. [10] S. Yun, S. Na, A. Babajayan, H. Kim, B. Friedman, K. Lee, Thin Solid Films 515 (2006) 1354. [11] T. Kasagi, T. Tsutaoka, K. Hatakeyama, Appl. Phys. Lett. 88 (2006) 172502. [12] B. Friedman, M. Gaspar, S. Kalachikov, K. Lee, R. Levisky, G. Shen, H. Yoo, J. Am. Chem. Soc. 127 (2005) 9666.