1: % Template article for preprint document class `elsart'

2: % SP 2001/01/05

3: 

4: \documentclass{elsart4}

5: 

6: % Use the option doublespacing or reviewcopy to obtain double line spacing

7: \usepackage{graphics}

8: \usepackage{graphicx}

9: \usepackage{epsfig}

10: \usepackage{amssymb}

11: 

12: \begin{document}

13: 

14: \begin{frontmatter}

15: 

16: \title{Measurement of mesoscopic High-$T_c$ superconductors using Si mechanical micro-oscillators}

17: 

18: \author{M. Dolz , D. Antonio, H. Pastoriza}

19: \address{Centro At\'omico Bariloche, Comisi\'on Nacional de Energ\'{\i}a At\'omica and CONICET, R8402AGP S.

20: C. de Bariloche, Argentina}

21: 

22: \begin{abstract}

23: 

24: In a superconducting mesoscopic sample, with dimensions comparable

25: to the London penetration depth, some properties are qualitatively

26: different to those found in the bulk material. These properties

27: include magnetization, vortex dynamics and ordering of the vortex

28: lattice. In order to detect the small signals produced by this

29: kind of samples, new instruments designed for the microscale are

30: needed. In this work we use micromechanical oscillators to study

31: the magnetic properties of a Bi$_2$Sr$_2$CaCu$_2$O$_{8 + \delta}$

32: disk with a diameter of $13.5$ microns and a thickness of $2.5$

33: microns. The discussion of our results is based on the existence

34: and contribution of inter and intra layer currents.

35: \end{abstract}

36: 

37: \begin{keyword}

38: MEMs \sep mesoscopic \sep high $T_c$

39: 

40: \PACS 85.85.+j \sep 74.78.Na \sep 74.72.Hs

41: \end{keyword}

42: \end{frontmatter}

43: 

44: % main text

45: \section{Introduction}

46: \label{}

47: 

48: The study of the vortex physics in mesoscopic samples \cite{Aksyuk98,Chan01} is difficult

49: due to the need of instruments sensible enough to detect their signals. Sensitive

50: instruments such as SQUIDs are not the best option because they are not designed to

51: measure microscopic samples. The use of mechanical oscillators as magnetometers is not a

52: new idea, they have been used successfully for this application for some time

53: \cite{Kleiman85}. Our approach for studying mesoscopic high $T_c$ samples is to use

54: silicon micro-oscillators (following the work of \cite{Bolle99}) which have a torsional

55: mode with a resonant frequency $\nu _r \approx 45$ kHz and a quality factor $Q > 10^4$ at

56: low temperatures. This instrument integrates high sensitivity and reduced size with a

57: small signal loss, which is an important factor in the measurement of micron sized

58: samples.

59: \begin{figure}[t]\begin{center}

60: \includegraphics[width=\linewidth]{dolzFig1.eps}

61: \caption{Scanning electron micrograph of a high-$Q$ mechanical

62: oscillator with a BSCCO disk with a diameter of $13.5$ microns and

63: a thickness of $2.5$ microns. The sample was glued by means of

64: micro-pipettes and micro-manipulators.} \label{oscillator}

65: \end{center}\end{figure}

66: 

67: 

68: In this work we present the response of a micro-oscillator with a

69: Bi$_2$Sr$_2$CaCu$_2$O$_{8 + \delta}$ (BSCCO) disk mounted on top

70: of it, in the presence of an external magnetic field $H$. The

71: system (oscillator and sample) is shown in Fig. \ref{oscillator}.

72: 

73: \section{Experimental details}

74: \label{}

75: 

76: The mechanical micro-oscillator was manufactured at MEMSCAP \cite{Memscap} following the

77: MUMPS specifications. It consists of a central plate connected to two serpentine springs

78: which are anchored to the substrate. Below the plate two separate electrodes carry the

79: electrical signals. The plate is electrically grounded and harmonically driven by one of

80: the electrodes. The other electrode is used to detect the amplitude and phase of the

81: mechanical oscillations capacitively. The plate and the detection electrode form a

82: capacitor of $\approx 10$ fF. The motion of the plate produces a variation in the

83: capacitance $\delta C < 1$ fF. A bias voltage $V_b = 1.6$ V is held constant in the

84: capacitor and the current, proportional to $\delta C$ is measured by means of a

85: transimpedance amplifier and a lockin amplifier. This method diminishes the effect of the

86: parasitic capacitances because $V_b$ is constant. A sketch of the circuit is shown in

87: Fig. \ref{circuit}. From the Lorentzian fit of the amplitude vs. frequency curve, $\nu_r$

88: and $Q$ are derived.

89: \begin{figure}[t]\begin{center}

90: \includegraphics[width=\linewidth]{dolzFig2.eps}

91: \caption{System under study. The plate is driven by a sinusoidal

92: voltage. The displacement is capacitively detected with a

93: Current-Voltage Converter and a Lock-in Amplifier.}

94: \label{circuit}

95: \end{center}\end{figure}

96: 

97: The resonant frequency depends on the springs constant $k$ and on

98: the system's moment of inertia $I$:

99: \begin{equation}

100: \nu_r = \frac{1}{2 \pi} \sqrt{\frac{k}{I}} .\label{fundamental}

101: \end{equation}

102: 

103: 

104: After the sample is glued to the oscillator, the system's moment of inertia is modified.

105: In our case, the sample is a single crystal BSCCO disk fabricated by lithographic

106: techniques and ion milling. The frequency variation was $\Delta \nu_r = 868$ Hz in

107: agreement with calculations (Fig. \ref{momentoinercia}).

108: 

109: On the other hand, $\nu_r$ is also modified by variations in the

110: effective $k$ of the system. If the sample is magnetic, an

111: external magnetic field $H$ produces an additional torque

112: \begin{equation}

113: \overrightarrow{\tau}=(\overrightarrow{M} \times

114: \overrightarrow{H})V = MHV \sin \theta,

115: \end{equation}

116: where $M$ is the magnetization, $V$ is the sample 's volume and

117: $\theta$ is the angle between $\overrightarrow{M}$ and

118: $\overrightarrow{H}$. For small $\theta$

119: \begin{equation}

120: \tau=MHV \theta =k_s \theta.

121: \end{equation}

122: 

123: The effective $k_e$ is obtained by adding $k_s$ to the spring

124: constant $k$. As a consequence, the resonant frequency increases

125: (decreases) if this torque is restoring (not restoring). The size,

126: geometry and anisotropy of the sample play an important role in

127: the magnitude and direction of its magnetization $M$. For example,

128: in a spherically symmetrical sample which is in the Meissner

129: state, $M$ remains antiparallel to $H$ during the whole period of

130: oscillation and the magnetic torque is zero ($k_s = 0$).

131: \begin{figure}[t]\begin{center}

132: \includegraphics[width=\linewidth]{dolzFig3.eps}

133: \caption{Response of the oscillator with and without sample. When the sample is mounted

134: on top of the oscillator the system moment of inertia increases and the resonant

135: frequency decreases $868$ Hz.} \label{momentoinercia}

136: \end{center}\end{figure}

137: \begin{figure}[t]\begin{center}

138: \includegraphics[width=\linewidth]{dolzFig4.eps}

139: \caption{Different behavior of the system when is applied an

140: external magnetic field. Currents in the layers of the sample are

141: depicted as a coil perpendicular to the plate and currents between

142: layers as a coil parallel to it} \label{coils}

143: \end{center}\end{figure}

144: 

145: A more complete analysis of the system response can be made by

146: supposing a constant current flowing in a coil placed on the

147: oscillator in the presence of $H$. The coil carries the original

148: current plus any current induced by the tilt of the oscillator and

149: generated by Lenz's Law. In this way, a change in the angle

150: between the coil and the magnetic field produces a change in the

151: induced current and therefore in the total current flowing in the

152: coil. In Fig. \ref{coils}.a and Fig. \ref{coils}.b we sketch the

153: case where $H$ and the coil are perpendicular to the plate. The

154: original current in the coil generates a magnetic moment $m$ that

155: interacts with $H$ exerting an additional torque, which is

156: restoring (not restoring) when $m$ and $H$ are parallel

157: (anti-parallel). The induced current in the coil produced by the

158: tilt of the oscillator is proportional to $1 - \cos \theta$. On

159: the other hand, when the coil is parallel to the plate and

160: perpendicular to $H$ as in Fig \ref{coils}.c the original $static$

161: current produces a change in the angle of equilibrium $\theta _0$.

162: However, the change in $k_e$ produced by this current is

163: negligible if $\theta _0$ is small as in this case. The current

164: induced by the alternating tilt of the oscillator produces a

165: restoring torque proportional to $\sin \theta$. Finally, when $H$

166: is parallel to the oscillator the torque is restoring (not

167: restoring) if $m$ and $H$ are parallel (anti-parallel). If $m$ and

168: $H$ are perpendicular the current induced by the alternating tilt

169: of the oscillator always produces a restoring torque.

170: 

171: 

172: The sample (BSCCO) has a layered structure. On mounting the disk,

173: the Cu-O layers ($ab$ planes) are parallel to the oscillator.

174: Following the previous description, currents in the layers can be

175: associated with a coil perpendicular to the oscillator and

176: currents between layers with a coil parallel to it.

177: 

178: \section{Results}

179: \label{}

180: 

181: We did ZFC (zero field cooling) and FC (field cooling)

182: measurements. The difference $\Delta \nu _r$ between $\nu _r$

183: measured with and without an applied $H$ reflects the magnetic

184: response of the sample, eliminating any intrinsic effect in the

185: oscillator produced by changes in temperature. When an external

186: magnetic field is applied perpendicular to the $ab$ planes of the

187: sample, Meissner currents appear. These currents screen $H$

188: causing a non-restitutive torque (the intra layers current behaves

189: as in Fig. \ref{coils}.a). In this case, when the sample is in the

190: vortex state, $H$ penetrates the BSCCO forming pancake vortices

191: (PVs), and PVs in different layers are coupled weakly via

192: Josephson vortices (JVs). In Fig. \ref{perpendicular} we show the

193: measured $\Delta \nu_r vs. T$ data with $H=245$ Oe. $\Delta \nu

194: _r$ is always negative. When the temperature increases, the sample

195: becomes less diamagnetic and the magnitude of $\Delta \nu _r$

196: decreases. This is due to the entering of vortex in the sample. If

197: $H$ is parallel to the sample and to the oscillator $\Delta \nu

198: _r$ is always positive (Fig. \ref{parallel}). This is because the

199: thickness of the sample ($2.5$ microns) is smaller than the

200: penetration length in the $c$ direction ($30$ microns

201: approximately) and the magnetic moment of the JVs is negligible .

202: The only currents present in the sample are the Meissner ones,

203: that shield the alternating perpendicular $H$.

204: \begin{figure}[t]\begin{center}

205: \includegraphics[width=\linewidth]{dolzFig5.eps}

206: \caption{Temperature dependence of the normalized resonant frequency for ZFC-FC

207: measurements with external magnetic field ($H=245$ Oe) perpendicular to the plate and to

208: the $ab$ planes of the sample. $\Delta \nu_r$ is always negative.} \label{perpendicular}

209: \end{center}\end{figure}

210: \begin{figure}[t]\begin{center}

211: \includegraphics[width=\linewidth]{dolzFig6.eps}

212: \caption{Temperature dependence of the normalized resonant frequency for ZFC-FC

213: measurements with external magnetic field ($H=245$ Oe) parallel to the plate and to the

214: $ab$ planes of the sample. $\Delta \nu _r$ is always positive.} \label{parallel}

215: \end{center}\end{figure}

216: \section{Discussion and Conclusion}

217: \label{}

218: 

219: In the first case, when $H$ is perpendicular to the layers of the sample, $m$ is

220: proportional to $1 - \cos \theta$ (as we already said in the previous section). The

221: maximum deflection of the oscillator depends on the width of the plate and the minimum

222: distance between the plate and the electrode before {\em snap-down} takes place. This

223: phenomenon consist in the collapse of the plate with the bottom electrode and it happens

224: when the deflection reaches a third of the equilibrium distance between them

225: \cite{microsystem}. In our case the maximum possible $\theta$ is $\approx 1.5$ degrees

226: giving a maximum change in $m$ of $0.035$ $\%$. On the other hand, when $H$ is parallel

227: to the layers $m$ is induced by the tilt of the oscillator (perpendicularly to the planes

228: $ab$ of the sample) and is proportional to $\sin \theta$.

229: 

230: 

231: It is well-known \cite{Kes90} that in BSCCO $2D$ samples, the

232: phase transition to the vortex state depends on the component of

233: $H$ perpendicular to the $ab$ planes of the sample

234: ($H_{c\parallel}$) and not on the component parallel to them.

235: Considering an error of $0.4$ Hz in the measurement of $\nu_r$ and

236: from the curves of Fig. \ref{perpendicular} we obtain $T_c (H_{c

237: \parallel}=245 Oe) = 78.7 \pm 0.7$ K. Due to the importance of

238: pinning, in the ZFC measurement $\nu_r$ is smaller than in the FC

239: measurement at low temperatures. When the temperature increases

240: these two curves become closer, and for $T > 46.4 \pm 0.7$ K the

241: behavior of the system is reversible. On the other hand, when $H$

242: is applied parallel to the sample (Fig. \ref{parallel}) the

243: measurements are almost reversible in all the range of

244: temperatures. In order to generate PVs, it is necessary to reach

245: the critical angle for which the perpendicular component of $H$

246: equals $H_{c1}$. The maximum component of $H$ perpendicular to the

247: $ab$ planes (appearing when the oscillator tilts) is of $6.5$ Oe

248: and, based on the observed reversibility, we can say that $H_{c1}$

249: is not reached.

250: 

251: These Si mechanical micro-oscillators are very sensitive magnetometers, suitable to study

252: mesoscopic high $T_c$ samples. The change in the resonant frequency ($\Delta \nu _r$) can

253: be positive or negative, depending on factors such as sample size, geometry and

254: anisotropy. These factors determine how the currents and magnetic moments are generated

255: in the sample.

256: \begin{thebibliography}{00}

257: 

258: \bibitem{Aksyuk98} V. Aksyuk, F. F. Balakirev, G. S. Boebinger, P. L. Gammel, R. C. Haddon,

259: D. J. Bishop, Science 280 (1998) 720.

260: \bibitem{Chan01} H. B. Chan, V. A. Aksyuk, R. N. Kleiman, D. J. Bishop, F.Capasso,

261:  Science 291 (2001) 1941.

262: \bibitem{Kleiman85} R. N. Kleiman, G. K. Kaminsky, J. D. Reppy, R. Pindak, D. J. Bishop, Rev. Sci. Instrum.

263: 56 (1985)2088-2091.

264: \bibitem{Bolle99} C. A. Bolle, V. Aksyuk, F. Pardo, P. L. Gammel, E. Zeldov, E. Bucher, R. Boie, D. J. Bishop,

265: D. R. Nelson, Nature 399 (1999) 43-46 .

266: \bibitem{Memscap} MEMSCAP Inc. ,Durham, NC 27703, USA. http://www.memscap.com

267: \bibitem{microsystem} S. D. Senturia, Microsystem Design, Kluwer Academic Publishers, Norwel, 2001, Chapter 6.

268: \bibitem{Kes90} P. H. Kes, J. Aarts, V. M. Vinokur, C. J. van der Beek, Phys. Rev. Lett. 64 (1990) 1063-1066.

269: 

270: 

271: \end{thebibliography}

272: 

273: \end{document}

274: