1: \documentclass{epl}
2: \usepackage{amssymb,graphicx,psfrag,amsmath,cite}
3: \title{The angular dependent magnetoresistance in
4: $\alpha$-(BEDT-TTF)$_2$KHg(SCN)$_4$}
5: \shorttitle{The angular dependent magnetoresistance\dots}
6: \author{Bal\'azs D\'ora\inst{1} \and Kazumi Maki\inst{2,3} \and Bojana
7: Korin-Hamzi\'c\inst{4} \and Mario Basletic\inst{5} \and Attila
8: Virosztek\inst{1,6}\and Mark V. Kartsovnik\inst{7} \and Harald
9: M\"uller\inst{8}}
10: \institute{
11: \inst{1} Department of Physics, Technical University of Budapest, H-1521 Budapest, Hungary \\
12: \inst{2} Max Planck Institute for the Physics of Complex Systems,
13: N\"othnitzer Str. 38, D-01187, Dresden, Germany\\
14: \inst{3} Department of Physics and Astronomy, University of Southern
15: California, Los Angeles
16: CA 90089-0484, USA \\
17: \inst{4} Institute of Physics, POB 304, HR-10001 Zagreb, Croatia\\
18: \inst{5} Department of Physics, Faculty of Science, POB 331, HR-10001
19: Zagreb, Croatia\\
20: \inst{6} Research Institute for Solid State Physics and Optics, P.O.Box
21: 49,
22: H-1525 Budapest, Hungary\\
23: \inst{7} Walther-Meissner Institute, D-85748 Garching, Germany\\
24: \inst{8} European Synchrotron Radiation Facility, F-38043 Grenoble,
25: France}
26: \shortauthor{Bal\'azs D\'ora \etal}
27:
28: \pacs{71.20.Rv}{Polymers and organic compounds}
29: \pacs{72.15.Gd}{Galvanomagnetic and other magnetotransport
30: effects}
31: \pacs{71.45.Lr}{Charge-density-wave systems}
32:
33: \date{}
34:
35: \begin{document}
36:
37: \maketitle
38:
39: \begin{abstract}
40: In spite of extensive experimental studies of the angular dependent
41: magnetoresistance (ADMR) of
42: the low temperature phase (LTP) of $\alpha$-(BEDT-TTF)$_2$KHg(SCN)$_4$ about a
43: decade ago, the
44: nature of LTP remains elusive. Here we present a new study of ADMR of LTP
45: in $\alpha-$(ET)$_2$
46: salts assuming that LTP is unconventional charge density wave (UCDW). In the
47: presence of magnetic field the quasiparticle spectrum in UCDW is quantized,
48: which gives rise to
49: striking ADMR in UCDW. The present model appears to account for many existing
50: ADMR data of
51: $\alpha$-(BEDT-TTF)$_2$KHg(SCN)$_4$ remarkably well.
52: \end{abstract}
53:
54: \section{Introduction}
55:
56: The series of the quasi-two-dimensional organic conductors
57: $\alpha$-(BEDT-TTF)$_2$MHg(SCN)$_4$ (where BEDT-TTF is
58: bis(ethylenedithio)tetrathiafulvalene and M=K, NH$_4$, Rb and Tl) have
59: attracted considerable attention in the last few years due to the richness
60: of physical phenomena observed \cite{singl}. Whereas the M=NH$_4$ compound becomes
61: superconducting below $1$ K, the other salts, at $T_c=8-10$ K, undergo a phase
62: transition into a specific low temperature phase (LTP), with associated
63: numerous anomalies in magnetic field. LTP is thought to be caused by a
64: density wave formation, but its nature appears still to be unsettled.
65: We have proposed recently that unconventional charge density wave can
66: account for a number of features of LTP in
67: $\alpha$-(BEDT-TTF)$_2$KHg(SCN)$_4$ including the threshold
68: electric field\cite{kuszobter,rapid,tesla}. Recently,
69: unconventional charge density wave (UCDW) and unconventional spin density
70: wave (USDW) have been proposed by several people as possible electronic
71: ground states in
72: quasi-one-dimensional and quasi-two-dimensional crystals\cite{Ners1,Ners2,benfatto,nagycikk,nayak}. Unlike conventional density wave\cite{gruner} the
73: order parameter in UCDW $\Delta(\bf k)$ depends on the quasiparticle wavevector $\bf k$.
74: In $\alpha$-(ET)$_2$ salts where the conducting plane lies in the a-c
75: plane and the quasi-one-dimensional Fermi surface is perpendicular to
76: the a-axis, we assume that $\Delta({\bf k})=\Delta\cos(ck_z)$ ( i.e.
77: $\Delta(\bf k)$ depends on $\bf k$ perpendicular to the most conducting direction),
78: where $c=9.778$ \AA $ $
79: is the lattice constant along the c axis\cite{endo}. It is known also that the
80: thermodynamics of UCDW and USDW is identical to the one in d-wave superconductors\cite{nagycikk,d-wave}.
81: Also, in spite of the clear thermodynamic signal, the first order term
82: in $\Delta(\bf k)$ usually vanishes when averaged over the Fermi surface. This implies no clear X-ray or
83: spin signal for UCDW and USDW. Due to this fact unconventional density wave ( i.e. UCDW and USDW) is sometimes
84: called the phase with hidden order parameter\cite{nayak}. In a magnetic field the quasiparticle spectrum is quantized
85: as first shown by Nersesyan et al.\cite{Ners1,Ners2}. This dramatic change in the quasiparticle spectrum is most readily seen
86: in ADMR, as demonstrated recently for SDW plus USDW in (TMTSF)$_2$PF$_6$ below $T=T^*$($\sim 4$K)\cite{romamaki,makitmtsf}.
87:
88: About a decade ago ADMR in LTP in $\alpha$-(ET)$_2$ salts has been studied intensively. However the origin
89: of ADMR has been hotly debated\cite{fermi,caulfield1,caulfield2,hanasaki}. At that time the Fermi surface reconstruction due to nesting has been accepted as the most likely solution\cite{hibas,blundell}.
90:
91: In the following we shall show that the quasiparticle spectrum in UCDW in $\alpha$-(ET)$_2$ salts is quantized
92: in the presence of magnetic field. The small energy gap which depends on both the direction and the strength
93: of the magnetic field can be seen through ADMR. Indeed we can describe most aspects of ADMR seen in LTP of $\alpha$-(BEDT-TTF)$_2$KHg(SCN)$_4$ consistently. Therefore we may conclude that ADMR in $\alpha$-(BEDT-TTF)$_2$KHg(SCN)$_4$
94: provides conclusive evidence that the LTP is UCDW.
95:
96: \section{Landau quantization}
97:
98: We shall consider the configuration shown in Fig. \ref{fig:koord},
99: where a magnetic field
100: $\bf B$ is applied within the a-b plane (i.e. $\phi=0$).
101: by angle $\theta$ tilted away from the b axis and in the
102: plane with angle $\phi$ from the a axis. For simplicity we shall limit our
103: analysis for $\phi=0$. Here the conducting plane is the a-c plane and the
104: quasi-one-dimensional Fermi surfaces lie perpendicular to the a axis. In
105: addition there is a quasi-two-dimensional Fermi surface with elliptical
106: cross-section in the a-c plane.
107:
108: \begin{figure}[h!]
109: \psfrag{B}[t][b][2][0]{$\bf B$}
110: \psfrag{c}[][][2][0]{$c$}
111: \psfrag{a}[][][2][0]{$a$}
112: \psfrag{b}[r][l][2][0]{$b$}
113: \psfrag{pp}[][][2][0]{$\phi$}
114: \psfrag{p}[][][2][0]{$\theta$}
115: \onefigure[width=7cm,height=7cm]{conf.eps}
116:
117: \caption{The geometrical configuration of the magnetic field with respect to
118: the conducting plane.}\label{fig:koord}
119: \end{figure}
120:
121: Therefore we assume that there are two conducting channels in this system:
122: the quasi-two-dimensional one stays
123: in the normal state while the quasi-one-dimensional one undergoes UCDW
124: transition around $T=8$K. Then the
125: quasiparticle spectrum in the quasi-1D channel is given by
126: \begin{equation}
127: E({\bf k})=\sqrt{\xi^2+\Delta({\bf k})^2}-\varepsilon_0\cos(2\bf b^\prime k),\label{elso}
128: \end{equation}
129: where $\xi\approx v_a k_a$, $\Delta({\bf k})=\Delta\cos(ck_c)$ and
130: $\varepsilon_0$ is the parameter describing
131: the imperfect nesting\cite{yamaji1,yamaji2,huang}. Finally $\bf b^\prime$
132: is the vector lying outside of the a-c plane.
133: In order to fit the dip in ADMR at $\theta=\theta_0$ it is necessary to
134: tilt $\bf b^\prime$ from the b axis
135: by an angle $\theta_0$ \cite{hanasaki}. As is seen from Eq. (\ref{elso}),
136: the quasiparticle
137: spectrum in the absence of magnetic field is gapless. When the magnetic field
138: tilted by an angle $\theta$ from the b axis is applied, the lowest Landau level
139: above the Fermi surface is given by
140: \begin{equation}
141: E(B,\theta)=\sqrt{2v_a\Delta c e
142: |B\cos\theta|}-\varepsilon_0\exp\left[-\frac{2\Delta c b^\prime}{v_a}e |B|
143: \frac{\sin^2(\theta-\theta_0)}{|cos{\theta}|}\right].
144: \label{ketto}
145: \end{equation}
146: The first term in Eq. (\ref{ketto}) is obtained following Nersesyan et al.
147: \cite{Ners1,Ners2}, while the second term in Eq.
148: (\ref{ketto}) comes from the spatial average of the second term in Eq.
149: (\ref{elso}) using the wavefunction of the Landau level at the Fermi
150: surface ($\sim exp(-v_a e |B\cos\theta|z^2/2c\Delta)$).
151:
152: Noting the fact that the system has two conducting channels ( i.e.
153: quasi-1D Fermi surface and quasi-2D Fermi surface) and that only the
154: quasi-1D Fermi surface is affected by the formation of UCDW, the ADMR is
155: written as \begin{eqnarray}
156: R(B,\theta)=\dfrac{1}{\dfrac{4\sigma_1}{1+e^x}+\sigma_2},\label{harom}
157: \\
158: x=\beta E(B,\theta),
159: \end{eqnarray}
160: where only the thermal excitations to the lowest Landau level are taken into
161: account explicitly, which is doubly degenerated.
162:
163: Similarly we obtain
164: \begin{equation}
165: \frac{\Delta R}{R(0,0)}=\frac{2\sigma_1(e^x-1)}{[4\sigma_1+\sigma_2(1+e^x)]}.
166: \end{equation}
167:
168: \section{Comparison with experiments}
169:
170: First, we compare Eq. (\ref{harom}) with $R(B,\theta)$ versus $T$ and
171: $R(B.\theta)$ versus $B$ in Figs. \ref{fig:mrtemp} and \ref{fig:mrfield}.
172: The temperature dependence of the magnetoresistance, for B=5T
173: perpendicular to the a-c plane, is presented in Fig \ref{fig:mrtemp}.
174: Solid line is the fit based on Eq. (\ref{harom}).
175: At low temperatures the consideration of the lowest
176: Landau level provides convincing agreement. But the higher the temperature the
177: higher Landau levels should be taken into account, and close to $T_c$ the
178: thermal
179: fluctuations play also an important role what we neglected here for simplicity.
180: The strength of the two conducting channels was found to be
181: $\sigma_2/\sigma_1=0.372$,
182: and by assuming the weak coupling value of
183: $\Delta=17K$ and using $c=9.778$\AA \cite{endo},
184: the Fermi velocity is obtained as $v_a=7\times 10^6$cm/s.
185:
186: The magnetic field dependence of the magnetoresistance at $T=2.2$ K and
187: $T=4.2$ K for magnetic field perpendicular to the a-c plane is shown in
188: Fig. \ref{fig:mrfield}. Solid line is the fit based on the Eq. (\ref{harom}).
189: At higher fields, where our simple
190: approximation is valid, the agreement looks perfect again. Here
191: $\sigma_2/\sigma_1=0.24$ and $0.48$ for $T=2.2$K and $4.2$K, respectively, and
192: the extracted Fermi velocities (assuming again the weak coupling value of
193: $\Delta$) are $v_a=3\times 10^6$ cm/s and
194: $7\times 10^6$cm/s.
195: %$\beta\sqrt{v_a\Delta ceB}=3.44$ at
196: %$T=2.2K$ and $B=5T$ we can extract $v_a\approx 6\times 10^6$m/s, where we
197: %assumed
198: %the weak-coupling value of $\Delta=17$K. A smaller
199: %$\Delta$ may result in a higher value of $v_a$.
200:
201: \begin{figure}[h!]
202: \psfrag{xx}[t][b][1][0]{$T(K)$}
203: \psfrag{yy}[b][t][1][0]{$\Delta R/R(0,0)$}
204: \psfrag{5T}[][][1][0]{$B=5$T}
205: \psfrag{x}[t][b][1][0]{$B(T)$}
206: \psfrag{y}[b][t][1][0]{$\Delta R/R(0,0)$}
207: \psfrag{2.2K}[r][l][1][0]{$T=2.2$K}
208: \psfrag{4.2K}[l][r][1][0]{$T=4.2$K}
209: \twofigures[width=7cm,height=7cm]{mrtemp.eps}{mrfield.eps}
210:
211: \caption{The temperature dependence of the magnetoresistance is plotted at
212: $B=5$T.}\label{fig:mrtemp}
213: \caption{The magnetic field dependence of the magnetoresistance is shown at
214: $T=2.2$K and $4.2$K.}\label{fig:mrfield}
215: \end{figure}
216:
217: Finally ADMR
218: is shown in Fig. \ref{fig:mrtheta} as a function of angle $\theta$ at $T=4.2$K,
219: $B=5$T. The solid line shows the fit to the theoretical model explained above.
220: As is seen from the figure
221: the global $\theta$ dependence is given by $x\approx \beta
222: \sqrt{2v_a \Delta ce|B\cos\theta|}$, since the data is taken at $T=4.2$K and
223: $B=5$T.
224: \begin{figure}[h!]
225: \psfrag{x}[t][b][1][0]{$\theta$}
226: \psfrag{y}[b][t][1][0]{$R(5T,\theta)$ $(\Omega)$}
227: \psfrag{s}[l][r][1][0]{$T=4.2$K, $B=5$T}
228: \onefigure[width=7cm,height=7cm]{mrtheta.eps}
229: \caption{The angle dependent magnetoresistance is shown at $T=4.2$K and $B=5$T.}\label{fig:mrtheta}
230: \end{figure}
231:
232: The ratio of the conductivities in the two channels is $\sigma_2/\sigma_1=0.36$.
233: The different values of this ratio might arise from the fact that we considered
234: only the explicit $B$ and $T$ dependence of the main mechanism coming from
235: the UCDW condensate, which according to us is responsible for the general
236: behaviour of the measured
237: magnetoresistance, and we neglected the magnetic field
238: and temperature dependence of the other possible conducting processes.
239: The Fermi velocity is obtained as $v_a=5\times 10^6$cm/s. From these
240: data, the Fermi velocity turned out to be of the order of $10^6$cm/s, and
241: its
242: uncertainty should also be affected by the exclusion of the other conducting
243: mechanisms.
244:
245: The dip structure in ADMR is described fairly well by assuming
246: $\varepsilon_0=0.132$K and $\theta_0=40^\circ$. This $\varepsilon_0$
247: value is an order of magnitude smaller than the one we needed to describe the
248: threshold electric field observed in $\alpha$-(BEDT-TTF)$_2$KHg(SCN)$_4$
249: \cite{kuszobter,rapid,tesla,physicaB}.
250: However, compared to other similar data (see \cite{hanasaki} ), it appears that there are
251: considerable variability in the magnitude of $\varepsilon_0$ and the angle
252: $\theta_0$ in different
253: crystals ($\theta$ varies between $35^\circ$ and $50^\circ$).
254: Therefore, we may conclude that UCDW in a magnetic field describes
255: a variety of features in ADMR in $\alpha$-(BEDT-TTF)$_2$KHg(SCN)$_4$
256: satisfactorily.
257:
258: \section{Concluding remarks}
259:
260: From the analysis of the temperature dependence of the threshold electric field we have concluded earlier
261: that the LTP in $\alpha$-(ET)$_2$ salts is most likely
262: UCDW\cite{rapid}. The present analysis of ADMR appears to confirm this
263: identification.
264: The quasiparticle spectrum in UCDW in magnetic field is quantized in
265: general\cite{Ners1,Ners2}.
266: This effect should be most readily seen by the angular dependent
267: magnetoresistance as we have demonstrated in
268: (TMTSF)$_2$PF$_6$\cite{romamaki,makitmtsf}.
269: We believe that ADMR will provide a powerful technique to explore other possible UCDW or USDW states in transition metal trichalogenate\cite{castroneto} and URu$_2$Si$_2$\cite{IO,roma} for example.
270:
271: \section{Acknowledgements}
272:
273: We thank Peter Thalmeier, Amir Hamzi\'c
274: and Silvia Tomi\'c for useful discussions on the related topics.
275: One of the authors (B. D.) gratefully
276: acknowledges the hospitality of the Max Planck Institute for the
277: Physics of Complex Systems, Dresden, where part of this work was done.
278: This work
279: was supported by the Hungarian National Research Fund under grant numbers
280: OTKA T032162 and T037451, and by the Ministry of Education under grant
281: number FKFP 0029/1999.
282:
283:
284: \bibliographystyle{apsrev}
285: \bibliography{mr}
286: \end{document}
287: