1: %\documentclass[preprint,showpacs,showkeys,preprintnumbers,amsmath,amssymb]{revtex4}
2: \documentclass[aps,prl,showpacs,showkeys,twocolumn,amsmath,amssymb]{revtex4}
3:
4: % Some other (several out of many) possibilities
5: %\documentclass[preprint,aps]{revtex4}
6:
7: \usepackage{graphicx}% Include figure files
8: \usepackage{dcolumn}% Align table columns on decimal point
9: \usepackage{natbib}
10: \usepackage{color,calc}
11: \usepackage{bm}% bold math
12:
13: \newcommand{\bvs}{BaVS$_{3-\delta}$}
14: \newcommand{\bsvs}{Ba$_{1-x}$Sr$_x$VS$_3$}
15:
16: \setlength{\textheight}{220mm}
17: \setlength{\textwidth}{170mm}
18:
19: \begin{document}
20:
21: \title{Evidence of ferromagnetic quantum phase transition in \bsvs~}
22: \author{Andrea Gauzzi$^1$}
23: \email{andrea.gauzzi@upmc.fr}
24: \author{Neven Bari\v{s}i\'{c}$^2$}
25: \author{Francesca Licci$^1$}
26: \author{Gianluca Calestani$^3$}
27: \author{Fulvio Bolzoni$^1$}
28: \author{Patrik Fazekas$^4$}
29: \author{Edi Gilioli$^1$}
30: \author{L\'{a}szl\'{o} Forr\'{o}$^2$}
31:
32: \affiliation{$^1$IMEM-CNR, Area delle Scienze, 43010 Parma, Italy\\
33: $^2$Institut de Physique de la Mati\`ere Complexe, EPFL, CH-1015 Lausanne,
34: Switzerland\\
35: $^3$Dipartimento di Chimica, Universit\`a di Parma, Area delle Scienze, 43100
36: Parma, Italy\\
37: $^4$Research Institute for Solid State Physics and Optics, P.O. Box 49, H-1525
38: Budapest, Hungary}
39:
40: \date{\today}
41:
42: \begin{abstract}
43: Resistivity under pressure and magnetization measurements on \bsvs~ single
44: crystals with 0 $\leq x \leq$ 0.18 and no sulphur deficiency show an abrupt
45: onset of ferromagnetic (FM) order at a critical value $x$=0.07, concomitant to a
46: change of the magnetic properties at the metal-insulator transition (MIT) and to
47: a collapse of the unit cell at ambient temperature. A reduction of the MIT
48: temperature to 50 K upon $x$ that scales as the V-S distance is also observed.
49: This gives evidence of a chemical pressure induced quantum phase transition that
50: stabilizes the incipient FM order of BaVS$_3$. The $0.07<x$ results suggest a
51: coexistence of FM and metallic phases at larger $x$.
52: \end{abstract}
53:
54: \pacs{71.30.+h,71.27.+a,75.50.-y}
55: % PACS, the Physics and Astronomy Classification Scheme.
56: %\keywords{Metal-insulator transitions and other electronic transitions;
57: %Strongly correlated electron systems; Studies of specific magnetic materials}
58: %Use showkeys class option if keyword
59: %display desired
60:
61: \maketitle
62:
63: The hexagonal perovskite \bvs \cite{gar} displays a manifold of magnetic,
64: transport and structural properties driven by 3$d$(V$^{4+}$) electron correlations.
65: In the high temperature metallic phase, it consists of hexagonal closed
66: packed chains of face-sharing VS$_6$ octahedra. The MI transition $T_{MI}$=69 K
67: is preceded by a zigzag of the chains at 240 K leading to an orthorhombic
68: structure \cite{tak}. $T_{MI}$ decreases under pressure and the insulating phase
69: is suppressed for $P>2$ GPa \cite{for}. In stoichiometric compounds
70: ($\delta$=0), the ambient pressure transition is concomitant to an AF-like cusp
71: of the magnetic susceptibility, $\chi$ \cite{gar}, and to a $2 \times c$
72: superstructure along the chains \cite{ina,fag}. A neutron scattering study
73: \cite{nak} reported a long range AF-like order in the $ab$-plane and an intrachain
74: ferromagnetic (FM) one at $T_X$=30 K, related to the cusp. Assuming a large number
75: of degenerate configurations of singlet pairs with different orbital symmetry in
76: the triangular plane, a picture of spin-orbital liquid was proposed \cite{nak,mih}.
77: A latent FM instability is apparent from the positive Weiss constant in the metallic
78: phase and the FM behavior of sulphur-deficient samples \cite{mas}. In agreement with
79: these experiments, \emph{ab initio} calculations \cite{xue,san} predict a
80: dominant FM coupling along $c$ and a comparatively weak AF one in the $ab$-plane,
81: thus suggesting a ground state of FM chains with a nearly frustrated AF interchain
82: coupling.
83:
84: The origin of the MI transition and its relation with the magnetic and orbital
85: orderings remain controversial. Overlapping wide and narrow $d$-bands were
86: recently observed by ARPES \cite{mitr}. A simple tight-binding model explaining
87: the physical properties at the MI transition includes a broad $A_{1g}$ band
88: arising from the direct overlap of $3d_{z^2}$ orbitals along the $c$-axis, and a
89: doubly degenerate $E_{g}$ narrow band created by the small V(3$d$)-S(3$p$)
90: hybridization \cite{mas,bar}. The width and relative occupancy of the bands
91: remain to be determined experimentally. An X-ray diffraction study \cite{fag}
92: suggests a nearly equal filling of the $A_{1g}$ and $E_{g}$ bands. A tuning
93: mechanism for the occupancy of these two bands was recently proposed \cite{lec}.
94:
95: In this letter, we study the effects of chemical pressure on the magnetic,
96: transport and structural properties of the system. At a critical value,
97: $x$=0.07, an abrupt onset of FM order is found concomitant to a disappearance of
98: the AF-like cusp at the MI transition and to a collapse of the unit cell at room
99: temperature. A progressive reduction of $T_{MI}$ and smearing of the AF-like
100: cusp with increasing $x$ is also observed. The observation of a quantum phase
101: transition (QPT) from a non-FM to a FM low-temperature state shows that the
102: incipient FM order in the parent compound is stabilized by chemical pressure.
103: The second result suggests a coexistence of FM and metallic phases at high
104: pressure \cite{bar}. We discuss the possibility that the two phases coincide in
105: the absence of disorder. We argue that the collapse of the unit cell is due
106: to a sudden redistribution of electron charge between the relevant $A_{1g}$ and
107: $E_{g1}$ bands.
108:
109: A series of \bsvs~ single crystals with $x$ = 0.032, 0.053, 0.064, 0.068, 0.097,
110: 0.126, and 0.183 were grown as described elsewhere using a solid state method
111: \cite{gau}. Attempts to grow crystals with larger $x$ values were unsuccessful.
112: At $x$ = 0.183, diffraction data indicate a significant amount of disorder and
113: stacking faults, thus $x \approx$ 0.18 appears to be close to the solubility
114: limit. The needle-shaped crystals, 0.1 - 0.3 mm long, are aligned along the $c$-
115: axis. The room temperature crystal structure, the Sr/Ba composition and the
116: sulfur content were determined using X-ray diffraction. Within the experimental
117: resolution, the crystal symmetry remains $P6_3/mmc$ for all $x$ values, as in the
118: unsubstituted compound. The structural refinement indicates no sulfur deficiency
119: for any $x$ with a statistical uncertainty $\delta \pm$ 0.05 or better. As a
120: further check of the sulfur stoichiometry, a post-sulfurisation treatment in
121: evacuated quartz ampoules at 450$^\circ$C for 3 days is found to affect neither
122: the structural nor the physical properties. Thus, the results reported are
123: exclusively ascribed to the Sr-induced chemical pressure.
124:
125: The crystals were investigated by means of dc electrical resistivity under
126: pressure, $\varrho(T,P)$, up to 2 GPa, and dc magnetization, $M(T)$. $\varrho$
127: was measured in a four contact bar configuration in self-clamping pressure cells
128: filled with kerosene as pressurizing agent. Since the needle-shaped crystals
129: grow along the $c$-axis, the $c$-axis resistivity, $\varrho_c$, was measured.
130: $M$ was measured in the zero-field and field-cooling (ZFC, FC) modes at 100
131: oersted using a commercial RF SQUID magnetometer. Because of the small crystal
132: size, the sample weight could not be determined. Yet, the sample sizes are
133: comparable and the relative magnitude of $M$ is significant.
134:
135: A set of $\varrho_c(T,P)$ curves is shown in Fig. 1 for $x$=0.126.
136: Similarly to the case of the unsubstituted compound \cite{for} ($x$=0), a
137: progressive reduction of $T_{MI}$ with increasing pressure is found. $T_{MI}$
138: was estimated as the peak of the $d$ln$\varrho/d(1/T)$ vs. $T$ plot \cite{for}.
139: The effect of $T_{MI}$ reduction with increasing $x$ and with increasing
140: pressure is apparent in the $P-T$ diagram of Fig. 2 that summarizes the
141: $\varrho$ data. For $x \approx$ 0.12, the insulating state is suppressed at
142: $\approx$ 1 GPa, i.e. half of the value for the unsubstituted compound \cite{for}.
143: By comparing the $\varrho_c(T,P)$ curves in the inset of Fig. 1, one notes that
144: the $T_{MI}$ reduction is accompanied by an increase of the slope, $d\varrho_c/dT$,
145: with increasing pressure. This trend is common to all samples and indicates a
146: pressure-induced enhancement of the metallic properties, although the system remains
147: a bad metal. By comparing the $\varrho(T)$ curves for different $x$, we observed
148: that the residual resistivity increases with increasing $x$ \cite{bar2}. This trend
149: indicates that the Sr-substitution creates a substantial amount of disorder,
150: in agreement with diffraction data.
151: \begin{figure}
152: \includegraphics[width=8cm]{bsvs_fig1.eps}
153: \caption{$c$-axis resistivity curves as a function of temperature and pressure
154: for $x$=0.126. One notes the steady reduction of the metal-insulator transition
155: temperature $T_{MI}$, as pressure increases.}
156: \label{rhovsTP}
157: \end{figure}
158: \begin{figure}
159: \includegraphics[width=8cm]{bsvs_fig2.eps}
160: \caption{Progressive reduction of $T_{MI}$ upon $x$ from the full set of the
161: $\varrho_c(T,P)$ data. The $x=0$ data are taken from ref. \cite{for}. Lines are
162: a guide to the eye.}
163: \label{T_MIvsx}
164: \end{figure}
165:
166: The $T_{MI}$ reduction is correlated with the Sr--induced structural changes of
167: the $a$- and $c$-axis as shown in Fig. 3. One notes that the smaller Sr ion
168: induces a shrinking of the unit cell, as expected. Two features are though
169: unusual: 1) at low substitution levels, $x \lesssim$ 0.05, the shrinking is
170: negligible; 2) at higher $x$ values, the shrinking is large and occurs suddenly,
171: first along the $c$-axis at $x \approx$ 0.07 and then in the $ab$-plane at $x
172: \approx$ 0.10. The latter shrinking is largest and corresponds to a collapse of
173: the chains of VS$_6$ octahedra one against the other. The $c$-axis increase at
174: $x \gtrsim 0.10$ is consistent with the previous observation of disorder and
175: stacking faults in this $x$ range. In spite of this, the cell volume continues
176: to decrease with $x$, for the shrinking of the $a$-axis dominates. The sudden
177: drop of the $a$- and $c$-axis at $x \approx$ 0.07 appears to be a room
178: temperature signature of a phase transition at low temperature, as discussed below.
179: \begin{figure}
180: \includegraphics[width=8cm]{bsvs_fig3.eps}
181: \caption{Top panel: $a$- and $c$-axis variations of the hexagonal unit cell of
182: \bsvs~ with $x$. Note the drop of the $c$-axis at $x \approx$ 0.07 and the
183: progressive reduction of the $a$-axis followed by a large drop at $x \approx$
184: 0.10. Error bars are standard deviations of the data refinement. Bottom panel:
185: correlation between cell volume reduction and Curie temperature, $T_C$, and $x$.
186: The vertical line at $x$ = 0.07 marks the onset of the cell volume drop and of
187: the FM order.}
188: \label{cellvsx}
189: \end{figure}
190:
191: By analyzing the structural data, the V-S bond distance in the VS$_6$ octahedra
192: is found to be the only parameter that scales with $T_{MI}$. A linear regression
193: of the $T_{MI}$ vs. V-S data enables to predict that the M-I transition would be
194: suppressed for V-S=2.372 \AA. This result is in agreement with the observation
195: of $T_{MI}$ reduction upon hydrostatic pressure \cite{for}. One concludes that
196: the volume reduction induced by either chemical or hydrostatic pressure
197: stabilizes the metallic state. The $T_{MI}$ reduction observed in our \bsvs~
198: samples is correlated to the magnetization data as follows. In Fig. 4 we report
199: a selection of $M(T)$ curves for three different $x$ values in the $x$=0.06-0.10
200: region, that turns out to be critical, as shown below. For $x \lesssim 0.07$, no
201: significant difference between ZFC and FC curves is observed, so only the former
202: case is considered. The AF-like cusp of $\chi$ characteristic of the unsubstituted
203: compound \cite{tak,mas} remains visible at $T_{MI}$, although smeared by disorder.
204: By taking the cusp as the magnetic signature of $T_{MI}$, the $T_{MI}$ vs. $x$
205: magnetization data are found to coincide with the resistivity ones discussed
206: before. The two data sets are reported in Fig. 5 and show the same monotonic
207: decrease of $T_{MI}$ with $x$. Also the upturn of the $M(T)$ curve at $T_X$
208: characteristic of the unsubstituted compound \cite{mih} remains visible for $x
209: \lesssim 0.07$. Contrary to the case of the AF-like cusp, the temperature of the
210: upturn hardly varies with $x$ but also the upturn becomes smeared with $x$. At
211: $x_{cr}=0.07$, both the cusp and the upturn disappear and a FM response suddenly
212: appears with $T_C$=12 K, determined as the maximum of the derivative of the
213: $M(T)$ curve.
214: \begin{figure}
215: \includegraphics[width=8cm]{bsvs_fig4}
216: \caption{ZFC magnetization for three different $x$-values showing the
217: disappearance of the AF-like cusp at $T_{MI}$ and of the upturn at $T_X$
218: concomitant to the onset of FM order at $x \approx 0.07$. Note the different
219: vertical scales.}
220: \label{MvsT}
221: \end{figure}
222:
223: The above observations lead to conclude that the MI transition is not
224: necessarily associated with the magnetic correlations. This contrasts to the
225: transition at $T_x$ \cite{bar,bar2}, characterized by a FM order along the chains,
226: in agreement with neutron diffraction data \cite{nak} and with the FM above
227: $x_{cr}$ presented in this letter. The analysis of several crystals with $x
228: \approx 0.07$ confirms the sudden character of the onset of the FM order and its
229: coincidence with the disappearance of the AF-like cusp of $\chi$ and an equally
230: sudden drop of cell volume. In the bottom panel of Fig. 3, we show the correlation
231: between volume reduction and onset of FM order upon $x$. At $x_{cr} > 0.07$, $T_C$
232: slowly increases and equals 22 K at $x \approx$ 0.18.
233:
234: Fig. 5 summarizes all data discussed above and serves as electronic phase
235: diagram. The AF-like borderline is derived from the estimation of $T_X$ as
236: upturn of the magnetization. The upturn being increasingly smeared with $x$, as
237: it is apparent from Fig. 4, we believe that a long range order in the ab-plane
238: no longer exists in the substituted compounds near $x_{cr}$. Thus, the sudden
239: stabilization of the FM order at $x_{cr}$ and at low temperature is a quantum
240: phase transition that distinguishes the FM order from a phase with AF-like
241: correlations in the ab-plane that become weaker with increasing $x$. The drop
242: of cell volume at $x_{cr}$ appears as a structural signature of the electronic
243: instability at room temperature.
244: \begin{figure}
245: \includegraphics[width=8cm]{bsvs_fig5.eps}
246: \caption{Electronic phase diagram of Ba$_{1-x}$Sr$_x$VS$_3$ at ambient pressure.
247: Full and open circles indicate the $T_{MI}$ values determined from,
248: respectively, resistivity and magnetization data. The smearing of the AF-like
249: cusp at $T_X$ upon $x$ discussed in the text is indicated as a broken line. The
250: vertical line at $x$=0.07 marks the abrupt onset of ferromagnetic (FM) order.
251: Lines are a guide to the eye.}
252: \label{phasediagram}
253: \end{figure}
254:
255: In conclusion, we studied the effects of chemical pressure on the MI and AF-FM
256: competitions in the BaVS$_3$ system on a series of \bsvs~ single crystals. The
257: data of resistivity under pressure show that the Sr-induced chemical pressure
258: stabilizes the metallic phase. Consequently, by increasing $x$ from 0 to 0.12
259: the $T_{MI}$ is reduced from 69K down to 50 and the critical pressure for
260: the suppression of the MI transition at 0K decreases from 2 to 1 GPa. At
261: $x_{cr}$ = 0.07, we found a sudden shrinking of the unit cell concomitant to the
262: disappearance of the AF-like susceptibility cusp at $T_{MI}$ and to the
263: appearance of a full FM order at low temperatures. This transition confirms the
264: small energy difference between the AF-like and FM orderings of the V$^{4+}$
265: ions calculated \emph{ab initio} \cite{san}. By extrapolating the $T_{MI}$ vs.
266: V-S bond data, a metallic state at 0 K is predicted for a distance of 2.372 \AA.
267: The present data suggest that, at higher $x$ or pressure, the FM and metallic
268: state should coexist \cite{bar}, as in BaVSe$_3$ \cite{yam,bar2}. The question
269: arises as to the mechanism leading to the FM order in our case. The absence of
270: sulphur deficiency rules out the Hund mechanism proposed for sulphur-deficient
271: compounds \cite{yam2}. The Sr/Ba chemical disorder would favor a double-exchange
272: mechanism, thus one should study samples without disorder. As disorder tends to
273: stabilize the insulating state, the reduction of $T_{MI}$ reported here is
274: expected to be even stronger in the absence of disorder. The MI quantum phase
275: transition predicted at $x \approx 0.15$ from the data of Fig. 5 would then
276: approach the FM QPT at $x=0.07$. In the extreme case, the two points may merge,
277: which would be important from a fundamental point of view.
278:
279: \begin{acknowledgments}
280: The authors thank M. Marezio for useful discussions and T. Besagni for technical
281: assistance and acknowledge financial support provided by the \textit{Consiglio
282: Nazionale delle Ricerche}.
283: \end{acknowledgments}
284:
285: \begin{thebibliography}{100}
286:
287: \bibitem{gar} R. A. Gardner, M. Vlasse and A. Wold, Acta Cryst. B \textbf{25},
288: 781 (1969).
289: \bibitem{for} L. Forr\'{o} \textit{et al.}, Phys. Rev. Lett. \textbf{85}, 1938
290: (2000).
291: \bibitem{tak} M. Takano \textit{et al.}, J. Phys. Soc. Jpn. \textbf{43}, 1101
292: (1977).
293: \bibitem{ina} T. Inami \textit{et al.}, Phys. Rev. B \textbf{66}, 073108 (2002).
294: \bibitem{fag} S. Fagot \textit{et al.}, Phys. Rev. Lett. \textbf{90}, 196401
295: (2003).
296: \bibitem{nak} H. Nakamura \textit{et al.}, J. Phys. Soc. Japan \textbf{69}, 2763
297: (2000).
298: \bibitem{mih} G. Mih\'{a}ly \textit{et al.}, Phys. Rev. B \textbf{61}, R7831
299: (2000).
300: \bibitem{mas} O. Massenet \textit{et al.}, Mat. Res. Bull. \textbf{13}, 187
301: (1978).
302: \bibitem{xue} X. Jiang and G. Y. Guo, Phys. Rev. B \textbf{70}, 35110 (2004).
303: \bibitem{san} A. Sanna, C. Franchini, S. Massidda, and A. Gauzzi, Phys. Rev. B
304: \textbf{70}, 235102 (2004).
305: \bibitem{mitr} S. Mitrovic \textit{et al.}, cond-mat/0502144.
306: \bibitem{lec} F. Lechermann, S. Biermann, and A. Georges, Phys. Rev. Lett.
307: \textbf{94}, 166402 (2004).
308: \bibitem{gau} A. Gauzzi \textit{et al.}, Int. J. Mod. Phys. B \textbf{17}, 3503
309: (2003).
310: \bibitem{bar2} N. Bari\v{s}i\'{c}, \emph{PhD Thesis}, (EPFL, Lausanne, 2004)
311: \emph{https://nanotubes.epfl.ch/nbarisic}.
312: \bibitem{bar} N. Bari\v{s}i\'{c} \textit{et al.}, unpublished (2005).
313: \bibitem{yam} T. Yamasaki, S. Giri, H. Nakamura, and M. Shiga, J. Phys. Soc.
314: Jpn. \textbf{70}, 1768 (2001).
315: \bibitem{yam2} T. Yamasaki, H. Nakamura, and M. Shiga, J. Phys. Soc. Jpn.
316: \textbf{69}, 3068 (2000).
317: \end{thebibliography}
318:
319: \end{document}
320:
321: % ****** End of file ******
322:
323:
324:
325:
326:
327:
328: