1: \documentclass[aps,showpacs,two column]{revtex4}
2: %\documentclass[preprint,showpacs,nobibnotes,epsf]{revtex4}
3:
4: \usepackage{graphicx}% Include figure files
5: \usepackage[squaren]{SIunits}
6:
7: \begin{document}
8: \title{Origin of superconductivity in nominally ``undoped"
9: T'-La$_{2-x}$Y$_{x}$CuO$_{4}$ films}
10:
11: \author{L. Zhao, R. H. Liu, G. Wu, G. Y. Wang, T. Wu, X. G. Luo and X. H. Chen}
12: \altaffiliation{Corresponding author} \email{chenxh@ustc.edu.cn}
13: \affiliation{Hefei National Laboratory for Physical Science at
14: Microscale and Department of Physics, University of Science and
15: Technology of China, Hefei, Anhui 230026, People's Republic of
16: China}
17:
18:
19: \begin{abstract}
20: We have systematically studied the transport properties of the
21: La$_{2-x}$Y$_{x}$CuO$_{4}$(LYCO) films of T'-phase ($0.05\leq x \leq
22: 0.30$). In this nominally ``undoped" system, superconductivity was
23: acquired in certain Y doping range ($0.10\leq x \leq 0.20$).
24: Measurements of resistivity, Hall coefficients in normal states and
25: resistive critical field ($H^\rho_{c2}$)in superconducting states of
26: the T'-LYCO films show the similar behavior as the known Ce-doped
27: n-type cuprate superconductors, indicating the intrinsic
28: electron-doping nature. The charge carriers are induced by oxygen
29: deficiency. Non-superconducting Y-doped Pr- or Nd-based T'-phase
30: cuprate films were also investigated for comparison, suggesting the
31: crucial role of the radii of A-site cations in the origin of
32: superconductivity in the nominally ``undoped" cuptates. Based on a
33: reasonable scenario in the microscopic reduction process, we put
34: forward a self-consistent interpretation of these experimental
35: observations.
36: \end{abstract}
37: \vskip 15 pt
38: \pacs{74.72.Dn, 74.78.Bz, 74.62.Dh, 74.90.+n}
39: \maketitle
40:
41:
42: \section{INTRODUCTION}
43:
44: After 20 years of intense research since the discovery of
45: high-temperature superconductivity(HTSC)\cite{bednorz}, many
46: interesting physical phenomena unique to the cuprate superconductors
47: are better understood. Although the underlying mechanism for HTSC
48: remains elusive and unresolved, it has been widely accepted
49: \cite{orenstein} that the parent compounds of all the known
50: superconducting cuprates are half-filled Mott antiferromagnetic
51: insulators, in which the strong electronic correlation exists. Upon
52: doping holes or electrons into CuO$_2$ sheets, the
53: antiferromagnetism is suppressed, and superconductivity can be
54: induced in a certain concentration range of charge carriers.
55:
56: Recently, Tsukada et al discovered a new class of cuprate
57: superconducting films, T'-(La, R)$_{2}$CuO$_{4}$ (LRCO, R=Sm, Eu,
58: Gd, Tb, Lu, and Y), which are nominally ``non-doped" because
59: R$^{3+}$ is the isovalent cation with La$^{3+}$ and the substitution
60: can not provide net charge carriers to CuO$_2$ planes \cite{naito2}.
61: In particular, the La-based T'-phase cuprates are metastable and
62: have not be acquired by conventional bulk synthesis now. These
63: T'-phase LRCO films were synthesized by molecular beam epitaxy
64: (MBE), and T$_c$ above 20K could be achieved in certain R doping
65: ranges. the authors claimed that the electron doping via oxygen
66: deficiencies was, at least, not a main source of charge carriers and
67: T'-LRCO are most plausibly conventional half-filled ``band
68: superconductors" with weak electronic correlation \cite{naito3,
69: naito4}. Their viewpoints arose strong skepticism because it
70: contradicts the acknowledged basic picture of HTSC in last 20 years.
71:
72: However, further study on the physical properties of T'-LRCO is
73: lacking because the sample preparation is quite difficult due to the
74: metastability of La-based T'-phase structure and the repeatability
75: far from satisfaction by other groups. For example, an unsuccessful
76: attempt was reported recently\cite{Idemoto}, indicating that the
77: superconductivity of films is very sensitive to preparation
78: conditions. By pulsed laser deposition(PLD), Greene and co-workers
79: also prepared T'-LYCO films for only x=0.15 with partial success
80: \cite{greenelyco}. At present, it is urgent to perform systematic
81: study on LRCO physical properties.
82:
83: Here using dc magnetron sputtering, we have been able to fabricate
84: the superconducting La$_{2-x}$Y$_{x}$CuO$_{4}$ (LYCO) films of pure
85: c-axis oriented T'-structure with high repeatability
86: \cite{lzhaoapl}. In this paper, we systematically investigated
87: in-plane electronic transport and Hall coefficients of T'- LYCO
88: films with different Y contents: $x=0.05, ~0.10, ~0.125, ~0.15,
89: ~0.20$ and $0.30$. The superconducting phase diagram and effect of
90: oxygen content on transport properties are studied. For films with
91: $x=0.125$ and $0.20$, the magnetic field dependence of
92: superconducting transitions was also measured to get the resistive
93: critical field $H^\rho_{c2}(T)$.
94:
95: All these results show the similar behavior as the known
96: electron-doped T'-214 R$_{2-x}$Ce$_{x}$CuO$_{4}$ (R=Pr, Nd, ...)
97: systems, suggesting the same electron-doping nature in LYCO. The
98: charge carriers in T'-LYCO are induced by oxygen deficiency during
99: the reduction process. For comparison, we also studied the isovalent
100: substitution in Nd$_2$CuO$_4$(NCO) and Pr$_2$CuO$_4$(PCO) systems.
101: No superconductivity was found. Combining our data and the results
102: in references \cite{naito2, naito3, naito4}, we present an approving
103: and self-consistent interpretation on the origin of
104: superconductivity in ``undoped" T'-214 cuprates, based on a
105: reasonable scenario about the microscopic depletion process of
106: oxygen upon annealing.
107:
108: \section{EXPERIMENT DETAILS}
109:
110: La$^{3+}$ ion is the largest among lanthanide series. The detailed
111: analysis on the perovskite tolerance factor $t$ \cite{goodenough1,
112: goodenough2}, has told us that La-based 214 cuprates adopt the
113: T-type phase by high-temperature bulk processes, while T'-214 phase
114: prefers stabilizing only at low synthesis temperature. By
115: extrapolation of the T/T' phase boundary in the
116: La$_{2-y}$Nd$_y$CuO$_4$ system to $y=0$, it has been predicted that
117: undoped T'-La$_2$CuO$_4$ can be stable below 425$\celsius$
118: \cite{goodenough2}. The partial substitution of La$^{3+}$ by smaller
119: Y$^{3+}$ can reduce the average ion radius of A-site, and shift
120: slightly the T/T' phase boundary to higher temperature. But it is
121: still too low for bulk synthesis of La-based T'-phase. In comparison
122: with the solid-state reaction process, the preparation of thin films
123: can usually be realized at lower temperature. Furthermore the
124: appropriate epitaxy strain through the substrates may stabilize some
125: metastable phases \cite{Tsukada}. Therefore, we grew c-axis oriented
126: LYCO films of pure T'-phase by dc magnetron sputtering. The
127: stoichiometric ceramic targets are prepared by conventional
128: solid-state reaction. Fig. 1 shows the x-ray diffraction(XRD)
129: pattern of a typical LYCO film with x=0.15, which is pure
130: T'-structure of c-axis orientation. Only peaks of ($0~0~2l$)
131: diffraction appear, and the c-axis lattice constant $c$ is around
132: 12.44\AA. The typical full width at half maximum (FWHM) of rocking
133: curves at the (006) reflection is only 0.28\degree, which confirms
134: the good epitaxial growth of the T'-LYCO films.
135:
136: \begin{figure}
137: \includegraphics[width=0.50\textwidth]{fig1.eps}
138: \caption{ XRD pattern of La$_{1.85}$Y$_ {0.15}$CuO$_{4}$ thin film
139: deposited under T$_D$=690$\celsius$. the substrate reflections of
140: SrTiO$_3$ have been removed for clarity. The insert shows the XRD
141: data of corresponding target sintered at 1050$\celsius$ (the
142: un-indexed peak marked with stars is from a small quantity of
143: un-reacted Y$_2$O$_3$).\\}
144: \end{figure}
145:
146: The XRD pattern of polycrystal LYCO target for $x=0.15$ is shown in
147: the insert of Fig. 1. Except the trace of un-reacted yttrium oxide,
148: all the peaks are well indexed assuming the La$_2$CuO$_4$-like
149: orthogonal T-214 structure. The calculated lattice constants of the
150: orthorhombic lattice are: a = 5.355\AA, b = 5.399\AA, and c =
151: 13.136\AA, which are slightly less than those of undoped
152: La$_2$CuO$_4$, consistent with the partial replacement of larger
153: La$^{3+}$ (ion radii is 1.216\AA) by smaller Y{$^{3+}$ (1.075\AA).
154: There is a great contrast between T'-films and T-bulk synthesized at
155: different temperatures, consistent with above analysis about phase
156: stability.
157:
158: The optimal growth conditions for x=0.15 LYCO films have been
159: discussed in detail elsewhere \cite{lzhaoapl}. Upon increasing Y
160: doping, the average radii of A-site ions decrease, so does the $t$
161: factor, which leads to the shift of T/T' phase boundary to higher
162: temperature, so that the corresponding optimal deposition
163: temperature T$_D$ is adjusted slightly to higher with increasing Y
164: doping. After deposition, all the films are annealed at
165: 600$\celsius$ in high vacuum (near $10^{-4}$Pa) and finally followed
166: by very slow cooling to room temperature.
167:
168: The Nd$_{2-x}$Y$_{x}$CuO$_{4}$ (NYCO) and Pr$_{2-x}$Y$_{x}$CuO$_{4}$
169: (PYCO) films were also prepared by dc sputtering for comparative
170: research, in which T'-214 structure is intrinsically stable. the
171: reduction process of films is the same as LYCO films.
172:
173: The thickness of the typical films for transport measurements is
174: about 3000\AA~ . The measurements of resistivity and Hall
175: coefficients are carried out by standard 6-lead geometry using an
176: ac resistance bridge(LR-700, Linear Research). The Ag electrodes are
177: made by Ag evaporation. Samples are mounted on the top of the
178: variable-temperature loading probe. Low temperature and high
179: magnetic field up to 14T were supplied by a Teslatron system (Oxford
180: Instruments)
181:
182:
183: \section{EXPERIMENTAL RESULTS}
184:
185: \subsection{Superconductivity and Resistivity of T'-LYCO}
186:
187: We first focus on the temperature-dependend resistivity $\rho(T)$ of
188: the T'-LYCO in zero magnetic field. Fig. 2 shows the the evolution
189: of $\rho(T)$ curves with different Y doping ($0.05\leq x \leq0.30$).
190:
191: The film of x=0.05 (plotted in the insert of Fig. 2) shows a
192: insulating behavior in the whole temperature range like heavily
193: under-doped or oxygenated cuprates as NCCO\cite{wujiang, wangchprb}.
194:
195: For all the $\rho(T)$ curves except for x=0.05, we observed metallic
196: behavior ($d\rho/dT>0$) in the high temperature range. In contrast
197: to the familiar linear temperature dependence in the hole-doped
198: cuprates like YBCO and LSCO, the nearly quadratic behavior of
199: $\rho$(T) in the normal state is observed (detailed analysis is
200: presented elsewhere \cite{lzhaoapl}). This kind of behavior is a
201: common feature in all the known electron-doped superconducting
202: cuprates near optimal and over doping, such as
203: Ln$_{2-x}$Ce$_{x}$CuO$_{4}$ (Ln=La, Pr, Nd, and Sm)\cite{Sawa,
204: Crusellas, Brinkmann, Greene}. It is usually considered as a quasi
205: 2D Landau-Fermi liquid behavior due to electron-electron
206: scattering\cite{Tsuei}.
207:
208: The metallic behaviors is maintained down to the resistivity minimum
209: temperature (T$_{min}$). Below T$_{min}$ , the resistivity shows a
210: upturn as in either hole- or electron- cuprates at low dopings. The
211: insulating upturn at low temperature are commonly interpreted in the
212: 2D weak localization \cite{fournier} or the Kondo scattering by the
213: uncompensated Cu$^{2+}$ spins\cite{sekitani}.
214:
215: \begin{figure}
216: \includegraphics[width=0.50\textwidth]{fig2.eps}
217: \caption{ Resistivity vs. temperature for T'-LYCO films with various
218: Y contents: x=0.10, 0.125, 0.15, 0.20, and 0.30. The insert shows
219: the $\rho(T)$ of the film with x=0.05. The reduction processes for
220: all these films after deposition is the same.}
221: \end{figure}
222:
223:
224: \begin{figure}
225: \includegraphics[width=0.40\textwidth]{fig3.eps}
226: \caption{ Phase diagram of T'-LYCO films as a function of Y content
227: x, determined from the resistivity data in Fig. 2. solid circles and
228: squares represent T$_c^{Zero}$ and T$_c^{Onset}$ respectively.}
229: \end{figure}
230:
231: Although the Y$^{3+}$ ions doping can't bring net charge by
232: substitution of La$^{3+}$, the resistivity of films reduces with
233: increasing Y content until to the minimum near x=0.20, suggesting
234: more electron doping by Y doping. As shown in the Fig.2,
235: superconductivity occurs when $0.10\leq x \leq 0.20$. But further
236: doping leads to the increase of $\rho$ and disappearance of
237: superconductivity. It is consistent with the evolution of Hall
238: coefficients (as discussed below). The superconducting phase diagram
239: for Y doping is plotted in Fig. 3. Superconductivity exists in
240: narrow Y doping range, i.e. $0.10\leq x \leq0.20$, which is in
241: agreement with previous results \cite{naito3}. The best T$_c^{Zero}$
242: is 13.5K near optimal doping x=0.15. T$_c^{Zero}$ does not vary much
243: with doping, differs from the usual superconducting ``dome" of other
244: cuprates in the diagram. One should note that the Y content is
245: indirectly relevant with the concentration of charge carriers. On
246: the contrary, heavy Y doping leads to larger resistivity and the
247: insulating behavior, instead of more metallic behavior induced by
248: overdoing in other cuprates.
249:
250:
251: \subsection{Resistive critical field $H^\rho_{c2}$}
252:
253: Magnetic field dependence of superconducting transitions of two
254: T'-LYCO films (x=0.125, 0.20) is shown in Fig. 4 and 5. The magnetic
255: field was applied parallel to c-axis (shown in the main figures) or
256: parallel to ab plane (shown in the upper-right inserts of Fig. 4 and
257: 5).
258:
259: \begin{figure}
260: \includegraphics[width=0.50\textwidth]{fig4.eps}
261: \caption{ Resistive transition of T'-LYCO with x=0.125 at different
262: magnetic fields. the derived temperature dependence of $H^\rho_{c2}$
263: are shown in the bottom-right insert.}
264: \end{figure}
265:
266: For $H // c$, the superconducting onset temperature (T$^{Onset}$)
267: for both films decreases and the width of the transition slightly
268: broadens at H=1T. Upon increasing the magnetic field, the transition
269: shifts to lower temperature and the $R(T)$ curves roughly parallel
270: to each other. no further broadening is found. The field of 7 Tesla
271: along c-axis is enough to kill the superconducting transition of
272: T'-LYCO film with x=0.125, but insufficient for x=0.20. The higher
273: upper critical field may be due to stronger vortex pinning by
274: in-plane oxygen deficiencies or local distortion of the lattice
275: induced by Y substitution in fluorite layers in the T'- structure.
276:
277:
278: \begin{figure}
279: \includegraphics[width=0.50\textwidth]{fig5.eps}
280: \caption{ Resistive transition of T'-LYCO with x=0.20 at different
281: magnetic fields. the derived temperature dependence of $H^\rho_{c2}$
282: are shown in the bottom-right insert.}
283: \end{figure}
284:
285: In Fig. 4 and 5, the most remarkable character is the strong
286: anisotropy in the effect of suppressing superconductivity by
287: magnetic field. When H// CuO$_2$ planes, fields up to 13.5 Tesla
288: just only slightly reduce Tc by a few Kelvins . This strong
289: anisotropy is a common feature in high-Tc cuprates due to their low
290: dimensionality and strong fluctuation effect.
291:
292: From these $R(T)$ data in different magnetic fields, one can derive
293: a characteristic field which is often referred to as the resistive
294: upper critical field $H^\rho_{c2}$. The choice of criterion to
295: determine $H^\rho_{c2}$ remains with some uncertainty and
296: arbitrariness, because of the broadened flux-flow resistivity
297: curves. But as pointed in \cite{pfournier} and \cite{naitopg}, the
298: general trend of $H^\rho_{c2}$ is very insensitive to the criterion
299: used in electron-doped NCCO and PCCO. Here we adopt the zero
300: resistivity temperature $T^{Zero}_{C}$. Other criterions (10\% and
301: 50\% of the transition), were tested, and also gave the similar
302: trend of($H^\rho_{c2}$) in the case of T'-LYCO.
303:
304: $H^\rho_{c2}$ for H//c and H$\bot$c are shown in the bottom-right
305: inserts of Fig. 4 and 5 (the difference in Tc at H=0 between two
306: kinds of field configurations are from the remounting of samples on
307: the probe). For H$\bot$c, the derived values are too close to Tc
308: because of our experimental limits. It is difficult to deduce the
309: accurate trends towards low temperature. For H//c, the anomalous
310: positive curvature of $H^\rho_{c2}$(T) were observed in both of the
311: films. It seems to deviate from the conventional theory \cite{whh}.
312:
313: The behavior of $H^\rho_{c2}(T)$ in PCCO\cite{pfournier} and
314: NCCO\cite{naitopg} have been found to have the similar
315: temperature-dependence. It has been pointed out by Ong and
316: collaborators\cite{ong, yayuwang} that the resistivity is a bad
317: diagnostic to determine intrinsic $H_{c2}(T)$. The $H^\rho_{c2}(T)$
318: in cupates determined by resistivity measurements, can be actually
319: correlated with irreversibility line defined at the onset of flux
320: flow and near the crosser from the vortex solid to liquid in $H-T$
321: phase diagram in superconducting state. Only a very rough criterion
322: corresponding to a full recovery of the normal-state resistivity for
323: x=0.15 brings the characteristic field temperature dependence close
324: to the expected description by WHH theory with minus curvature and
325: saturate at low temperature. Because of the absence of pseudogap
326: above Tc in electron-doped cuprates, the Nernst measurements are
327: expected to serve as a much more accurate probe for $H_{c2}(T)$ in
328: T'-LYCO, as have been done in NCCO\cite{yayuwang}.
329:
330: \subsection{Hall coefficients and origin of charge carriers in T'-LYCO}
331:
332: At present. T'-LYCO films behave very similarly to the Ce-doped
333: n-type cuprates. No distinct difference is observed in above
334: experiments. To clarify the signature and concentration of charge
335: carriers in T'-LYCO, we examined the Hall coefficients of T'-LYCO
336: with different Y doping, as shown in Fig. 6.
337:
338: \begin{figure}
339: \includegraphics[width=0.50\textwidth]{fig6.eps}
340: \caption{ Temperature dependence of Hall coefficients of T'-LYCO
341: films with different Y content x.}
342: \end{figure}
343:
344: All the films show negative R$_{H}$ near room temperature. It
345: confirms the intrinsic electron-doped nature in T'-214 LYCO films.
346: For non-superconducting film with x=0.30, R$_{H}$ decreases upon
347: lowering temperature and becomes divergent approaching T=0. It is
348: typical behavior observe in lightly doped n-type cuprates, as NCCO
349: \cite{wangchprb}, PCCO\cite{Brinkmann} and LCCO \cite{Sawa}. The
350: R$_{H}$ of x=0.125 resembles the behavior of x=0.30, except the
351: upturn at low temperature and its smaller absolute value, suggesting
352: more concentration of electron carriers.
353:
354: For x=0.15, R$_{H}$ decreases gradually with decreasing temperature
355: from room temperature down to about 100K. A slight upturn toward
356: zero occurs upon decreasing temperature further, which is analogical
357: to slightly underdoes NCCO or LCCO near the optimal doping in the
358: phase diagram. While for x=0.20, the upturn in low temperature is
359: observed, and the R$_{H}$ changes from negative in high temperature
360: to positive in low temperature. Such behavior is similar to that in
361: electron-doped cuprates near optimal doping, which can been
362: considered as a direct evidence of two types of charge carriers
363: \cite{wujiang}.
364:
365:
366: From above analysis on R$_{H}$, the variation of electron doping
367: n$_e$ with different Y contents is clear. n$_e|_{x=0.20}$ $>$
368: n$_e|_{x=0.15}$ $>$ n$_e|_{x=0.125}$ $>$n$_e|_{x=0.30}$, which is
369: consistent with previous resistivity data in Fig. 2.
370:
371: Because Y$^{3+}$ and La$^{3+}$ are isovalent, no net charge can be
372: provided by Y doping. the most natural assumption is that the charge
373: carrier arise from oxygen deficiency since the superconductivity can
374: only be acquired by annealing films in high vacuum. But there is
375: still no way to determine the oxygen content in thin films
376: accurately and straightforwardly at present.
377:
378: To confirm this assumption, we deposited LYCO(x=0.15) films at the
379: optimal temperature (T$_{D}$ =690$\celsius$), but annealed them in
380: lower vacuum (P$_{O_2}\sim 0.6\times 10^{-2}$Pa). These films are
381: still of pure T'-214 phase, but with less oxygen deficiency. Their
382: resistivity become much larger and insulating behavior are observed
383: at low temperature, as shown in the insert of Fig. 7. The
384: corresponding Hall coefficient R$_H$ was shown in Fig. 7, together
385: with R$_H$ of optimal T'-LYCO(x=0.15) for comparison. R$_H$ of
386: oxygenated films is negative in the whole temperature range and
387: decreases sharply with lowering temperature. The absolute value of
388: R$_H$ is much larger than that for optimal films, indicating much
389: less effective carrier density. It is consistent with its larger
390: $\rho$. Similar phenomena were observed during the oxygenation of
391: optimal NCCO \cite{wujiang} and PCCO\cite{Brinkmann}. It suggest
392: that oxygen deficiency is the source of electron doping in T'-LYCO
393: films.
394:
395: \begin{figure}
396: \includegraphics[width=0.50\textwidth]{fig7.eps}
397: \caption{ Temperature dependence of Hall coefficient R$_{H}$ for (a)
398: the optimal film annealed in high vacucum(near $10^{-4}$Pa), and (b)
399: the film annealed in lower vacuum(P$_{O_2}\sim 0.6\times
400: 10^{-2}$Pa). (a) and (b) are of the same Y doping x=0.15. The inset
401: shows the $\rho$-T data for the film (b).}
402: \end{figure}
403:
404: \subsection{Comparative research on Y-doped Pr- and Nd-based T'-phases}
405:
406: From above results, we can conclude that ``undoped" T'-LYCO is
407: actually intrinsically electron doped and its charge carriers and
408: subsequent superconductivity in T'-LYCO film are induced by oxygen
409: deficiency. T'-LYCO seems to be an electron-doped counterpart of the
410: hole-doped superconductor La$_2$CuO$_{4+\delta}$ whose
411: superconductivity is induced by excess oxygenation.
412:
413: Besides reducing \emph{t} factor to improve the stability of
414: T'-structure slightly, the role of Y in superconducting T'-LYCO is
415: still elusive. It is found that the superconductivity occurs only at
416: certain Y doping range($0.10\leq x \leq 0.20$). Does there exist
417: superconductivity in other ``undoped" T'-structure systems? To find
418: answer to these problems, comparative research on other T'-214
419: lanthanide cuprates is necessary.
420:
421: It is known that in n-type superconducting cuprates the CuO$_2$
422: network is under tensile strain because of the bond-length mismatch
423: between the CuO$_2$ sheets and Ln$_2$O$_2$ fluorite layers
424: \cite{ytzhu, ytzhu2}. The tensile CuO$_2$ is apt to receive electron
425: doping and be reduced below the formal oxidation state
426: (CuO$_2$)$^{2-}$. Doping electrons into CuO$_2$ planes will increase
427: the Cu-O bond length and release the tension. Larger cations can
428: introduce more tensile strain. Naturally more stretch will increase
429: the capability to receive electron doping by CuO$_2$ sheets.
430: Countrawise, if the Cu-O bond length in the T' structures is too
431: small, as in Gd214 system(a$\sim $3.89\AA), the planes do not
432: readily accept election doping, no superconductivity is found with
433: Ce substitution \cite{smith}.
434:
435: In Ref \cite{naito2}, the series of R$_2$CuO$_4$ films with
436: increasing radii of the R$^{3+}$ ions (R=Tb, Gd, Eu, Sm, Nd, Pr, La)
437: are fabricated by MBE. Their resistivity decreases gradually with
438: increasing ionic radius (shown in Fig. 3 in \cite{naito2}), the
439: trend is consistent with above analysis.
440:
441: The Cu ions are reduced by in-plane deficiency and the stretch can
442: be partly released. Since the bond energies between Cu-O is less
443: than R-O \cite{bondingenergy}, It suggests more likely that main
444: oxygen deficit is mainly introduced at O(1) sites while O(2) sites
445: remain intact in the reduction. The in-plane oxygen defect created
446: by reduction have been evidenced in infrared transmission, Raman
447: and ultrasonic studies in NCCO\cite{Raman, ultrasonic}and
448: PCCO\cite{infrared} systems .
449:
450: In our work, we have prepared Nd$_{2-x}$Y$_{x}$CuO$_{4}$(NYCO) and
451: Pr$_{2-x}$Y$_{x}$CuO$_{4}$(PYCO) films with Y content x=0.15. The
452: temperature dependences of resistivity are shown in Fig. 8, in which
453: T'-214 structure is intrinsically stable even for x=0 and the role
454: of Y to stablize T'-structure is needlessly. It is convenient to
455: study the size effect at A-site cation in T'-phase. Furthermore, to
456: study the effect of cation disorder, the undoped
457: Pr$_{2}$CuO$_{4}$(PCO) films were also prepared for comparison. All
458: the films are of pure c-axis oriented T'-phase, and annealed in the
459: high vacuum.
460:
461: \begin{figure}
462: \includegraphics[width=0.50\textwidth]{fig8.eps}
463: \caption{Temperature dependence of resistivity of
464: Pr$_{1.85}$Y$_{0.15}$CuO$_{4}$(PYCO) and
465: Nd$_{1.85}$Y$_{0.15}$CuO$_{4}$(NYCO), Comparison of resistivity of
466: undoped Pr$_{2}$CuO$_{4}$(PCO) and PYCO is shown in the insert.}
467: \end{figure}
468:
469: All films show insulating-like behavior and no superconductivity is
470: found, the resistivity of PYCO films is smaller than of NYCO, which
471: is consistent with above analysis because of the larger average ion
472: radii at A-site in PYCO. In T'-phase, the larger cations at A-site,
473: the stronger driving force for the acceptance of electrons into
474: CuO$_2$ planes and thereafter the more electron doping by oxygen
475: vacancies are created during the reduction, as has been observed in
476: many early experiments on the preparation of T'-214 cuprate ceramics
477: \cite{kato, tao, ytzhu2}.
478:
479: To find out why PYCO and NYCO are not superconducting in contrast to
480: LYCO, we measured their Hall coefficients from 280K down to 20K, as
481: shown in Fig. 9. The R$_H$ of superconducting LYCO films(x=0.15) is
482: also plotted for comparison.
483:
484: \begin{figure}
485: \includegraphics[width=0.50\textwidth]{fig9.eps}
486: \caption{Temperature dependence of Hall coefficients R$_{H}$ of
487: LYCO, PYCO and NYCO fims with the same Y content, x=0.15.}
488: \end{figure}
489:
490: The Hall coefficients R$_{H}$ of PYCO and NYCO films are both
491: negative in the whole temperature range and decrease gradually upon
492: lowering temperature with a slight upturn below 50K. The absolute
493: value of R$_H$ is much larger than that for LYCO films, indicating
494: much less of effective carrier density, consistent with resistivity
495: data. As pointed out in \cite{ytzhu}, the critical amount of
496: electron doping necessary to induce a transition from
497: antiferromagnetic insulator to superconductor is nearly fixed in
498: cuprates. The low carrier density should be the main reason for the
499: absence of superconductivity in NYCO and PYCO.
500:
501: For comparison, the resistivity of PYCO and undoped PCO is shown in
502: the insert of Fig. 8. The partial substitution of Y for Pr greatly
503: reduces the resistivity, suggesting more oxygen vacancies induced by
504: Y doping. Although at present there is no the quantitative theoretic
505: calculation and direct experimental evidence about the oxygen
506: deficiency, we think it a reasonable microscopic scenarios that the
507: local distorted fluorite structure weaken the local binding of Ln-O
508: and make the further loss of oxygen in fluorite layer. Therefore, Y
509: doping leads to the additional oxygen deficiency, and the increase
510: of the effective charge carriers, consequently.
511:
512: \section{DISCUSSION}
513:
514: We have shown the T'-LYCO is in nature electron-doped, like other
515: Ce-doped cuprate superconductors as NCCO and PCCO. The electron
516: doping is considered to origin from oxygen deficiency.
517:
518: Although the strain effect by the substrates may help to stabilize
519: the T' structure to some degree, this kind of effect can be
520: negligible because the films we used in measurements are much
521: thicker than 1000\AA. In addition, two kinds of substrates,
522: SrTiO$_3$(STO) and
523: [(LaAlO$_3$)$_{0.3}$(Sr$_2$AlTaO$_6$)$_{0.7}$](LSAT) are used and no
524: apparent difference is found in the transport properties in our
525: experiments. Since we are working with thin films, the composition
526: inhomogeneity have also been ruled out.
527:
528:
529: The apical or interstitial oxygen problem, which is thought
530: deleterious and even deadly to superconductivity, are also excluded
531: in our discussion, because these films are annealed sufficiently in
532: the vacuum where the oxygen partial pressure is far below the level
533: in reducing single crystals or ceramics in flowing inert gas. The
534: removal of apical oxygen is complete and actual stoichiometric ratio
535: for oxygen should be less than 4.0, according to previous
536: determination of the oxygen content in bulk after sufficient
537: reduction\cite{ytzhu2}
538:
539: Based on our present results and data from Tsukada et al
540: \cite{naito2, naito3}, we present a most likely and reasonable
541: mechanism of the superconductive in T'-(La, RE)$_{2}$CuO$_{4}$. The
542: electron doping origins from two different microscopic processes of
543: oxygen deficiency, and each is indispensable to trigger
544: superconductivity.
545:
546: \textbf{Firstly}, the large average radius of cations at A-site
547: should make the CuO$_2$ sheets in enough tensile, which become the
548: internal driving force for the acceptance of electrons into CuO$_2$
549: sheets to release the stretch strain. Driven by the internal stress,
550: oxygen deficiency is produced in reduce process, mainly at the O(1)
551: position in the CuO$_2$ planes because of the lower binding energy
552: at the O(1) site than that at the O(2) site in La$_2$O$_2$ fluorite
553: layers. The larger cations at A-site will introduce more oxygen
554: vacancies which are responsible for more charge carriers and the
555: suppression of the long-range antiferromagnetism.
556:
557: People have found in previous experiments that there are more oxygen
558: vacancies produced as the average ion radius of A-site increases in
559: T'-214 systems during a constant reduction process \cite{ytzhu,
560: tao}. It also consistent with the results that the resistivity
561: decreases gradually with increasing lanthanide ionic radius in
562: annealed undoped Ln$_2$CuO$_4$ (Ln=Tb, Gd, Sm, Eu,La, Pr, Nd) as
563: shown in Fig. 3 in Ref.\cite{naito2}.
564:
565: Furthermore the difference from ion size effect also leads to
566: different additional doping by oxygen deficiency, and modify the
567: superconducting phase diagram as a function of Ce content x in known
568: Ce-doped n-type cuprates films (as shown in Fig. 1 in Ref
569: \cite{naito2}). The shift of the highest Tc towards lower Ce
570: concentrations is also observed as increasing the substitution of La
571: to PCCO\cite{koike} or NCCO\cite{tao,kato} bulk systems. This
572: evolution with average size of A-site cations is consistent with
573: our above analysis.
574:
575: Nevertheless, in undoped T'-214 cuprates the charge carriers induced
576: only by in-plane vacancies are not enough to drive the CuO$_2$ into
577: superconducting states. The additional oxygen deficiency in fluorite
578: layers is needed. But in undoped Ln$_2$CuO$_4$ , the binding energy
579: in Ln-O is always larger than that in in-plane Cu-O, so that O(2)
580: can be too hard to be removed before the excess in-plane oxygen
581: vacancies destabilize the T'-structure and lead to chemical
582: decomposition of samples during the further reduction.
583:
584: \textbf{Secondly}, the partial substitution of smaller Y$^{3+}$ ions
585: for La$^{3+}$ leads to local strong distortion in the fluorite
586: Ln$_2$O$_2$ layers. It will be apt to further produce oxygen
587: vacancies during the reduction. The strong distortion in fluorite
588: layers by Y substitution is thought to decrease the local binding
589: energy to be overcome, and then creates additional oxygen vacancies
590: in fluorite layers in reduction. It has been demonstrated in the
591: comparison of doped and undoped PCO films.
592:
593: Therefore the total oxygen deficiency, created in two ways proposed
594: above, will be adequate to induce superconductivity in T'-LRCO. As
595: proposed by Zhu and Manthiram \cite{ytzhu}, the transition from
596: antiferromagnetic insulator to superconductor occurs at the critical
597: amount of electron concentration, $n_c$, which is nearly fixed in
598: n-type cuprates. Although the Madelung energy or charge transfer gap
599: $\Delta$ decreases upon increasing ion radius at A-site, its effect
600: on $n_c$ is negligible.
601:
602: Increasing Y doping will introduce more charge carriers. However,
603: excess Y doping induces the decrease of the average size of Ln at
604: A-site, which will hinder the oxygen depletion at O(1)-site.
605: Furthermore, the disorder induced by Y doping localizes the charge
606: carriers. Therefore excess Y will kill the superconductivity in
607: extreme case. The increasing resistivity and insulating upturn in
608: the T'-LYCO films with high doping(x=0.30), are observed in our
609: experiments.
610:
611: Our analysis above also naturally explains other results of Tsukada
612: and cooperators \cite{naito2}. Superconductivity has been found in
613: many ``non-doped" T'-(La, RE)$_{2}$CuO$_{4}$ cuprates as RE=Sm, Eu,
614: Gd, Tb, Lu, and Y, but neither Pr nor Nd. Pr and Nd are more
615: adjacent to La in lanthanide series, the difference of ion radii is
616: smaller than other lanthanide. It suggests that the local distortion
617: caused by Pr or Nd substitution is not as remarkable as by Y etc.
618: The fluorite structure remains almost intact during the usual
619: annealing process. Oxygen vacancies mainly exist in CuO$_2$ network,
620: and they can't afford necessary doping to trigger superconductivity
621: unless there is extra charge induce by aliovalent Ce substitution in
622: charge reservoir layers as in the (La,Nd,Ce)$_2$CuO$_4$ systems
623: \cite{tao}.
624:
625:
626: Our present results provide a more comprehensive understanding of
627: this new class of cuprates as well as the microscopic process of
628: oxygen reduction which is indispensable to acquire superconductivity
629: in electron-doped cuprates. Because of the metastability of La-based
630: T'-214 structure, people can only obtain superconducting T'-LRCO in
631: the form of thin films so far. Therefore, it is difficult and even
632: impossible to determine directly the oxygen occupancy at O(1) and
633: O(2) sites in films within present technologies such as conventional
634: X-ray or neutron diffraction. Bulk T'-La$_2$CuO$_{4}$ \cite{chou}and
635: even supercooling La$_{2-x}$Ce$_{x}$CuO$_{4}$ \cite{yamada} have
636: been synthesized in special precursor route. It indicates that the
637: bulk synthesis of superconducting T'-LYCO seems to be possible. Such
638: attempts is underway. Due to low sintering temperature, it will be
639: very difficult to achieved the good crystallization and homogeneity
640: in bulk. For the second processes of oxygen reduction we prepose
641: above, other experimental methods such as Raman, infrared
642: transmission and Extended x-ray absorption fine structure (EXAFS)
643: x-ray absorption near edge structure (XANES) etc is expected to
644: detect the distortion of the local atomic structure and thereafter
645: its effect on oxygen depletion. Corresponding studies are expected
646: in future study and are beyond our present research.
647:
648:
649: \section{CONCLUSION}
650:
651: In summary, we systematically investigated the transport properties
652: of newly discovered nominal ``non-doped" La$_{2-x}$Y$_{x}$CuO$_{4}$
653: ($0.05\leq x \leq 0.30$). Detailed studies on the resistivity and
654: Hall coefficient of T'-LYCO films, strong anisotropic
655: H$^{\rho}_{c2}$(T) and its positive curvature, suggest the
656: similarity to the known cation-doped election-type cuprates
657: superconductors as NCCO or PCCO. This kind of so-called ``non-doped"
658: superconductor is intrinsically electron-doped and the effect of
659: oxygen content indicates that the charge carriers mainly arise from
660: oxygen deficiency obtained during the reduction process. The
661: fundamental picture \cite{anderson} depicted since the discover of
662: HTSC still holds at present. Further studies on the effect of the
663: in-plane partial substitution of magnetic or non-magnetic impurities
664: are in process.
665:
666: Comparative experiments on non-superconducting Y-doped Pr- and Nd-
667: based T'-214 films exhibit the the crucial role of cation radius
668: played in the oxygen depletion during the reduction. The cooperation
669: of large-sized La$^{3+}$ ions with smaller Y$^{3+}$ enable the
670: sufficient oxygen deficiency in T'-structure to induce
671: superconductivity in T'-LYCO. We propose a reasonable scenario about
672: the microscopic reduction process. The large La$^{3+}$ ions at
673: A-site keep the CuO$_2$ in adequate tensile strain which enhances
674: in-plane oxygen deficiency during reduction. And the appropriate
675: local distortion induced by the substitution of small Y$^{3+}$ in
676: fluorite layers lowers the binding energy to be overcome in the
677: oxygen depletion in fluorite layers. They are two indispensable keys
678: to induce enough electron doping to trigger superconductivity in the
679: nominally ``undoped" cuprates.
680:
681: The work is supported by the Nature Science Foundation of China and
682: by the Ministry of Science and Technology of China (973 project No:
683: 2006CB601001 and 2006CB0L1205), and by the Knowledge Innovation
684: Project of Chinese Academy of Sciences.
685:
686: \begin{thebibliography}{}
687:
688: \bibitem{bednorz}
689: J. G. Bednorz and K. A. M\"{u}ller, Z. Phys. B {\bf 64}, 189 (1986)
690: \bibitem{orenstein}
691: J. Orenstein and A. J. Millis, Science {\bf 288}, 468 (2000)
692: \bibitem{naito2}
693: A. Tsukada, Y. Krockenberger, M. Noda, H. Yamamoto, D. Manske, L.
694: Alff, and M. Naito, Solid State Commun. {\bf 133} 427 (2005)
695: \bibitem{naito3}
696: M. Noda, A. Tsukad, H. Yamamoto, and M. Naito, Physica C {\bf
697: 426-431} 220 (2005)
698: \bibitem{naito4}
699: A. Tsukada, M. Noda, H. Yamamoto, and M. Naito, Physica C {\bf
700: 426-431} 459 (2005)
701: \bibitem{Idemoto}
702: Y. Idemoto, T. Miyabe, and N. Koura, J. Ceram. Soc. Jpn. {\bf 114},
703: 359 (2006)
704:
705: \bibitem{greenelyco}
706: W. Yu, B. Liang, P. Li, S. Fujino, T. Murakami, I. Takeuchi, and R.
707: L. Greene, cond-mat/0610889
708:
709: \bibitem{lzhaoapl}
710: L. Zhao, G. Wu, R. H. Liu and X. H. Chen, accepted by Appl. Phys.
711: Lett.
712:
713: \bibitem{goodenough1}
714: A. Manthiram and J. B. Goodenough, J. Solid State Chem. {\bf 87},
715: 402 ( 1990)
716:
717: \bibitem{goodenough2}
718: A. Manthiram and J. B. Goodenough, J. Solid State Chem. {\bf 92},
719: 231 (1991)
720:
721: \bibitem{Tsukada}
722: A. Tsukada, T. Greibe and M. Naito, Phys. Rev. B {\bf 66}, 184515
723: (2002)
724:
725: \bibitem{wujiang}
726: Wu Jiang et al, Phys. Rev. Lett. {\bf 73}, 1291 (1994)
727:
728: \bibitem{wangchprb}
729: C. H. Wang, G. Y. Wang, T. Wu, Z. Feng, X. G. Luo and X. H. Chen,
730: Phys. Rev. B, {\bf 72}, 132506 (2005)
731:
732: \bibitem{Sawa}
733: A. Sawa, M. Kawasaki, H. Takagi and Y. Tokura, Phys. Rev. B {\bf
734: 66}, 014531 (2002)
735:
736: \bibitem{Crusellas}
737: M. A. Crusellas, J. Fontcuberta, S. Pinol, T. Grenet and J. Beille,
738: Physica C {\bf180}, 313 (1991)
739:
740: \bibitem{Brinkmann}
741: M. Brinkmann, H. Bach, and K. Westerholt, Phys. Rev. B {\bf 54},
742: 6680 (1996)
743:
744: \bibitem{Greene}
745: S.J. Hagen, J.L. Peng, and R.L. Greene, et al, Phys. Rev. B {\bf
746: 43}, 13606(1991)
747:
748: \bibitem{Tsuei}
749: C.C. Tsuei, A. Gupta and G. Koren, Physica C {\bf 161}, 415 (1989)
750:
751: \bibitem{fournier}
752: P. Fournier, P. Mohanty, E. Maiser, S. Darzens, T. Venkatesan, C.J.
753: Lobb, G. Czjzek, R.A. Webb, and R.L. Greene, Phys. Rev. Lett. {\bf
754: 81}, 4720 (1998)
755:
756: \bibitem{sekitani}
757: T. Sekitani, M. Naito, and N. Miura, Phys. Rev. B {\bf 67}, 174503
758: (2003)
759:
760: \bibitem{pfournier}
761: P. Fournier, and R.L. Greene, Phys. Rev. B {\bf 68}, 094507 (2003)
762:
763: \bibitem{naitopg}
764: S. Kleefisch, B. Welter, A. Marx, L. Alff, R. Gross and M. Naito,
765: Phys. Rev. B {\bf 63}, 100507(2001)
766:
767: \bibitem{whh}
768: N.R. Werthamer, E. Helfand, and O.C. Hohenberg, Phys. Rev. {\bf
769: 147}, 295 (1966)
770:
771: \bibitem{ong}
772: Y.Y. Wang, L. Li,and N. P. Ong, Phys. Rev. B {\bf 73}, 024510 (2006)
773:
774: \bibitem{yayuwang}
775: Y.Y. Wang, "Nernst effect in high temperature supercondcutors", PhD
776: thesis (2004)
777:
778: \bibitem{ytzhu}
779: Y. T. Zhu and A. Manthiram, Phys. Rev. B {\bf 49}, 6293 (1994)
780: \bibitem{ytzhu2}
781: Y. T. Zhu and A. Manthiram, Physica C {\bf 224}, 256(1994)
782:
783:
784: \bibitem{smith}
785: M.G. Smith, A. Manthiram, J. Zhou, J.B. Goodenough, J.T. Markert,
786: Nature, {\bf 351}, 549 - 551 (1991)
787:
788: \bibitem{bondingenergy}
789: J.A. Kerr, in: R.C. Weast (Ed.), CRC Handbook of Chemistry and
790: Physics, 66th ed., CRC Press Inc., Florida, 1985, p. F174.
791:
792: \bibitem{Raman}
793: P. Richard, G. Riou, I. Hetel, S. Jandl, M. Poirier, and P.
794: Fournier, Phys. Rev. B {\bf 70}, 064513 (2004)
795: \bibitem{ultrasonic}
796: P. Richard, M. Poirier, S. Jandl and P. Fournier, Phys. Rev. B {\bf
797: 72}, 184514 (2005)
798:
799: \bibitem{infrared}
800: G. Riou, P. Richard, S. Jandl, M. Poirier, and P. Fournier,
801: Phys. Rev. B {\bf 69}, 024511 (2004)
802:
803: \bibitem{koike}
804: Y. Koike, A. Kakimoto, M. Mochida, H. Sato, T. Noji, M. Kato and Y.
805: Saito, Jpn. J. Appl. Phys, {\bf 31}, 2721(1992)
806:
807: \bibitem{tao}
808: Y.K. Tao, M. Bonvalot, Y.Y. Sun, R.L. Meng, P.H. Hor, and C.W. Chu,
809: Physica C {\bf 165}, 13 (1990)
810:
811: \bibitem{kato}
812: E. Takayama-Muromachia, Y. Uchidaa and K. Kato Physica C {\bf 165},
813: 147-151 (1990)
814:
815: \bibitem{chou}
816: F. C. Chou, J. H. Cho, L. L. Miller, and D. C. Johnston, Phys. Rev.
817: B {\bf 42}, 6172 (1990)
818:
819: \bibitem{yamada}
820: T. Yamada, K. Kinoshita, and H. Shibata, Jpn. J. Appl. Phys., Part2
821: {\bf 33}, L168 (1994)
822:
823: \bibitem{anderson}
824: P. W. Anderson, Science {\bf 235}, 1196 (1987)
825:
826: \end{thebibliography}
827:
828: \clearpage
829:
830: \end{document}
831: