1: \documentstyle[twocolumn,aps,epsf]{revtex}
2: %\documentstyle[manuscript,aps]{revtex}
3: \newcommand{\kvec}{\mbox{{\scriptsize {\bf k}}}}
4: \newcommand{\pvec}{\mbox{{\scriptsize {\bf p}}}}
5: \newcommand{\qvec}{\mbox{{\scriptsize {\bf q}}}}
6: \newcommand{\Qvec}{\mbox{{\scriptsize {\bf Q}}}}
7: \newcommand{\zvec}{\mbox{{\scriptsize {\bf 0}}}}
8: \newcommand{\qv}{\mbox{{\bf q}}}
9: \newcommand{\bv}{\mbox{{\bf b}}}
10: \newcommand{\Qv}{\mbox{{\bf Q}}}
11: \newcommand{\kv}{\mbox{{\bf k}}}
12: \newcommand{\pv}{\mbox{{\bf p}}}
13: \newcommand{\Av}{\mbox{{\bf A}}}
14: \newcommand{\Hv}{\mbox{{\bf H}}}
15: \newcommand{\Rv}{\mbox{{\bf R}}}
16: \newcommand{\sv}{\mbox{{\bf s}}}
17: \newcommand{\lv}{\mbox{{\bf l}}}
18: \newcommand{\jv}{\mbox{{\bf j}}}
19: \newcommand{\ev}{\mbox{{\bf e}}}
20: \begin{document}
21: %\baselineskip=10mm
22: \draft
23: \title{Upper critical field for underdoped high--$T_c$ superconductors. Pseudogap and
24: stripe--phase.}
25: \author{Marcin Mierzejewski and Maciej M. Ma\'ska\cite{email}}
26: \address{
27: Department of Theoretical Physics, Institute of Physics,
28: University of Silesia, 40-007 Katowice,
29: Poland}
30: \maketitle
31: \begin{abstract}
32:
33: We investigate the upper critical field in a
34: stripe--phase and in the presence of a phenomenological pseudogap.
35: Our results indicate that the formation of stripes affects the Landau
36: orbits and results in an enhancement of $H_{c2}$. On the other hand,
37: phenomenologically introduced pseudogap leads to a reduction of the
38: upper critical field. This effect is of particular importance when
39: the magnitude of the gap is of the order of the superconducting
40: transition temperature. We have found that a suppression of the
41: upper critical field takes place also for the gap that originates
42: from the charge--density waves.
43:
44:
45: \end{abstract}
46: \pacs{74.25.Ha,74.60.Ec,71.70Di}
47: \section{Introduction}
48:
49: The high--temperature superconductors (HTSC) exhibit qualitative
50: differences with respect to the classical
51: superconducting systems. Normal state properties
52: of underdoped superconductors and the upward curvature of the
53: upper critical field ($H_{c2}$) belong to one of the most spectacular
54: examples. The presence of a normal--state pseudogap has been
55: confirmed with the help of different experimental techniques like:
56: angle--resolved photoemission \cite{arpes1,arpes2,arpes3},
57: intrinsic tunneling spectroscopy \cite{krasnov1,krasnov2},
58: NMR \cite{nmr1,nmr2},
59: infrared \cite{infra} and transport \cite{transport} measurements.
60: Despite a wide spectrum of experimental data the underlying
61: microscopic mechanism if far from being understood.
62: A tempting hypothesis that the pseudogap is a precursor of the
63: superconducting gap has not definitively been confirmed.
64: In particular, the neutron scattering experiments
65: \cite{muller} reveal qualitative differences between the
66: isotope effects observed for the superconductivity
67: and the pseudogap. Moreover, results obtained with the help of
68: intrinsic tunneling spectroscopy \cite{krasnov1,krasnov2}
69: speak against the superconducting origin of the pseudogap.
70: The coexistence of superconductivity and the charge--density waves
71: \cite{eremin,seibold,markiewicz,kramer}
72: can be considered as a possible scenario, that accounts
73: for differences between superconductivity and the normal--state
74: gap. Inhomogeneous distribution of holes which enter the
75: copper-oxygen planes in the doping process can give rise to the
76: formation of stripes. There is a convincing experimental
77: and theoretical argumentation speaking in favor of the
78: stripe--phase \cite{str1,str2,str3,str4,str5,str6,str7}
79: with an intimate connection between
80: superconducting and stripe correlations.
81: This phase consists of antiferromagnetic domains
82: which are separated by the hole--rich domain walls.
83:
84: Differences between the high--temperature
85: superconductors and classical systems show also up in the
86: magnetic properties. The high--$T_c$ compounds are
87: characterized by large values of the upper critical field
88: and its unusual temperature dependence. The resistivity
89: measurements clearly indicate an upward
90: curvature of the upper critical field with no evidence
91: of saturation even at genuinely low temperatures
92: \cite{osofsky,mackenzie}. These results remain
93: in disagreement with the conventional, microscopic
94: approach \cite{gorkov}. This discrepancy can be
95: explained as a result of the Josephson tunneling between
96: superconducting clusters \cite{gesh,wen} produced by a
97: macroscopic phase separation.
98:
99: Due to the complexity of the Gor'kov equations one usually assumes
100: that the normal--state properties of the system under consideration
101: can properly be described by three-- \cite{hw} or two--dimensional
102: \cite{kresin} electron gas. Recently, we have proposed an approach
103: that enables calculation of the upper critical field for
104: a two--dimensional lattice gas \cite{my1,my2}. This method
105: allows one to derive $H_{c2}$ in a similar way as one
106: calculates the
107: critical temperature in the standard
108: BCS formalism. Therefore, any extension of the analysis of
109: the upper critical field is rather straightforward.
110: In the present paper we calculate the upper critical field
111: in a system that exhibits some important properties
112: of hole--doped cuprates: stripe--phase
113: and the presence of the pseudogap.
114: In the latter case we discuss
115: the coexistence of superconductivity and charge--density--wave
116: as well as a phenomenological pseudogap. We show that
117: the anomalous properties of the high--temperature
118: superconductors are reflected in
119: the upper critical field.
120:
121: \section{$H_{c2}$ in the presence of charge--density--waves}
122:
123: We consider a two--dimensional square lattice immersed
124: in a perpendicular uniform magnetic field of magnitude $H_z$.
125: We assume
126: the nearest--neighbor pairing interaction, $H_V$, that is responsible
127: for anisotropic superconductivity and the interaction term,
128: $H_{\rm CDW}$,
129: which leads to the charge--density--waves. The relevant
130: Hamiltonian reads \cite{my1,my2}
131: \begin{equation}
132: H=H_0+H_V+H_{\rm CDW},
133: \end{equation}
134: where
135: \begin{eqnarray}
136: H_0&=&\sum_{\langle ij\rangle,\sigma} t_{ij}\left(\Av \right)
137: c^{\dagger}_{i\sigma}c_{j\sigma} \nonumber \\
138: &&
139: +g\mu_{B}H_{z}\sum_{i}
140: \left(c^{\dagger}_{i\uparrow}c_{i\uparrow}
141: -c^{\dagger}_{i\downarrow}c_{i\downarrow} \right).
142: \end{eqnarray}
143: Here, $c^{\left(\dagger\right)}_{i\sigma}$ annihilates
144: (creates) an electron with spin $\sigma$ at the lattice
145: site $i$,
146: $g$ is the gyromagnetic ratio and $\mu_B$ is the Bohr
147: magneton.
148: $t_{ij}\left(\Av \right)$ is the nearest--neighbor
149: hopping integral that in the presence of the magnetic field
150: acquires the Peierls phase--factor \cite{peir,my1}
151: \begin{equation}
152: t_{ij}\left(\Av \right)= t\,
153: \exp\left(\frac{ie}{\hbar c} \int^{\Rv_{i}}_{\Rv_{j}}
154: \Av\cdot d\lv\right).
155: \end{equation}
156: In the mean-field approximation the pairing
157: interaction and the CDW coupling take on the form
158: \begin{eqnarray}
159: H_{V}&=&-\:
160: V\sum_{\langle ij\rangle}\left(c^{\dagger}_{i\uparrow}c^{\dagger}_{j\downarrow}
161: \Delta_{ij}
162: +c_{i\downarrow}c_{j\uparrow} \Delta^{*}_{ij} \right), \\
163: H_{\rm CDW}&=&-\delta_{\rm CDW} \sum_{j \sigma}
164: e^{i \Qv \cdot \Rv_{j}}
165: c^{\dagger}_{j \sigma} c_{j \sigma},
166: \end{eqnarray}
167: where $\Delta_{ij}=\langle c_{i\downarrow}c_{j\uparrow}
168: -c_{i\uparrow}c_{j\downarrow}\rangle$ is the superconducting singlet order
169: parameter and $\delta_{\rm CDW}$ represents the magnitude of the
170: CDW gap.
171: The complexity of calculations strongly depends on the
172: CDW modulation vector $\Qv$.
173: For the sake of simplicity we consider a commensurate
174: charge--density wave with
175: $\Qv=\left(\pi,\pi\right)$.
176: This choice of the modulation vector results in the
177: gap in the density of states that opens in the middle of the band
178: (in our case at the Fermi level) independently on the magnitude
179: of the external magnetic field.
180: Since the pseudogap hardly depends on the magnitude of the
181: magnetic field \cite{krasnov2}, $\delta_{\rm CDW}$ will be taken as
182: a model parameter.
183:
184: In order
185: to calculate the upper critical field we make use of the unitary
186: transformation that diagonalizes the kinetic part of the
187: Hamiltonian \cite{my1,my2}. In the case of the Landau gauge,
188: $\Av=H_z\left(0,x,0\right)$,
189: this transformation is determined by a plane--wave function
190: in $y$ direction
191: and an eigenfunction of the Harper equation
192: \cite{harper}:
193: \begin{eqnarray}
194: && g\left(\bar{k},p,m+1\right)+
195: 2\cos\left(h m -pa\right)g\left(\bar{k},p,m\right)
196: \nonumber \\
197: &&+ g\left(\bar{k},p,m-1\right)=
198: t^{-1}E\left(\bar{k},p\right)
199: g\left(\bar{k},p,m\right).
200: \end{eqnarray}
201: Here, $m$ is an integer number that enumerates
202: the lattice sites in $x$ direction,
203: whereas $h$ is the reduced magnetic field, $h= 2 \pi \Phi/\Phi_0$,
204: that is expressed by the ratio of the flux $\Phi$
205: through the lattice cell and the flux quantum $\Phi_0$.
206: $p$ is the wave--vector in $y$ direction and $\bar{k}$
207: is an additional quantum number, that in the absence
208: of the magnetic field is the wave--vector in
209: $x$ direction.
210: In the new basis the normal--state Hamiltonian takes on the form:
211: \begin{eqnarray}
212: H_{0} &=& \sum_{\bar{k},p,\sigma} E_{\bar{k}p\sigma}
213: a^{\dagger}_{\bar{k}p\sigma}a_{\bar{k}p\sigma}, \\
214: H_{\rm CDW}&=& -\delta_{\rm CDW}\!\! \sum_{\bar{k},\bar{l},p,m,\sigma}
215: g^*\left(\bar{k},p+\pi,m\right)g\left(\bar{l},p,m\right) \nonumber \\
216: && \times\: e^{i \pi m }
217: a^{\dagger}_{\bar{k}p+\pi,\sigma}a_{\bar{l}p\sigma},
218: \end{eqnarray}
219: where
220: \begin{equation}
221: E_{\bar{k}p\sigma}=E\left(\bar{k},p\right)+\sigma
222: g\mu_{B}H_{z}.
223: \end{equation}
224:
225: One can prove that if $E$ represents the eigenvalue of
226: the Harper equation obtained for the wave--vector $p$ then
227: $-E$ is one of the eigenvalues corresponding to $p+\pi$.
228: It can also be shown that
229: $ \tilde g\left(\bar{k},p,m\right)=
230: g\left(\bar{k},p+\pi,m\right)\exp\left(i \pi m \right)$
231: represents an eigenfunction of the Harper equation
232: calculated for momentum $p$
233: with the eigenvalue $-E\left(\bar{k},p+\pi\right)$.
234: With the help of these relations one can obtain analytically
235: the energy spectrum of the Hamiltonian in the normal state,
236: $H_0+H_{\rm CDW}$, provided that the eigenvalues of the Harper
237: equation are known. In particular, one can calculate the
238: anomalous Green functions which are related to the
239: superconducting order parameter:
240: \begin{eqnarray}
241: && \langle\langle a_{\bar{l}p\uparrow}\mid a_{\bar{k},-p\downarrow}\rangle\rangle =
242: \nonumber \\
243: && -\sum_{m}\left[ X^{*}_{\bar{l}\bar{k}p}\left(m\right)
244: \Delta^{x}\left(m\right)+Y^{*}_{\bar{l}\bar{k}p}\left(m\right)
245: \Delta^{y}\left(m\right)\right] K_{\bar{l}\bar{k}p}(\omega), \nonumber \\
246: \end{eqnarray}
247: where $X$ and $Y$ are determined by the solution of the Harper equation
248: \begin{eqnarray}
249: X_{\bar{l}\bar{k}p}\left(m\right) &=&
250: g\left(\bar{l},p,m\right)g\left(\bar{k},-p,m+1\right) \nonumber \\
251: && + g\left(\bar{l},p,m+1\right)g\left(\bar{k},-p,m\right), \\
252: Y_{\bar{l}\bar{k}p}\left(m\right) &=& 2 \cos(p)
253: g\left(\bar{l},p,m\right) g\left(\bar{k},-p,m\right).
254: \end{eqnarray}
255: Due to the plane--wave behavior in $y$ direction the superconducting
256: order parameter depends explicitly only on the position in $x$ direction
257: (see Ref. \cite{my2} for the details)
258: \begin{eqnarray}
259: \Delta^{x}\left(m\right) &=&
260: \frac{V}{N} \sum_{\bar{l}\bar{k} p}
261: X_{\bar{l}\bar{k}p}\left(m\right) \langle
262: a_{\bar{k},-p\downarrow}a_{\bar{l}p\uparrow}
263: \rangle,
264: \\
265: \Delta^{y}\left(m\right) &=&
266: \frac{V}{N} \sum_{\bar{l}\bar{k} p}
267: Y_{\bar{l}\bar{k}p}\left(m\right) \langle
268: a_{\bar{k},-p\downarrow}a_{\bar{l}p\uparrow}
269: \rangle.
270: \end{eqnarray}
271: $K_{\bar{l}\bar{k}p}(\omega)$, when
272: integrated over $\omega$ with the Fermi function, gives the Cooper pair
273: susceptibility
274: \begin{equation}
275: K_{\bar{l}\bar{k}p}(\omega)=
276: \frac{
277: \left(\omega+E_{\bar{l}p\uparrow}\right)\left(\omega-E_{\bar{k},-p\downarrow}\right)
278: +\delta_{\rm CDW}^2
279: }{
280: \left(\omega^2-E^2_{\bar{l}p\uparrow}-\delta_{\rm CDW}^2\right)
281: \left(\omega^2-E^2_{\bar{k},-p\downarrow}-\delta_{\rm CDW}^2\right)
282: }.
283: \end{equation}
284: One can see that the impact of the charge--density waves on superconductivity
285: is brought about only by the modification of this quantity. Equations (10),
286: (13) and (14) allow one to calculate the upper critical field. It is determined
287: as the highest magnitude of the magnetic field for which there exists
288: a non--zero solution for $\Delta^{x}(m)$ and $\Delta^{y}(m)$.
289:
290: \begin{figure}
291: \epsfxsize=9cm
292: \centerline{\epsffile{fig1.eps}}
293: \caption{Temperature dependence of the upper critical field for different
294: magnitudes of the CDW gap.}
295: \end{figure}
296:
297: Figure 1. shows the reduced upper critical field,
298: $h_{c2}$, as a function of temperature calculated for different magnitudes
299: of the CDW order parameter, $\delta_{\rm CDW}$. We have adjusted the magnitude
300: of the pairing potential $V$ that gives the same superconducting
301: transition temperature $kT_{c}=0.02t$ for all values of $\delta_{\rm CDW}$.
302: These results have been obtained
303: for $120 \times 120$ cluster that at temperatures $kT \sim 10^{-2} t$
304: gives convergent results (we refer to Ref. \cite{my2} for details
305: of the cluster calculations).
306: One can see that even for small magnitudes of the CDW order parameter
307: the upper critical field is significantly reduced. However, qualitative
308: temperature dependence of $H_{c2}$
309: is not affected by the charge--density wave correlations.
310: The reduction of the upper critical field due to the charge--density waves
311: may be brought about by a direct coupling between CDW and superconducting order
312: parameters as well as by the modification of the density of states. In order to
313: distinguish these contributions we investigate the upper critical field
314: in the presence of a phenomenological normal--state gap of arbitrary
315: magnitude and depth. This problem will be discussed in the next section.
316: %%%
317:
318: \section{$H_{c2}$ in the presence of a phenomenological gap}
319:
320: In this section we investigate modification of the upper critical
321: field that originates only from the normal--state gap in the density of states.
322: In contradistinction to the analysis presented
323: in the previous section, the density of states may remain finite despite
324: the presence of the gap. Here, the normal--state gap is characterized by
325: the width $2\delta$ and the relative depth
326: $\left(\rho_0-\rho_{\rm PG}\right)/\rho_0$, where $\rho_{\rm PG}$ and
327: $\rho_0$ denote the density of states in the presence and without the
328: pseudogap, respectively. It is visualized in the inset in Fig. 2.
329: In the absence of the CDW order the upper critical field is determined
330: by Eq. (10) with
331: \begin{equation}
332: K_{\bar{l}\bar{k}p}(\omega)=
333: \frac{
334: 1
335: }{
336: \left(\omega-E_{\bar{l}p\uparrow}\right)
337: \left(\omega+E_{\bar{k},-p\downarrow}\right)
338: }.\end{equation}
339: In order to account for the modification of the density of states
340: we renormalize the normal--state propagators which give rise
341: to the Cooper-pair susceptibility
342: \begin{eqnarray}
343: && \frac{1}{\omega-E_{\bar{l}p\uparrow}} \longrightarrow
344: \frac{\rho_{\rm PG}}{\rho_0} \frac{1}{\omega-E_{\bar{l}p\uparrow}}
345: +\frac{\rho_0-\rho_{\rm PG}}{2\rho_0} \nonumber \\
346: && \times
347: \left[ \left(1+\frac{E_{\bar{l}p\uparrow}}{\sqrt{E^2_{\bar{l}p\uparrow}+\delta^2}}
348: \right)\frac{1}{\omega-\sqrt{E^2_{\bar{l}p\uparrow}+\delta^2}} \right.
349: \nonumber \\
350: &&+\left.
351: \left(1-\frac{E_{\bar{l}p\uparrow}}{\sqrt{E^2_{\bar{l}p\uparrow}+\delta^2}}
352: \right)\frac{1}{\omega+\sqrt{E^2_{\bar{l}p\uparrow}+\delta^2}}
353: \right] \nonumber. \\
354: \end{eqnarray}
355: In the limiting case $\rho_{\rm PG}=\rho_0$ one obtains the standard density of states
356: as determined by the Hofstadter spectrum, whereas
357: for $\rho_{\rm PG}=0$ the density of states vanishes in the vicinity
358: of the Fermi energy. Substituting the renormalized
359: propagators into Eq. (16) one can calculate the upper critical
360: field in the same way as described in the previous section.
361:
362: \begin{figure}
363: \epsfxsize=9cm
364: \centerline{\epsffile{fig2.eps}}
365: \caption{Critical temperature as a function of the pseudogap width $\delta$
366: for different values of the external magnetic field. $\rho_{\rm PG} =
367: \frac{1}{2} \rho_0$ was used. The inset shows a schematic
368: density of states in the vicinity of the Fermi level.}
369: \end{figure}
370:
371: Figure 2. shows the superconducting transition temperature
372: obtained for different values of the reduced magnetic field
373: with $\rho_{\rm PG}=\frac{1}{2} \rho_0$. As before, the intersite coupling
374: $V$ have been adjusted to obtain $kT_c=0.02t$ in the absence of magnetic field.
375: One can see that the upper critical
376: field is reduced due to the presence of the normal--state gap.
377: The most significant lowering of
378: $H_{c2}$ takes place for finite values of the $\delta$
379: which are comparable to the magnitude
380: of the superconducting gap. This result originates from the fact
381: that the Cooper pair susceptibility is strongly peaked at the Fermi level with
382: a characteristic energy scale that is determined by temperature.
383: Therefore, for $\delta \gg kT_{c}$ the pseudogap results in a global lowering
384: of the density of states which can be compensated by an enhancement of the
385: pairing potential. It means that assuming stronger pairing potential $V$
386: we can reproduce $H_{c2}(T)$ calculated in the absence of the gap.
387:
388: \begin{figure}
389: \epsfxsize=9cm
390: \centerline{\epsffile{fig3.eps}}
391: \caption{Upper critical field as a function of temperature for different
392: densities of states at the Fermi level. The half width of the pseudogap
393: is $\delta=0.1t$. The inset shows temperature dependence of upper critical
394: field calculated for the phenomenological pseudogap with $\rho_{\rm PG}=0$
395: (solid line) and for the CDW gap (dashed line). In both cases the half
396: width is $\delta=\delta_{\rm CDW}=0.01t$.}
397: \end{figure}
398:
399: We have found that
400: the reduction of the upper critical field increases
401: with the depth of the gap as depicted in Fig. 3.
402: The inset in Fig. 3 shows a comparison of $H_{c2}$ obtained for
403: the phenomenological pseudogap with $\rho_{\rm PG}=0$ and for the
404: charge--density waves. A comparison of these results clearly
405: indicate that the coupling between CDW and superconducting
406: order parameters results in a small decrement of the
407: upper critical field.
408:
409:
410: \section{$H_{c2}$ in a stripe phase}
411: Other unusual feature of HTSC, that we discuss in the present section,
412: is related to inhomogeneous distribution of holes.
413: It results in a stripe--phase which consists of antiferromagnetic
414: domains separated by hole--rich domain walls.
415: We study how the upper critical field is affected by this specific
416: distribution of carriers.
417: In order to simulate the presence
418: of a stripe--phase we carry out the calculations for a long and narrow
419: rectangular--shape clusters.
420: We assume that the isolating, antiferromagnetic domains can
421: be simulated by fixed boundary conditions in the direction
422: perpendicular to the stripes (along the $x$ axis).
423: The spatial organization of the stripe structure has been
424: intensively investigated on experimental \cite{trank,stripee} and
425: theoretical grounds \cite{scalap1,scalap2}.
426: Experimental data for HTSC show that
427: the width of stripes depends on the concentration of holes and is
428: of the order of a few lattice constants.
429: The neutron--scattering study of the stripe phase \cite{trank}
430: suggests that the hole-rich domain walls are only single cell wide.
431: On the other hand, the numerical study of the two--dimensional $t$--$J$ model
432: \cite{scalap1} shows that the domain walls may have a significant
433: density of holes over three rows of sites.
434: According to these results we consider $150 \times n$
435: finite systems, where $n=2,3$ and $7$.
436: Our simplified approach does not restore the
437: actual structure of the stripe--phase. In particular, for $n=1$ one
438: obtains an unphysical, purely one--dimensional system, that hardly depends
439: on the external magnetic field. Therefore, we investigate the
440: rectangular--shape clusters with the width as a free parameter.
441: Since we neglect the correlations between different stripes,
442: the upper critical field is determined by Eqs. (10) and (16).
443:
444: In the case of free electron gas external magnetic field leads
445: to the occurrence of rotationally invariant states corresponding
446: to the Landau orbits.
447: However, the geometry of the stripe--phase may seriously
448: affect the formation of the Landau orbits. This effect is of particular
449: significance if the radii of the Landau orbits, $R_L$, exceed the width
450: of the stripe, $an$, ($a$ is the lattice constant).
451: In order to visualize the impact of magnetic field on electrons in
452: the stripe--phase we have calculated the resulting current distribution.
453: Within the framework of the linear--response theory the current operator
454: is given by $\hat{J}_l(x,y)=-\partial \hat{H}/\partial A_l(x,y)$,
455: where $(x,y)$ denotes spatial coordinates and $l$'s are unit
456: vectors in the lattice axes directions. Results obtained
457: in the normal state ($V=0$)
458: on a $150 \times 7$ cluster with applied magnetic field
459: $h=0.1$ are presented in Figure 4.
460:
461: \begin{figure}
462: \epsfxsize=10cm
463: \centerline{\epsffile{fig4.eps}}
464: \caption{Current distribution in a piece of stripe of width of 7 sites,
465: calculated for the reduced magnetic field $h=0.1$. The lengths of the arrows
466: are proportional to the currents. Such a pattern is periodically repeated
467: along the stripe.}
468: \end{figure}
469:
470: Modification of the Landau orbits affects
471: the diamagnetic pair--breaking mechanism. Therefore,
472: one may expect that superconductivity survives in
473: the presence of much stronger magnetic fields than in the
474: homogeneous phase. This observation is confirmed by the numerical
475: calculations, as depicted in Figure 5.
476: \begin{figure}
477: \epsfxsize=9cm
478: \centerline{\epsffile{fig5.eps}}
479: \caption{Upper critical field as a function of temperature calculated
480: for stripes of different width. We have chosen appropriate values of $V$
481: which give the same transition temperature in the absence of magnetic field.
482: The inset shows a comparison of upper critical field for $150\times 150$
483: and $150\times 7$ systems with the same value of the pairing strength
484: $V=0.244t$
485: }
486: \end{figure}
487:
488: \noindent
489: Here, $150 \times 150 $
490: cluster corresponds to an infinite system. The enhancement of $H_{c2}$
491: is of particular
492: importance for weak magnetic fields, when $R_L/na \rightarrow \infty$.
493: One can observe a dramatic change of the
494: slope, ${\rm d}H_{c2}/{\rm d}T$, calculated at $T=T_c$. Here,
495: the impact of the magnetic field on the superconducting transition
496: temperature is much less than in the homogeneous two--dimensional case.
497:
498: The pseudogap and stripes affect the superconducting
499: properties of the system both in the presence and in the absence of
500: the magnetic field. Modification of the density of states changes the
501: effective coupling constant, $\lambda=\rho_{\rm FS} V$, that enters the
502: standard BCS gap equation.
503: Therefore, we have directly compared the $H_{c2}$ for systems, which in the
504: absence of magnetic field are characterized by the same transition
505: temperature (one can roughly say that $\lambda={\rm const}$).
506: In order to complete the discussion, we have also calculated
507: the $H_{c2}(T)$ for the case when the pairing potential does not
508: depend on the pseudogap and the stripe structure ($V={\rm const}$). Since, the
509: opening
510: of the pseudogap reduces $T_c$ it results also in an additional decrement
511: of the upper critical field, when compared to the results presented in Figs. (1-3).
512: However, an enhancement of the $H_{c2}$ in the stripe phase can take place
513: despite
514: the reduction of the superconducting transition temperature, as depicted in the
515: inset in Fig. 5.
516:
517: \section{Concluding remarks}
518:
519: In order to clarify some physical aspects of our method one can
520: compare it with approaches, which are commonly used
521: to investigate $H_{c2}$. Previously, we have applied
522: the same method to discuss the upper critical
523: field for isotropic superconductivity \cite{my1,my2}.
524: Then, one ends up with the gap equation that can be written
525: in the form
526: \begin{equation}
527: \Delta_{i}= \frac{V}{\beta }\sum_{j,\omega_n}
528: \Delta_{j} G(i,j,\omega_n)G(i,j,-\omega_n).
529: \end{equation}
530: Here, $\Delta_i=\left<c_{i\downarrow} c_{i \uparrow} \right>$ and
531: $G(i,j,\omega_n)$ is the one--electron Green's function
532: in the presence of a uniform and static magnetic field.
533: It is clear that the above equation is a
534: lattice version of the linearized Gor'kov equations
535: \cite{gorkov}, which determine
536: the critical field at a second-order transition,
537: where the superconducting gap $\Delta_i$ is
538: vanishing \cite{hw1964}. However, our method
539: does not allow to discuss the superconducting properties
540: below the $H_{c2}$ (e.g. the vortex state). In our approach
541: the electron Green's functions have been calculated exactly,
542: whereas in the standard case one makes use of
543: the semiclassical approximation that neglects the Landau
544: level quantization.
545:
546: To conclude,
547: we have investigated how the upper critical field is connected with different
548: features of high--temperature superconductors. In particular, we have
549: discussed $H_{c2}$ in the presence of charge--density waves, phenomenological
550: pseudogap and stripes. Our results suggest that a gap in the density of states
551: reduces the upper critical field, independently on the underlying
552: microscopic mechanism. For finite density of states at the Fermi level
553: this reduction is mostly pronounced when the width of the gap is of the
554: order of the superconducting transition temperature. In the phase with
555: isotropic CDW gap the density of states at the Fermi level vanishes.
556: Then, as one can expect, the upper critical field is strongly reduced
557: even by a relatively small gap. Here, the coupling between the CDW
558: and superconducting order parameters results in an additional reduction
559: of $H_{c2}$. On the other hand, in the presence of stripes the upper
560: critical field is enhanced, especially close to $T_c$. We attribute
561: this effect to the reduction of the orbital pair--breaking
562: mechanism since the radii of the Landau orbits are much larger than
563: the width of the stripes.
564:
565:
566: The presented investigation of $H_{c2}$
567: is restricted to the simplest case of the uniform magnetic field and
568: neglects a possible disorder in the vortex system.
569: It can originate from fluctuations close to the phase transition
570: or inhomogeneous charge and spin
571: distribution in the stripe--phase.
572: However, as we are concerned exclusively with the critical field
573: at the second--order transition, these effects are of minor importance.
574: We have also not discussed
575: the reentrance of superconductivity in the strong--magnetic field.
576: This effect has been investigated in the continuum
577: model \cite{tesanovic,macdonal} as well as in the case of lattice
578: gas \cite{my1}, when the structure of fractal energy spectrum is reflected
579: in phase diagram. Theoretical argumentation that supports the reentrance of
580: superconductivity remains valid also in the presence of pseudogap, at least
581: on the simplest level that has been used in the present paper.
582: However, in the genuinely strong magnetic field the
583: assumption of the field--independent gap is unphysical and microscopic
584: investigation of this phenomenon is needed.
585:
586: \acknowledgments
587: This work has been supported by the Polish State Committee
588: for Scientific Research, Grant No. 2 P03B 01819.
589: We acknowledge a fruitful discussion with Janusz Zieli{\'n}ski.
590:
591: \begin{thebibliography}{99}
592: \bibitem[*]{email} Electronic address: maciek@phys.us.edu.pl
593: \bibitem{arpes1}
594: D.S. Marshall, D.S. Dessau, A.G. Loeser, C.-H. Park, A.Y. Matsuura,
595: J.N. Eckstein,
596: I. Bozovic, P. Fournier, A. Kapitulnik, W.E. Spicer, and Z.-X. Shen,
597: Phys. Rev. Lett., {\bf 76}, 4841 (1996).
598: \bibitem{arpes2}
599: H. Ding,T. Yokoya, J.C. Campuzano, T. Takahashi, M. Randeria,
600: M.R. Norman, T. Mochiku,
601: K. Hadowaki, and J. Giapintzakis, Nature (London) {\bf382}, 51 (1996).
602: \bibitem{arpes3}
603: M.R. Norman, H. Ding, M. Randeria, J.C. Campuzano, T. Yokoya,
604: T. Takeuchi, T. Takahashi, T. Mochiku, K. Kadowaki, P. Guptasarma, D. Hinks,
605: Nature (London) {\bf392}, 157 (1998).
606: \bibitem{krasnov1} V.M. Krasnov, A. Yurgens, D. Winkler,
607: P. Delsing, and T. Claeson, Phys. Rev. Lett.
608: {\bf84}, 5860 (2000).
609: \bibitem{krasnov2}
610: V.M. Krasnov, A.E. Kovalev, A. Yurgens, and D. Winkler,
611: Phys. Rev. Lett. {\bf86}, 2657 (2001).
612: \bibitem{nmr1}
613: G.V.M. Williams, J.L. Tallon, E.M. Haines, R. Michalak, and R. Dupree,
614: Phys. Rev. Lett. {\bf78}, 721 (1997).
615: %
616: \bibitem{nmr2}
617: G.V.M. Williams, J.L. Tallon, J.W. Quilty, H.J. Trodahl, and N.E Flower,
618: Phys. Rev. Lett. {\bf80}, 337 (1998).
619: \bibitem{infra}
620: D.N. Basov, T. Timusk, B. Dabrowski, and J.D. Jorgensen,
621: Phys. Rev B{\bf50}, 3511 (1994).
622: %
623: \bibitem{transport}
624: J.L. Tallon, J.R. Cooper, P.S.I.P.N de Silva, G.V.M. Williams, and J.W. Loram,
625: Phys. Rev. Lett. {\bf75}, 4114 (1995).
626:
627: \bibitem{muller}
628: D. Rubio Temprano, J. Mesot, S. Janssen, K. Conder, A. Furrer,
629: H. Mutka, and K.A. M{\"u}ller, Phys. Rev. Lett. {\bf84}, 1990 (2000).
630: \bibitem{eremin}
631: I. Eremin, M. Eremin, S. Varlamov, D. Brinkmann, M. Mali, and J. Roos,
632: Phys. Rev. B{\bf56}, 11305 (1997).
633: \bibitem{seibold}
634: G. Seibold and S. Varlamov, Phys. Rev. B{\bf60}, 13056 (1999).
635: \bibitem{markiewicz} R. S. Markiewicz, C. Kusko, and V. Kidambi,
636: Phys. Rev. B{\bf60}, 627 (1999).
637: \bibitem{kramer} S. Kr{\"a}mer and M. Mehring, Phys. Rev. Lett. {\bf 83}, 396 (1999).
638: \bibitem{str1}
639: S. -W. Cheong, G. Aeppli, T. E. Mason, H. Mook, S. M. Hayden, P. C. Canfield,
640: Z. Fisk, K. N. Clausen, and J. L. Martinez , Phys. Rev. Lett. {\bf 67}, 1791 (1991).
641: \bibitem{str2} T. E. Mason, G. Aeppli, and H. A. Mook,
642: Phys. Rev. Lett. {\bf 68}, 1414 (1992).
643: \bibitem{str3} T. R. Thurston, P. M. Gehring, G. Shirane, R. J. Birgeneau,
644: M. A. Kastner, Y. Endoh, M. Matsuda, K. Yamada
645: H. Kojima and I. Tanaka , Phys. Rev. B {\bf 46}, 9128 (1992).
646: \bibitem{str4} K. Yamada, S. Wakimoto, G. Shirane, C. H. Lee, M. A. Kastner, S. Hosoya,
647: M. Greven, Y. Endoh, and R. J. Birgeneau , Phys. Rev. Lett. {\bf 75}, 1626 (1995).
648: \bibitem{str5} S. M. Hayden, G. Aeppli, H. A. Mook, T. G. Perring, T. E. Mason,
649: S.-W. Cheong, and Z. Fisk , Phys. Rev. Lett. {\bf 76}, 1344 (1996).
650: \bibitem{str6} U. L{\"o}w, V. J. Emery, K. Fabricius, and S. A. Kivelson,
651: Phys. Rev. Lett. {\bf 72}, 1918 (1994).
652: \bibitem{str7} J. Zaanen and O. Gunnarsson, Phys. Rev. B {\bf 40}, 7391 (1989).
653: \bibitem{osofsky} M.S. Osofsky, R.J. Soulen, Jr.,
654: S.A. Wolf, J.M. Broto, H. Rakoto, J.C. Ousset, G. Coffe,
655: S. Askenazy, P. Pari, I. Bozovic, J.N. Eckstein, and G.F. Virshup,
656: Phys. Rev. Lett. {\bf71}, 2315 (1993).
657: \bibitem{mackenzie}
658: A.P. Mackenzie, S.R. Julian, G.G. Lonzarich,
659: A. Carrington, S.D. Hughes, R.S. Liu, and D.C. Simclair,
660: Phys. Rev. Lett. {\bf71}, 1238 (1993).
661: \bibitem{gorkov} L.P. Gor'kov, Zh. Eksp, Teor. Fiz. {\bf36}, 1918 (1959)
662: [Sov. Phys. JETP {\bf9}, 1364 (1960)].
663: \bibitem{gesh} V. B. Geshkenbein, L. B. Ioffe and A. J. Millis,
664: Phys. Rev. Lett. {\bf 80}, 5778 (1998).
665: \bibitem{wen} H. H. Wen, W. L. Yang and Y. M. Ni, Phys. Rev. Lett. {\bf 82},
666: 410 (1999).
667: \bibitem{hw} E. Helfand and N.R. Werthamer, Phys. Rev. Lett. {\bf13},
668: 686 (1964).
669: \bibitem{kresin} Yu. N. Ovchinnikov and V. Z. Kresin,
670: Phys. Rev. B {\bf52}, 3075 (1995).
671: \bibitem{my1}
672: M. Mierzejewski and M.M. Ma{\'s}ka, Phys. Rev. B {\bf60},
673: 6300 (1999).
674: \bibitem{my2}
675: M. M. Ma{\'s}ka and M. Mierzejewski, Phys. Rev. B {\bf64}, 064501 (2001).
676: \bibitem{peir}
677: R. E. Peierls, Z. Phys. {\bf80}, 763 (1933),
678: J.M. Luttinger, Phys. Rev. {\bf84}, 814 (1951).
679: \bibitem{harper} P.G. Harper, Proc. Phys. Soc. London, Sect. A{\bf68},
680: 874 (1955).
681: \bibitem{trank} J. M. Tranquada, J. D. Axe
682: N. Ichikawa, Y. Nakamura, S. Uchida, and B. Nachumi,
683: Phys. Rev. B {\bf54}, 7489 (1999).
684: \bibitem{stripee}
685: A. W. Hunt, P. M. Singer, K. R. Thurber, and T. Imai,
686: Phys. Rev. Lett. {\bf 82}, 4300 (1999);
687: %
688: M. Roepke, E. Holland-Moritz, B. B{\"u}chner, H. Berg,
689: R. E. Lechner, S. Longeville, J. Fitter,
690: R. Kahn G. Coddens,
691: and M. Ferrand,
692: Phys. Rev. B {\bf60}, 9793 (1999);
693: %
694: P. M. Singer, A. W. Hunt, A. F. Cederstr{\"o}m, and T. Imai,
695: Phys. Rev. B {\bf60}, 15345 (1999);
696: %
697: N. J. Curro, P. C. Hammel, B. J. Suh, M. H{\"u}cker, B. B{\"u}chner, U. Ammerahl,
698: and A. Revcolevschi, Phys. Rev. Lett. {\bf85}, 642 (2000);
699: %
700: B. J. Suh, P. C. Hammel, M. H{\"u}cker B. B{\"u}chne, U. Ammerahl,
701: and A. Revcolevschi, Phys. Rev. B {\bf61}, R9265 (2000).
702: \bibitem{scalap1} S. R. White and D. J. Scalapino,
703: Phys. Rev. Lett. {\bf 80}, 1272 (1998).
704: \bibitem{scalap2} S. R. White and D. J. Scalapino,
705: Phys. Rev. Lett. {\bf 81}, 3227 (1998).
706: \bibitem{hw1964} E. Helfand and N. R. Werthamer, Phys. Rev. Lett.
707: {\bf 13}, 686 (1964).
708: \bibitem{tesanovic} M. Rasolt and Z. Tesanovic, Rev. Mod. Phys.
709: {\bf 64}, 709 (1992).
710: \bibitem{macdonal}
711: A. H. MacDonald, H. Akera, and M. R. Norman, Phys. Rev. B{\bf 45},
712: 10147 (1992);
713: M. R. Norman, H. Akera, and A. H. MacDonald, Physica C {\bf 196},
714: 43, (1992); H. Akera, A. H. MacDonald and M. R. Norman,
715: Physica B {\bf184}, 337 (1993).
716: \end{thebibliography}
717: \end{document}
718:
719:
720:
721:
722:
723:
724:
725:
726:
727:
728:
729:
730:
731:
732: