1: \documentclass[aps,prb,twocolumn,showpacs,showkeys,groupedaddress,floatfix,amsmath,amssymb]{revtex4}

2: %\documentclass[aps,prb,preprint,showpacs,showkeys,groupedaddress,floatfix,amsmath,amssymb]{revtex4}

3: 

4: \usepackage{graphicx}

5: \usepackage{dcolumn}

6: \usepackage{bm}

7: 

8: 

9: \begin{document}

10: 

11: 

12: \title{Stationary phase slip state in quasi-one-dimensional rings.}

13: 

14: \author{D. Y. Vodolazov}

15: \author{B. J. Baelus}

16: \author{F. M. Peeters}

17: \email{peeters@uia.ua.ac.be} \affiliation{Departement Natuurkunde,

18: Universiteit Antwerpen (UIA), Universiteitsplein 1,  B-2610

19: Antwerpen, Belgium}

20: 

21: 

22: %\date{\today}

23: 

24: 

25: \begin{abstract}

26: The nonuniform superconducting state in a ring in which the order

27: parameter vanishing at one point is studied. This state is

28: characterized by a jump of the phase by $\pi$ at the point where

29: the order parameter becomes zero. In uniform rings such a state is

30: a saddle-point state and consequently unstable. However, for

31: non-uniform rings with e.g. variations of geometrical or physical

32: parameters or with attached wires this state can be stabilized and

33: may be realized experimentally.

34: \end{abstract}

35: 

36: \pacs{74.60Ec, 74.20.De, 73.23.-b} \keywords{phase slip,

37: mesoscopic superconductivity}

38: 

39: \maketitle

40: 

41: In recent years \cite{Berger,Horane,Berger1} the existence of a

42: single-connected state in a ring was proposed theoretically. Such

43: a state implies that the relation between the phase $\phi$ of the

44: order parameter $\psi=fe^{{\rm i}\phi}$, the current density $j$

45: and the magnetic flux $\Phi$ through the ring (which follows from

46: the single valuedness of $\psi$ - see for example Ref.

47: \cite{Tinkham})

48: \begin{equation}

49: \oint \frac{j}{f^2}ds=2\pi n-\Phi,

50: \end{equation}

51: is no longer valid (the flux is expressed in $\Phi_0/2\pi$, the

52: current density in $j_0=c\Phi_0/4\pi^2\lambda^2\xi$, $\lambda$ is

53: the London penetration length, $\xi$ the coherence length and

54: $\Phi_0$ is the quantum of magnetic flux). The reason is that the

55: order parameter in such a single-connected state is zero at one

56: point. Moreover, it was claimed that under certain conditions,

57: i.e. radius of the ring less than $\xi$ and the flux through the

58: ring is about $(n+1/2)\Phi_0$ this state may become metastable in

59: some range of magnetic fields \cite{Horane}.

60: 

61: In this paper we revisited this problem and we will show that a

62: state where the order parameter vanishes in one point is still

63: double-connected and that Eq. (1) is valid for such a state. We

64: propose to call such a state a one-dimensional vortex state (ODV

65: state), because like for an ordinary two-dimensional Abrikosov

66: vortex, there is a jump in the phase of the order parameter of

67: $\pi$ at the point where $\psi=0$. In contrast to a

68: two-dimensional Abrikosov vortex the 1D vortex is an unstable

69: structure in an uniform ring. However, if some inhomogeneities are

70: present in the ring (defects, nonuniform thickness or nonuniform

71: width of the ring, attached superconducting wires) the ODV

72: structure can be stabilized and may consequently be realized

73: experimentally.

74: 

75: Consider a uniform ring with thickness $d \lesssim \lambda$ and

76: width $w \lesssim \xi$. In addition, let the radius of the ring

77: $R$ be much larger than $w$. Under these conditions we can neglect

78: the screening effects and the problem is reduced to a

79: one-dimensional one. The distribution of the current density $j$

80: and the order parameter $\psi$ of the system at a temperature not

81: far from the critical temperature $T_c$ is described by the 1D

82: Ginzburg-Landau equation (plus the condition ${\rm div} {\bf

83: j}=0$)

84: \begin{subequations}

85: \begin{equation}

86:  \frac{d^2f}{ds^2}+f(1-f^2-p^2)=0, %2a

87: \end{equation}

88: \begin{equation}

89:  \frac{dj}{ds}=\frac{d}{ds}f^2p=0, %2b

90: \end{equation}

91: \end{subequations}

92: where the gauge-invariant momentum $p=\nabla \phi -A$ is scaled in

93: units of $\Phi_0/(2\pi\xi)$, the length of the ring is $L=2\pi R$

94: and the circular coordinate $s$ is in units of the coherence

95: length $\xi$. In these units, the magnetic field is scaled in

96: units of the second critical field, $H_{c2}$, and the magnetic

97: flux in $\Phi_0/2\pi$.

98: 

99: The coupled Eqs. (2a) and (2b) have to be solved with the boundary

100: condition $\psi(-L/2)=\psi(L/2)$. We  use the method proposed in

101: \cite{Langer} (see also \cite{Zhang}). These equations have the

102: first integral

103: \begin{equation}

104: \frac{1}{2}\left(\frac{df}{ds}\right)^2+\frac{f^2}{2}-\frac{f^4}{4} %3

105: +\frac{j^2}{2f^2}=E.

106: \end{equation}

107: From a formal point of view, Eq. (3) is nothing else then the

108: condition of "energy", $(E)$, conservation for some "particle"

109: with "coordinate" $f$ and "momentum" $j$. The role of "time" is

110: played \cite{Langer} by the circular coordinate $s$ . In Fig. 1 we

111: show the dependence of the "potential energy" of this system

112: \begin{equation}

113: U(f)=\frac{f^2}{2}-\frac{f^4}{4}+\frac{j^2}{2f^2}, %4

114: \end{equation}

115: for different values of $j$. Possible solutions of Eqs. (2a,b) are

116: in the region where confined "trajectories" of our virtual

117: particle exist. This is possible for currents $0<j\leq j_c$

118: ($j_c=\sqrt{4/27}j_0$ is the depairing current density ).

119: \begin{figure}[ht]

120: \includegraphics[width=0.48\textwidth]{fig1.eps}

121: \caption{Dependence of the "potential energy" $U(f)$ for different

122: values of the current $j$: 1- $j$=$0.01$, 2- $j$=$0.05$, 3-

123: $j$=$0.15$, 4- $j$=$j_c$=$\sqrt{4/27}$.}

124: \end{figure}

125: Using Eq. (3) we immediately can write the solution of Eqs. (2a,b)

126: for the ring

127: \begin{eqnarray}

128: \sqrt{2}s=\int_{t_0}^{t}\frac{dt}{\sqrt{(t-t_0)(t_1-t_0)(t_2-t_0)}}= \nonumber\\

129: \frac{2}{\sqrt{t_2-t_0}}F\left({\rm sin^{-1}}\sqrt{\frac{t-t_0} %5

130: {t_1-t_0}},\sqrt{\frac{t_1-t_0}{t_2-t_0}}\right),

131: \end{eqnarray}

132: where $t(s)=f^2(s)$, $F(\theta,m)$ is the elliptic integral of the

133: first kind and $t_0\leq t_1\leq t_2$ are the solutions of the

134: cubic equation

135: \begin{equation}

136: t^3-2t^2+4Et-2j^2=0. %6

137: \end{equation}

138: Using the boundary condition for $f$ we may conclude, that for a

139: given current there exist a maximum of three solutions. There are

140: two uniform solutions (in the points of the minimum $E=E_{min}$

141: and the maximum $E=E_{max}$ of the "potential energy"(4)) and one

142: nonuniform solution with energy $E_{min}<E<E_{max}$ which is

143: defined by the equation

144: \begin{equation}

145: L=\sqrt{\frac{8}{t_2-t_0}}K\left(\sqrt{\frac{t_1-t_0}{t_2-t_0}}\right), %7

146: \end{equation}

147: where $K(m)$ is the complete elliptic integral of the first kind.

148: Numerical analysis of Eq. (7) for $0<j<j_c$ shows that there is a

149: minimal value for the ring length $L_{min}$ for which there exists

150: a solution for this equation. When $j\to j_c$, $L_{min} \to

151: \infty$ and in the opposite limit $j \to 0$ one can show that

152: $L_{min} \to \pi$. The latter corresponds to a ring radius $R=1/2$

153: (or $\xi/2$ in dimensional units). For a radius $R>1/2$ metastable

154: states may exist and superconductivity is present for any value of

155: the magnetic flux (at least in our one-dimensional model)

156: \cite{Horane,Buzdin}.

157: 

158: In principle, Eqs. (5-7) define the nonuniform solution of Eqs.

159: (2a,b) for a ring. Unfortunately, even using the explicit Kordano

160: expressions for the roots of Eq. (6) the results are rather

161: complicated for arbitrary values of the current $j$. However a

162: tractable analytical solution is possible if we consider the case

163: of low currents $j \ll 1$ for which the roots of Eq. (6) are

164: simplified to

165: \begin{subequations}

166: \begin{equation}

167:  t_0 \simeq j^2/2E ,  %8a

168: \end{equation}

169: \begin{equation}

170:  t_1\simeq 1-\sqrt{1-4E} , %8b

171: \end{equation}

172: \begin{equation}

173: t_2\simeq 1+\sqrt{1-4E} . %8c

174: \end{equation}

175: \end{subequations}

176: After inserting these results into Eq. (7) we obtain the "energy".

177: For $L \gg \pi$ the "energy" $E\simeq 1/4$ is practically

178: independent of $L$ (e.g. for $L=8$ the difference $1/4-E(L=8)$ is

179: less than 0.002). Also in the limit $L-\pi \ll 1$ we found that

180: the "energy" $E$ is independent of $j$

181: \begin{equation}

182: E(L) \simeq \frac{1}{2}\left(\frac{L-\pi}{\pi/8}\right)^2. %9

183: \end{equation}

184: In the limit $j \ll 1$ we can also find \cite{comm1} the

185: dependence of $t(s)$ near the minimum point of $t(s)$

186: \begin{subequations}

187: \begin{eqnarray}

188: t(s)=2j^2+s^2/2, & L\gg \pi, s\ll 1 \label{a}%10a

189: \\

190: t(s)=j^2/2E+2Es^2, &  L-\pi \ll 1,  s\ll 1 \label{b} %10b

191: \end{eqnarray}

192: \end{subequations}

193: Using Eqs. (10a,b) it is easy to show, that in the limit $j \to 0$

194: the gauge-invariant momentum is given by

195: \begin{equation}

196: p(s)=j/t(s)=\frac{j}{|j|}(\pi+2\pi m)\delta(s), %11

197: \end{equation}

198: where $\delta(s)$ is the Dirac function and $m$ is a integer. As a

199: result the integral

200: \begin{eqnarray}

201: \lim\limits_{j\to 0}\oint \frac{j}{f^2}ds=\lim\limits_{j \to

202: 0}\oint p(s)ds= \nonumber

203: \\

204: =\phi(+\epsilon)-\phi(-\epsilon)=\Delta \phi=\pm %12

205: (\pi+2\pi m),

206: \end{eqnarray}

207: for arbitrary ring size \cite{comm2}. Combining Eq. (12) with Eq.

208: (1) it is easy to show that if $\Phi=(k+1/2)\Phi_0$ ($k$ is an

209: integer) a solution of Eq. (2a) can be found which vanishes at one

210: point.

211: 

212: In previous \cite{Berger,Horane,Berger1} studies of this state Eq.

213: (11) was not taken into account (only the absolute value of the

214: order parameter was found). Furthermore, in Ref. \cite{Horane} it

215: was claimed that for small rings there is a magnetic field region

216: in which such a state exist and is metastable. But we find that in

217: this region this state even does not exist (except in one point)

218: because Eq. (1) is not fulfilled!

219: 

220: The distribution of the order parameter as obtained in the present

221: paper may also be applied to a ring in which some part of the ring

222: consists of a normal metal. Then at the boundary between the

223: normal metal and the superconductor the condition $\psi \simeq 0$

224: is fulfilled (if the normal part is longer than $\xi$) and the

225: distribution of the density $|\psi(s)|^2$ coincides with the one

226: obtained in Ref. \cite{Horane} (but the phase $\phi(s)$ will be

227: different for our previous ring geometry). Besides we cannot call

228: such a state single-connected as in Ref.

229: \cite{Berger,Horane,Berger1} because the relation (1) is valid

230: even for this case. Therefore, it is better to call this state a

231: one-dimensional vortex state because like the two-dimensional

232: Abrikosov vortex there is a point where $|\psi|=0$ and the phase

233: of the order parameter exhibits a jump equal to $k\pi$ (as the

234: Abrikosov vortex with orbital momentum $2\pi k$). A numerical

235: calculation shows that, as in the case for an ordinary Abrikosov

236: vortex in superconductors where such a state is stable, the phase

237: jump is always equal to $\pi$, i.e. $k=1$.

238: \begin{figure}[h]

239: \includegraphics[width=0.5\textwidth]{fig2.eps}

240: \caption{Distribution of the absolute value (a) and phase (b) of

241: the order parameter at different values of the magnetic flux: 1-

242: $\Phi$=$0.45\Phi_0$, 2- $\Phi$=$0.49\Phi_0$, 3-

243: $\Phi$=$0.498\Phi_0$, 4- $\Phi$=$(0.5-0)\Phi_0$, 5-

244: $\Phi$=$(0.5+0)\Phi_0$, 6- $\Phi$=$0.502\Phi_0$, 7-

245: $\Phi$=$0.51\Phi_0$, 8- $\Phi$=$0.55\Phi_0$. Note that the order

246: parameter for $\Phi=\Phi_0/2+\alpha$ is the same as for

247: $\Phi=\Phi_0/2-\alpha$ when $\alpha<\Phi_0/2$. In the insets (a)

248: $f_{min}$ and (b) the phase difference $\Delta \phi$ near the

249: point $s/L=3/4$ are shown as function of $\Phi$.}

250: \end{figure}

251: 

252: Our numerical analysis of the time-dependent Ginzburg-Landau

253: equations showed that this state is completely unstable for a

254: uniform ring. However for rings with nonuniform width(thickness)

255: or attached wires a nonuniform distribution of the order parameter

256: becomes possible \cite{Berger,Berger1,Fink1,Fink2} and for small

257: rings with $L \sim \pi$ this state is realized in practice

258: (besides for the case of a ring, a ODV state can also exist in the

259: Wheatstone bridge - see Refs. \cite{Fink3,Ammann}). In Fig. 2 the

260: distribution of the absolute value and the phase of the order

261: parameter is shown for different values of the magnetic field for

262: a nonuniform ring \cite{comm3}. We used the following model where

263: the width of the ring was varied as

264: \begin{equation}

265: w(s)=1+w_0{\rm sin}\left(\frac{2\pi s}{L}\right), %13

266: \end{equation}

267: The parameters used are: $w_0=0.1$ and $L=3.25$. It is seen that

268: with increasing magnetic flux the magnitude of the order parameter

269: $f$ decreases in the thinner part of the ring (at $s/L=3/4$) and

270: becomes zero when $\Phi/\Phi_0=1/2$. In Ref. \cite{Berger1} it was

271: found that the ODV state is only possible at $\Phi=(n+1/2)\Phi_0$.

272: The reason is now clear - only at this values of the magnetic flux

273: the phase difference near the point where the order parameter is

274: zero will be compensated by the term $2\pi n-\Phi$ and the current

275: density will be equal to zero. It is interesting to note that the

276: current density is also zero for $\Phi=n\Phi_0$ even in the case

277: of a nonuniform ring, but for values of the flux where the ODV

278: state does not exist. At these values of the magnetic flux the

279: order parameter is uniform along the ring and the term $2\pi n$ is

280: completely compensated by the term $\Phi$ in Eq. (1).

281: 

282: \begin{figure}[hbtp]

283: \includegraphics[width=0.48\textwidth]{fig3.eps}

284: \caption{The current in the ring (with attached wire of length

285: $5\xi$) as function of the flux. The different curves are for

286: different circumference of the ring: 1- $L$=1, 2- $L$=2, 3- $L$=3,

287: 4- $L$=3.5, 5- $L$=4. For $L$=4 the ODV state does not exist in

288: the ring and hysteresis appears. In the inset the dependence of

289: $dj/d\Phi$ on the flux $\Phi$ is shown for rings with lengths

290: $L$=1, 2, 3, 3.5 (curves 1,2,3,4 respectively). Note that there

291: are two maxima in the range $(0,1)$. With increasing length the

292: two maxima merge into one and when one approaches the critical

293: length $dj/d\Phi$ diverges at $\Phi=\Phi_0/2$.}

294: \end{figure}

295: 

296: Above we showed that the ODV state can be realized by varying the

297: geometrical parameters of the ring. But there are two alternative

298: approaches to realize the ODV state in the ring experimentally.

299: First, it is possible to include an other phase in the ring. This

300: leads to the appearance of a weak link in the sample and if the

301: radius of the ring is less than some critical value (about $\xi$)

302: it also leads to the existence of the ODV state in the ring at

303: $\Phi=(n+1/2)\Phi_0$. We modelled this situation by introducing an

304: additional term $\rho(s)f$ in Eq. (2a), and, as an example, we

305: took $\rho(s)=-\alpha \delta(s)$ and $\alpha=1$ (it should be

306: noted that qualitatively the results do not depend on the specific

307: value of $\alpha$). We will not present the numerical results of

308: the modified Ginzburg-Landau equations but we found that they are

309: qualitatively similar to the behavior of $f$ and $\phi$ shown in

310: Fig. 2. When approaching the flux $\Phi=(n+1/2)\Phi_0$ the order

311: parameter reaches zero in the defect point and a jump in the phase

312: equal to $\pi$ occurs. Secondly, such a ODV state should also

313: appear in small rings with attached wire(s). As was shown in Refs.

314: \cite{Fink1,Fink2} an attached wire leads to a nonuniform

315: distribution of the order parameter in the ring. In Fig. 3 the

316: dependence of the current $j$ in such a ring on the flux through

317: the ring is shown (see also Ref. \cite{Fink2}). As for nonuniform

318: rings, where the parameters are chosen such that the ODV state may

319: exist, there is no hysteresis in such a system \cite{Berger1}.

320: \begin{figure}[t]

321: \includegraphics[width=0.48\textwidth]{fig4.eps}

322: \caption{The Gibbs free energy and the magnetization (right lower

323: inset) of the asymmetric ring (upper inset) as function of the

324: applied magnetic field.}

325: \end{figure}

326: Small inhomogeneities in rings gives us the unique possibility to

327: study states close to the ODV state because there is no sharp

328: transition from the state with finite order parameter to a state

329: with vanishing order parameter at one point (see Fig. 2). The

330: latter is, in some sense, a 'frozen' slip phase state. A small

331: variation of the flux through the ring near $\Phi=\Phi_0/2$ leads

332: to a change of $\Delta \phi$ by $2\pi$ (from $-\pi$ to $+\pi$ with

333: a concomitant change of the current from $-0$ to $+0$ - see Fig.

334: 2(b)). As a result an additional phase circulation $\oint \nabla

335: \phi ds$=$2\pi$ appears in the system, not because the magnetic

336: flux $\Phi$ is changed, but because the term $\oint j/f^2 ds$

337: changes from $-\pi$ to $+\pi$ (see Eq. (1)). When such a jump

338: occurs $\nabla \phi$ does not change in the ring, except near the

339: point where $f=0$, and as a result the current density in the

340: system changes continuously. In the usual case transitions of the

341: vorticity from a state with phase circulation $2\pi n$ to a state

342: with phase circulation $2\pi (n+1)$ leads to a jump in the current

343: {\it everywhere} in the ring because $\oint \nabla \phi ds$

344: changes by $2\pi$ and the order parameter and the current density

345: are finite {\it everywhere} in the system.

346: 

347: To supplement the above study we made also a numerical study of a

348: nonuniform ring of finite width and thickness by implementing our

349: previous finite difference solution of the coupled nonlinear

350: Ginzburg-Landau equations \cite{Baelus1}. As an example we took

351: the following parameters: outer radius of the ring $R_o=2\xi$,

352: radius of the hole $R_i=1.5\xi$, displacement of the hole from the

353: center $a=0.4\xi$, ring thickness $d=0.001\xi$ and Ginzburg-Landau

354: parameter $\kappa=2$.

355: 

356: In Fig. 4 the dependence of the Gibbs free energy $F$ and the

357: magnetization $M=-\partial F/\partial H$ of this system on the

358: magnetic field are shown. As in the case of our one-dimensional

359: ring these dependencies are reversible and there are magnetic

360: fields at which the magnetization is equal to zero (in the points

361: of local maximum and minimum of the free energy). We found that

362: the distribution of the order parameter and the phase in the ring

363: (see inset of Fig. 5) is similar to analogical distributions for

364: the above one-dimensional ring (Fig. 2) at low magnetic fields.

365: But different with the one-dimensional system the phase

366: circulation increase of $2\pi$ (between the points (2) and (3) in

367: Fig. 5) does not occur at the magnetic field value where the free

368: energy has a local maximum (and zero magnetization). With

369: decreasing width of the ring this point shifts towards the local

370: maximum in the free energy.

371: \begin{figure}[t]

372: \includegraphics[width=0.35\textwidth,angle=-90]{fig5.eps}

373: \caption{The Gibbs free energy $F(H)$ near the first maximum. In

374: the inset the phase and the order parameter distribution at

375: different values of the applied magnetic field (indicated by the

376: squares on the $F(H)$ curve) are shown. The phase was calculated

377: along the outer perimeter of the ring. The open circle on $F(H)$

378: indicates the position at which the vorticity increases from 0 to

379: 1.}

380: \end{figure}

381: It is interesting to note that for the system corresponding to

382: Fig. 4 the vortex enters through the thinnest part of the ring in

383: the case of the first three maxima in the free energy while for

384: the highest magnetic field maxima (i.e. $H/H_{c2}\simeq 4.6$) the

385: vortex enters through the thickest part of the ring (see Fig. 6).

386: Because the width of thicker part of the ring ($=0.9\xi$) is

387: considerable larger than the thinner part ($=0.1\xi$) and is of

388: order $\xi$ the ODV for $H/H_{c2}\simeq 4.6$ becomes the usual

389: two-dimensional vortex. As a result the order parameter is equal

390: to zero only in one point along the radial coordinate and the

391: circulation of the phase of the order parameter now is also a

392: function of this coordinate (see Fig. 6). When the vortex

393: enters/exits the ring at low magnetic fields there is also a slow

394: dependence of the order parameter on the radial coordinate along

395: the thinnest part of the ring. Therefore, we can conclude that in

396: a ring with finite width the one-dimensional vortex is transformed

397: into the usual two-dimensional vortex.

398: 

399: \begin{figure}[h]

400: \includegraphics[width=0.36\textwidth,angle=-90]{fig6.eps}

401: \caption{The same as Fig. 5 but now for the last maximum in Fig.

402: 4. In the inset a contourplot of the order parameter is shown for

403: different values of the magnetic field which are given by the

404: squares in the main figure. $L_{outer}$ is the vorticity as

405: calculated along the outer perimeter of the ring and $L_{inner}$

406: is the vorticity calculated along the perimeter of the hole.}

407: \end{figure}

408: 

409: The process of the appearance of a stable vortex in a ring with

410: finite width is similar to the creation of a phase slip in wires

411: of finite width in the presence of a transport current. In the

412: latter case the distribution of the gauge-invariant momentum is

413: not exactly uniform over the wire width. The maximum value of $p$

414: is obtained at the edges and as a result the order parameter

415: vanishes first in these points. When a phase slip is created the

416: distribution of the order parameter is not uniform over the width.

417: With decreasing wire width this nonuniformity decreases but it

418: will be uniform, strictly speaking, only in the limit $w \to 0$.

419: With increasing wire width the phase slip transforms to the

420: ordinary process of vortex/antivortex pair nucleating at the

421: edges, penetrating deep into the superconductor and annihilating.

422: In nonuniform rings the distribution of $p$ over the width is

423: nonuniform and in contrast to wires with transport current it is

424: nonsymmetric with respect to the ring width. Suppression of the

425: order parameter first occurs only on the external side (when

426: increasing the magnetic field) or on the internal side (when

427: decreasing the magnetic field). As in the case of an

428: one-dimensional ring we may call this state a stable ("frozen")

429: phase slip state if the width of the ring, where the vortex

430: penetrates, is less than $\xi$.

431: 

432: The ODV state of a nonuniform ring is very similar to the

433: saddle-point state as found in Ref. \cite{Baelus1} in case of an

434: uniform ring of finite width and in Ref. \cite{Schweigert} for the

435: case of a disk. Note that for an uniform ring with zero width the

436: nonuniform solution Eq. (6) corresponds to a saddle-point point of

437: $F(\psi)$. In Fig. 7 the gradual decrease of the hysteresis with

438: increasing displacement $a$ of the hole from the center of the

439: ring is shown. It is seen that with increasing $a$ the region

440: where metastable states exist decreases and ultimately vanishes

441: for some critical displacement $a_c$. The saddle point states

442: (only shown for $a=0$, by the thin full curve) approaches the

443: stable phase slip state when $a=a_c$.

444: \begin{figure}[t]

445: \includegraphics[width=0.36\textwidth,angle=-90]{fig7.eps}

446: \caption{The Gibbs free energy $F(H)$ at low magnetic fields for

447: different displacements $a$ of the hole from the center of the

448: ring. The thin solid curve (for $a=0$) corresponds to the saddle

449: point state.}

450: \end{figure}

451: In Ref. \cite{Baelus2} also nonuniform superconducting rings of

452: finite width were investigated (but with larger ring radius). It

453: was found that for low magnetic fields $F(H)$ is irreversible but

454: for high magnetic fields the dependence $F(H)$ was reversible. In

455: high fields the transition from a state with vorticity one to

456: another state occured through the same scenario as discussed

457: above. A similar state was also considered in Ref. \cite{Berger3}

458: in the framework of the linearized Ginzburg-Landau equations for a

459: nonuniform ring of finite width. They found that in such a system

460: the vortex may be stable in some magnetic field region. In low

461: magnetic field this vortex enter through the thinner part of the

462: ring and at high magnetic fields through the thicker part. Our

463: calculations generalizes this result to lower temperatures (i.e.

464: where the nonlinear Ginzburg-Landau equations have to be used) and

465: with inclusion of the non-zero demagnetization factor of the ring.

466: We found the dependence of the Gibbs free energy (and

467: magnetization) of this sample on the applied magnetic field. Our

468: results also allowed to compare the results of the one-dimensional

469: model with the full two-dimensional one.

470: 

471: The ODV (or stable phase slip) state may be observed by magnetic

472: experiments. Magnetic susceptibility is proportional \cite{Zhang}

473: to $dj/d\Phi$ and the ODV state exhibits some characteristic

474: peculiarities as was shown in the inset of Fig. 3. Magnetization

475: $M$ is proportional to the current $j$ and hence $M(H)$ is

476: reversible, for samples where the ODV state exists (see Fig. 3 and

477: inset of Fig. 5). Furthermore $M=0$ at $\Phi\simeq (n+1/2)\Phi_0$.

478: Alternatively, because at $\Phi=(n+1/2)\Phi_0$ there is a point in

479: the ring where the order parameter is equal to zero, this state

480: may be found by transport measurements. For example in Ref.

481: \cite{Liu} the dependence of the resistance of a hollow cylinder

482: with radius of order $\xi$ was studied at temperature $T<T_c$ far

483: from $T_c$. At $\Phi=\Phi_0/2$ the resistance $(R)$ exhibited a

484: maximum but the value of $R$ was a factor three less than the

485: normal state resistance $R_n$. If the cylinder has a nonuniform

486: thickness the ODV state can be realized as a stable state when

487: $\Phi=\Phi_0/2$ and it will lead to a resistive (but

488: superconducting) state even for very small currents and naturally

489: the resistance of such a state will be less than $R_n$ as was

490: found by Liu et al \cite{Liu}.

491: 

492: In conclusion, we studied the nonuniform state in a

493: superconducting ring in which the order parameter vanishes at one

494: point. It was shown that this state is characterized by a jump in

495: the phase of the order parameter by $\pi$ near the point $\psi=0$.

496: It allows, in correspondence with the ordinary two-dimensional

497: Abrikosov vortex, to call such a state a one-dimensional vortex.

498: This state is unstable in an uniform ring. In case of a nonuniform

499: ring (with variations of the geometrical ($d$, $w$) or physical

500: parameters ($\xi$, $\lambda$) along the ring) or for a ring with

501: an attached wire, this state may be realized in practise. For

502: rings with non-zero width the ODV state transforms into the usual

503: two-dimensional vortex. We also showed that this state is the

504: remnant of the saddle point state connecting two superconducting

505: states with different vorticity as found in a uniform ring. The

506: latter state becomes stable in a nonuniform ring.

507: 

508: The work was supported by the Flemish Science Foundation (FWO-Vl),

509: the "Onderzoeksraad van de Universiteit Antwerpen," the

510: "Interuniversity Poles of Attraction Program - Belgian State,

511: Prime Minister's Office - Federal Office for Scientific, Technical

512: and Cultural Affairs," and the European ESF-Vortex Matter. One of

513: us (D.Y.V.) is supported by a postdoctoral fellowship of FWO-Vl

514: and partially by a RFBR grant N01-02-16593. Discussions with Prof.

515: A. Geim and comments from Prof. J. Berger are gratefully

516: acknowledged.

517: 

518: \begin{references}

519: 

520: \bibitem{Berger} J. Berger and J. Rubinstein, Phys. Rev. Lett. {\bf

521: 75}, 320 (1995).

522: 

523: \bibitem{Horane} E. M. Horane, J. I. Castro, G. C. Buscaglia, and A.

524: Lopez,  Phys. Rev. B {\bf 53}, 9296 (1996).

525: 

526: \bibitem{Berger1} J. Berger and J. Rubinstein, Phys. Rev. B {\bf 56},

527: 5124 (1997).

528: 

529: \bibitem{Tinkham} M. Tinkham, Phys. Rev. {\bf 129}, 2413 (1963).

530: 

531: \bibitem{Langer} J. S. Langer and V. Ambegaokar, Phys. Rev. {\bf 164},

532:  498 (1967).

533: 

534: \bibitem{Zhang} X. Zhang and J. Price, Phys. Rev. B {\bf 55}, 3128

535: (1997).

536: 

537: \bibitem{Buzdin} A. Bezryadin, A. Buzdin, and B. Pannetier, Phys. Rev.

538: B {\bf 51}, 3718 (1995).

539: 

540: \bibitem{comm1}Numerical analysis shows that for large rings $L\gg \pi$ the solution

541: $t(s)=t_1-(t_1 -t_0)/{\rm cosh}^2(s\sqrt{(t_1-t_0)/2})$ found in

542: \cite{Langer} works very well for arbitrary currents $j<j_c$. The

543: reason is that for rings with $R \gg 1$ the "energy" $E$ is close

544: to $E_{max}$ and the roots $t_1$ and $t_2$ practically coincide.

545: As a result the situation is analogous to the one considered in

546: Ref. \cite{Langer}.

547: 

548: \bibitem{comm2} This result is a generalization of a result obtained

549: in Ref. \cite{Langer} for the saddle-point of Eq. (2) for the case

550: of an infinite long wire with vanishing transport current.

551: 

552: \bibitem{Fink1} H. J. Fink and V. Grunfeld, Phys. Rev. B {\bf 33},

553: 6088 (1986).

554: 

555: \bibitem{Fink2} H. J. Fink and V. Grunfeld, Phys. Rev. B {\bf 31},

556: 600 (1985).

557: 

558: \bibitem{Fink3} H. J. Fink, Phys. Rev. B {\bf 45}, 4799 (1992); {\it ibid.}

559: {\bf 48}, 3579 (1993).

560: 

561: \bibitem{Ammann} C. Ammann, P. Erd\"os, and S. B. Haley,

562: Phys. Rev. B {\bf 51}, 11739 (1995).

563: 

564: \bibitem{comm3} It is interesting to note that the results shown in

565: Fig. 2 correspond qualitatively to the time evolution of the order

566: parameter for an uniform ring in the case the vorticity increases

567: with one unit. The only difference is that for an uniform ring the

568: current density, generally speaking, is not equal to zero at the

569: point where the order parameter vanishes. The reason is that for

570: time-dependent processes the full current density is equal to the

571: sum of the superconducting $j_s$ and the normal $j_n$ parts. In

572: the point $f=0$, $j_s=0$ but $j_n\neq 0$.

573: 

574: \bibitem{Baelus1} B. J. Baelus, F. M. Peeters, and V. A.

575: Schweigert, Phys. Rev. B {\bf 63}, 144517 (2001).

576: 

577: \bibitem{Schweigert} V. A. Schweigert and F. M. Peeters, Phys.

578: Rev. Lett. {\bf 83}, 2409 (1999).

579: 

580: \bibitem{Baelus2} B. J. Baelus, F. M. Peeters, and V. A.

581: Schweigert, Phys. Rev. B {\bf 61}, 9734 (2000).

582: 

583: \bibitem{Berger3} J. Berger and J. Rubinstein, Phys. Rev. B {\bf 59},

584: 8896 (1999).

585: 

586: \bibitem{Liu} Y. Liu, Yu. Zadorozhny, M. M. Rosario, B. Y. Rock,

587: P. T. Carrigan, and H. Wang, Science {\bf 294}, 2332 (2001).

588: 

589: \end{references}

590: 

591: \end{document}

592: