1: %\documentclass[preprint,showpacs,aps]{revtex4}
2: \documentclass[twocolumn,showpacs,aps]{revtex4}
3:
4: \usepackage{graphicx}% Include figure files
5: \usepackage{dcolumn}% Align table columns on decimal point
6: \usepackage{bm}% bold math
7:
8: \begin{document}
9: \title {Comments on "Vortex phase diagram of HgBa$_2$Ca$_2$Cu$_3$O$_{8+\delta}$ thin films from magnetoresistance measurements"}
10: \author {Y. Z. Zhang}
11: \author{Z. Wang}
12: \affiliation{National Laboratory for superconductivity, Institute
13: of Physics \& Center for Condensed Matter Physics, Chinese Academy
14: of Sciences, P. O. Box 603, 100080, Beijing, China}
15:
16:
17: \pacs{74.25.Fy, 74.25.Ha, 74.25.Qt}
18: \begin{abstract}
19: We make comments on the paper presented by Kim \textit{et al.}
20: [Phys. Rev. B \textbf{61}, 11317 (2000)]. The authors analyzed activation energies of
21: HgBa$_2$Ca$_2$Cu$_3$O$_{8+\delta}$ thin films with a scaling
22: relation, and defined four vortex regions for thin films. We
23: find that the definitions of four vortex regions and the
24: scaling relation are questionable when studying their definition
25: of the resistivity range for thermally activated flux flow. Using
26: the empirical activation energy form suggested by Zhang \textit{et
27: al.} [Phys. Rev. B \textbf{71}, 052502 (2005)] for lower resistivity data, we find that the
28: form is not only in good agreement with the resistivity
29: $\rho(T,H)$, but also with the apparent activation energy
30: $-\partial \ln \rho(T,H)/\partial T^{-1}$.
31: \end{abstract}
32:
33: \maketitle Previously, Kim \textit{\textit{et al.}} \cite{Kim}
34: reported an investigation of resistive behaviors of
35: HgBa$_2$Ca$_2$Cu$_3$O$_{8+\delta}$ (Hg-1223) thin films in the
36: mixed state. A scaling relation was proposed for analyzing the
37: resistive behavior and was found in good agreement with the
38: apparent activation energy $-\partial \ln \rho(T,H)/\partial
39: T^{-1}$ (the so-called effective activation energy in this paper)
40: in a specialized resistivity range. With this scaling, they
41: suggested that the vortex system could be divided into four
42: different vortex regions corresponding to the flux flow (FF)
43: region, the thermally activated flux flow (TAFF) region, the
44: critical state region, and the vortex solid (VS) region. Three
45: characteristic temperatures corresponding to $T_{ff}$, $T^*$, and
46: $T_{irr}$ were defined for the boundaries of these regions. They
47: claimed that the TAFF behavior was limited in the region of
48: $T^*<T<T_{ff}$ and the corresponding activation energy was
49: expressed as $U_0(T,H) \sim H^{-1.1}(1-T/T_c)^{1.5}$ for the
50: magnetic field range from 1 to 9 T. In this comment, we point out
51: that the scaling relation and the definitions of the four
52: different vortex regions are questionable. We propose that the
53: activation energies of Hg-1223 thin films relate to lower resistivity where $T<T^*$. Using
54: lower resistivity data, we find that the activation energy is
55: expressed as the form suggested by Zhang \textit{et al.} \cite
56: {Zhang}.
57:
58: After the discovery of high temperature superconductors
59: (HTSCs), activation energies of different HTSCs have been
60: widely studied in theories and experiments. According to Palstra
61: \textit{et al.} \cite{Palstra}, activation energies of HTSCs
62: should be determined in the temperature interval over which the
63: resistivity changes from $10^{-4}$ to $1$ $\mu\Omega$ cm or for
64: the resistivity ratio $\rho(T,H)/\rho_n$ approximately below
65: $1\%$, where $\rho_n$ is normal-state resistivity. The temperature
66: interval is widely accepted for studying activation energies of
67: HTSCs; besides, it is widely accepted that TAFF resistivity shall
68: be ohmic (linear $I$-$V$ relation), and the activation energy is
69: independent of the applied current density $j$ for $j\to 0$ \cite
70: {Palstra, Geshkenbein, Vinokur, Blatter, Brandt, Cohen}. This
71: means that non-ohmic behavior ought to be observed for
72: decreasing temperature out of the TAFF region.
73:
74: Using the data of Fig. 1 and Fig. 4 in Ref.~\cite{Kim}, we roughly
75: redraw the $\rho (T^*,H)$ data in Arrhenius plot with gray circles as
76: shown in Figure.~\ref{f1}(a), for which the data in the range of
77: $\mu_0H(T^*)\le 7.0$ T were presented due to $H(T^*)$ data being
78: only presented for $\mu_0H(T^*)\le 7.0$ T in Fig. 4 of
79: Ref.~\cite{Kim}. One will easily find that $T^*(H)$ results in
80: that TAFF behavior is related to the resistivity value $\rho>1.4$
81: $\mu \Omega$ cm with resistivity ratio $\rho/\rho_n> 1.5 \%$,
82: where $\rho_n=\rho(140$ K$, H=0)\approx 93$ $\mu \Omega$ cm. In
83: this case, Kim \textit{et al.} gave the TAFF temperature interval
84: over which the resistivity approximately changes from $1.4 $ to
85: $10$ $\mu \Omega$ cm, or $\rho(T,H)/\rho_n$ approximately above
86: $1.5\%$. This interval is apparently mismatched with the interval
87: suggested by Palstra \textit{et al.} Below temperature $T^*$, Kim
88: \textit{et al.} defined the critical state region, but they did
89: not further present physical meaning for the region. In
90: comparing with other HTSCs, it is very dubious that non-ohmic
91: resistivity can be found around the so-called $T^*$. As a
92: result, we argue that definitions of the TAFF and the
93: critical state regions in the article are questionable and the
94: TAFF region ought to relate to the temperature interval as
95: suggested by Palstra \textit{et al.}; besides, we argue that the
96: definition of $T_{ff}$ is incorrect in the paper \cite
97: {Zhang}.
98:
99: %%%%%%%%%%%%%%%
100: \begin{figure}
101: \includegraphics
102: [width= .80\columnwidth] {fig1} \caption{\label{f1} (a) Arrhenius
103: plot of $\rho(T,H)$ from Ref.~\cite{Kim}. The gray circles
104: correspond to $\rho(1/T^*,H)$ data where $H(T^*)$ was determined
105: from Fig.4 of Ref.~\cite{Kim}. (b)$-\partial \ln
106: \rho(T,H)/\partial T_{ff}^{-1}$ curves from Ref.~\cite{Kim}. The
107: black lines, the black dashed lines, arrows, $T^*$, and
108: $T_{ff}$ are presented by the authors \cite{Kim}. The gray
109: lines in (a) and (b) are our regressions using $\rho_{0f}=8$
110: $\mu\Omega$ cm for the curves (see text for more details).}
111: \end{figure}
112: %%%%%%%%%%%%%%%
113: %%%%%%%%%%%%%%%
114:
115:
116:
117: Normally, the TAFF resistivity is expressed as
118: $\rho=\rho_0\exp(-U_0/T)$ with the activation energy
119: $U_0(T,H)=U_0(0,H)(1-t)^\beta$, where $t=T/T_c$, $\beta$ is
120: constant, and $T_c$ is the critical temperature. Considering the
121: interlayer decoupling in high fields, Zhang \textit{et al.} \cite
122: {Zhang} suggested an empirical relation
123: %%%%%%%%%%%%%%%
124: \begin{equation}
125: \rho=\rho_{0f}\exp[-U(T, H)/T], \label{eq1}
126: \end{equation}
127: %%%%%%%%%%%%%%%
128: with
129: %%%%%%%%%%%%%%%
130: \begin{equation}
131: U(T,H)=g(H)f(t), \label{eq2}
132: \end{equation}
133: %%%%%%%%%%%%%%%
134: and
135: %%%%%%%%%%%%%%%
136: \begin{equation}
137: f=(1-t)^\beta, \label{eq3}
138: \end{equation}
139: %%%%%%%%%%%%%%%
140: where $\rho_{0f}$ is constant, and $g$ and $\beta$ are magnetic
141: field dependent. Using the progression $(1-t)^\beta= 1-\beta
142: t+\beta(\beta-1)t^2/2!-\beta(\beta-1)(\beta-2)t^3/3!+\ldots$, we
143: have $\ln \rho \approx (\ln \rho_{0f}+
144: g\beta/T_c)-(g/T)[1+\beta(\beta-1)t^2/2!-\beta(\beta-1)(\beta-2)t^3/3!+\ldots]$,
145: where the term $(\ln \rho_{0f}+ g\beta/T_c)\approx \ln \rho_0$ is
146: temperature independent. With $\beta= 1$, we have $\ln
147: \rho_0\approx \ln \rho_{0f}+ g/T_c$. However, we find if $\beta$
148: largely deviates from $\beta= 1$, the relation of $\ln
149: \rho_0\approx \ln \rho_{0f}+ g\beta/T_c$ will bring in large
150: uncertainty and errors for determining $\rho_{0f}$ and $T_c$ in
151: the relation of $\ln \rho_0= \ln \rho_{0f}+ g\beta/T_c$ plot (see Ref.\cite
152: {Zhang}) , as the value of the local slope $-\partial \ln
153: \rho(T,H)/\partial T^{-1}$ largely changes from one local
154: temperature to the other as shown in Fig.~\ref{f1}(b). This means
155: that we shall determine the $\rho_{0f}$ value in the other way
156: when $\beta$ largely deviates from $\beta = 1$.
157:
158: Accordingly, we must determine four free parameters $T_c$,
159: $\rho_{0f}$, $g(H)$, and $\beta(H)$ using Eqs.(\ref{eq1}),
160: (\ref{eq2}), and (\ref{eq3}) with the experimental data.
161: Generally, $T_c$ and $\rho_{0f}$ are magnetic field independent
162: and therefore can be eliminated from the free parameter list in
163: each magnetic field. In simplicity, we follow the selection of
164: Kim \textit{et al.} to take $T_c=131$ K (this will not lead to
165: large errors as the transition width is less than 2 K in zero
166: magnetic field and $T_c$ shall be determined around the transition
167: interval). We consider the resistivity value for $\rho_{0f}$ (by
168: trial and error) which is not only getting better regression for
169: each $\rho(T,H)$ curve in TAFF region, but also is in good
170: agreement with each $-\partial \ln \rho(T,H)/\partial T^{-1}$.
171: Hence, this will leave only two free parameters $g$ and $\beta$
172: for each magnetic field. At first, we use the formula
173: %%%%%%%%%%%%%%%
174: %%%%%%%%%%%%%%%
175: \begin{equation}
176: U(T,H)\approx T\ln [\rho_{0f}/\rho(T,H)] \label{eq4}
177: \end{equation}
178: %%%%%%%%%%%%%%%
179: with $\rho(T,H)$ data to determine the parameters, and then check
180: the parameters with corresponding regressions for the curves in
181: Figs.~\ref{f1}(a) and (b). The gray solid lines in
182: ~Figs.~\ref{f1}(a) and (b) show the results which we take
183: $\rho_{0f}=8$ $\mu\Omega$ cm for regressions. One will find that
184: the regressions are in good agreement with the data in the
185: temperature interval where the resistivity roughly changing from
186: $10^{-2}$ to $1$ $\mu\Omega$ cm. Note that the interval matches
187: the temperature interval suggested by Palstra \textit{et al}.
188: \cite{Palstra}.
189:
190:
191:
192: In fact, a $\rho_{0f}$ value can be selected in a broad range without changing the
193: consistent matches of all the fits for $U(T,H)$ and $\rho(T,H)$ curves in TAFF region. However, changing
194: $\rho_{0f}$ value will lead to changes of fitting parameters and
195: still results in uncertainty in analysis. This means that we can
196: not decide the $\rho_{0f}$ value which is only good for the fits
197: of $U$ or $\rho$ curves, and we have to check the results with
198: $-\partial \ln \rho(T,H)/\partial T^{-1}$ relation. We find that
199: the value of $\rho_{0f}$ will change the consistent matches of the fits for
200: $-\partial \ln \rho(T,H)/\partial T^{-1}$ curves in TAFF region.
201:
202: %%%%%%%%%%%%%%%
203: \begin{figure}
204: \includegraphics
205: [width= .80\columnwidth] {fig2} \caption{\label{f2} The gray lines
206: in (a) and (b) are our regressions using (a)$\rho_{0f}=6$
207: $\mu\Omega$ cm, and (b)$\rho_{0f}=10$ $\mu\Omega$ cm for
208: $-\partial \ln \rho(T,H)/\partial T_{ff}^{-1}$ curves.}
209: \end{figure}
210: %%%%%%%%%%%%%%%
211: %%%%%%%%%%%%%%%
212:
213:
214: Figures~\ref{f2}(a) and (b) show the fitting results of $-\partial
215: \ln \rho(T,H)/\partial T^{-1}$ with $\rho_{0f}=6$ $\mu\Omega$ cm
216: and 10 $\mu\Omega$ cm, respectively. The fits are not good in
217: Fig.~\ref{f2}(a) for low fields and in Fig.~\ref{f2}(b) for high
218: fields, respectively, while they are in good agreement with
219: experimental data in other field ranges. The analysis suggests
220: that we have $\rho_{0f}= 8.0\pm 1.5$ $\mu\Omega$ cm for getting
221: better fits for all $-\partial \ln \rho(T,H)/\partial T^{-1}$
222: curves. The deviations between the regressions and
223: experimental data in low temperature range are possibly due to
224: competing relations between coupling and decoupling and between
225: pinning and depinning.
226:
227:
228: Figures~\ref{f3}(a) and (b) represent $g(H)$ and $\beta(H)$ data
229: for $\rho_{0f}=6$, 7, 8, 9, 10 $\mu\Omega$ cm, respectively. With
230: $\rho_{0f}=8$ $\mu\Omega$ cm, we find that $g\sim H^{-1.94}$ and
231: $\beta\approx 3$ for $1\le \mu_0H\le 3$ T, while $g\sim
232: H^{-0.73}$ and $\beta$ approximately linear increases with $H$ for
233: $5\le \mu_0H\le 9$ T. The characteristic changes of $g$ and
234: $\beta$ around $\mu_0H=4$ T are probably due to the crossover from
235: 3D to 2D as discussed in Ref.~\cite{Zhang}.
236:
237: %%%%%%%%%%%%%%%
238: \begin{figure}
239: \includegraphics
240: [width= .80\columnwidth] {fig3} \caption{\label{f3} (a)$g(H)$ and
241: (b)$\beta(H)$ determined with different $\rho_{0f}$ values.}
242: \end{figure}
243: %%%%%%%%%%%%%%%
244: %%%%%%%%%%%%%%%
245:
246:
247: In summary, we suggest that the analysis of activation energies
248: and definitions of the four different vortex regions are
249: incorrect in Ref.~\cite{Kim}, as the TAFF temperature interval
250: presented in it does not match the interval suggested by Palstra
251: \textit{et al.} \cite{Palstra}. Using the temperature interval
252: suggested by Palstra \textit{et al.}, we find that the empirical
253: activation energy form suggested by Zhang \textit{et al.}
254: \cite{Zhang} is in good agreement with the lower resistivity
255: data.
256:
257: This work has been financially supported by the National Science
258: Foundation of China (Grant No. 10174091).
259:
260:
261: \begin{thebibliography}{}
262: \bibitem{Kim} W. S. Kim, W. N. Kang, M. S. Kim, and S. I. Lee, Phys. Rev. B \textbf{61}, 11317 (2000).
263: \bibitem{Zhang} Y. Z. Zhang, Z. Wang, X. F. Lu, H. H. Wen, J. F. de Marneffe, R. Deltour, A. G. M. Jansen, and P. Wyder, Phys. Rev. B \textbf{71}, 052502 (2005).
264: \bibitem{Palstra}T. T. M. Palstra, B. Batlogg, R. B. van Dover, I. F. Schneemeyer, and J. V. Waszczak, Phys. Rev. B \textbf{41}, 6621 (1990).
265: \bibitem{Geshkenbein}V. B. Geshkenbein, M. V. Feigel\rq man, A. I. Larkin, and V. M. Vinokur, Physica C \textbf{162}, 239 (1989).
266: \bibitem{Vinokur}V. M. Vinokur, M. V. Feigel'man, V. B. Geshkenbein, and A. I. Larkin, Phys. Rev. Lett. \textbf{65}, 259 (1990); J. Kierfeld, H. Nordborg, and V. M. Vinokur, Phys. Rev. Lett. \textbf{85}, 4948 (2000).
267: \bibitem{Blatter} G. Blatter, M. V. Feigel\rq man, V. B. Geshkenbein, A. I. Larkin, and V. M. Vinokur, \rmp \textbf{66}, 1125 (1994).
268: \bibitem{Brandt}E. H. Brandt, Rep. Prog. Phys. {\bf 58},
269: 1465 (1995).
270: \bibitem{Cohen} L. F. Cohen and H. J. Jensen, Rep. Prog. Phys. {\bf 60}, 1581 (1997).
271:
272: \end{thebibliography}
273:
274:
275:
276: \end{document}
277: