- Email: [email protected]
Electron spin resonance study of Fe doping effect in La0.67Ca0.33MnO3
ARTICLE IN PRESS Journal of Magnetism and Magnetic Materials 321 (2009) 1159–1162
Contents lists available at ScienceDirect
Journal of Magnetism and Magnetic Materials journal homepage: www.elsevier.com/locate/jmmm
Electron spin resonance study of Fe doping effect in La0.67Ca0.33MnO3 Wei Ning, Xiang-Qun Zhang, Zhao-Hua Cheng, Young Sun Beijing National Laboratory for Condensed Matter Physics, Institute of Physics, Chinese Academy of Sciences, Beijing 100080, PR China
a r t i c l e in fo
abstract
Article history: Received 4 March 2008 Received in revised form 28 September 2008 Available online 14 November 2008
Electron spin resonance (ESR) study was carried out on La0.67Ca0.33Mn1xFexO3 (x ¼ 0.0, 0.04) samples. The temperature dependence of the ESR spectra indicates the presence of phase separation above and below TC in x ¼ 0.0 and 0.04 sample, respectively. The increase of the g-value in the high-temperature region indicates the existence of local spin correlations even in the paramagnetic state. The activation energy obtained from both the temperature dependence of the ESR intensity and linewidth exhibits a smaller value in the Fe-doped sample. Our study suggests that the ferromagnetic spin correlations would be significantly weakened by a slight doping of Fe ions on Mn sites. & 2008 Elsevier B.V. All rights reserved.
Keywords: Electron spin resonance Manganite Doping effect
1. Introduction
2. Experiments
The magnetic and resistive behavior in the colossal magnetoresistance manganites is well known to be sensitive to the doping on Mn sites [1,2]. A number of studies have been made on the effects of the replacement of Mn by various transition elements, such as Mg [3], Cu [4], Co [5], Cr [6], Fe [7,8], etc. Attention has been paid especially to the doping of Fe because Mn3+ and Fe3+ have similar ionic radius and only minor lattice distortions are expected by the substitution. It has been generally observed that the Curie temperature TC decreases with increasing Fe doping level [8]. Especially, Ogale et al. [9] found that the Fe-doped samples showed a marked decrease in TC at the 4% doping level and attributed this to the average Fe–Fe separation approaching the size of the charge carriers (polarons) at this concentration. Electron spin resonance (ESR) is a powerful tool for the study of magnetic correlation in manganites. Valuable information can be obtained regarding the interplay of different interactions through a study of the temperature dependence of various ESR parameters, such as the resonance field, g-value, peak-to-peak linewidth and resonance intensity. A number of ESR studies have been carried out on some manganites. Although several ESR studies on La0.67Ca0.33MnO3 have been reported [10], there have been no detailed ESR studies on the Fe-doping effect. Here we report the ESR study of La0.67Ca0.33Mn1xFexO3 (x ¼ 0.0, 0.04) at temperatures between 100 and 400 K. The analysis and comparison of the ESR spectra of both samples clarify the effect of Fe doping.
The ESR experiments were carried out on a JEOL JES-FA200 ESR spectrometer at X-band frequencies (vE9.4 GHz) with temperature range from 100 to 400 K. The samples used for the experiment were single crystals prepared by the floating-zone technique in an optical image four-mirror furnace. The magnetization was measured using a superconducting quantum interference device magnetometer (Quantum Design, MPMS-7). The magnetizations show sharp changes at TC ¼ 218 and 214 K, for x ¼ 0.0 and 0.04, respectively, as shown in the inset of Figs. 2(b) and 3(b), being consistent with the reported results. In order to eliminate the demagnetization effect and internal field, which could distort the ESR spectra as temperature approaches TC, we grind the single-crystal samples into powders and measured the ESR on the powder samples.
Corresponding author.
E-mail address: [email protected] (Y. Sun). 0304-8853/$ - see front matter & 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.jmmm.2008.10.045
3. Results and discussion Fig. 1 shows the temperature dependence of the ESR spectra for x ¼ 0.0 (left panel) and x ¼ 0.04 (right panel). For the x ¼ 0.0 sample, the ESR spectra consist of a single paramagnetic resonance (PMR) line at high temperature. Below 250 K, a ferromagnetic resonance (FMR) line appears at the low-field side and shifts to lower fields with further cooling. Below Curie temperature TC220 K, the PMR signal becomes very weak and undetectable. The coexistence of FMR and PMR signals between 220 and 250 K suggests a phase separation in the vicinity of TC, in which the FM phase and PM phase coexist. This result is in agreement with the neuron scattering result of Teresa et al. [11]. Their small-angle neutron scattering experiments on La0.67Ca0.33MnO3 demonstrated that small FM metallic clusters start to form in the paramagnetic insulating matrix slightly above TC.
ARTICLE IN PRESS 1160
W. Ning et al. / Journal of Magnetism and Magnetic Materials 321 (2009) 1159–1162
400K
400K
298K
250K
ESR signal (arb.units)
251K
190K
242K
180K 170K
232K
163K 221K
148K
199K
124K
150K 0
2
4
6 0 H (kOe)
2
4
6
Fig. 1. Temperature dependence of the ESR spectra: (a) x ¼ 0.0 sample and (b) x ¼ 0.04 sample. The arrows mark the ferromagnetic resonance signals.
(1)
where DHpp is the peak-to-peak linewidth, Hr is the resonance field and A is the area under the absorption curve. The temperature dependence of various ESR parameters for both samples is shown in Figs. 2 and 3. Figs. 2(a) and 3(a) show the temperature dependence of g-value for both samples. In the high-temperature region, the g-value obtained from the resonance field Hr shows weak temperature dependence. As temperature decreases close to TC, the g-value increases gradually in the PM state. This indicates that the spin correlations are present even above TC and gradually develops with decreasing temperature. The temperature dependence of ESR intensity obtained through double integration of the spectra is shown in Figs. 2(b) and 3(b). For both samples, the intensity initially increases slowly with decreasing temperature, and then increases rapidly as the temperature approaches TC. This behavior is qualitatively similar to the susceptibility data.
g-value
2.02
2.00
M (emu/g)
5.0
I (arb.units)
! DHpp DHpp dP d ¼ A þ , dH dH DH2pp þ ðx Hr Þ2 DH2pp þ ðx þ Hr Þ2
2.04
2.5
0.0 0
100
200 T (K)
300
0.9
ΔHpp (kOe)
For the Fe-doped sample, the spectra show a single Lorentzian resonance line above 180 K. The FMR signal appears just below TC174 K, which indicates that the FM clusters do not form above TC. Meanwhile, the PMR signals persist well below TC. The coexistence of FMR and PMR signal also indicates a phase separation in the Fe-doped sample. However, in contrast to La0.67Ca0.33MnO3, where the phase separation occurs above TC, the phase separation in La0.67Ca0.33Mn0.96Fe0.04O3 occurs mainly below TC in the ferromagnetic state. On comparing the spectra for x ¼ 0.0 and 0.04, we found that the doping of Fe decreases the TC and lowers the phase separation temperature range. It indicates the doping of Fe ions disrupts the FM correlation above TC and it is consistent with the general trend in these systems, where doping with Fe tends to result in a weakened double-exchange interaction, thereby lowering TC. In order to obtain more information from the ESR spectra and further investigate the effect of Fe doping in La0.67Ca0.33MnO3, we analyzed the ESR parameters after fitting the spectra by the equation
0.6
0.3 250
300 T(K)
350
400
Fig. 2. Temperature dependence of the ESR spectra parameter for the x ¼ 0.0 sample: (a) g-value, (b) intensity I and (c) linewidth DHpp. The inset of (b) shows the temperature dependence of magnetization for the x ¼ 0.0 sample.
ARTICLE IN PRESS W. Ning et al. / Journal of Magnetism and Magnetic Materials 321 (2009) 1159–1162
2.10
1161
x = 0.0
ln (IPMR) (arb.units)
g-value
x = 0.04 2.05
2.00
I (arb.units)
M (emu/g)
4
H = 200 Oe
Ea (x = 0.0) =121.1 meV Ea (x = 0.04) = 82.5 meV
2 0
3.0 0
100
200 T (K)
4.5
6.0
1000/T (K-1)
300
0.6
Ea (x = 0.0) = 111.7 meV
6.5
Ea (x = 0.04) = 98.0 meV
ln (Hpp)
ΔHpp (kOe)
Fig. 4. ESR intensity plotted as ln(I) vs 1000/T for x ¼ 0.0 (filled circles) and 0.04 (open circles) samples. The solid line represents the best fit to Eq. (2).
0.3 200
300 T (K)
400
Fig. 3. Temperature dependence of the ESR spectra parameter for the x ¼ 0.04 sample: (a) g-value, (b) intensity I and (c) linewidth DHpp. The inset of (b) shows the temperature dependence of magnetization for x ¼ 0.04 sample.
The resonance intensity in the paramagnetic state can be described by the expression I ¼ I0 expðEa =kB TÞ,
(2)
where I0 is a fitting parameter, kB the Boltzmann constant and Ea the activation energy. From the plot of ln(I) vs temperature we obtain, Ea ¼ 121 and 83 meV for x ¼ 0.0 and 0.04, respectively, as shown in Fig. 4. According to the model proposal by Oseroff et al. [12], the origin of the resonance line in manganites is due to a spin complex, which is some form of short-range ordering among the spin-polarized carriers and Ea is the activation energy needed for the dissociation of the clusters. Therefore, the decrease of Ea in the Fe-doped sample suggests that the strength of short-range FM correlations decreases after the Fe doping. Figs. 2(c) and 3(c) show the temperature dependence of the linewidth for both compositions. The linewidth decreases linearly with decreasing temperature and shows a minimum near 250 and 200 K for x ¼ 0.0 and 0.04, respectively, and then increases dramatically on further cooling. The temperature Tmin where the linewidth shows a minimum is about TminE1.15TC. In order to explain the temperature dependence of DHpp above Tmin, we take the model of adiabatic hopping motion of small polarons to fit it. Similar work has been done by Shengelaya et al. [13]. In the PM phase, the behavior of the linewidth DHpp is similar to the electrical conductivity observed in manganites. The conduction is dominated by the adiabatic hopping motion of
6.0
x = 0.0 5.5
x = 0.04
2.5
3.0
3.5
4.0
4.5
5.0
1000/T (K-1) Fig. 5. ESR linewidth plotted as ln(DHPP) vs 1000/T for x ¼ 0.0 (filled circles) and 0.04 (open circles) samples. The solid line represents the best fit to Eq. (3).
small polarons. Thus, the DHpp is given by the equation A T
DHPP ¼ DH0 þ expðEa =kB TÞ,
(3)
where DH0 and A are constants and kB and Ea are the Boltzmann constant and the activation energy for the dissociation of the FM spin clusters, respectively. The following fitting parameters were obtained: Ea ¼ 112 meV for x ¼ 0.0 and Ea ¼ 98 meV for x ¼ 0.04, as shown in Fig. 5. The smaller energy in the Fe-doped sample implies that the FM clusters is smaller after Fe doping. Therefore, the activation energy obtained from both the intensity and linewidth indicates that the doping of Fe ions in La0.67Ca0.33MnO3 suppresses the short-range FM correlations in the PM state. Similar ESR studies have been reported. The early work performed by Rettori et al. [10] has shown that the temperature dependence of linewidth for La0.67Ca0.33MnO3 shows a minimum value at TminE1.25TC. Their results also indicate the existence of spin clusters well above TC. Recently, Zajac et al. [14] have studied the Fe-doped sample with ESR: the temperature dependence of the linewidth exhibited a minimum value at TminE1.25TC. They found that the ferromagnetic metallic clusters play a dominant role in the resonance absorption. Our results are similar with
ARTICLE IN PRESS 1162
W. Ning et al. / Journal of Magnetism and Magnetic Materials 321 (2009) 1159–1162
these reported ones. In addition, there are many Mo¨essbauer studies concerning the iron-doped samples. In these studies [15–20] they found that the paramagnetic contributions persist down to the ferromagnetic metallic regime (77 K) and provides evidence of the two-phase character of the metallic state. These studies also give evidence that spin clusters start to form above TC in the undoped samples and below TC in the doped samples. Our results are similar with these and prove that ESR is a powerful tool in the study of magnetic correlations in perovskite manganites.
4. Conclusions The temperature dependence of the ESR spectra of La0.67 Ca0.33Mn1xFexO3 (x ¼ 0, 0.04) suggests a phase separation in both samples. The g-value increases gradually with decreasing temperature in the paramagnetic state, which indicates the existence of weak spin correlations above TC. The activation energies obtained from the temperature dependence of resonance intensity and linewidth are smaller in the Fe-doped sample, which indicates that the doping of Fe ions weakens the double-exchange interaction and the ferromagnetic correlations. These results prove that ESR is a powerful tool in the study of perovskite manganites.
Acknowledgments This work was supported by the Natural Science Foundation of China and the National Key Basic Research Program of China.
References [1] M. Tokunaga, N. Miura, Y. Tomioka, Y. Tokura, Phys. Rev. B 57 (1998) 5259. [2] Q. Huang, A. Santoro, J.W. Lynn, R.W. Erwin, J.A. Borchers, J.L. Peng, K. Ghosh, R.L. Greene, Phys. Rev. B 58 (1998) 2684. [3] A. Maignan, C. Martin, B. Raveau, Z. Phys. B: Condens. Matter 102 (1997) 19. [4] K. Ghosh, S.B. Ogale, R. Ramesh, R.L. Greene, T. Venkatesan, K.M. Gapchup, Ravi Bathe, S.I. Patil, Phys. Rev. B 59 (1999) 533. [5] Mark Rubinstein, D.J. Gillespie, John E. Snyder, Terry M. Tritt, Phys. Rev. B 56 (1997) 5412. [6] Y. Sun, X. Xu, Y.H. Zhang, Phys. Rev. B 63 (2000) 054404. [7] K.H. Ahn, X.W. Wu, K. Liu, C.L. Chien, Phys. Rev. B 54 (1996) 15299. [8] J.R. Sun, G.H. Rao, B.G. Shen, H.K. Wong, Appl. Phys. Lett. 73 (1998) 2998. [9] S.B. Ogale, R. Shreekala, R. Bathe, S.K. Date, S.I. Patil, B. Hannoyer, F. Petit, G. Marest, Phys. Rev. B 57 (1998) 7841. [10] C. Rettori, D. Rao, J. Singler, D. Kidwell, S.B. Oseroff, M.T. Causa, J.J. Neumeier, K.J. McClellan, S.-W. Cheong, S. Schultz, Phys. Rev. B 55 (2000) 3083. [11] J.M. De Teresa, M.R. Ibarra, P.A. Algarabel, C. Ritter, C. Marquina, J. Blasco, J. Garcia, A. del Moral, Z. Arnold, Nature (London) 386 (1997) 256. [12] S.B. Oseroff, M. Torikachvili, J. Singley, S. Ali, S.-W. Cheong, S. Schultz, Phys. Rev. B 53 (1996) 6521. [13] A. Shengelaya, G.M. Zhao, H. Keller, K.A. Mu¨ller, B.I. Kochelaev, Phys. Rev. B 61 (2000) 5888. [14] T. Zajac, L. Folcik, A. Kolodziejczyk, H. Drulis, K. Krop, G. Gritzner, J. Magn. Magn. Mater. 272 (2004) 120. [15] S.B. Ogale, R. Shreekala, Ravi Bathe, S.K. Date, S.I. Patil, B. Hannoyer, F. Petit, G. Marest, Phys. Rev. B 57 (1998) 7841. [16] A. Tkachuk, K. Rogacki, D.E. Brown, B. Dabrowski, A.J. Fedro, C.W. Kimball, B. Pyles, X. Xiong, Daniel Rosenmann, B.D. Dunlap, Phys. Rev. B 57 (1998) 8509. [17] V. Chechersky, A. Nath, I. Isaac, J.P. Franck, K. Ghosh, H. Ju, R.L. Greene, Phys. Rev. B 59 (1999) 497. [18] V. Chechersky, A. Nath, I. Isaac, J.P. Franck, K. Ghosh, R.L. Greene, J. Phys.: Condens. Matter 11 (1999) 8921. [19] B. Hannoyer, G. Marest, J.M. Greneche, Ravi Bathe, S.I. Patil, S.B. Ogale, Phys. Rev. B 61 (2000) 9613. [20] A. Simopoulos, M. Pissas, G. Kallias, E. Devlin, N. Moutis, I. Panagiotopoulos, D. Niarchos, C. Christides, R. Sonntag, Phys. Rev. B 59 (1999) 1263.
Contents lists available at ScienceDirect
Journal of Magnetism and Magnetic Materials journal homepage: www.elsevier.com/locate/jmmm
Electron spin resonance study of Fe doping effect in La0.67Ca0.33MnO3 Wei Ning, Xiang-Qun Zhang, Zhao-Hua Cheng, Young Sun Beijing National Laboratory for Condensed Matter Physics, Institute of Physics, Chinese Academy of Sciences, Beijing 100080, PR China
a r t i c l e in fo
abstract
Article history: Received 4 March 2008 Received in revised form 28 September 2008 Available online 14 November 2008
Electron spin resonance (ESR) study was carried out on La0.67Ca0.33Mn1xFexO3 (x ¼ 0.0, 0.04) samples. The temperature dependence of the ESR spectra indicates the presence of phase separation above and below TC in x ¼ 0.0 and 0.04 sample, respectively. The increase of the g-value in the high-temperature region indicates the existence of local spin correlations even in the paramagnetic state. The activation energy obtained from both the temperature dependence of the ESR intensity and linewidth exhibits a smaller value in the Fe-doped sample. Our study suggests that the ferromagnetic spin correlations would be significantly weakened by a slight doping of Fe ions on Mn sites. & 2008 Elsevier B.V. All rights reserved.
Keywords: Electron spin resonance Manganite Doping effect
1. Introduction
2. Experiments
The magnetic and resistive behavior in the colossal magnetoresistance manganites is well known to be sensitive to the doping on Mn sites [1,2]. A number of studies have been made on the effects of the replacement of Mn by various transition elements, such as Mg [3], Cu [4], Co [5], Cr [6], Fe [7,8], etc. Attention has been paid especially to the doping of Fe because Mn3+ and Fe3+ have similar ionic radius and only minor lattice distortions are expected by the substitution. It has been generally observed that the Curie temperature TC decreases with increasing Fe doping level [8]. Especially, Ogale et al. [9] found that the Fe-doped samples showed a marked decrease in TC at the 4% doping level and attributed this to the average Fe–Fe separation approaching the size of the charge carriers (polarons) at this concentration. Electron spin resonance (ESR) is a powerful tool for the study of magnetic correlation in manganites. Valuable information can be obtained regarding the interplay of different interactions through a study of the temperature dependence of various ESR parameters, such as the resonance field, g-value, peak-to-peak linewidth and resonance intensity. A number of ESR studies have been carried out on some manganites. Although several ESR studies on La0.67Ca0.33MnO3 have been reported [10], there have been no detailed ESR studies on the Fe-doping effect. Here we report the ESR study of La0.67Ca0.33Mn1xFexO3 (x ¼ 0.0, 0.04) at temperatures between 100 and 400 K. The analysis and comparison of the ESR spectra of both samples clarify the effect of Fe doping.
The ESR experiments were carried out on a JEOL JES-FA200 ESR spectrometer at X-band frequencies (vE9.4 GHz) with temperature range from 100 to 400 K. The samples used for the experiment were single crystals prepared by the floating-zone technique in an optical image four-mirror furnace. The magnetization was measured using a superconducting quantum interference device magnetometer (Quantum Design, MPMS-7). The magnetizations show sharp changes at TC ¼ 218 and 214 K, for x ¼ 0.0 and 0.04, respectively, as shown in the inset of Figs. 2(b) and 3(b), being consistent with the reported results. In order to eliminate the demagnetization effect and internal field, which could distort the ESR spectra as temperature approaches TC, we grind the single-crystal samples into powders and measured the ESR on the powder samples.
Corresponding author.
E-mail address: [email protected] (Y. Sun). 0304-8853/$ - see front matter & 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.jmmm.2008.10.045
3. Results and discussion Fig. 1 shows the temperature dependence of the ESR spectra for x ¼ 0.0 (left panel) and x ¼ 0.04 (right panel). For the x ¼ 0.0 sample, the ESR spectra consist of a single paramagnetic resonance (PMR) line at high temperature. Below 250 K, a ferromagnetic resonance (FMR) line appears at the low-field side and shifts to lower fields with further cooling. Below Curie temperature TC220 K, the PMR signal becomes very weak and undetectable. The coexistence of FMR and PMR signals between 220 and 250 K suggests a phase separation in the vicinity of TC, in which the FM phase and PM phase coexist. This result is in agreement with the neuron scattering result of Teresa et al. [11]. Their small-angle neutron scattering experiments on La0.67Ca0.33MnO3 demonstrated that small FM metallic clusters start to form in the paramagnetic insulating matrix slightly above TC.
ARTICLE IN PRESS 1160
W. Ning et al. / Journal of Magnetism and Magnetic Materials 321 (2009) 1159–1162
400K
400K
298K
250K
ESR signal (arb.units)
251K
190K
242K
180K 170K
232K
163K 221K
148K
199K
124K
150K 0
2
4
6 0 H (kOe)
2
4
6
Fig. 1. Temperature dependence of the ESR spectra: (a) x ¼ 0.0 sample and (b) x ¼ 0.04 sample. The arrows mark the ferromagnetic resonance signals.
(1)
where DHpp is the peak-to-peak linewidth, Hr is the resonance field and A is the area under the absorption curve. The temperature dependence of various ESR parameters for both samples is shown in Figs. 2 and 3. Figs. 2(a) and 3(a) show the temperature dependence of g-value for both samples. In the high-temperature region, the g-value obtained from the resonance field Hr shows weak temperature dependence. As temperature decreases close to TC, the g-value increases gradually in the PM state. This indicates that the spin correlations are present even above TC and gradually develops with decreasing temperature. The temperature dependence of ESR intensity obtained through double integration of the spectra is shown in Figs. 2(b) and 3(b). For both samples, the intensity initially increases slowly with decreasing temperature, and then increases rapidly as the temperature approaches TC. This behavior is qualitatively similar to the susceptibility data.
g-value
2.02
2.00
M (emu/g)
5.0
I (arb.units)
! DHpp DHpp dP d ¼ A þ , dH dH DH2pp þ ðx Hr Þ2 DH2pp þ ðx þ Hr Þ2
2.04
2.5
0.0 0
100
200 T (K)
300
0.9
ΔHpp (kOe)
For the Fe-doped sample, the spectra show a single Lorentzian resonance line above 180 K. The FMR signal appears just below TC174 K, which indicates that the FM clusters do not form above TC. Meanwhile, the PMR signals persist well below TC. The coexistence of FMR and PMR signal also indicates a phase separation in the Fe-doped sample. However, in contrast to La0.67Ca0.33MnO3, where the phase separation occurs above TC, the phase separation in La0.67Ca0.33Mn0.96Fe0.04O3 occurs mainly below TC in the ferromagnetic state. On comparing the spectra for x ¼ 0.0 and 0.04, we found that the doping of Fe decreases the TC and lowers the phase separation temperature range. It indicates the doping of Fe ions disrupts the FM correlation above TC and it is consistent with the general trend in these systems, where doping with Fe tends to result in a weakened double-exchange interaction, thereby lowering TC. In order to obtain more information from the ESR spectra and further investigate the effect of Fe doping in La0.67Ca0.33MnO3, we analyzed the ESR parameters after fitting the spectra by the equation
0.6
0.3 250
300 T(K)
350
400
Fig. 2. Temperature dependence of the ESR spectra parameter for the x ¼ 0.0 sample: (a) g-value, (b) intensity I and (c) linewidth DHpp. The inset of (b) shows the temperature dependence of magnetization for the x ¼ 0.0 sample.
ARTICLE IN PRESS W. Ning et al. / Journal of Magnetism and Magnetic Materials 321 (2009) 1159–1162
2.10
1161
x = 0.0
ln (IPMR) (arb.units)
g-value
x = 0.04 2.05
2.00
I (arb.units)
M (emu/g)
4
H = 200 Oe
Ea (x = 0.0) =121.1 meV Ea (x = 0.04) = 82.5 meV
2 0
3.0 0
100
200 T (K)
4.5
6.0
1000/T (K-1)
300
0.6
Ea (x = 0.0) = 111.7 meV
6.5
Ea (x = 0.04) = 98.0 meV
ln (Hpp)
ΔHpp (kOe)
Fig. 4. ESR intensity plotted as ln(I) vs 1000/T for x ¼ 0.0 (filled circles) and 0.04 (open circles) samples. The solid line represents the best fit to Eq. (2).
0.3 200
300 T (K)
400
Fig. 3. Temperature dependence of the ESR spectra parameter for the x ¼ 0.04 sample: (a) g-value, (b) intensity I and (c) linewidth DHpp. The inset of (b) shows the temperature dependence of magnetization for x ¼ 0.04 sample.
The resonance intensity in the paramagnetic state can be described by the expression I ¼ I0 expðEa =kB TÞ,
(2)
where I0 is a fitting parameter, kB the Boltzmann constant and Ea the activation energy. From the plot of ln(I) vs temperature we obtain, Ea ¼ 121 and 83 meV for x ¼ 0.0 and 0.04, respectively, as shown in Fig. 4. According to the model proposal by Oseroff et al. [12], the origin of the resonance line in manganites is due to a spin complex, which is some form of short-range ordering among the spin-polarized carriers and Ea is the activation energy needed for the dissociation of the clusters. Therefore, the decrease of Ea in the Fe-doped sample suggests that the strength of short-range FM correlations decreases after the Fe doping. Figs. 2(c) and 3(c) show the temperature dependence of the linewidth for both compositions. The linewidth decreases linearly with decreasing temperature and shows a minimum near 250 and 200 K for x ¼ 0.0 and 0.04, respectively, and then increases dramatically on further cooling. The temperature Tmin where the linewidth shows a minimum is about TminE1.15TC. In order to explain the temperature dependence of DHpp above Tmin, we take the model of adiabatic hopping motion of small polarons to fit it. Similar work has been done by Shengelaya et al. [13]. In the PM phase, the behavior of the linewidth DHpp is similar to the electrical conductivity observed in manganites. The conduction is dominated by the adiabatic hopping motion of
6.0
x = 0.0 5.5
x = 0.04
2.5
3.0
3.5
4.0
4.5
5.0
1000/T (K-1) Fig. 5. ESR linewidth plotted as ln(DHPP) vs 1000/T for x ¼ 0.0 (filled circles) and 0.04 (open circles) samples. The solid line represents the best fit to Eq. (3).
small polarons. Thus, the DHpp is given by the equation A T
DHPP ¼ DH0 þ expðEa =kB TÞ,
(3)
where DH0 and A are constants and kB and Ea are the Boltzmann constant and the activation energy for the dissociation of the FM spin clusters, respectively. The following fitting parameters were obtained: Ea ¼ 112 meV for x ¼ 0.0 and Ea ¼ 98 meV for x ¼ 0.04, as shown in Fig. 5. The smaller energy in the Fe-doped sample implies that the FM clusters is smaller after Fe doping. Therefore, the activation energy obtained from both the intensity and linewidth indicates that the doping of Fe ions in La0.67Ca0.33MnO3 suppresses the short-range FM correlations in the PM state. Similar ESR studies have been reported. The early work performed by Rettori et al. [10] has shown that the temperature dependence of linewidth for La0.67Ca0.33MnO3 shows a minimum value at TminE1.25TC. Their results also indicate the existence of spin clusters well above TC. Recently, Zajac et al. [14] have studied the Fe-doped sample with ESR: the temperature dependence of the linewidth exhibited a minimum value at TminE1.25TC. They found that the ferromagnetic metallic clusters play a dominant role in the resonance absorption. Our results are similar with
ARTICLE IN PRESS 1162
W. Ning et al. / Journal of Magnetism and Magnetic Materials 321 (2009) 1159–1162
these reported ones. In addition, there are many Mo¨essbauer studies concerning the iron-doped samples. In these studies [15–20] they found that the paramagnetic contributions persist down to the ferromagnetic metallic regime (77 K) and provides evidence of the two-phase character of the metallic state. These studies also give evidence that spin clusters start to form above TC in the undoped samples and below TC in the doped samples. Our results are similar with these and prove that ESR is a powerful tool in the study of magnetic correlations in perovskite manganites.
4. Conclusions The temperature dependence of the ESR spectra of La0.67 Ca0.33Mn1xFexO3 (x ¼ 0, 0.04) suggests a phase separation in both samples. The g-value increases gradually with decreasing temperature in the paramagnetic state, which indicates the existence of weak spin correlations above TC. The activation energies obtained from the temperature dependence of resonance intensity and linewidth are smaller in the Fe-doped sample, which indicates that the doping of Fe ions weakens the double-exchange interaction and the ferromagnetic correlations. These results prove that ESR is a powerful tool in the study of perovskite manganites.
Acknowledgments This work was supported by the Natural Science Foundation of China and the National Key Basic Research Program of China.
References [1] M. Tokunaga, N. Miura, Y. Tomioka, Y. Tokura, Phys. Rev. B 57 (1998) 5259. [2] Q. Huang, A. Santoro, J.W. Lynn, R.W. Erwin, J.A. Borchers, J.L. Peng, K. Ghosh, R.L. Greene, Phys. Rev. B 58 (1998) 2684. [3] A. Maignan, C. Martin, B. Raveau, Z. Phys. B: Condens. Matter 102 (1997) 19. [4] K. Ghosh, S.B. Ogale, R. Ramesh, R.L. Greene, T. Venkatesan, K.M. Gapchup, Ravi Bathe, S.I. Patil, Phys. Rev. B 59 (1999) 533. [5] Mark Rubinstein, D.J. Gillespie, John E. Snyder, Terry M. Tritt, Phys. Rev. B 56 (1997) 5412. [6] Y. Sun, X. Xu, Y.H. Zhang, Phys. Rev. B 63 (2000) 054404. [7] K.H. Ahn, X.W. Wu, K. Liu, C.L. Chien, Phys. Rev. B 54 (1996) 15299. [8] J.R. Sun, G.H. Rao, B.G. Shen, H.K. Wong, Appl. Phys. Lett. 73 (1998) 2998. [9] S.B. Ogale, R. Shreekala, R. Bathe, S.K. Date, S.I. Patil, B. Hannoyer, F. Petit, G. Marest, Phys. Rev. B 57 (1998) 7841. [10] C. Rettori, D. Rao, J. Singler, D. Kidwell, S.B. Oseroff, M.T. Causa, J.J. Neumeier, K.J. McClellan, S.-W. Cheong, S. Schultz, Phys. Rev. B 55 (2000) 3083. [11] J.M. De Teresa, M.R. Ibarra, P.A. Algarabel, C. Ritter, C. Marquina, J. Blasco, J. Garcia, A. del Moral, Z. Arnold, Nature (London) 386 (1997) 256. [12] S.B. Oseroff, M. Torikachvili, J. Singley, S. Ali, S.-W. Cheong, S. Schultz, Phys. Rev. B 53 (1996) 6521. [13] A. Shengelaya, G.M. Zhao, H. Keller, K.A. Mu¨ller, B.I. Kochelaev, Phys. Rev. B 61 (2000) 5888. [14] T. Zajac, L. Folcik, A. Kolodziejczyk, H. Drulis, K. Krop, G. Gritzner, J. Magn. Magn. Mater. 272 (2004) 120. [15] S.B. Ogale, R. Shreekala, Ravi Bathe, S.K. Date, S.I. Patil, B. Hannoyer, F. Petit, G. Marest, Phys. Rev. B 57 (1998) 7841. [16] A. Tkachuk, K. Rogacki, D.E. Brown, B. Dabrowski, A.J. Fedro, C.W. Kimball, B. Pyles, X. Xiong, Daniel Rosenmann, B.D. Dunlap, Phys. Rev. B 57 (1998) 8509. [17] V. Chechersky, A. Nath, I. Isaac, J.P. Franck, K. Ghosh, H. Ju, R.L. Greene, Phys. Rev. B 59 (1999) 497. [18] V. Chechersky, A. Nath, I. Isaac, J.P. Franck, K. Ghosh, R.L. Greene, J. Phys.: Condens. Matter 11 (1999) 8921. [19] B. Hannoyer, G. Marest, J.M. Greneche, Ravi Bathe, S.I. Patil, S.B. Ogale, Phys. Rev. B 61 (2000) 9613. [20] A. Simopoulos, M. Pissas, G. Kallias, E. Devlin, N. Moutis, I. Panagiotopoulos, D. Niarchos, C. Christides, R. Sonntag, Phys. Rev. B 59 (1999) 1263.