- Email: [email protected]
FTIR study of cerium doped electron superconductors R1.85Ce0.15CuO4 (R=Nd, Pr, Sm, Eu and Gd)
Talanta 53 (2001) 733 – 739 www.elsevier.com/locate/talanta
FTIR study of cerium doped electron superconductors R1.85Ce0.15CuO4 (R =Nd, Pr, Sm, Eu and Gd) R. Kannan a, S. Mohan b,* a
Department of Physics, Pondicherry Engineering College, Pondicherry 605 014, India b Department of Physics, Pondicherry Uni6ersity, Pondicherry 605 014, India
Received 10 August 1999; received in revised form 20 June 2000; accepted 5 July 2000
Abstract Systematic FTIR study of electron doped compounds of the T%-series are presented in this paper. Spectroscopic techniques are more sensitive than any other technique to the disorder due to doping in the structure. Polycrystalline ‘as prepared’ bulk samples, allows one to observe c-axis vibrations clearly, avoiding in-plane (a – b) polarizability due to charge carriers. Even though all R2CuO4(R =Nd, Pr, Sm, Eu and Gd) doped compounds belong to T% structure, due to the different sizes of the rare earth ionic radii, the structure stabilization is very difficult to achieve since copper–oxygen distance vary from compound to compound and hence the phonon characteristic spectrum of the series. The role of apical oxygen in T-structure is discussed in detail from the point of factor group analysis and compared with the V-structure. The IR study of R2CuO4 (214) compound with Ce doping reveals the difference and similarities of the electronic properties between hole doping and electron doping when compared with T-structure compounds. © 2001 Elsevier Science B.V. All rights reserved. Keywords: Electron doped superconductors; FTIR spectra; R1.85Ce0.15CuO4 (R =Nd, Pr, Sm, Eu and Gd); Phonon properties
1. Introduction Tokura et al. [1] were the first to reveal that the charge reservoir or block layers above and below the CuO2 sheet are not only able to accept electrons from the CuO2 sheet but can also donate electrons and with these mobile electrons the com
Paper presented at the Colloquium Spectroscopicum Internationale XXXI, September 5–10, 1999, Ankara, Turkey. * Corresponding author. Tel.: +91-413-655382; fax: +91413-655265. E-mail address: s – [email protected] (S. Mohan).
pound Ln2CuO4 becomes superconductor. Superconducting properties of these T% compounds were elaborately studied by Maple et al. [2–4]. The charge balance [5] also indicates that electrons are the charge carriers in Ln1.85Ce0.15 CuO4 rather than holes as in the case of La1.85(Sr, Ba)0.15CuO4. It has been proposed that the dopability of R2CuO4 (214) to produce T or T%structure with holes or electrons depends on the presence or absence of apical oxygen atom near the CuO2 layers. For hole doping, the negatively charged apical oxygen atoms make these layers more negatively charged and hence favours hole
0039-9140/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S0039-9140(00)00497-5
734
R. Kannan, S. Mohan / Talanta 53 (2001) 733–739
doping, while the opposite is true in the absence of apical oxygen atoms for the T%-structure. However, this does not apply to Y-123 or bismuth system which also possesses apical oxygen. Muller et al. [6] attempted to correlate the ratio of apex oxygen to the total number of oxygen, with superconductivity. When compared to T-structure compounds, both have single isolated Cu – O planes perpendicular to the c-axis and in both structures the La/R atoms are along the c-axis directly above and below the copper atoms. The difference between these structures is the arrangement of additional oxygen atoms at (4d-D2d) (T%-structure) and (4e-C4v) (T-structure) positions, respectively. XRD measurements reveal that the crystal structure and lattice parameters were poorly affected [7] for the ‘as prepared (R1.85 Ce0.15 CuO4) material’ as well as for the reduced compound (R1.85 Ce0.15 CuO4 − y ). It is yet to be observed, whether the atomic structure shows any change upon the onset of superconductivity in assessing the role of the lattice in the pairing mechanism. While the lattice constant indicates only a very small change near Tc, the local atomic or electronic structure appears to undergo some significant changes. Weber et al. [8] reported that the Cu–O plane atoms are major contributors to the high frequency vibration modes in La2 − x (Ba, Sr)x CuO4. Whatever the mechanism proposed for the high temperature superconductivity, the role of oxygen can not be denied. Abnormal behavior is commonly found not only in transport and magnetic properties of high T superconductors, but also in the frequency and linewidth of phonons around the superconducting transition. Hence the study of doped 214 compounds, is considered important since this is the only system in which electronic structure, lattice dynamics, isotope effect and superconductivity were intrinsically connected with each other. From the point of factor group analysis of T-structure and T%structure compounds, it was found that the copper or the oxygen atoms in the Cu – O sheet do not participate in the Raman active [9] vibrations. Hence it will be appropriate to study the IR vibrations of these series, with reference to oxygen vibrations in the ungerade mode.
When Ln2CuO4 changes from an insulator to conductor and superconductor by Ce doping, the optical spectra also changes by the transition. The vibrational modes are strongly affected by the holes/electrons introduced into the CuO planes. The qualitative symmetry in doping of CuO2 planes with electrons or holes in the same parent compound, which is a main discriminating factor, is expected to throw some light on the mechanism responsible for high temperature superconductivity. In light of this, we present a systematic spectroscopic investigation of these series. It is essential to understand their unique electronic structure in the normal state, by spectroscopic investigation, because the superconductivity is obtained when the electronic structure is slightly modified. The far IR spectra is mainly used for the determination of energy gap of the superconductors. As the low energy peaks involve vibrating large metal ions and are independent of the oxygen coordination number on the in-plane copper atoms, we chose to record mid IR spectra of the series. Low energy elementary excitations in solids using IR reveal a highly rich phonon structure. As the possible contribution from phonons for weak interactions is not yet ruled out for the mechanism of superconductivity, the study of phonon properties are considered important.
2. Experimental The preparation of single phase compounds of these perovskite related structures is a delicate process in comparison with the other high Tc oxides since structural stability and chemical dopability depends upon the ionic radii of the dopant. Xue et al. [10] have established that there must be proper matching between block layer and charge reservoir layers. In La2CuO4, the R–O bond is under tension while the Cu–O bond is under great compression. The length of the R–O bond can be increased through proper doping and that of Cu–O can be shortened to produce more stable T% structure from T structure (La2CuO4) compound. Hence ionic radii of Nd (0.983 A°), Pr (1.09°), Sm (0.965 A°), Eu (0.947 A°) and Gd (0.938 A°) less than that of La (1.15 A°) produces
R. Kannan, S. Mohan / Talanta 53 (2001) 733–739
stable T% structure even upon doping/partial replacement of R3 + by tetravalent Ce4 + impurity. As the Cu– O bond is under compression, beyond R =Gd, T%-phase is no longer stable since R–O bond length and Cu – O bond length do not match each other. When R varies from bigger Pr, Nd to the smaller Eu and Gd, the Cu – O bond changes from the being stretched to compressed indicating a reduction in the unit cell volume. The samples were prepared by solid state reaction technique [11]. High purity (Cerac, USA)
Fig. 1. Powder XRD pattern of R1.85Ce0.15CuO4 (R =Nd, Pr, Sm, Eu and Gd).
735
oxides were ground thoroughly and calcined in air at 950°C for 12 h. The resulting powders were reground and pressed into pellets and fired in the air at 1050°C for 15 h. The preparation of single phase materials of T%-structure is very difficult in comparison, with the other high To oxides since uniform distribution of Ce in the bulk sample is very difficult. Repeated grinding to achieve required particle size, consequently larger diffusion path length and homogeneous distribution of the components are needed to obtain the desired phase. Care was taken to avoid impurity phases like unreacted CeO2 and CuO by choosing a low annealing temperature, since solubility of Ce is a function of temperature [12]. All the five compounds of the series were prepared under identical heating conditions to study the effect of ambient conditions of the FTIR spectra. All the ceramic bulk samples prepared by solid state reaction were characterised by XRD spectra (Fig. 1) using copper Ka radiation (Rich Seifert, Germany). Tetragonal lattice cell parameters were refined using Autox [13] programme and they are presented in Table 1 and the values are in good agreement with the available literature for the ‘as prepared samples’ [14]. Following Lopez et al. [5], lattice parameters were also calculated for characteristic reflection planes 103 and 110. Tetravalent Ce impurity doping in 214 produces these splitting and the lattice parameters calculated for these characteristic reflection planes 103 and 110 were given in Table 2. The ‘c’ lattice parameter is found to be rapidly decreasing with respect to decreasing ionic radii (da/dr = 2.1) and the ‘a’ lattice parameter is less sensitive (da/dr =0.392) to substitution is found (Fig. 2). The c/a ratio for this series is found to be 3.033–3.084 similar to Y-123 series (3.055). Substitution of smaller ionic radii like Ce4 + suggests shortening of the Cu–O bond length which induces lattice distortion in the end member of the series resulting in pseudo T-structure due to buckling of Cu–O layer. Mid IR measurements were carried out on the polycrystalline samples at room temperature. Sintered pellets were reground and mixed with KBr, then pressed into pellets. Transmission mid IR spectra was recorded for the samples using Bruker IFS 66v between 400 and 4000 cm − 1. The IR
736
R. Kannan, S. Mohan / Talanta 53 (2001) 733–739
3. Results and discussion
Fig. 2. Change in lattice parameters a, c and unit cell volume on rare earth dependence.
spectrum of pure KBr pellet of the same size was recorded and then used to eliminate the background from the spectra of the mixed sample. The transmission minima which represents the absorption maxima correspond to phonon frequencies [15] are tabulated in Table 3. Moreover, analytical absorption peaks identified with a species are concentration dependent, hence the relative concentration of the peaks are estimated from the IR transmission spectra.
IR spectra measurements of the series are depicted in Fig. 3 which exhibit distinct structure is attributed to the presence of single phase formation of the compounds. The characteristic change and the dependence of the phonon frequency on the Cu–O bond length and ionic radii are discussed. The characteristic phonon frequency occurring at 5059 10 cm − 1 is found to be decreasing from Gd to Pr indicating a decreasing trend in the stiffness constant as the bond length increases except Nd (non-typical T% behaviour, due to the appearance of the of the peak at 683 cm − 1) and Gd ( the average absorbance is not between 0.2 and 0.5). Clear-cut and unambiguous assignments are not possible in the case of IR spectra, some general points were made and it is compared with the well known T-structural compounds. The phonon frequencies at and around 5059 10 cm − 1 is found to be common in all the five members of T% phase and it is attributed to the similarity of the structure of the different members of the same phase. Nd2CuO4 when doped with Ce4 + impurity reveals more and strong phonon features. The spectra is divided into three broad peaks, each asymmetric peak having largest multiplicity. Phonon features are observed at 406, 418 and 431 while the peak at 438 cm − 1 appears as side band of a main peak. Peak 2 comprises of the wavenumbers ranging from 458 to 506 cm − 1 which includes characteristic peak of the series. Another broad band ( peak 3) composed of several peaks is detected in the region 539–567 cm − 1 omitting fundamentals and overtones. Weak and broad features, non-typical of T%-structure is also observed at 673 and 692 cm − 1, respectively. The appearance of the broad peak at 683 cm − 1 (average of 673 and 692 cm − 1) indicates T type behaviour. Singh and Ganguly [16] have assigned the IR bands at 685 and 515 cm − 1 to copper– oxygen stretching vibrations of La2CuO4. However, the peak occurring at 683 cm − 1 is not found in Pr, Sm, Eu and Gd. Mortimer et at. [17] also reported the presence of relatively two narrow bands at 685 and 515 cm − 1 in La2CuO4 and the pressure studies indicate that peak at 515 cm − 1 is
R. Kannan, S. Mohan / Talanta 53 (2001) 733–739
insensitive whereas the peak at 685 cm − 1 is found to be changing with pressure and disappears under high pressure. The peak at 683 cm − 1 appears
737
only in Nd and not in the compounds with Pr, Sm, Eu and Gd. This indicates the formation of T% structure in the Pr, Sm, Eu and Gd but not in
Fig. 3. FTIR spectra of R1.85Ce0.15CuO4 (R =Nd, Pr, Sm, Eu and Gd).
738
R. Kannan, S. Mohan / Talanta 53 (2001) 733–739
Nd. The appearance of the peak at 683 cm − 1 may be due to disorder caused in the oxygen position, which is of non-typical T%-phase. The pronounced, one of the multiplicity occurring at 506 cm − 1 is found to be the characteristic phonon mode of all the five members of the family. However, the peaks at and around 505 cm − 1 can be ascribed to the phonons associated with the displacement of oxygen in the Cu – O plane. This vibration frequency when compared with other Cu – O vibration is found to be different due to the presence of extra positive charge of Ce4 + , together with smaller ionic radius (0.978 A°) reducing the compression in the R – O bound and thus by shortening of the Cu – O bond, produces an inward relaxation of oxygen surrounding. Pr2CuO4 when doped with Ce4 + exhibits a few IR phonon structures at 400, 456, 467, 508 and 584 cm − 1, respectively. The sharp infrared transition at 508 cm − 1 is the characteristic member of the series. The absorbance of the peak is found to 0.355. Due to large ionic radius (1.09 A°) of Pr, low temperature annealing results in uniform solubility of Ce over the bulk volume indicating the pure phase formation. Similar phonon structure was found in SM1.85Ce0.15CuO4 and EU1.85Ce0.15CuO4 with remarkable differences in the transmittance of the pronounced phonon structure at 523 and 517 cm − 1, respectively. The IR spectra of Gd1.85Ce0.15CuO4 − y reveals a very rich IR phonon structure at 412, 434, 453, 482, 508, 524, 547, 578, 610, and 633 cm − 1, respectively. The characteristic peak appears as a broad band with largest multiplicity at 508, 524, and 547 cm − 1 having an average absorbance 0.726. Since the peak absorbance is not lying between 0.2 and 0.5, the contribution of energy by the detector is almost nil. It is very difficult to assign the behaviour of the high energy portion of the peaks which contain unresolved side bands. This being the end member of the series and beyond which no stable T%-structure could be formed ensuring a bond matching between R–O bond and Cu – O bond, much attention is paid to the spectroscopic information of the compound. But the observed spectra for Gd, seems to be a noise pattern rather than the transmitted intensity pattern. This is because T% phase is stabilised over
a very narrow compositional range and also the structural distortion involving the displacement of oxygen atom increases for the end member of the series. The strong IR transition peak occurring at 5059 10 cm − 1 is found uniformly in the compounds with Pr, Sm and Eu and whereas the compounds with Nd and Gd are found to be deviating from the general trend. The amplitude of the peak identified with a species is supposed to reflect the hypothetical carrier concentration of the species to which it belongs. The ratio of amplitude of the peak occurring at 505 9 10 cm − 1, is found to be increasing from Sm, Pr to Eu omitting the members which do not have sharp peak or absorbance between 0. 2 and 0. 5. The factor group analysis yields one A2u phonon, two Eu phonons and one silent B2u phonon due to the oxygen atoms in the CuO plane same as that of T%-structure compounds. The difference between these structures is the arrangement of the additional oxygen atoms Op (4d-D2d) and Oz (4e-C4v) respectively. In the case of T% structure the oxygen atom at the 4d position, which is not chemically bonded to the copper atom gives rise to both an A2u and Eu phonon, making the total contribution same as 2A2u + B2u + 3Eu, for both T structure and T% structure in ungerade mode. Even though the contribution of oxygen atoms is the same, due to different bond length and oxygen co-ordination number, the observed IR spectra for the T%-structure is different. The phonon with similar Eigen vectors in La2CuO4 compared to R2CuO4 (R=Nd Pr, Sm, Eu and Gd) will have large energies due to large lattice parameters (a, b, c) values and unit cell volume. This could be the reason for the shifting of the peaks in the range 445–450 cm − 1 in T structure to 460–470 cm − 1 in T% structure. Also shifting of the peak at 490 cm − 1 in T structure to 505910 cm − 1 in T% structure, could be due to short Cu–O distance in the T%-structure due to doping. In T%-structure, the peak at 680 cm − 1 is found to be missing.. When compared to T%-structure, the peak at 680 cm − 1 mode appears as zone centre phonon and that becomes a zone boundary phonon when the symmetry is raised to tetragonal
R. Kannan, S. Mohan / Talanta 53 (2001) 733–739
by doping [18]. The phonon mode at and around 5059 10 cm − 1 arising due to c axis vibration of copper and oxygen atoms is found to be common for both T and T%-structures.
4. Conclusion FTIR spectra of the series of Cerium doped R2CuO4 are analysed based on the oxygen related vibrations of Cu – O planes in the high frequency region The similarity of the structure of the series is exhibited by the commonly found peaks occurring at 505 910 cm − 1 which is attributed to the in plane oxygen displacement of Cu – O plane. The intensity of these peak are found to be increasing from Sm, Pr to Eu. Similarity in the frequency of modes indicates that its configuration is the same despite differences in intensity. The peak absorbance between 0.2 and 0.5 is considered, otherwise it may due to very little contribution of energy by detector. The change in the Cu – O bond length is exhibited by the change in the plane stretching frequency mode. This stretching frequency of the series is found to be increasing for the end member of the series. IR spectra indicate dramatic changes in intensity as a consequence of change in carrier concentration, due to the buckling of Cu–O planes, as some of the plane oxygen atoms take up T-phase apical oxygen position, which in turn related to the differences in the ionic radii and the non matching of R – O and Cu – O bonds. Out of five member of the series, Pr, Sm, and Eu shows extraordinarily similar phonon structure in the IR spectra with remarkable differences in the transmittance of the pronounced phonon structure at 505910 cm − 1. This peak is considered very important, since even under very high pressure conditions it does not show any change [17] which is common to both T and T% structure. When compared with T structure, the peak at 680 cm − 1 disappears for Pr, Nd, Sm, Eu and Gd. However, in our spectra the compound with Nd shows a very strong band at 683 cm − 1. This may be due to the displacement of oxygen, in the verge of a structural transition since superconductivity also occurs for these series, in a very narrow compositional range. Besides, the inhomogeneous distribution of Ce4 +
739
over the bulk volume and the presence of impurity phases also, may not be ruled out. Thus a very simple study of IR transmission spectra of T% compounds reveal, the structural disorder due to doping. Acknowledgements The authors gratefully acknowledges the help offered by Professor U.V. Varadaraju and M.S. Ramachandra Rao, of M.S.R.C, I I T, Madras, without which this work would not have been possible. References [1] Y. Tokura, H. Takagi, S. Uchida, Nature 337 (1989) 345. [2] J.T. Markert, M.B. Maple, Solid State Commun. 70 (1989) 145. [3] J.T. Markert, E.A. Early, T. Bjomholm, S. Gharnaty, B.W. Lee, J.J. Neumeier, R.D. Price, C.L. Seaman, M.B. Maple, Phys. C 158 (1989) 178. [4] E.A. Early, N.Y. Ayoub, J. Beille, J.T. Markert, M.B. Maple, Phys. C 160 (1989) 320. [5] M.E. Lopez-Morales, R.J. Savoy, P.M. Grant, Solid State Commun. 71 (1989) 1079. [6] A. Muller, Z. Phys. B 80 (1990) 193. [7] M.L. Sanjuan, M.A. Laguna, S. Pinol, P. Canfield, Z. Fisk, Phys. Rev. B 46 (1992) 8683. [8] W. Weber, Phys. Rev. Lett. 58 (1987) 1371. [9] M.K. Crawford, G. Burns, G.V. Chandrashekhar, F.H. Dacol, W.E. Fameth, E.M. McCarron, III, R.T. Smalley, Phys. Rev. B 41 (1990) 8933. [10] Y.Y. Xue, P.H. Hor, R.L. Meng, Y.K. Tao, Y.Y. Sun, Z.J. Huang, L. Gao, C.W. Chu, Phys. C 165 (1990) 357. [11] V. Radhakrishnan, C.K. Subramanian, V. Sankaranarayanan, G.V. Subba Rao, R. Srinivasan, Phys. C 167 (1990) 53. [12] J.M. Tarascon, E. Wang, L.H. Greene, B.G. Bagley, G.W. Hull, S.M. D’Egidio, P.T. Mceli, Z.Z. Wang, T.W. Jing, J. Clayhold, D. Brawner, N.P. Ong, Phys. Rev. B 40 (1989) 4494. [13] V.B. Zlokazov Mriaau, J. Appl. Cryst. 25 (1992) 69. [14] J.L. Peng, R.N. Shelton, H.B. Radousky, Phys. Rev. B 41 (1990) 187. [15] G. Burus, F.H. Dacol, P. Freitas, T.S. Plaskett, W. Konig, Solid State Commun. 64 (1987) 471. [16] K.K. Singh, P. Ganguly, Spectrochim. Acta A 40 (1984) 539. [17] R. Mortimer, J.G. Powell, N.Y. Vasnthacharya, J. Phys. 9 (1997) 11209. [18] V.A. Maroni, T.O. Brun, M. Grimsditch, K.C. Loong, Phys. Rev. B 39 (1989) 4127.
FTIR study of cerium doped electron superconductors R1.85Ce0.15CuO4 (R =Nd, Pr, Sm, Eu and Gd) R. Kannan a, S. Mohan b,* a
Department of Physics, Pondicherry Engineering College, Pondicherry 605 014, India b Department of Physics, Pondicherry Uni6ersity, Pondicherry 605 014, India
Received 10 August 1999; received in revised form 20 June 2000; accepted 5 July 2000
Abstract Systematic FTIR study of electron doped compounds of the T%-series are presented in this paper. Spectroscopic techniques are more sensitive than any other technique to the disorder due to doping in the structure. Polycrystalline ‘as prepared’ bulk samples, allows one to observe c-axis vibrations clearly, avoiding in-plane (a – b) polarizability due to charge carriers. Even though all R2CuO4(R =Nd, Pr, Sm, Eu and Gd) doped compounds belong to T% structure, due to the different sizes of the rare earth ionic radii, the structure stabilization is very difficult to achieve since copper–oxygen distance vary from compound to compound and hence the phonon characteristic spectrum of the series. The role of apical oxygen in T-structure is discussed in detail from the point of factor group analysis and compared with the V-structure. The IR study of R2CuO4 (214) compound with Ce doping reveals the difference and similarities of the electronic properties between hole doping and electron doping when compared with T-structure compounds. © 2001 Elsevier Science B.V. All rights reserved. Keywords: Electron doped superconductors; FTIR spectra; R1.85Ce0.15CuO4 (R =Nd, Pr, Sm, Eu and Gd); Phonon properties
1. Introduction Tokura et al. [1] were the first to reveal that the charge reservoir or block layers above and below the CuO2 sheet are not only able to accept electrons from the CuO2 sheet but can also donate electrons and with these mobile electrons the com
Paper presented at the Colloquium Spectroscopicum Internationale XXXI, September 5–10, 1999, Ankara, Turkey. * Corresponding author. Tel.: +91-413-655382; fax: +91413-655265. E-mail address: s – [email protected] (S. Mohan).
pound Ln2CuO4 becomes superconductor. Superconducting properties of these T% compounds were elaborately studied by Maple et al. [2–4]. The charge balance [5] also indicates that electrons are the charge carriers in Ln1.85Ce0.15 CuO4 rather than holes as in the case of La1.85(Sr, Ba)0.15CuO4. It has been proposed that the dopability of R2CuO4 (214) to produce T or T%structure with holes or electrons depends on the presence or absence of apical oxygen atom near the CuO2 layers. For hole doping, the negatively charged apical oxygen atoms make these layers more negatively charged and hence favours hole
0039-9140/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S0039-9140(00)00497-5
734
R. Kannan, S. Mohan / Talanta 53 (2001) 733–739
doping, while the opposite is true in the absence of apical oxygen atoms for the T%-structure. However, this does not apply to Y-123 or bismuth system which also possesses apical oxygen. Muller et al. [6] attempted to correlate the ratio of apex oxygen to the total number of oxygen, with superconductivity. When compared to T-structure compounds, both have single isolated Cu – O planes perpendicular to the c-axis and in both structures the La/R atoms are along the c-axis directly above and below the copper atoms. The difference between these structures is the arrangement of additional oxygen atoms at (4d-D2d) (T%-structure) and (4e-C4v) (T-structure) positions, respectively. XRD measurements reveal that the crystal structure and lattice parameters were poorly affected [7] for the ‘as prepared (R1.85 Ce0.15 CuO4) material’ as well as for the reduced compound (R1.85 Ce0.15 CuO4 − y ). It is yet to be observed, whether the atomic structure shows any change upon the onset of superconductivity in assessing the role of the lattice in the pairing mechanism. While the lattice constant indicates only a very small change near Tc, the local atomic or electronic structure appears to undergo some significant changes. Weber et al. [8] reported that the Cu–O plane atoms are major contributors to the high frequency vibration modes in La2 − x (Ba, Sr)x CuO4. Whatever the mechanism proposed for the high temperature superconductivity, the role of oxygen can not be denied. Abnormal behavior is commonly found not only in transport and magnetic properties of high T superconductors, but also in the frequency and linewidth of phonons around the superconducting transition. Hence the study of doped 214 compounds, is considered important since this is the only system in which electronic structure, lattice dynamics, isotope effect and superconductivity were intrinsically connected with each other. From the point of factor group analysis of T-structure and T%structure compounds, it was found that the copper or the oxygen atoms in the Cu – O sheet do not participate in the Raman active [9] vibrations. Hence it will be appropriate to study the IR vibrations of these series, with reference to oxygen vibrations in the ungerade mode.
When Ln2CuO4 changes from an insulator to conductor and superconductor by Ce doping, the optical spectra also changes by the transition. The vibrational modes are strongly affected by the holes/electrons introduced into the CuO planes. The qualitative symmetry in doping of CuO2 planes with electrons or holes in the same parent compound, which is a main discriminating factor, is expected to throw some light on the mechanism responsible for high temperature superconductivity. In light of this, we present a systematic spectroscopic investigation of these series. It is essential to understand their unique electronic structure in the normal state, by spectroscopic investigation, because the superconductivity is obtained when the electronic structure is slightly modified. The far IR spectra is mainly used for the determination of energy gap of the superconductors. As the low energy peaks involve vibrating large metal ions and are independent of the oxygen coordination number on the in-plane copper atoms, we chose to record mid IR spectra of the series. Low energy elementary excitations in solids using IR reveal a highly rich phonon structure. As the possible contribution from phonons for weak interactions is not yet ruled out for the mechanism of superconductivity, the study of phonon properties are considered important.
2. Experimental The preparation of single phase compounds of these perovskite related structures is a delicate process in comparison with the other high Tc oxides since structural stability and chemical dopability depends upon the ionic radii of the dopant. Xue et al. [10] have established that there must be proper matching between block layer and charge reservoir layers. In La2CuO4, the R–O bond is under tension while the Cu–O bond is under great compression. The length of the R–O bond can be increased through proper doping and that of Cu–O can be shortened to produce more stable T% structure from T structure (La2CuO4) compound. Hence ionic radii of Nd (0.983 A°), Pr (1.09°), Sm (0.965 A°), Eu (0.947 A°) and Gd (0.938 A°) less than that of La (1.15 A°) produces
R. Kannan, S. Mohan / Talanta 53 (2001) 733–739
stable T% structure even upon doping/partial replacement of R3 + by tetravalent Ce4 + impurity. As the Cu– O bond is under compression, beyond R =Gd, T%-phase is no longer stable since R–O bond length and Cu – O bond length do not match each other. When R varies from bigger Pr, Nd to the smaller Eu and Gd, the Cu – O bond changes from the being stretched to compressed indicating a reduction in the unit cell volume. The samples were prepared by solid state reaction technique [11]. High purity (Cerac, USA)
Fig. 1. Powder XRD pattern of R1.85Ce0.15CuO4 (R =Nd, Pr, Sm, Eu and Gd).
735
oxides were ground thoroughly and calcined in air at 950°C for 12 h. The resulting powders were reground and pressed into pellets and fired in the air at 1050°C for 15 h. The preparation of single phase materials of T%-structure is very difficult in comparison, with the other high To oxides since uniform distribution of Ce in the bulk sample is very difficult. Repeated grinding to achieve required particle size, consequently larger diffusion path length and homogeneous distribution of the components are needed to obtain the desired phase. Care was taken to avoid impurity phases like unreacted CeO2 and CuO by choosing a low annealing temperature, since solubility of Ce is a function of temperature [12]. All the five compounds of the series were prepared under identical heating conditions to study the effect of ambient conditions of the FTIR spectra. All the ceramic bulk samples prepared by solid state reaction were characterised by XRD spectra (Fig. 1) using copper Ka radiation (Rich Seifert, Germany). Tetragonal lattice cell parameters were refined using Autox [13] programme and they are presented in Table 1 and the values are in good agreement with the available literature for the ‘as prepared samples’ [14]. Following Lopez et al. [5], lattice parameters were also calculated for characteristic reflection planes 103 and 110. Tetravalent Ce impurity doping in 214 produces these splitting and the lattice parameters calculated for these characteristic reflection planes 103 and 110 were given in Table 2. The ‘c’ lattice parameter is found to be rapidly decreasing with respect to decreasing ionic radii (da/dr = 2.1) and the ‘a’ lattice parameter is less sensitive (da/dr =0.392) to substitution is found (Fig. 2). The c/a ratio for this series is found to be 3.033–3.084 similar to Y-123 series (3.055). Substitution of smaller ionic radii like Ce4 + suggests shortening of the Cu–O bond length which induces lattice distortion in the end member of the series resulting in pseudo T-structure due to buckling of Cu–O layer. Mid IR measurements were carried out on the polycrystalline samples at room temperature. Sintered pellets were reground and mixed with KBr, then pressed into pellets. Transmission mid IR spectra was recorded for the samples using Bruker IFS 66v between 400 and 4000 cm − 1. The IR
736
R. Kannan, S. Mohan / Talanta 53 (2001) 733–739
3. Results and discussion
Fig. 2. Change in lattice parameters a, c and unit cell volume on rare earth dependence.
spectrum of pure KBr pellet of the same size was recorded and then used to eliminate the background from the spectra of the mixed sample. The transmission minima which represents the absorption maxima correspond to phonon frequencies [15] are tabulated in Table 3. Moreover, analytical absorption peaks identified with a species are concentration dependent, hence the relative concentration of the peaks are estimated from the IR transmission spectra.
IR spectra measurements of the series are depicted in Fig. 3 which exhibit distinct structure is attributed to the presence of single phase formation of the compounds. The characteristic change and the dependence of the phonon frequency on the Cu–O bond length and ionic radii are discussed. The characteristic phonon frequency occurring at 5059 10 cm − 1 is found to be decreasing from Gd to Pr indicating a decreasing trend in the stiffness constant as the bond length increases except Nd (non-typical T% behaviour, due to the appearance of the of the peak at 683 cm − 1) and Gd ( the average absorbance is not between 0.2 and 0.5). Clear-cut and unambiguous assignments are not possible in the case of IR spectra, some general points were made and it is compared with the well known T-structural compounds. The phonon frequencies at and around 5059 10 cm − 1 is found to be common in all the five members of T% phase and it is attributed to the similarity of the structure of the different members of the same phase. Nd2CuO4 when doped with Ce4 + impurity reveals more and strong phonon features. The spectra is divided into three broad peaks, each asymmetric peak having largest multiplicity. Phonon features are observed at 406, 418 and 431 while the peak at 438 cm − 1 appears as side band of a main peak. Peak 2 comprises of the wavenumbers ranging from 458 to 506 cm − 1 which includes characteristic peak of the series. Another broad band ( peak 3) composed of several peaks is detected in the region 539–567 cm − 1 omitting fundamentals and overtones. Weak and broad features, non-typical of T%-structure is also observed at 673 and 692 cm − 1, respectively. The appearance of the broad peak at 683 cm − 1 (average of 673 and 692 cm − 1) indicates T type behaviour. Singh and Ganguly [16] have assigned the IR bands at 685 and 515 cm − 1 to copper– oxygen stretching vibrations of La2CuO4. However, the peak occurring at 683 cm − 1 is not found in Pr, Sm, Eu and Gd. Mortimer et at. [17] also reported the presence of relatively two narrow bands at 685 and 515 cm − 1 in La2CuO4 and the pressure studies indicate that peak at 515 cm − 1 is
R. Kannan, S. Mohan / Talanta 53 (2001) 733–739
insensitive whereas the peak at 685 cm − 1 is found to be changing with pressure and disappears under high pressure. The peak at 683 cm − 1 appears
737
only in Nd and not in the compounds with Pr, Sm, Eu and Gd. This indicates the formation of T% structure in the Pr, Sm, Eu and Gd but not in
Fig. 3. FTIR spectra of R1.85Ce0.15CuO4 (R =Nd, Pr, Sm, Eu and Gd).
738
R. Kannan, S. Mohan / Talanta 53 (2001) 733–739
Nd. The appearance of the peak at 683 cm − 1 may be due to disorder caused in the oxygen position, which is of non-typical T%-phase. The pronounced, one of the multiplicity occurring at 506 cm − 1 is found to be the characteristic phonon mode of all the five members of the family. However, the peaks at and around 505 cm − 1 can be ascribed to the phonons associated with the displacement of oxygen in the Cu – O plane. This vibration frequency when compared with other Cu – O vibration is found to be different due to the presence of extra positive charge of Ce4 + , together with smaller ionic radius (0.978 A°) reducing the compression in the R – O bound and thus by shortening of the Cu – O bond, produces an inward relaxation of oxygen surrounding. Pr2CuO4 when doped with Ce4 + exhibits a few IR phonon structures at 400, 456, 467, 508 and 584 cm − 1, respectively. The sharp infrared transition at 508 cm − 1 is the characteristic member of the series. The absorbance of the peak is found to 0.355. Due to large ionic radius (1.09 A°) of Pr, low temperature annealing results in uniform solubility of Ce over the bulk volume indicating the pure phase formation. Similar phonon structure was found in SM1.85Ce0.15CuO4 and EU1.85Ce0.15CuO4 with remarkable differences in the transmittance of the pronounced phonon structure at 523 and 517 cm − 1, respectively. The IR spectra of Gd1.85Ce0.15CuO4 − y reveals a very rich IR phonon structure at 412, 434, 453, 482, 508, 524, 547, 578, 610, and 633 cm − 1, respectively. The characteristic peak appears as a broad band with largest multiplicity at 508, 524, and 547 cm − 1 having an average absorbance 0.726. Since the peak absorbance is not lying between 0.2 and 0.5, the contribution of energy by the detector is almost nil. It is very difficult to assign the behaviour of the high energy portion of the peaks which contain unresolved side bands. This being the end member of the series and beyond which no stable T%-structure could be formed ensuring a bond matching between R–O bond and Cu – O bond, much attention is paid to the spectroscopic information of the compound. But the observed spectra for Gd, seems to be a noise pattern rather than the transmitted intensity pattern. This is because T% phase is stabilised over
a very narrow compositional range and also the structural distortion involving the displacement of oxygen atom increases for the end member of the series. The strong IR transition peak occurring at 5059 10 cm − 1 is found uniformly in the compounds with Pr, Sm and Eu and whereas the compounds with Nd and Gd are found to be deviating from the general trend. The amplitude of the peak identified with a species is supposed to reflect the hypothetical carrier concentration of the species to which it belongs. The ratio of amplitude of the peak occurring at 505 9 10 cm − 1, is found to be increasing from Sm, Pr to Eu omitting the members which do not have sharp peak or absorbance between 0. 2 and 0. 5. The factor group analysis yields one A2u phonon, two Eu phonons and one silent B2u phonon due to the oxygen atoms in the CuO plane same as that of T%-structure compounds. The difference between these structures is the arrangement of the additional oxygen atoms Op (4d-D2d) and Oz (4e-C4v) respectively. In the case of T% structure the oxygen atom at the 4d position, which is not chemically bonded to the copper atom gives rise to both an A2u and Eu phonon, making the total contribution same as 2A2u + B2u + 3Eu, for both T structure and T% structure in ungerade mode. Even though the contribution of oxygen atoms is the same, due to different bond length and oxygen co-ordination number, the observed IR spectra for the T%-structure is different. The phonon with similar Eigen vectors in La2CuO4 compared to R2CuO4 (R=Nd Pr, Sm, Eu and Gd) will have large energies due to large lattice parameters (a, b, c) values and unit cell volume. This could be the reason for the shifting of the peaks in the range 445–450 cm − 1 in T structure to 460–470 cm − 1 in T% structure. Also shifting of the peak at 490 cm − 1 in T structure to 505910 cm − 1 in T% structure, could be due to short Cu–O distance in the T%-structure due to doping. In T%-structure, the peak at 680 cm − 1 is found to be missing.. When compared to T%-structure, the peak at 680 cm − 1 mode appears as zone centre phonon and that becomes a zone boundary phonon when the symmetry is raised to tetragonal
R. Kannan, S. Mohan / Talanta 53 (2001) 733–739
by doping [18]. The phonon mode at and around 5059 10 cm − 1 arising due to c axis vibration of copper and oxygen atoms is found to be common for both T and T%-structures.
4. Conclusion FTIR spectra of the series of Cerium doped R2CuO4 are analysed based on the oxygen related vibrations of Cu – O planes in the high frequency region The similarity of the structure of the series is exhibited by the commonly found peaks occurring at 505 910 cm − 1 which is attributed to the in plane oxygen displacement of Cu – O plane. The intensity of these peak are found to be increasing from Sm, Pr to Eu. Similarity in the frequency of modes indicates that its configuration is the same despite differences in intensity. The peak absorbance between 0.2 and 0.5 is considered, otherwise it may due to very little contribution of energy by detector. The change in the Cu – O bond length is exhibited by the change in the plane stretching frequency mode. This stretching frequency of the series is found to be increasing for the end member of the series. IR spectra indicate dramatic changes in intensity as a consequence of change in carrier concentration, due to the buckling of Cu–O planes, as some of the plane oxygen atoms take up T-phase apical oxygen position, which in turn related to the differences in the ionic radii and the non matching of R – O and Cu – O bonds. Out of five member of the series, Pr, Sm, and Eu shows extraordinarily similar phonon structure in the IR spectra with remarkable differences in the transmittance of the pronounced phonon structure at 505910 cm − 1. This peak is considered very important, since even under very high pressure conditions it does not show any change [17] which is common to both T and T% structure. When compared with T structure, the peak at 680 cm − 1 disappears for Pr, Nd, Sm, Eu and Gd. However, in our spectra the compound with Nd shows a very strong band at 683 cm − 1. This may be due to the displacement of oxygen, in the verge of a structural transition since superconductivity also occurs for these series, in a very narrow compositional range. Besides, the inhomogeneous distribution of Ce4 +
739
over the bulk volume and the presence of impurity phases also, may not be ruled out. Thus a very simple study of IR transmission spectra of T% compounds reveal, the structural disorder due to doping. Acknowledgements The authors gratefully acknowledges the help offered by Professor U.V. Varadaraju and M.S. Ramachandra Rao, of M.S.R.C, I I T, Madras, without which this work would not have been possible. References [1] Y. Tokura, H. Takagi, S. Uchida, Nature 337 (1989) 345. [2] J.T. Markert, M.B. Maple, Solid State Commun. 70 (1989) 145. [3] J.T. Markert, E.A. Early, T. Bjomholm, S. Gharnaty, B.W. Lee, J.J. Neumeier, R.D. Price, C.L. Seaman, M.B. Maple, Phys. C 158 (1989) 178. [4] E.A. Early, N.Y. Ayoub, J. Beille, J.T. Markert, M.B. Maple, Phys. C 160 (1989) 320. [5] M.E. Lopez-Morales, R.J. Savoy, P.M. Grant, Solid State Commun. 71 (1989) 1079. [6] A. Muller, Z. Phys. B 80 (1990) 193. [7] M.L. Sanjuan, M.A. Laguna, S. Pinol, P. Canfield, Z. Fisk, Phys. Rev. B 46 (1992) 8683. [8] W. Weber, Phys. Rev. Lett. 58 (1987) 1371. [9] M.K. Crawford, G. Burns, G.V. Chandrashekhar, F.H. Dacol, W.E. Fameth, E.M. McCarron, III, R.T. Smalley, Phys. Rev. B 41 (1990) 8933. [10] Y.Y. Xue, P.H. Hor, R.L. Meng, Y.K. Tao, Y.Y. Sun, Z.J. Huang, L. Gao, C.W. Chu, Phys. C 165 (1990) 357. [11] V. Radhakrishnan, C.K. Subramanian, V. Sankaranarayanan, G.V. Subba Rao, R. Srinivasan, Phys. C 167 (1990) 53. [12] J.M. Tarascon, E. Wang, L.H. Greene, B.G. Bagley, G.W. Hull, S.M. D’Egidio, P.T. Mceli, Z.Z. Wang, T.W. Jing, J. Clayhold, D. Brawner, N.P. Ong, Phys. Rev. B 40 (1989) 4494. [13] V.B. Zlokazov Mriaau, J. Appl. Cryst. 25 (1992) 69. [14] J.L. Peng, R.N. Shelton, H.B. Radousky, Phys. Rev. B 41 (1990) 187. [15] G. Burus, F.H. Dacol, P. Freitas, T.S. Plaskett, W. Konig, Solid State Commun. 64 (1987) 471. [16] K.K. Singh, P. Ganguly, Spectrochim. Acta A 40 (1984) 539. [17] R. Mortimer, J.G. Powell, N.Y. Vasnthacharya, J. Phys. 9 (1997) 11209. [18] V.A. Maroni, T.O. Brun, M. Grimsditch, K.C. Loong, Phys. Rev. B 39 (1989) 4127.