Magnetic properties of the ferrimagnetic cobaltite CaBaCo4O7
aa r X i v : . [ c ond - m a t . s t r- e l ] A p r Magnetic properties of the ferrimagnetic cobaltiteCaBaCo O a, ∗ , Langsheng Ling a , Lei Zhang a , Li Pi b,a , Yuheng Zhang a,b a High Magnetic Field Laboratory, Chinese Academy of Sciences,Hefei, Anhui, 230031, China b Hefei National Laboratory for Physical Sciences at the Microscale,University of Science and Technology of China, Hefei, Anhui, 230026, China
Abstract
The magnetic properties of the ferrimagnetic cobaltite CaBaCo O are sys-tematically investigated. We find that the susceptibility exhibits a downwarddeviation below ∼
360 K, suggesting the occurrence of short range magneticcorrelations at temperature well above T C . The effective moment is de-termined to be 4.5 µ B /f.u, which is consistent with that expected for theCo /Co high spin species. Using a criterion given by Banerjee [Phys.Lett. , 16 (1964)], we demonstrate that the paramagnetic to ferrimagnetictransition in CaBaCo O has a first order character. Keywords:
A. magnetically ordered materials, D. phase transitions
1. Introduction
Transition metal oxides with geometry frustration have attracted consid-erable interest over decades. [1, 2, 3, 4] They commonly exhibit the per-sistence of strong spin fluctuations at low temperatures. As a result, thelong-range magnetic order is at least partially suppressed and various shortrange correlated phases such as spin liquid, spin glass or spin ice develop.In some cases, frustration can be partially or entirely released, either bystructural distortions that lift the ground-state degeneracy, or by the ”order-by-disorder” mechanism, [5] resulting in the establishment of a long-rangemagnetic order. ∗ Corresponding author. Tel: +86-551-559-5640; Fax: +86-551-559-1149.
Email address: [email protected] (Zhe Qu)
Preprint submitted to Solid State Communications October 11, 2018 wo well-known structural topology causing the presence of geometryfrustration are two-dimensional triangular lattice and two-dimensional kagomelattice. Compositions whose structural motif embraces triangular or kagomelayers are of great interest as model systems and have been the focus ofnumerous studies. In this respect, the recently discovered ”114” cobaltiteCaBaCo O [6, 7] is particularly interesting because its crystal structure isbuilt up of an alternate stacking of triangular or kagome layers formed by theCoO tetrahedra (shown in the inset to Fig. 1). There is very large distortionin the crystal, characterized by a strong buckling of the kagome layers. [6, 7]In addition, it exhibits charge ordering, with Co sitting on two sites and”mixed valent” cobalt Co /Co L sitting on two other sites. [7] Due to thelarge structural distortion and the charge ordering, the geometry frustrationis lifted, resulting in a ferrimagnetic ground state at low temperatures. [6, 7]Although significant progress has been made in understanding the mag-netic properties in CaBaCo O , a few questions remain to be answered. Forexample, does the system shows short-range magnetic correlations above T C like their ”114” cousins such as YBaCo O ? [8, 9] Why the obtained effec-tive moment differs significantly from the expected value in CaBaCo O ? [6]What’s the nature of the paramagnetic to ferrimagnetic transition?To address these questions, we systematically measured the magneticproperties of CaBaCo O . It is found that the susceptibility exhibits andownward deviation below ∼
360 K, suggesting the occurrence of short rangemagnetic correlations at temperature well above T C . By extending the mag-netization measurement up to 800 K, the effective moment is determined tobe 4.5 µ B /f.u through a Curie-Weiss analysis, which is consistent with thatexpected for the Co /Co high spin species. Using a criterion given byBanerjee, [10] we demonstrate that the paramagnetic to ferrimagnetic phasetransition in CaBaCo O is a first order one.
2. Experiment
Polycrystalline sample of CaBaCo O was prepared by using the con-ventional solid-state reaction method described in Ref. [6]. Stoichiometricproportions of high purity CaCO , BaCO and Co O were mixed and heatedat 900 o C in air to decarbonation. They are then pelletized, and then sin-tered at 1100 o C in air for 12 hours and quenched to room temperature. Thestructure and the phase purity of the samples were checked by powder X-raydiffraction (XRD) at room temperature. Magnetization measurements were2erformed with a commercial superconducting quantum interference device(SQUID) magnetometer (Quantum Design MPMS 7T-XL) and a PhysicalProperty Measurement System (Quantum Design PPMS-16T) equipped witha vibrating sample magnetometer (VSM).
3. Results and Discussion
Figure 1 displays the powder XRD pattern of CaBaCo O at room tem-perature. Rietveld refinement [11, 12] of the XRD pattern confirms that thesample is single phase with an orthorhombic structure ( P bn space group).The lattice parameters are determined to be a = 6.2871 ˚ A , b = 11.0106˚ A and c = 10.1945 ˚ A , which agree well with previous reports within theexperimental error. [6, 7]The temperature dependence of the magnetization M ( T ) between 2 Kand 400 K under 0.1 T is shown in the upper panel of Fig. 2. They are mea-sured during field cooling sequence (FCC), during warming after field coolingsequence (FCW) and during warming after zero field cooling sequence (ZFC),respectively. All curve shows a rapid increase below ∼
70 K, suggesting theoccurrence of the transition into a magnetically ordered state. At 5 K, thesaturated magnetic moment is still relatively small, only ∼ µ B /f.u. under14 T (see the lower panel of Fig. 2), agreeing with a ferrimagnetic groundstate. The Curie temperature, defined as the temperature corresponding tothe maximum in the dM/dT curve, is determined to be 60 K (see the inset toFig. 2). Below T C , we observe significant irreversibility between the magne-tization curve measured after ZFC and FCC histories. This is attributed tothe large coercive field compared to the applied field. [6, 13] As shown in thelower panel of Fig. 2, the coercive field of CaBaCo O is about 2 T at 5 K,which is much larger than the applied field of 0.1 T. Therefore, the magneticdomains will be locked in random direction during ZFC sequence while bealigned to the same direction during FCC or FCW sequences, resulting in thelarge irreversibility below T C . All these results are consistent with previousreport, [6] confirming that our sample is of high quality.A close look on the temperature dependence of the magnetization revealsmore information. The inset to Fig. 3 displays the enlarged view of the M ( T )curves between T C and 400 K. One can see that while the magnetizationdecreases with increasing temperature above T C the slope of the M ( T ) curvedoes not decrease monotonously as that expected for a pure paramagneticstate where the Curie-Weiss law predicts χ ∝ C/ ( T − T CW ) (Here C is3he Curie constant and T CW is the Curie-Weiss temperature). In order tounderstand this behavior, we extend the measurement of the M ( T ) duringFCC sequence up to 800 K and perform the Curie-Weiss analysis. As shownin Fig. 3, 1/ χ shows an upward deviation from the linearity below ∼
360 K.Since the deviation temperature is much higher than the Curie temperature,this behavior could not be attributed to the critical enhancement of thespin fluctuations due to the approach to the paramagnetic to ferrimagnetictransition but suggests the occurrence of short range magnetic correlations.The Curie-Weiss temperature T CW is determined to be ∼ -890 K. This gives f = T CW /T C ∼ . O is strongly frustrated. Theeffective moment is determined to be ∼ µ B /f.u. , which agrees well withthe value of a 1:1 combination of Co /Co high spin species expected basedon the chemical formula. The Curie-Weiss temperature and the effectivemoment obtained here are different from previous report. [6] This should beunderstood because short-range magnetic correlations might appears in theirfitting temperature region.In order to obtain further information on the paramagnetic to ferrimag-netic transition in CaBaCo O , we use the criteria proposed by Banerjee todetermine the order of this transition. By considering the essential similaritybetween the Landau-Lifshitz [14] and Bean-Rodbell [15] criteria, Banerjeeshows that the slope of the H/M versus M curves near the critical tem-perature can distinguish the first-order magnetic transition from the secondorder ones: a negative slope means the former and a positive slope means thelatter. [10] We then measured the initial isothermal magnetization curves attemperatures in the vicinity of the Curie temperature. Before each run, thesample is warmed up to 200 K and then cooled to the measuring temperatureunder zero field to ensure a perfect demagnetization of the sample. The dataare summarized in the inset to Fig. 4. It is noted that the M ( H ) curve ex-hibits a peculiar behavior that its slope shows a decrease before an increase atintermediate fields. This behavior was also observed in MnAs, where a firstorder transition occurs at its Curie temperature and is used to test Banerjee’scriteria. [10, 15] We replotted the M ( H ) curves as H/M vs.M in Fig. 4.Negative slope is clearly observed between 64 and 70 K, which confirms thatthe paramagnetic to ferrimagnetic transition occurred in CaBaCo O has afirst order character according to the criterion.4 . Conclusion In conclusion, we systematically investigate the magnetic properties ofCaBaCo O . The Curie-Weiss temperature is determined to be ∼ -890 K andthe effective moment be ∼ µ B /f.u. . The susceptibility shows downwarddeviation from the Curie-Weiss law below ∼
360 K, hinting that short rangemagnetic correlations might occur at temperature much higher than T C =60 K. The paramagnetic to ferrimagnetic transition in CaBaCo O is foundto have the first order character.
5. Acknowledgments
This work is financially supported by the National Key Basic Research ofChina under Grant No. 2007CB925001 and 2010CB923403, and by NationalNatural Science Foundation of China under contract No. 11004198.
References [1] A. P. Ramirez, Annu. Rev. Mater. Sci. , (1994) 453.[2] P. Schiffer and A. P. Ramirez, Comments Condens. Matter Phys. ,(1996) 21.[3] J. E. Greedan, J. Alloys Compd. , (2006) 444.[4] R. Moessner and A. P. Ramirez, Phys. Today
59 (2) , (2006) 24.[5] J. Villain, R. Bidaux, J. P. Carton, and R. Conte, J. Phys. (France) ,(1980) 1263.[6] V. Caignaert, V. Pralong, A. Maignan, and B. Raveau, Solid State Com-mun. , (2009) 453.[7] V. Caignaert, V. Pralong, V. Hardy, C. Ritter, and B. Raveau, Phys.Rev. B , (2010) 094417.[8] M. Soda, Y. Yasui, T. Moyoshi, M. Sato, N. Igawa, and K. Kakurai, J.Phys. Soc. Jpn. , (2006) 054707.[9] P. Manuel, L. C. Chapon, P. G. Radaelli, H. Zheng, and J. F. Mitchell,Phys. Rev. Lett. , (2009) 054707.510] Phys. Lett. , (1964) 16; see also, for example, S. V. Vonsovskii, Mag-netism (Wiley, New York, 1974), Vol. 2, Chap. 25.[11] A. C. Larson and R. B. Von Dreele,
General Structure Analysis System(GSAS) (Los Alamos National Laboratory Report LAUR 86-748, 2000).[12] B. H. Toby, J. Appl. Cryst. , 2001 (210).[13] Z. R. Yang, S. Tan, and Y. H. Zhang, Appl. Phys. Lett. , (2001) 3645.[14] L. D. Landau, Zh. Eksp. Teor. Fiz. , (1937) 19; ibid. , (1937) 627;ibid.E. M. Lifshitz, , (1941) 269; ibid.V. L. Ginzburg, , (1947) 833;S. V. Vonsovskii, Izv. Akad. Nauk SSSR, Ser. Fiz. , (1947) 485.[15] C. P. Bean and D. S. Rodbell, Phys. Rev. , (1962) 104.6 I n t e n s it y ( a r b . un it s ) q (degree) BaCaK CoT Co ac
Figure 1: (Color online) Powder XRD patterns of CaBaCo O . The solid curve is thebest fit from the Rietveld refinement using GSAS, with R p = 11.68% and R wp = 10.08%.The vertical marks indicate the position of Bragg peaks and the bottom curves show thedifference between the observed and calculated intensities. Inset shows the structure ofCaBaCo O viewed along b axis. K Co and T Co represent kagome layer and triangularlayer of CoO tetrahedra, respectively.
100 200 3000.00.20.40.60.8 -10 -5 0 5 10-1.0-0.50.00.51.0 H = 0.1 T ZFC FCC FCW M ( m b / f . u . ) T (K)
40 60 80 100 d M / d T T (K) M ( m b / f . u . ) H (T) T = 5 K Figure 2: (Color online) Upper panel: The magnetization as function of the temperatureunder an applied field of 0.1 T. inset shows the dM/dT as function of the temperature.Lower panel: the magnetization as function of the field measured at 5 K.
200 400 600 80010 -2 -1 / c ( T f . u . / m B ) c ( m B / f . u . T ) T (K) 0204060 H = 0.1T
100 200 300 4002.02.53.03.5 M ( - m b / f . u . ) T (K) Figure 3: (Color online) The susceptibility and the reciprocal of the susceptibility asfunction of temperature between 2 and 800 K under 0.1 T measured in FCC sequence.The solid lines represent the Curie-Weiss fitting. Inset shows the enlarged view of the M ( T ) curve to highlight the deviation from the Curie-Weiss fitting. H / M ( T f . u . / m B ) M ( m B2 /f.u. )70 K
70 K50 K M ( m B / f . u . ) H (T) Figure 4: Inset shows the initial isothermal magnetization curves at temperatures in thevicinity of the Curie temperature T C = 60 K at an interval of 1 K. The main panel showsthese curves replotted as H/M vs.M ..