Flux growth of multiferroic Cu3Nb2O8 single crystals

Flux growth of multiferroic Cu3Nb2O8 single crystals

Accepted Manuscript Flux growth of multiferroic Cu3Nb2O8 single crystals R. Chen, M.M. Shi, Y.J. Liu, C.B. Liu, H.P. Zhu, C. Dong, Y. Liu, J. Shi, Z.C...

796KB Sizes 8 Downloads 86 Views

Accepted Manuscript Flux growth of multiferroic Cu3Nb2O8 single crystals R. Chen, M.M. Shi, Y.J. Liu, C.B. Liu, H.P. Zhu, C. Dong, Y. Liu, J. Shi, Z.C. Xia, Z.W. Ouyang, J.F. Wang PII: DOI: Reference:

S0022-0248(17)30449-9 http://dx.doi.org/10.1016/j.jcrysgro.2017.07.004 CRYS 24235

To appear in:

Journal of Crystal Growth

Received Date: Revised Date: Accepted Date:

14 April 2017 3 July 2017 5 July 2017

Please cite this article as: R. Chen, M.M. Shi, Y.J. Liu, C.B. Liu, H.P. Zhu, C. Dong, Y. Liu, J. Shi, Z.C. Xia, Z.W. Ouyang, J.F. Wang, Flux growth of multiferroic Cu3Nb2O8 single crystals, Journal of Crystal Growth (2017), doi: http://dx.doi.org/10.1016/j.jcrysgro.2017.07.004

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Flux growth of multiferroic Cu3Nb2O8 single crystals R. Chen a,b, M.M. Shi a,b, Y.J. Liu a,b, C.B. Liu a,b, H.P. Zhu a,b, C. Dong a,b, Y. Liu c, J. Shi c, Z.C. Xia a, Z.W. Ouyang a, J.F. Wang a,* a

Wuhan National High Magnetic Field Center, Huazhong University of Science and Technology, Wuhan 430074, PR China

b

School of Physics, Huazhong University of Science and Technology, Wuhan 430074, PR China c

School of Physics and Technology, Wuhan University, Wuhan 430072, PR China

Abstract: Cu3Nb2O8 belongs to a new multiferroic family with coexisting structure and magnetic chiralities. In this work, we report on the flux growth of Cu3Nb2O8 single crystals using a V2O5-K2MoO4 mixture as the flux with a 5:1 ratio. The obtained crystals have an average size of 3×3×2 mm3 and high quality determined from X-ray diffraction, specific heat and susceptibility measurements. The experimental results on the single crystals reveal robust anomalies at 26.2 K and 24.5 K in specific heat and magnetic anisotropy along different crystallographic axes, which provide a better understanding of this fascinating multiferroic compound.

Keywords: A1. Characterization A2. Single crystal growth A2. Flux method B1. Inorganic compounds B2. Multiferroic material *

Corresponding author. E-mail address: [email protected] (J.F. Wang)

1. Introduction Multiferroic materials, in which spontaneous ferroelectricity and magnetic order coexist and correlate with each other, have been one of the most intriguing issues in the past decade. In particular, magnetic ordering driven multiferroics, possessing a strong magnetoelectric coupling, have attracted considerable interests due to the rich physics and enormous potential applications [1-4]. Different theoretical models have been proposed to explain their microscopic mechanisms, such as the inverse Dzyaloshinskii-Moriya (DM) model (or the spin-current model) [5-7], exchange striction mechanism [8], and the p-d hybridization theoretical model. These mechanisms have been successfully applied to studies on the spiral spin structure compounds RMnO3 (R=Gd, Tb, Dy) [5, 9, 10], Ni3V2O8 [11, 12], LiCuVO4 [13], MnWO4 [14, 15], E-type antiferromagnetic RMnO3 (R = Ho, Tm and Lu) [16], Ca3CoMnO6 [17], and the proper screw magnetic ordering multiferroics Ba2CoGe2O7 [18] and CuFeO2 [19]. Recently RbFe(MoO4)2 [20-22], CaMn7O12 [23, 24], MnSb2O6 [25] and Cu3Nb2O8 [26] have been discovered to be a new multiferroic family with coexisting structure and magnetic chiralities. These unique characteristics may produce exotic quantum states such as the recently discovered Skyrmion phases [25]. In addition, a giant improper ferroelectricity as large as 2870 C/m2 was found in CaMn7O12 [23]. This polarization value is about 4-5 times larger than those of RMnO3, and almost the largest in all magnetic ordering driven multiferroics. Johnson et al. proposed a 2

ferroaxial mechanism for the ferroelectricity in CaMn7O12, in which the polarization couples to the ferroaxial component of the crystal structure [23]. Alternatively, Lu et al. proposed that the polarization originates from a combination of both the DM interaction and the exchange striction based on their first-principles calculations [24]. Similar disagreement also occurred for the ferroelectricity in Cu3Nb2O8. This compound has a triclinic symmetry with a space group of

. Fig.1 shows its crystal

structure viewed along the a and b axis [26]. There are two nonequivalent Cu2+ sites called Cu1 and Cu2. They form sawtooth Cu-O chains along the a axis and Cu-O layers in the ac plane, which is separated by the nonmagnetic Nb atoms along the b axis. As the temperature is lowered, Cu3Nb2O8 undergoes two phase transitions at T1=26.5 K and T2=24 K, which are associated with an antiferromagnetic ordering below T1 and the appearance of a novel ferroelectric polar phase below T2 [26, 27]. Li et al. attributed this ferroelectricity to symmetric exchange striction with slight spin canting. The direction of ferroelectric polarization is not determined by the orientation of the spin rotation plane, in contradiction to the ferroaxial explanation proposed by Johnson et al. [28]. Nevertheless, the underlying mechanism for this kind of compounds remains unclear and is still an open issue. Growth of high quality single crystals is crucial for the investigations of novel multiferroic behaviors. So far, there is no detailed information for single crystal growth of Cu3Nb2O8. In Ref. [26], Johnson et al. only briefly mentioned the growth of a millimeter-sized single crystal (~2×2×1mm3) using an optical floating-zone furnace. 3

Another way to grow Cu3Nb2O8 crystals is the flux method as reported in the early literature using CuO, V2O5, Nb2O5, MnO3, K2CO3 as the starting materials [29]. However, the crystal quality has not been examined by either the X-ray diffraction or magnetic property measurements. Furthermore, the decomposed CO2 gas from K2CO3 during the experiments may have an essential influence on the quality of the obtained crystals and therefore their physical properties. In this work, we use an effective and simple way to resolve this problem instead of the above methods. We report in details the single crystal growth of Cu3Nb2O8 using the mixtures of V2O5 and K2MoO4 as the flux. The as-synthesized crystals have a large average size of 3×3×2 mm3. These crystals were then found to possess good quality through X-ray diffraction, specific heat and magnetic property measurements. 2. Crystal growth and characterization Prior to single crystal growth, we first prepared polycrystalline Cu 3Nb2O8 using a standard solid-state reaction method. CuO (99.9%) and Nb2O5 (99.95%) with a molar ratio of 3:2 were mixed and ground carefully for 5 hours. Then the mixture was packed into an alumina crucible and sintered at 900 ℃ for 40 hours in ambient air with several intermediate grindings. Brown powders of pure Cu3Nb2O8 were obtained as seen from X-ray powder diffraction measurements. The flux of V2O5 and K2MoO4 has a molar ratio of 5:1. A certain amount of CuO is also added to suppress the formation of impurity phases and decomposition of the target product at high temperature. Cu3Nb2O8 powders, the flux and excess CuO with a mass ratio of 4:3:0.5 4

were mixed in a platinum crucible fired at 1200 ℃ for 15 hours, and then cooled slowly to 950 ℃ with a cooling rate of 2 ℃/h. The resulting crystals were washed several times with dilute nitric acid to remove residues that remained on the crystal surface. Finally, black prism-like Cu3Nb2O8 single crystals and by-product CuO single crystals were obtained. Selected Cu3Nb2O8 single crystals are shown in Fig.2. All the crystals display 2-3 flat natural facets and corresponding crystallographic axes. The typical size of the crystals grown in our experiments is about 3×3×2 mm3, which is even slightly larger than the size of 2×2×1 mm3 that was grown using an optical floating-zone furnace [26]. The largest one we obtained has dimensions of 4×3× 3 mm3. Small single crystals of Cu3Nb2O8 were crushed for X-ray powder diffraction (XRD, Philips X’pert pro). Data were collected in the angular range 2θ = 10°- 90° with a scan step width of 0.01313 at room temperature using Cu Kα radiation (1.54 Å). The result is shown in Fig.3, which is in good agreement with the diffraction peaks of Cu3Nb2O8 (Ref.: ICSD Code 193472). Single crystal X-ray diffraction (XRD) measurements were performed at 173 K on a Bruker D8 QUEST machine equipped with a CMOS camera (Bruker AXS Inc., Germany). The multi-scan method was used for absorption corrections. The structures were solved by the direct method and were refined with SHELXL-97. Crystal data are given in Table 1 and Table 2, respectively. We also have supplied cif files as supplementary information. To determine the crystallographic axis further, we oriented the largest single crystals using X-ray 5

scattering. The diffraction patterns of natural (011), (001) and (100) planes of the single crystals are shown in Fig.4. All these results are in good agreement with those previously reported [26] and demonstrate the as-grown single crystals of Cu3Nb2O8 are of high quality. 3. Physical property measurements Specific heat was measured using a commercial Quantum Design Physical Property Measurement System (PPMS). Magnetic properties were measured by means of a vibrating sample magnetometer on a commercial superconducting quantum interference device (SQUID-VSM Quantum Design). Fig.5 presents the specific heat data of a Cu3Nb2O8 single crystal cooled from 100 to 10 K in a zero field. As the temperature is reduced, the specific heat exhibits two sharp anomalies in the vicinity of T1 = 26.2 K and T2 = 24.5 K, indicating two successive phase transitions. This interesting result is in good agreement with previous experimental investigations on a polycrystalline sample [27] and the crystal grown using an optical floating-zone furnace [26]. Note that these two Cp peaks of our data are much sharper and robust than those previously reported, also showing that the grown crystals have high quality. Fig.6 shows the temperature dependence of the magnetic susceptibilities along different crystallographic axes. A characteristic feature for all the three curves is given by a broad maximum at ~40 K, indicating a typical antiferromagnetic ordering below this temperature. The inset of Fig.6a shows a close inspection of the susceptibility along the [111] direction and its derivative for 15-40 K. Two anomalies are clearly 6

seen at 26.2 K and 24.2 K that are in agreement with our specific heat data. The inverse magnetic susceptibilities are shown in Fig.6b. It is found that the data follow a perfect linear relation in the high temperature region of 120-300 K. Using the Curie-Weiss law: χ = C / (T - θcw), where θcw is the Curie-Weiss temperature and C denotes the Curie constant, we obtained the effective moments for the three crystallographic = 1.05

axes:

/Cu2+,

= 1.24

= 1.22

/Cu2+

and

/Cu2+. These values are smaller than the theoretical effective moment = 1.73

temperatures are

/Cu2+ (S = 1/2). The corresponding Curie-Weiss

= -70.7 K,

= -59.3 K and

= -51.2 K,

respectively, showing the existence of strong antiferromagnetic interactions in Cu3Nb2O8. We find that the obtained values are not sensitive to the fitting temperature range between 120 K and 300 K, indicating a high precision in the determination of the effective moments and Weiss temperatures in this work. In Cu3Nb2O8, the magnetic Cu2+ ions form sawtooth chains along the a axis, where magnetic frustration may occur between the nearest and the next-nearest Cu2+ magnetic ions. The calculated frustration factor (f = θcw/TN) is ~3, showing that Cu3Nb2O8 is not a strong frustrated antiferromagnet. The deviation of the Curie-Weiss law of the susceptibility begins to appear at 120 K, which is far away from T1, indicating the onset of some possible exchange interactions from this temperature. In Fig.6a, we find less magnetic anisotropy for the [100] and [001] axes in the paramagnetic phase. This result is consistent with the isotropic nature of the CuO layers within the ac plane, as 7

illustrated in Fig.1. It is worth noting that in Ref. [27] a clear anomaly is also seen near 60 K in both the specific heat and the magnetization data measured on the polycrystalline sample. However, our experimental data on the single crystals do not show such a 60 K- anomaly as seen in Fig.5 and Fig.6. This means that the 60 K- anomaly is not an intrinsic behavior and likely due to magnetic impurities in the polycrystalline sample. Magnetization processes at 2 K for the three crystallographic axes are shown in Fig.7. The magnetization for H//[100] increases linearly and does not saturate up to 7 T, while it increases slowly for H//[001] and H//[111] with a change of slope above 5 T. No hysteresis and remnant magnetization at H = 0 T are observed consistent with an antiferromagnetic ground state. In addition, the magnetic moment at 7 T for each curve is much smaller than the expected value of Cu2+ (gBS=1B for S=1/2 and g=2) for the saturation magnetization. We suggest that the upturn of the magnetization above 5 T is possibly a sign of the meta-magnetic transition in higher magnetic fields. 4. Summary We have grown large single crystals of the magnetic multiferroic compound Cu3Nb2O8 using a flux method. The obtained crystals have high quality as seen from X-ray diffraction, specific heat and magnetization measurements. Our experimental data also demonstrate two magnetic ordering states below 26.2 K and 24.5 K and the absence of a possible phase transition at ~60 K as reported in Ref. [27]. Further investigations in high magnetic fields are necessary to explore magnetic phase 8

transitions and to reveal the underlying novel multiferroic mechanism in this complex frustrated system. Acknowledgment This work was supported by the National Natural Science Foundation of China (Grant Nos. 11574098, 11474110 and 51571152).

References [1] S. Dong, J.M. Liu, S.W. Cheong, Z. Ren, Multiferroic materials and magnetoelectric physics: symmetry, entanglement, excitation, and topology, Adv. Phys. 64 (2015) 519-626. [2] Y. Tokura, S. Seki, N. Nagaosa, Multiferroics of spin origin. Rep. Prog. Phys. 77 (2014) 076501. [3] R.D. Johnson, P.G. Radaelli, Diffraction Studies of Multiferroics, Annu. Rev. Mater. Res. 44 (2014) 269-298. [4] K.F. Wang, J.M. Liu, Z.F. Ren, Multiferroicity: the coupling between magnetic and polarization orders, Adv. Phys. 58 (2009) 321-448. [5] H. Katsura, N. Nagaosa, A.V. Balatsky, Spin current and magnetoelectric effect in noncollinear magnets, Phys. Rev. Lett. 95 (2005) 057205. [6] I. Dzyaloshinsky, A thermodynamic theory of "weak” ferromagnetism of antiferromagnetics, J. Phys. Chem. Solids 4 (1958) 241-255. [7] A. Crépieux, C. Lacroix, Dzyaloshinsky—Moriya interactions induced by symmetry, J. Magn. Magn. Mater. 182 (1988) 341-349. [8] N. Lee, Y.J. Choi, M. Ramazanoglu, W. Ratcliff, V. Kiryukhin, S.W. Cheong, Mechanism of exchange striction of ferroelectricity in multiferroic orthorhombic HoMnO3 single crystals, Phys. Rev. B 84 (2011) 020101 [9] M. Mostovoy, Ferroelectricity in spiral magnets, Phys. Rev. Lett. 96 (2006) 067601. [10] I.A. Sergienko, E. Dagotto, Role of the Dzyaloshinskii-Moriya interaction in multiferroic perovskites, Phys. Rev. B 73 (2006) 094434. [11] F. Fabrizi, H.C. Walker, L. Paolasini, F. de Bergevin, T. Fennell, N. Rogado, R.J. Cava, T. Wolf, M. 9

Kenzelmann, D.F. McMorrow, Electric field control of multiferroic domains in Ni3V2O8 imaged by x-ray polarization-enhanced topography, Phys. Rev. B 82 (2010) 024434. [12] G. Lawes, A.B. Harris, T. Kimura, N. Rogado, R.J. Cava, A. Aharony, O. Entin-Wohlman, T. Yildirim, M. Kenzelmann, C. Broholm, A.P. Ramirez, Magnetically driven ferroelectric order in Ni 3V2O8, Phys. Rev. Lett. 95 (2005) 087205. [13] L.A. Prozorova, S.S. Sosin, L.E. Svistov, N. Büttgen, J.B. Kemper, A.P. Reyes, S. Riggs, A. Prokofiev, O.A. Petrenko, Magnetic field driven 2D-3D crossover in the S = 1/2 frustrated chain magnet LiCuVO4, Phys. Rev. B, 91 (2015) 174410. [14] K. Taniguchi, N. Abe, T. Takenobu, Y. Iwasa, T. Arima, Ferroelectric polarization flop in a frustrated magnet MnWO4 induced by a magnetic field, Phys. Rev. Lett. 97 (2006) 097203. [15] H. Nojiri, S. Yoshii, M. Yasui, K. Okada, M. Matsuda, J.S. Jung, T. Kimura, L. Santodonato, G.E. Granroth, K.A. Ross, J.P. Carlo, B.D. Gaulin, Neutron Laue diffraction study on the magnetic phase diagram of multiferroic MnWO4 under pulsed high magnetic fields, Phys. Rev. Lett. 106 (2011) 237202. [16] S. Ishiwata, Y. Kaneko, Y. Tokunaga, Y. Taguchi, T. Arima, Y. Tokura, Perovskite manganites hosting versatile multiferroic phases with symmetric and antisymmetric exchange strictions, Phys. Rev. B 81 (2010) 100411. [17] Y.J. Choi, H.T. Yi, S. Lee, Q. Huang, V. Kiryukhin, S.W. Cheong, Ferroelectricity in an ising chain magnet, Phys. Rev. Lett. 100 (2008) 047601. [18] H. Murakawa, Y. Onose, S. Miyahara, N. Furukawa, Y. Tokura, Ferroelectricity induced by spin-dependent metal-ligand hybridization in Ba2CoGe2O7, Phys. Rev. Lett. 105 (2010) 137202. [19] H.J. Xiang, P.S. Wang, M.H. Whangbo, X.G. Gong, Unified model of ferroelectricity induced by spin order, Phys. Rev. B 88 (2013) 054404. [20] A.J. Hearmon, F. Fabrizi, L.C. Chapon, R.D. Johnson, D. Prabhakaran, S.V. Streltsov, P.J. Brown, P.G. Radaelli, Electric field control of the magnetic chiralities in ferroaxial multiferroic RbFe(MoO 4)2, Phys. Rev. Lett. 108 (2012) 237201. [21] J.S. White, C. Niedermayer, G. Gasparovic, C. Broholm, J.M.S. Park, A.Y. Shapiro, L.A. Demianets, M. Kenzelmann, Multiferroicity in the generic easy-plane triangular lattice antiferromagnet RbFe(MoO4)2, Phys. Rev. B 88 (2013) 060409. [22] K. Cao, R.D. Johnson, F. Giustino, P.G. Radaelli, G.C. Guo, L. He, First-principles study of multiferroic 10

RbFe(MoO4)2, Phys. Rev. B 90 (2014) 024402. [23] R.D. Johnson, L.C. Chapon, D.D. Khalyavin, P. Manuel, P.G. Radaelli, C. Martin, Giant improper ferroelectricity in the ferroaxial magnet CaMn 7O12, Phys. Rev. Lett. 108 (2012) 067201. [24] X.Z. Lu, M.H. Whangbo, S. Dong, X.G. Gong, H.J. Xiang, Giant ferroelectric polarization of CaMn 7O12 induced by a combined effect of Dzyaloshinskii-Moriya interaction and exchange striction, Phys. Rev. Lett. 108 (2012) 187204. [25] R.D. Johnson, K. Cao, L.C. Chapon, F. Fabrizi, N. Perks, P. Manuel, J.J. Yang, Y.S. Oh, S.W. Cheong, P.G. Radaelli, MnSb2O6: a polar magnet with a chiral crystal structure, Phys. Rev. Lett. 111 (2013) 017202. [26] R.D. Johnson, S. Nair, L.C. Chapon, A. Bombardi, C. Vecchini, D. Prabhakaran, A.T. Boothroyd, P.G. Radaelli, Cu3Nb2O8: a multiferroic with chiral coupling to the crystal structure, Phys. Rev. Lett. 107 (2011) 137205. [27] G. Sharma, J. Saha, S.D. Kaushik, V. Siruguri, S. Patnaik, Improper ferroelectricity in helicoidal antiferromagnet Cu3Nb2O8, Solid State Commun. 203 (2015) 54-57. [28] Z.L. Li, M.H. Whangbo, X.G. Gong, H.J. Xiang, Helicoidal magnetic structure and ferroelectric polarization in Cu3Nb2O8, Phys. Rev. B 86 (2012) 174401. [29] B.M. Wanklyn, F.R. Wondre, W. Davison, Flux growth of crystals of some complex oxides, J. Mate. Sci. Lett. 3 (1984) 539-543.

11

Fig.1 Crystal structure of Cu3Nb2O8 viewed along the a and b axis.

Fig.2 As-grown Cu3Nb2O8 single crystals. The scale of the coordinate paper is 1 mm

12

-102

30

35

40

113 003 212

110

25

222

101 002 -120

-210 1-12 020 -112 -1-12 -220

1-21 -111 1-11 -101

100 001

-110

20

-1-11

011

Intensity ( a. u .) 15

45

50

Angle ( 2 theta )

Fig.3 Characteristic XRD pattern of crushed Cu3Nb2O8 single crystals at room temperature.

13

Table 1. Crystal data and structure refinement for Cu3Nb2O8. Formula

Cu3Nb2O8

Fw

507.47

T, K

173(2)

λ, Å

0.71073

Space group

P

a, Å

5.1725(14)

b, Å

5.4774(15)

c, Å

6.0042(16)

α, deg

72.472(5)

β, deg

83.444(5)

γ, deg

65.680(4)

V, Å3

147.81(7) 1

Z -3

5.667

Dcalcd, g cm -1

1.4384

µ, cm

2

GOF on F

1.029 a

R1,wR2 [I>2σ(I)]

0.0276, 0.0714

R1,wR2 (all data)

0.0276, 0.0714

a

2 22 2 2 1/2 R1 = ∑||Fo| – |Fc||/∑|F o|, wR2 = {∑w[(Fo) – (Fc) ] /∑w[(Fo) ] }

Table 2. Atomic coordinates (×104) and equivalent isotropic displacement parameters (Å 2×103) for Cu3Nb2O8. U(eq) is defined as one third of the trace of the orthogonalized U ij tensor.

atom

x

y

z

U(eq)

Nb(1)

7215(1)

416(1)

6513(1)

2(1)

Cu(1)

403(1)

4271(1)

7643(1)

3(1)

Cu(2)

5000

5000

0

4(1)

O(1)

3263(6)

1925(6)

5945(6)

5(1)

O(2)

8572(6)

2378(6)

3626(6)

4(1)

O(3)

7641(6)

2443(6)

8297(6)

4(1)

O(4)

7377(6)

-2973(6)

9001(5)

3(1)

14

001

002

(b)

003

11 1

001 004

200

(c)

100 001

Intensity( a.u.)

033

011

Intensity( a.u.)

011

001

100

20

40

60

400

300

100

Intensity( a.u.)

022

(a)

80

2(deg)

Fig.4 Orientation of the crystal surfaces using X-ray scattering analysis: (a) for the (011) plane, (b) for the (001)

plane and (c) for the (100) plane. The illustrations show the corresponding crystal planes.

15

T1 = 26.2 K T2 = 24.5 K

100

T2

Cp/T (J mg-1K-2)

Cp (J mg-1K-1)

200

T1

2

1

0

20

30

40

50

60

70

T (K)

0 20

40

60

80

100

T (K) Fig.5 Specific heat capacity of a Cu3Nb2O8 single crystal measured in a zero magnetic field. The insets show the

data of Cp/T vs T.

16

14 

[111]

10 d/dT

( emu/mol)

(a) 12

8

20

25

30

35

T (K)

6 4

H=0.1 T

 (103 mol/emu)

(b) 0.2

0.1

[001] [100] [111]

0.0 0

50

100

150

200

250

300

T (K) Fig.6 Temperature dependence of the susceptibilities (a) and the inverse curve (b) of Cu3Nb2O8 measured in an

external field H = 0.1 T. The inset of Fig.6a shows enlarged zooms of the susceptibility and its derivative along the

[111]-axis. Solid dots, open circles and hollow triangles represent extra field H parallel to the [001]-axis,

[100]-axis and [111]-axis, respectively.

17

M (B/Cu2+)

[001] [100] [111]

0.04

0.02

T=2K 0.00

0

1

2

3

4

5

6

7

H (T) Fig.7 Magnetization as a function of applied field H at 2 K. Solid dots, open circles and hollow triangles represent

extra field H parallel to the [001]-axis, [100]-axis and [111]-axis, respectively.

18

Highlights

1. We first report the flux growth of large-sized and high-quality single crystals of Cu3Nb2O8. 2. The experimental results on the single crystals reveal robust anomalies at 26.2 K and 24.5 K in specific heat and magnetic anisotropy along different crystallographic axes.

19