RoomtemperatureagglomerationforthegrowthofBiTeIsinglecrystalswithagiantRashbaeffect PAPER CrystEngComm

(1)

PAPER

Cite this: CrystEngComm, 2014, 16, 8678

Received 14th May 2014, Accepted 4th July 2014 DOI: 10.1039/c4ce01006j www.rsc.org/crystengcomm

Room temperature agglomeration for the growth of BiTeI single crystals with a giant Rashba effect

R. Sankar,âI. Panneer Muthuselvam,âChristopher John Butler,^bS.-C. Liou,â B. H. Chen,âM.-W. Chu,âW. L. Lee,^cMinn-Tsong Lin,^dR. Jayavelê

and F. C. Chou*^afg

We report a room temperature agglomeration (RTA) procedure to grow highly homogeneous and impurity-free BiTeI single crystals safely. The proposed four-step procedure of mixing and heating is able to prevent severe iodine loss and avoid the danger of explosion during large scale crystal growth. Following the RTA treatment of the precursor, the single crystals obtained from three different growth methods, including vertical Bridgman, melt growth and chemical vapour transport (CVT), were compared. Crystals grown using the Bridgman method showed the highest residual-resistance ratio (RRR) and mobility, and the largest domain size among the three. The crystal quality and purity have been confirmed using X-ray diffraction, Electron Probe Microanalysis (EPMA), resistivity, TEM, and STM. Additionally, Mn-intercalated and -substituted BiTeI crystals have also been investigated.

Introduction

The Rashba effect, the spin-up and spin-down band splitting phenomenon in an external electric field, can be observed when the spin degeneracy originally protected by the time- reversal and inversion symmetries is lifted.¹A surface energy gradient on structural inversion asymmetry in bulk materials can also provide a similar effect. Because it is desirable to fabricate devices that can be switched more quickly with spin, the Rashba effect in solids has a potential use in spintronics devices. However, the reported Rashba effect on material surfaces is usually too weak for practical applications.²The recent finding of a giant Rashba effect in the layered material of BiTeI with strong inversion asymmetry has generated strong interest in the research community.^3–5 BiTeI has a layered structure in the space group of P3m1 (no. 156) (see Fig. 1), which lacks inversion symmetry as indicated by the ion ordering of Te- and I-layers on the opposite sides of Bi with unequal Bi–Te and Bi–I bond lengths. In particular, the spin band splitting of BiTeI, described by the Rashba energy

ER (~100 meV) and Rashba parameterαR(~3.8 eV Å), is at a record high level compared to that induced mostly by the surface energy gradient as reported previously. Very recently, the Rashba effect has also been observed in the layered compound LaOBiS2, which has a similar spin splitting to

aCenter for Condensed Matter Sciences, National Taiwan University, Taipei 10617, Taiwan. E-mail: [email protected]

bDepartment of Physics, National Taiwan University, Taipei 10617, Taiwan

cInstitute of Physics, Academia Sinica, Taipei 11529, Taiwan

dInstitute of Atomic and molecular Sciences, Academia Sinica, Taipei 10617, Taiwan

eAnna University, Crystal Growth Centre, Chennai-600025, India

fNational Synchrotron Radiation Research Center, Hsinchu 30076, Taiwan

gTaiwan Consortium of Emergent Crystalline Materials, National Science Council, Taipei 10622, Taiwan

Fig. 1 (a) BiTeI has a layered crystal structure of space group P3m1, which lacks inversion symmetry due to the ion ordering of Te and I on the opposite sides of Bi in hcp packing. (b) Te- and I-layers in c-projection for a single tri-layer unit of BiTeI are shown to be a 2D triangular lattice.

Published on 07 July 2014. Downloaded by National Taiwan University on 17/09/2014 07:15:56.

View Article Online

View Journal | View Issue

(2)

BiTeI that originates from the polar field induced by the strong ionic bonding between atoms.⁶While Bi2Te3has been demonstrated to show a high thermoelectric figure-of-merit (ZT) for commercial applications,⁷ BiTeI with a similar layered structure has also been explored for its potential for thermopower improvement.⁸Temperature dependencies of the Seebeck coefficient have been explored in BiTeBr and BiTeI,^9,10 but the thermal properties reported for the compound BiTeI are still limited. Bi2Se3and Bi2Te3have an identical crystal structure and are characterized as second- generation topological insulators, which have a surface band structure similar to that of graphene with a Dirac cone at the

-point.¹¹ Both Bi2Te3 and Bi2Se3 have a quintuple layer building block with an identical Te/Se surface near the van der Waals gap.¹²On the other hand, BiTeI has an inter- esting heterogeneous van der Waals gap that is formed by two different surfaces composed of Te and I atoms separately after the van der Waals gap is cleaved open.

The growth of impurity-free BiTeI crystals is challenging, mostly due to the nature of the solid iodine precursor to easily sublime at room temperature, which becomes worse during the vacuum sealing of quartz tubing for an oxygen- free growth environment. The BiTeI samples used in previously published works often suffered from impurity phase inclusion. For example, the single crystal BiTeI sample prepared by Ishizaka et al. contains a persistent BiI3impurity phase, unless a small amount of Mn was substituted into the Bi sites to eliminate the BiI3impurity phase and to improve crystal quality.³However, the impact of Mn substitution has not been explored fully, and the introduced Mn ions may have served as nucleation centers for better growth. The difficulty of avoiding the BiI3 impurity phase in BiTeI has even led to a complete thermoelectric property study on BiTeI samples with different amounts of BiI3 inclusion.⁸ The published phase diagram of Bi2Te3–BiI3 indicates that nonstoichiometry close to the exact BiTeI ratio can be complicated by the existence of Bi/Te antisites and iodine defects.¹³Clearly, high purity BiTeI single crystals are highly desirable for both fundamental and applied research fields.

Herein, we report a unique growth recipe that grows impurity-free BiTeI single crystals through a newly designed room temperature agglomeration (RTA) procedure. The newly designed RTA procedure enhances the iodine handling safety significantly to avoid the danger of explosion of sealed quartz tubing during the heating process.

Recently Wang et al. reported the synthesis of submicrometer hollow spheres of BiTeI by a hydrothermal method,¹⁴and a report of BiTeI single crystal growth by the Bridgman method indicated several phase transitions under high pressure.¹⁵ In large scale single crystal growth, the handling of iodine of low boiling point is very important during the heating and mixing process. We report a four-step procedure of mixing and heating to prevent iodine loss and avoid the danger of explosion, which is a novel method to grow highly homogenous and impurity-free single crystals of BiTeI.

Material characterization

Powder X-ray diffraction was performed with crushed crystal samples on a Bruker-AXS D8 ADVANCE X-ray diffractometer which was equipped with a diffracted beam monochromator set for Cu K_α1radiation (λ = 1.54056 Å) at room temperature.

The structure refinement used data obtained from the synchrotron X-ray source in NSRRC-Taiwan. Chemical analysis was performed using electron microprobe analysis (EPMA). The transport properties of the BiTeI crystals were measured using a four-probe method for the in-plane resistivity as a function of temperature. Two types of speci- men with the crystalline c-axis perpendicular and parallel to the sample plane were prepared for investigation. The sample with the plane perpendicular to the c-axis was easily prepared by mechanical cleavage, whereas the sample with the plane parallel to the c-axis was prepared by a standard cross- sectional technique followed by Ar⁺ ion milling at 3 keV (Gatan PIPS). Both specimens were finally cleaned by low- energy ion milling at 0.3 keV to remove the possible surface amorphous layers. All the electron diffraction patterns were acquired on a FEI field-emission electron microscope (TecnaiF20) operated at 200 kV. High-angle annular dark- field (HAADF) imaging in a scanning transmission electron microscope (STEM) was performed on a JEOL-2100F microscope equipped with a probe spherical aberration corrector.

Cleavage of all the BiTeI crystal for investigations by scanning tunneling microscopy (STM) was performed in a preparation chamber with a base pressure lower than 5 × 10⁻¹¹ mbar.

Crystals were cleaved on a liquid He cooled cryostat, allowing a cleavage temperature of around 8 K. All STM measurements were performed at 4.5 K on an Omicron low temperature scanning tunneling microscope using a chemically etched tungsten tip. dI/dV intensity maps were acquired simultaneously with constant current topography images using a lock-in amplifier, with a bias modulation of 40 mV at frequency of 5.9 kHz.

Experimental details

In order to obtain the best method for the BiTeI growth, we have tried and compared three reaction routes using different precursor mixtures. As the iodine solid is easy to sublime during the vacuum sealing process, two types of precursor avoiding direct iodine handling have also been tried, including the use of the intermediate compounds BiI3and Bi2Te3. The three tested reaction routes are:

Bi2Te3+ BiI3→ 3BiTeI (1) BiI3+ 2Bi + 3Te→ 3BiTeI (2)

Bi + Te + I→ BiTeI (3)

Reaction (1) requires the use of both pre-reacted Bi2Te3

and BiI3. Reaction (2) uses pre-reacted BiI3, plus Bi and Te

(3)

metal. Reaction (3) uses the raw materials of I2crystals, and Te and Bi solids directly, but it must be cautioned that a purification process is necessary before mixing, especially for the iodine.

Melt growth of BiTeI single crystals

We applied the same preliminary steps of a room temperature agglomeration (RTA) procedure and high temperature homogenization for all three of the reaction routes shown above. Single crystal growth of BiTeI containing elemental iodine must be separated into several steps, instead of mixing and melting in one step, to avoid potential explosion and impurity formation. The reagent grade ingot of Bi metal and Te metal crystal must be purified several times before weighing and mixing. A purification process has been applied to each element sealed in evacuated quartz tubing using a zone refining method five times. Each element is first melted at 650°C in zone-I for 6 h and recrystalized in zone-II, which is kept at 550°C by keeping the tube tilted at an angle of 40°. A sublimation process is used to purify the iodine crystals. The BiI3 and Bi2Te3 precursors were prepared starting from the reaction of stoichiometric mixtures of Bi, I2, and Te at 450°C (for BiI3) and 600°C (for Bi2Te3) for 24 hours in an evacuated quartz ampoule. These mixtures were added to a quartz ampoule with a conical tip in a glovebox and were evacuated under pure argon atmosphere at 20 mTorr. Two different annealing temperatures of 550 and 600 °C were tested in order to compare the potential iodine nonstoichiometry. For the melt growth method, final cooling was carried out from 550/600 to 400°C at a rate of 1.5 °C h⁻¹ followed by furnace cooling to room temperature. The obtained crystals were plate-like with the c-axis parallel to the vertical direction. The as-grown crystals were found to be easy to cleave along the (001) plane with a mirror-like quality, as shown in Fig. 3. The typical crystal size is ~3 cm in length and ~1 cm in diameter. We can summarize the melt growth conditions in the four steps below (see Fig. 2):

Step 1. Room temperature agglomeration:

A total weight of ~5 grams of stoichiometric and purified precursors are mixed and sealed in a quartz ampoule of 1.8 cm inner diameter and ~20 cm length in a glovebox. The sealed ampoule is placed in a horizontal furnace at room temperature for 24 hours to allow the iodine to intercalate into the Bi/Te and react homogeneously, which is the key procedure to avoid quartz tube explosion upon heating, especially for reaction (3).

Step 2. Horizontal high temperature annealing:

The ampoule is placed in a horizontal furnace at a small tilt angle, with annealing at 550/600°C for 24 hours at a heating rate of 50 °C h⁻¹. The BiTeI melt can be observed at the bottom of the ampoule and small iodine crystals can be seen on the wall of the quartz ampoule after cooling.

Step 3. Vertical high temperature annealing:

The same ampoule is placed in a vertical tube furnace with the conical tip pointing downward, and heated at 550/600°C for 24 hours at a heating rate of 50°C h⁻¹. This step ensures that the precipitated iodine crystals on the wall fully react with the flux.

Step 4. Slow cooling crystal growth:

After the completion of Step 3 of high temperature annealing, slow cooling from 600–400 °C at a rate of 1.5 °C h⁻¹ensures that crystal growth occurs starting from the nucleation at the relatively colder point tip below. Finally, the ampoule is cooled down to room temperature at a rate of 10°C h⁻¹after the crystal growth is completed below 400°C.

Bridgman growth of BiTeI single crystals

BiTeI single crystals were prepared from 5N purity bismuth, tellurium and iodine. The synthesis of the compound was carried out in conical quartz ampoules evacuated to 20 mTorr.

The homogenization of the batches and synthesis of the compound was carried out identically to the melt growth method up to Step 3 described above. Before applying the Bridgman pulling method, the ampoules containing the melt were heat-treated at 600°C for 24 h to ensure the melt filled the tip of ampoule; the ampoules were then lowered through the zone heated to 600°C with gradient 1 °C cm⁻¹at a rate of Fig. 2 Melt growth procedures for single crystal BiTeI. (a) Step 1:

room temperature agglomeration, (b) step 2: horizontal high temperature annealing, (c) step 3: vertical high temperature annealing, and (d) step 4: slow cooling crystal growth.

(4)

0.1 mm h⁻¹. The obtained BiTeI single crystals were 3 cm long and 1.2 cm in diameter, which were easy to cleave to expose the (001) plane. This growth method gave homogeneous single crystals and the (001) plane is always perpendicular to the pulling direction.

Chemical vapour transport growth of BiTeI single crystals

CVT growth of BiTeI crystals was carried out with a quartz ampoule of 45 cm length and 18 mm diameter. The solidified melt precursor was prepared following the Step 3 treatment identical to the melt growth procedure described above. The quartz ampoules charged with the melt precursor were rinsed with pure argon and evacuated to a pressure of 26.7 Pa.

Single crystals were transported from the hot zone between 550°C (T1) and 525°C (T2) over 15 cm, to the cold zone at 500°C (T3). The crystals were easy to cleave and gave large shiny surfaces of the (001) crystal plane.

Results and discussion

Single crystal samples grown using the three different methods are shown in Fig. 3. The as-grown pure BiTeI crystal is stable in air and can maintain its cleaved shiny surface for hours under air without degradation. The room temperature powder X-ray diffraction patterns for the melt growth crystals from the three different reaction routes are compared in Fig. 4. It is clear that, except for reaction (3) of direct initial mixing of Bi + Te + I, the other routes contain Bi2Te3and BiI3

impurity phases. We find that the MnxBiTeI crystals become hygroscopic after being stored under air for one day, as evidenced from the build-up of a layer of brown colored film which is confirmed to be MnI2·4H2O. In addition, a MnTe2

impurity phase is commonly observed in the growth of the Bi_1−xMnxTeI crystals, although it has been reported that minor Mn substitutions could improve the crystal quality of BiTeI.³

The lattice parameters from X-ray diffraction and EPMA chemical analysis for the crystals grown using the Bridgman, melt growth, and CVT methods are summarized in Tables 1 and 2, respectively. The lattice parameters indexed with the space group P3m1 agree with those published in the litera- ture, especially for the Bridgman growth method.³There is no significant difference in the chemical composition for the crystals grown by the three different growth methods, although the melt growth crystal shows a slightly larger c-axis, which indicates possible iodine intercalation. The homogeneity and stoichiometry has been confirmed from EPMA chemical analysis through five point averaging and normalization to Bi.

The in-plane resistivity data for the single crystal BiTeI samples obtained from the three growth methods are compared in Fig. 5. The temperature dependence of the

Fig. 3 Single crystals grown from three different growth methods: (a) melt growth, (b) chemical vapor transport (CVT), and (c) Bridgman growth. Bridgman growth of Mn_xBiTeI (x = 0.1) is shown in (d) after it has been stored under air for one day. A thin brown film is coated on the surface which indicates that the sample of Mn_xBiTeI is hygroscopic.

Fig. 4 Powder X-ray diffraction patterns of melt growth BiTeI crystals grown using three different reaction routes are compared. As indicated in the inset, only the reaction (3) of direct Bi + Te + I mixing can avoid the formation of the BiI3impurity phase.

Table 2 The EPMA results obtained from the average of five selected points and normalized to Bi, indicated as a ratio of Bi : Te : I

Growth method Bi Te I

Bridgman 1 1.0000(0.0010) 1.000(0.003)

Melt growth 1 0.9985(0.0019) 0.993(0.002)

CVT 1 0.9997(0.0014) 1.093(0.007)

Table 1 Lattice parameters of BiTeI single crystals from different growth methods

Growth method a (Å) b (Å) c (Å)

Bridgman 4.3435 4.3435 6.8547

Melt growth 4.3421 4.3421 6.8835

CVT 4.3482 4.3482 6.8358

(5)

resistivity ρ indicates that all the samples show metallic behaviour. High residue resistivity below 10 K is found for the crystals grown from the melt growth and CVT methods.

On the other hand, the Bridgman growth crystals have aρ value at least one order of magnitude lower and the RRR of

~4 is the highest among the three methods, which suggests that the Bridgman growth method gives the best quality crystals with the least amounts of impurities and crystallographic defects. Details of the Hall effect and Shubnikov–de Haas quantum oscillation measurement results have been reported separately.¹⁶

The selected-area electron diffraction (SAED) patterns of single crystal BiTeI are shown in Fig. 6(a) and (b), which were acquired with the incident electron beam parallel to the

c-axis along the [001] direction and perpendicular to the c-axis along the [−110] direction, respectively. Both SAED patterns indicate that crystallinity is expected for the single phase single crystal. The STEM-HAADF imaging with atomic column resolution has been used to investigate the real-space structural arrangement of BiTeI. The representative HAADF image along the [−110] direction of BiTeI is shown in Fig. 6(c), and the corresponding intensity profile measured along the red-colored dashed line in Fig. 6(c) is shown in Fig. 6(d). The triple-stacking-layer structure of BiTeI and the van der Waals gaps are displayed clearly in agreement with the assigned space group. Because the bismuth has a much larger atomic number than those of tellurium and iodine (Bi:

Z = 83, Te: Z = 52 and I: Z = 53), the central columns of the triple stacking layers with brighter contrast correspond to the Bi atomic columns, and the others two columns are Te and I, respectively.

Scanning tunneling microscopy (STM) topography and dI/dV conductance images for the crystals grown using the three different methods are shown in Fig. 7. STM images acquired from the vacuum-cleaved (001) surfaces of the BiTeI crystals reveal a step-mesa morphology of two distinct domains separated by a step-height of around 1.8 Å. These steps are accompanied by a sharp contrast in conductance between the domains, observed in the simultaneously acquired dI/dV maps. This contrast is attributed to a difference in the chemical termination between the two domains. As cleavage of the crystal is only expected to occur at the van der Waals gap between the Te and I atomic layers, the regions of different termination are thought to correspond to domains of an inversely stacking sequence, resulting in a different atomic layer (Te or I) being uppermost at the surface after cleavage.

The domains found at the surfaces of the CVT and melt grown crystals have typical scales of around 100 nm, and Fig. 5 The in-plane resistivity as a function of temperature indicates

that the Bridgman growth method provides crystals of the lowest resistivity and the highest RRR, indicating the best quality.

Fig. 6 (a) and (b) The selected-area electron diffraction (SAED) patterns of single-crystal BiTeI acquired with the incident electron beam parallel to thec-axis along the [001] direction and perpendicular to thec-axis along the [−110] direction. (c) The HAADF image along the [−110] direction, and (d) the intensity profile measured along the red dashed line in (c).

Fig. 7 STM topography and dI/dV conductance maps of BiTeI, taken on the vacuum cleaved surfaces of melt growth crystal ((a) and (b)), CVT crystal ((c) and (d)), and Bridgman crystal ((e) and (f)). All images are shown at the same scale. (g) Topographic line profile taken along the red dotted line in panel (a).

(6)

those found at the surfaces of Bridgman grown crystals are much larger, with a scale of around 1 micro meter. The much larger domain size for the Bridgman growth crystals is consistent with their much lower resistivity (Fig. 5) due to less scattering from the domain boundaries.

Conclusions

In summary, we have presented a safe and reliable growth method for BiTeI single crystals to handle the iodine element which easily sublimes during handling. The room temperature agglomeration procedure directly using Bi + Te + I pristine elements is crucial to prevent potential explosions during the heating period of the reaction in a sealed quartz ampoule. Single crystal BiTeI samples grown using three different methods have been fully characterized using X-ray diffraction, EPMA chemical analysis, electron diffraction, STM, and resistivity measurements. Among the samples, the Bridgman growth crystal has been shown to have the highest quality. We propose that the growth method with the added room temperature agglomeration procedure can also be applied to the preparation of similar compounds that contain large amounts of iodine in the initial mixing process.

Acknowledgements

FCC acknowledges support from NSC-Taiwan under project number NSC 101-2119-M-002-007.

References

1 E. I. Rashba, Sov. Phys. Solid State, 1960, 2, 1109.

2 P. King, R. Hatch, M. Bianchi, R. Ovsyannikov, C. Lupulescu, G. Landolt, B. Slomski, G. Balakrishnan, B. Iversen, J. Osterwalder, W. Eberhardt, F. Baumberger and P. Hofmann, Phys. Rev. Lett., 2011, 107, 096802.

3 K. Ishizaka, M. S. Bahramy, H. Murakawa, M. Sakano, T. Shimojima, T. Sonobe, K. Koizumi, S. Shin, H. Miyahara, A. Kimura, K. Miyamoto, T. Okuda, H. Namatame,

M. Taniguchi, R. Arita, N. Nagaosa, K. Kobayashi, Y. Murakami, R. Kumai, Y. Kaneko, Y. Onose and Y. Tokura, Nat. Mater., 2011, 10, 521.

4 V. Gnezdilov, P. Lemmens, D. Wulferding, A. Moller, P. Recher, H. Berger, R. Sankar and F. C. Chou, Phys. Rev. B:

Condens. Matter Mater. Phys., 2014, 89, 195117.

5 C. J. Butler, H. Yang, J. Y. Hong, S. H. Hsu, R. Sankar, C. Lu, H. Y. Lu, K. H. O. Yang, H. W. Shiu, C. H. Chen, C. C. Kaun, G. J. Shu, F. C. Chou and M. T. Lin, Nat. Commun., 2014, 5, 4066.

6 Q. Liu, Y. Guo and A. J. Freeman, Nano Lett., 2013, 13, 5264.

7 S. Scherrer and H. Scherrer, in Bismuth Telluride, Antimony Telluride, and Their Solid Solutions, CRC Handbook of thermoelectrics, ed. D. M. Rowe, CRC Press, 1995, ch. 19.

8 V. A. Kulbachinskii, V. G. Kytin, A. A. Kudryashov, A. N. Kuznetsov and A. V. Shevelkov, J. Solid State Chem., 2012, 193, 154.

9 V. A. Kulbachinskii, V. G. Kytin, Z. V. Lavrukhina, A. N. Kuznetsov and A. V. Shevelkov, Semiconductors, 2010, 44, 1548.

10 V. A. Kulbachinskii, V. G. Kytin, A. A. Kudryashov, A. N. Kuznetsov and A. V. Shevelkov, J. Solid State Chem., 2012, 193, 154.

11 M. Z. Hasan and C. L. Kane, Rev. Mod. Phys., 2010, 82, 3045.

12 F.-T. Huang, M.-W. Chu, H. H. Kung, W. L. Lee, R. Sankar, S.-C. Liou, K. K. Wu, Y. K. Kuo and F. C. Chou, Phys. Rev. B:

Condens. Matter Mater. Phys., 2012, 86, 081104.

13 A. Tomokiyo, T. Okada and S. Kawano, Jpn. J. Appl. Phys., 1977, 16, 291.

14 C. Wang, K. Tang, Q. Yang, J. Hu and Y. Qian, J. Mater.

Chem., 2002, 12, 2426.

15 Y. Chen, X. Xi, W. Yim, F. Peng, Y. Wang, H. Wang, Y. Ma, G. Liu, C. Sun, C. Ma, Z. Chen and H. Berger, J. Phys. Chem.

C, 2013, 117, 25677.

16 C. Wang, J. Tung, R. Sankar, C. Hsieh, Y. Chien, G. Guo, F. C. Chou and W. Lee, Phys. Rev. B: Condens. Matter Mater.

Phys., 2013, 88, 081104.