Absence of van der Waals Gap in Ternary Thorium Nitride ThNF and

Jul 10, 2019 - Present address: Shenzhen Suntak multilayer circuit board Co., Ltd, Shenzhen, Guangdong,. 518054, China. School of Materials and Energy...
0 downloads 0 Views 2MB Size
Download PDF
Article Cite This: Inorg. Chem. XXXX, XXX, XXX−XXX

pubs.acs.org/IC

Absence of van der Waals Gap in Ternary Thorium Nitride ThNF and ThNCl Naizhou Wang,† Guojun Ye,†,‡ Guobing Hu,† Fanbao Meng,† Chao Shang,† Mengzhu Shi,† Jianjun Ying,† and Xianhui Chen*,†,‡,§,∥

Downloaded via KEAN UNIV on July 19, 2019 at 02:18:10 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

†

Key Laboratory of Strongly Coupled Quantum Matter Physics, Chinese Academy of Sciences, Hefei National Laboratory for Physical Sciences at Microscale, and Department of Physics, University of Science and Technology of China, Hefei, Anhui 230026, China ‡ Collaborative Innovation Center of Advanced Microstructures, Nanjing University, Nanjing 210093, China § CAS Center for Excellence in Superconducting Electronics (CENSE), Shanghai 200050, China ∥ CAS Center for Excellence in Quantum Information and Quantum Physics, Hefei, Anhui 230026, China S Supporting Information *

ABSTRACT: Two kinds of ternary thorium nitride compounds, ThNF and ThNCl, are synthesized. Via the refinement of X-ray diffraction patterns, the accurate crystal structure of the two compounds is solved. Although ThNF and ThNCl share a similar structure with MNX (M = Ti, Zr, Hf; X = Cl, Br) compounds, the interaction between adjacent ThNF and ThNCl layers is not a van der Waals gap. For ThNF, the strong electronegativity of F ions leads to the bonding of Th to the F both in the nearest neighbor layer and the next nearest neighbor layer, which results in the absence of a van der Waals gap between ThNF layers. However, for ThNCl, the reason for the absence of a van der Waals gap could be attributed to the large Th−Cl bond length due to the partially covalent Th−Cl bond as well as the flat ThN layer. It is the absence of van der Waals gap that results in the failure of intercalating cations into ThNF and ThNCl. Our result reveals the reason for unsuccessful intercalation in ThNF and ThNCl, thereby providing a deeper understanding for the interlayer interaction in ternary layer structures in metal nitride halides.

■

conductivity in β-type MNX compounds does not conform to the conventional BCS theory and may be unconventional. First, the Tc analysis shows a weak dependence with doping level in a moderate range.8 Besides, the Tc is not strongly affected by the intercalated metal ion types, whether they are alkali metals (Li, Na, K),5−7 alkaline earth metals (Ca, Sr, Ba),9,10 or even a magnetic rare earth metal (Yb).11 Moreover, in the same doping level, via intercalating different organic molecules to alter the distance of the van der Waals gap, the Tc in β-type MNX compounds could even increase.12 Second, the electron density of states at the Fermi level in superconducting β-type MNX compounds is quite low, which is revealed by magnetic susceptibility and heat capacity measurements.13,14 Third, the tunneling-current measurements and specific heat revealed a quite large superconducting gap with 2Δ/kBT = 4.6−5.2.14,15 Finally, heat capacity and isotope experiments suggest weak electron−phonon coupling and a small isotope effect.14,16 All the experiments mentioned above suggest the unconventional pairing mechanism other than the BCS interaction in electron-doped β-type MNX compounds.12,17

INTRODUCTION The ternary layered metal nitride halides have attracted much attention due to their abundant physical properties, especially the discovery of superconductivity in electron-doped ZrNCl and HfNCl. Besides, many features in electron-doped ZrNCl and HfNCl resemble the ones in high-Tc (superconducting transition temperature) cuprates and iron-based superconductors,1−3 which suggests that the superconductivity in these series of compounds may be unconventional. The mother compounds of ternary layered structures metal nitride halides are band insulators and could be divided into two categories according to different crystal structures, namely, the α-type with FeOCl structure and the β-type with SmSI structure.4 The α-type structure consists of an orthogonal two-dimensional MN (M = Ti, Zr, Hf) network sandwiched by halide layers. Via intercalating alkali-metals or organic molecules into the van der Waals gap, the α-type KxTiNCl could be a superconductor with the highest Tc of 16 K.5,6 The β-type structure is constituted by double honeycomb MN (M = Zr, Hf) layers sandwiched by close packed halide layers. The highest Tc in β-type HfNCl could reach as high as 25.5 K by intercalating Li into the van der Waals gap.7 Recently, more and more physical measurements suggest that the super© XXXX American Chemical Society

Received: April 11, 2019

A

DOI: 10.1021/acs.inorgchem.9b01050 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

thoroughly washed with deionized water and dilute nitric acid and dried under vacuum, and then high purity thorium metal powder is acquired. Second, the Th3N4 is synthesized by the reaction between thorium metal and high purity nitrogen gas (99.9999%) with 20% excess at 1000 °C for 24 h in a sealed quartz tube. The obtained Th metal and Th3N4 is checked by X-ray diffraction. Then, Th3N4 and ThF4 powders (provided by Wuping Liao) are stoichiometrically weighted and mixed. The mixture is wrapped with gold foil and sealed in an evacuated quartz tube in order to prevent potential oxidation. The reaction is conducted at 850 °C for 12 h. Then, the ThNF sample is acquired. The synthesis of ThNCl is similar to that of ThNF. The ThCl4 precursor is synthesized according a method report in previous literature.22 Then, the mixture of Th3N4 and ThCl4 powders is wrapped with gold foil, sealed in an evacuated quartz, and sintered at 750 °C for 12 h. The structure of ThNF and ThNCl is characterized by a powder Xray diffractometer (SmartLab-9, Rigaku Corp.) with Cu Kα radiation and a fixed graphite monochromator. The lattice parameters and the crystal structure are refined using the Rietveld method with the program GSAS package.23,24 The magnetic susceptibility was measured by a SQUID magnetometer (Quantum Design MPMS-5). The UV−vis absorption spectroscopy was collected with Shimadzu DUV-3700 UV−vis−FTIR spectrometer. The X-ray photoelectron spectroscopy (XPS) data were obtained by an ESCAlab250 X-ray photoelectron spectroscopy instrument using a monochromatized Al Kα X-ray source. The binding energies are calibrated against the C 1s peak (284.6 eV) of the residual carbon absorbed on the surface of the sample. WARNING! 232Th is a radioactive element with a half-life time of 1.41 × 1010 years. Although its own activity is low, it is still highly harmf ul by inhalation and if swallowed.

Moreover, the superconductivity in ZrNCl and HfNCl is induced by the electron doping by intercalating alkali-metals or organic molecules into the van der Waals gap. Thus, it is of great significance to explore other MNX type compounds as well as to seek indications for possible superconductivity. In unconventional superconductors, there is another kind of superconductor, i.e., the heavy Fermion superconductors with f-electrons, except for high Tc cuprates and iron-based superconductors. In heavy Fermion superconductors, the quasiparticles that participated in the superconducting pairing are much heavier than free electrons with a quite large effective mass (100−1000 me).18 The reason for producing such a large effective mass could be attributed to the Kondo effect from the hybridization between itinerant electrons and the local moment of f-electrons in heavy Fermion systems. The heavy Fermion superconductor system shows abundant phenomenons, such as quantum critical behavior, nonfermi liquid behavior, hidden order, etc.19 Moreover, a lot of heavy Fermion superconductors show a quasi two-dimensional structure, which resembles that of cuprates, iron-based superconductors, and MNX compounds.20 ThNF and ThNCl are the MNX compounds with 5f electrons.21 Thus, the discovery of heavy-Fermion superconductivity in them is expected, since both ZrNCl and HfNCl could show superconductivity by intercalating alkali-metals or organic molecules into the van der Waals gap. However, up to now, all of the available intercalation methods (i.e., the liquid ammonia method, electrochemical intercalation, n-buli/hexane intercalation, etc.) have been tried by us, yet all have failed. There have been no related reports on the successful intercalation of ThNF and ThNCl. In order to reveal the underlying reason, in this paper, we report the detailed characterization of ternary thorium nitride ThNF and ThNCl. Via the refinement of their crystal structure, although ThNF and ThNCl share a similar structure with MNX (M = Ti, Zr, Hf; X = Cl, Br) compounds, the interaction between adjacent ThNF and ThNCl layers is not a van der Waals gap, which results in the unsuccessful intercalation along with the absence of superconductivity. Through the combination of X-ray photoelectron spectroscopy and band calculation results, the Th−N bond in ThNF is covalent, whereas the Th−N bond in ThNCl is ionic, in contrast to the ionic bond in Zr(Hf)−N in ZrNCl and HfNCl. Besides, the Th−F bond in ThNF is an ionic bond, while the Th−Cl bond in ThNCl is a partially covalent bond. For ThNF, the strong electronegativity of F ions leads to the bonding of Th to the F both in the nearest neighbor layer and in the next nearest neighbor layer, which results in the absence of a van der Waals gap between ThNF layers. For ThNCl, the long bond length of Th−Cl due to partially covalent bonding as well as the much more flat ThN layer promotes the idea that the Th atoms could bond with the Cl atoms in different layers. Our result provides a deeper understanding for the interlayer interaction in ternary layer structures metal nitride halides.

■

■

RESULTS AND DISCUSSIONS Figure 1 shows the Rietveld refinement result and crystal structure of ThNF. The X-ray diffraction pattern was collected at room temperature with the 2θ range of 5−120°. The Bragg peaks in the X-ray diffraction pattern could be well indexed by a hexagonal structure with space group of R3̅m. No obvious impurity phase is observed. The atom positions and lattice constants of ThNF could be found in Table 1, which is quite similar to those in previous reports.21,25 The more detailed structure parameters of ThNF could be found in the Supporting Information. As is shown in Figure 1(a), ThNF shares a similar crystal structure with β-type HfNCl and ZrNCl. The structure of ThNF is constituted by the double honeycomb ThN layers sandwiched by close-packed F layers. The stacking sequence along the c axis is F−Th−N−N−Th-F. As is shown in the right side of Figure 1(a), every Th atom could bond with three N atoms in the same plane and one N atom in the different plane, with bond distances of 2.610 and 2.575 Å, respectively. When compared with the Th−N bond in thorium nitride Th3N4 (2.308 Å, from ICSD 9052), the length of the Th−N bond (2.58 Å, 2.61 Å) in ThNF is much longer. Considering that the Th−N bond in Th3N4 is ionic, we infer that the Th−N bond in ThNF is a covalent type. The bond angle of N1−Th−N1 and N1−Th−N2 is 100.48° and 62.58°, respectively. In addition, every Th atom is surrounded by three F atoms in the nearest neighbor plane and one F atom in the next nearest neighbor plane, with a distance of 2.361 and 2.509 Å, respectively. Considering the little deviation of the two Th− F distances, the Th atom could bond with the F atoms in both sites, which is also proved by bond valence calculations (see the Supporting Information). The bond angle of F1−Th−F1 and F1−Th−F2 is 116.40° and 78.93°, respectively. Although ThNF shares a similar crystal structure with ZrNCl, there are many differences between them. First, compared with ZrNCl,

EXPERIMENTAL SECTION

ThNF and ThNCl samples are synthesized by a solid-state reaction. First, thorium metal is synthesized by the reduction of thorium dioxide by calcium. The powder of ThO2 (Micxy, 99.5%), calcium granules (Alfa Aesar, 99.5%), and CaCl2 (Alfa Aesar, 99.9%, anhydrous) are weighted according to the weight ratio of 5:2:2 and thoroughly ground. Then, the mixture is sealed into a stainless vessel under argon, heated to 950 °C, and kept for 24 h. The production is B

DOI: 10.1021/acs.inorgchem.9b01050 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

reasonable bond distance and confirms the existence of a van der Waals gap between adjacent ZrNCl layers. All the results mentioned above suggest that the interlayer interaction in ThNF is not a van der Waals gap, whereas ThNF shares a similar crystal structure with β-type MNX compounds. We also successfully synthesized another ternary thorium nitride, ThNCl. The Rietveld refinement result and crystal structure of ThNCl are shown in Figure 2. The X-ray

Figure 1. (a) The schematic view of the crystal structure of ThNF. (b) Measured (crosses) and calculated (red solid line) XRD patterns for ThNF. Bragg peak positions are indicated by short vertical bars. The bottom of the figure shows the differences between measured and calculated intensities.

Table 1. Crystallographic Parameters from the Powder XRD Refinement of ThNF at 300 Ka atom

site

x

y

z

occup.

Uiso (× 100 Å2)

Th N F

6c 6c 6c

0 0 0

0 0 0

0.23992(3) 0.3672(6) 0.1156(4)

1.00 1.00 1.00

0.87(1) 1.836(21) 4.520(35)

Figure 2. (a) The schematic view of the crystal structure of ThNCl. (b) Measured (crosses) and calculated (red solid line) XRD patterns for ThNF. Bragg peak positions are indicated by short vertical bars. The bottom of the figure shows the differences between measured and calculated intensities.

a

Space group: R3̅m (no. 166); a = b = 4.01246(5) Å, c = 20.2176(5) Å, Rwp = 0.1085, Rp = 0.0787, χ2 = 2.452.

diffraction pattern is collected at room temperature with the 2θ range of 5−120°. The Bragg peaks in the X-ray diffraction pattern could be well indexed by a tetragonal structure with space group of P4/nmm. A minor phase of ThO2 could be observed, which may originate from the oxidation during the synthesis process. The atom positions and lattice constants of ThNCl could be found in Table 2. The more detailed structure parameters of ThNCl could be found in the Supporting Information. Our refined crystal structure is quite similar to

the a-axis parameter of ThNF is enlarged to 4.013 Å, while the c-axis lattice parameter of ThNF is much shortened to 20.218 Å (the a-axis and c-axis parameter for ZrNCl is 3.597 and 27.548 Å, respectively), which suggests a much shorter distance between the adjacent X−M−N−N−M−X layers in ThNF relative to other β-MNX compounds. Indeed, the distance between adjacent F−Th−N−N−Th−F layers in ThNF is only 1.753 Å, while the distance between adjacent layers in ZrNCl could reach 3.07 Å.26 The much shorter distance between adjacent layers may indicate that the interlayer interaction in ThNF is not a van der Waals gap. Second, in ThNF, every Th atom could bond with three F atoms in the nearest neighbor plane and one F atom in the next nearest neighbor plane with a quite close bond distance, which further indicates that the interlayer interaction is not a van der Waals gap. However, in ZrNCl, every Zr atom could bond only with three Cl atoms in the nearest plane with bond distance of 2.745 Å.26 The distance between Zr atom and the next nearest neighbor Cl plane is 5.25 Å, which exceeds the

Table 2. Crystallographic Parameters from the Powder XRD Refinement of ThNCl at 300 Ka atom

site

x

y

z

occup.

Uiso (× 100 Å2)

Th N Cl

2c 2a 2c

0.25 0.75 0.25

0.25 0.25 0.25

0.16720(7) 0 0.6137(4)

1.00 1.00 1.00

0.012(6) 1.16(25) 0.14(7)

a

Space group: P4/nmm (no. 129); a = b = 4.09486(4) Å, c = 6.90128(15) Å, Rwp = 0.0734, Rp = 0.0529, χ2 = 1.932.

C

DOI: 10.1021/acs.inorgchem.9b01050 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry that in a previous report.21 ThNCl, whose crystal structure is composed of tetragonal ThN layers sandwiched by Cl layers, shares a similar crystal structure with α-type TiNCl.27 The stacking sequence along the c-axis is Cl−Th−N−N−Th−Cl. It is worth noting that the structure of ThNCl is quite similar to the 111 iron-based superconductors (LiFeAs and NaFeAs). As shown in Figure 2 (a), every Th atom could bond with four N atoms to form a Th−N tetrahedron with a bond distance of 2.35 Å and a N−Th−N angle of 121.19° (α) and 76.052° (β). Compared with ThNF (Th−N of 2.58 or 2.61 Å), the length of Th−N bond in ThNCl is much shorter. Moreover, the height of the ThN layer in ThNCl is only 2.359 Å, while the height in ThNF could reach 3.811 Å. This result shows that the ThN layer in ThNCl is much flater than that of ThNF. Furthermore, every Th atom is surrounded by five Cl atoms, four in the nearest neighbor plane and one in the next nearest neighbor plane, with a distance of 3.267 and 3.081 Å, respectively. The distance between Th and Cl in the next nearest neighbor plane is even shorter, which proves the existence of bonding between adjacent ThNCl layers and the absence of a van der Waals gap in ThNCl. Moreover, compared with the Th−Cl length in ThCl4 (2.797 Å, from ICSD 6055), the Th−Cl length in ThNCl is much longer (3.081 Å, 3.267 Å), which may suggest that the Th−Cl bond in ThNCl is partially covalent. Besides, the distance between adjacent ThNCl layers is only 1.812 Å, while the distance in TiNCl could reach 2.718 Å.27 The much shorter interlayer spacing also indicates that the interlayer interaction in ThNCl is not a van der Waals gap. After the analysis of the crystal structure of ThNF and ThNCl, we carried out a series of physical measurements on them. Both ThNF and ThNCl are insulators, and the color of their powders is white. In order to measure the band gap of the two compounds, UV−vis absorption spectroscopy is utilized, as shown in Figure 3. The band gap is calculated according to the Tauc equation: (αhν)2 = A(hν − Eg), in which α is the absorption factor, hν is the photon energy, A is a constant number, and Eg is the band gap.28 For ThNF, the measured band gap is 3.0 eV, which is in accordance with the band calculation.27 For ThNCl, the measured band gap is 3.79 eV. In contrast, the band gap of β-ZrNCl, HfNCl, and α-TiNCl is 2.1, 2.5, and 0.5 eV, respectively.27,29 Figure 4 shows the temperature-dependent magnetic susceptibility of ThNF and ThNCl under an external field of 1 T. Both of the curves show a weak temperature dependence, and there is no long-range magnetic order observed in ThNF and ThNCl at the temperature range of 2−300 K. In low temperature, the magnetic susceptibility shows a small Curie−Weiss tail, which may originate from the impurities in the sample. In order to further investigate the electronic structure of ThNF and ThNCl, X-ray photoelectron spectra measurements were carried out. Figure 5 is the Th 4f, N 1s, and F 1s regions of the XPS spectrum from ThNF. First, for the Th 4f region, the Th 4f7/2 peak is apparently asymmetric and contains more than one peak. After the background was deduced with the Shirley function, two peaks can be obtained. One is the high binding energy peak at 334.206 eV, and the other is the low binding energy peak at 333.26 eV. For the N 1s region, only one peak could be obtained at 395.14 eV. For the F 1s region, only a peak at 684.9 eV is obtained. For the high binding energy peak of Th 4f7/2 at 334.206 eV, the peak could be assigned to ThO2 due to slight oxidation in the surface of the sample, since the XPS spectra are surface sensitive. The

Figure 3. (a) The UV−vis absorption spectrum and calculated band gap for ThNF. (b) The UV−vis absorption spectrum and calculated band gap for ThNCl.

binding energy of N 1s (395.14 eV) is much lower than that of other ternary metal nitrides (for example, 396.67 eV for ZrNCl and 396.6 eV for ZrNBr).30,31 This is in accordance with the larger bond length of Th−N in ThNF, indicating that the Th− N is a covalent bond. Moreover, for other thorium compounds, the binding energy peak of Th 4f7/2 ranges from 334.3 eV in ThBr4 to 336.9 eV in ThF4. Thus, the binding energy of Th 4f7/2 is lower than those of these compounds, suggesting weak bonding between Th−N in ThNF. The binding energy of F 1s (684.9 eV) is in accordance with the binding energy of ionic bonding fluoride (for example, 684.9 eV for CaF2 and 684.9 eV for ThF4). Thus, from the analysis of XPS data of ThNF, we can conclude that the Th−N bond in ThNF is a covalent bond, while the Th−F bond is in accordance with the ionic bond. Figure 6 shows the Th 4f, N 1s, and Cl 2p regions of XPS spectrum from ThNCl. With a similar approach, we obtained two peaks from the Th 4f7/2 peak. One is the high binding energy peak at 334.76 eV, and the other is the low binding energy peak at 333.49 eV. For the N 1s region, only one peak could be obtained at 396.58 eV. For the Cl 2p3/2 region, one peak at 197.9 eV is obtained. For the high binding energy peak of Th 4f7/2 at 334.76 eV, the peak could be assigned to ThO2 due to slight oxidation in the surface of the sample as well as D

DOI: 10.1021/acs.inorgchem.9b01050 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

being consistent with the larger Th−Cl length in ThNCl and suggesting a partially covalent chemical picture of Th−Cl bond. Moreover, the binding energy peak of Th 4f7/2 (333.49 eV) is lower than that of other thorium compounds, suggesting weak bonding between Th−Cl in ThNCl. Thus, based on the analysis of XPS data of ThNCl, we can conclude that the Th− Cl bond in ThNCl is partially covalent, while the Th−N bond is ionic. Compared with β-type ZrNCl and HfNCl, the electronic structure of ThNF and ThNCl shows significant difference.29,32 Taking the ThNF as an example, the conduction band has a majority contribution from the 5f-orbitals of thorium rather than the d-orbitals as in the case of the Zr or Hf counterparts. This difference is caused by two unique properties. First, in the actinide series, the electronic configurations of thorium and protactinium are known to be 6d27s2 and 5f26d7s2, indicating a very small energy gap between the 6d and 5f orbitals. Second, compared with the Zr or Hf counterparts, where the (Zr, Hf)−(N, F) bonds are ionic, the Th−N bonding in the ThNF is within a covalent chemical picture, which creates a large crystal field and pushes the 6d-orbitals to high energy. Moreover, another significant difference between ThNF (ThNCl) and β-type ZrNCl and HfNCl is the different interlayer interaction. For β-type ZrNCl and HfNCl, the interlayer interaction is a van der Waals gap, while this is not true for ThNF and ThNCl. For ThNF, the strong electronegativity of F ions leads to the bonding of Th to the F both in the nearest neighbor layer and in the next nearest neighbor layer, which results in the absence of a van der Waals gap between ThNF layers. For ThNCl, since the ThN bond is of ionic type, the ThN layer is much more flat. Besides, the large length of partially covalent Th−Cl bond promotes the idea that the Th atoms could bond with the Cl atoms in the nearest neighbor layer and in the next nearest neighbor layer.

Figure 4. (a, b) The field cooling (FC) temperature-dependent magnetic susceptibility of ThNF and ThNCl, respectively. The external magnetic field is 1 T.

tiny impurity phase (ThO2) in the ThNCl sample. For ThNCl, the binding energy for N 1s is determined to be 396.58 eV, which is basically the same with that of ZrNCl (396.67 eV) and ZrNBr (396.6 eV).30,31 Besides, the length of the Th−N bond (2.35 Å) in ThNCl is also similar to that of Th3N4 (2.308 Å) and shorter than that of ThNF. Thus, we can conclude that the bond between Th−N in ThNCl is of ionic type. Moreover, the binding energy of the Cl 2p3/2 peak (197.7 eV) is lower than the binding energy in ionic bonding chloride (for example, 198.6 eV for NaCl and 199.2 eV for CuCl2). This results indicates the weak bonding between Th−Cl in ThNCl,

■

CONCLUSIONS Finally, in order to reveal the underlying reason for the unsuccessful intercalation, we conducted detailed characterization of ternary thorium nitride ThNF and ThNCl. Via the analysis of their crystal structure, although ThNF and ThNCl

Figure 5. XPS spectrum for ThNF. (a) The region of Th 4f5/2 and Th 4f7/2. (b) The region of N 1s. (c) The region of F 1s. E

DOI: 10.1021/acs.inorgchem.9b01050 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 6. XPS spectrum for ThNCl. (a) The region of Th 4f5/2 and Th 4f7/2. (b) The region of N 1s. (c) The region of Cl 2p1/2 and Cl 2p3/2.

■

share a similar structure with MNX (M = Ti, Zr, Hf; X = Cl, Br) compounds, the interaction between adjacent ThNF and ThNCl layers is not a van der Waals gap. Further, X-ray photoelectron spectroscopy reveals that the Th−N bond in ThNF is covalent, whereas the Th−N bond in ThNCl is ionic, in contrast to the ionic bond in Zr(Hf)−N in ZrNCl and HfNCl. The Th−F bond in ThNF is an ionic bond, while the Th−Cl bond in ThNCl is a partially covalent bond. For ThNF, the strong electronegativity of F ions leads to the bonding of Th to the F both in the nearest neighbor layer and in the next nearest neighbor layer, which results in the absence of a van der Waals gap between ThNF layers. For ThNCl, the long bond length of the partially covalent Th−Cl bond as well as the much more flat ThN layer promotes the idea that the Th atoms could bond with the Cl atoms in the nearest neighbor layer and in the next nearest neighbor layer, which results in the absence of a van der Waals gap between ThNCl layers. Thus, it is the absence of a van der Waals gap that leads to the failure of intercalating alkali-metals or organic molecules in ThNF and ThNCl, thereby to the unsuccessful attempt to realize the high-Tc superconductivity in the 5f-electron thorium-based MNX compounds. Our result provides a deeper understanding for the interlayer interaction in ternary layer structure metal nitride halides and motivates the further exploration of new metal nitride halides.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Xianhui Chen: 0000-0001-6947-1407 Present Address ⊥

Shenzhen Suntak Multilayer Circuit Board Co., Ltd., Shenzhen, Guangdong, 518054, China or the School of Materials and Energy, University of Electronic Science and Technology of China, Chengdu 610054, China.

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank Wuping Liao for providing us with the ThF4 sample. This work was supported by the National Key R&D Program of China (2016YFA0300201 and 2017YFA0303001), the National Natural Science Foundation of China (11534010 and 11888101), the Strategic Priority Research Program (B) of the Chinese Academy of Sciences (XDB25010100), the Key Research Program of Frontier Sciences, CAS, China (QYZDYSSW-SLH021), the Science Challenge Project (TZ2016004), and the Hefei Science Center CAS (2016HSC-IU001).

■

ASSOCIATED CONTENT

S Supporting Information *

REFERENCES

(1) Bednorz, J. G.; Muller, K. A. Possible High-Tc Superconductivity in the Ba-La-Cu-O System. Z. Phys. B: Condens. Matter 1986, 64 (2), 189−193. (2) Kamihara, Y.; Watanabe, T.; Hirano, M.; Hosono, H. Iron-based layered superconductor LaO1‑xFxFeAs (x = 0.05−0.12) with Tc = 26 K. J. Am. Chem. Soc. 2008, 130 (11), 3296−7. (3) Chen, X. H.; Wu, T.; Wu, G.; Liu, R. H.; Chen, H.; Fang, D. F. Superconductivity at 43 K in SmFeAsO1‑xFx. Nature 2008, 453 (7196), 761−2. (4) Yamanaka, S. High-T-c superconductivity in electron-doped layer structured nitrides. Annu. Rev. Mater. Sci. 2000, 30, 53−82. (5) Yamanaka, S.; Kawaji, H.; Hotehama, K.; Ohashi, M. A new layer-structured nitride superconductor. Lithium-intercalated betazirconium nitride chloride, LixZrNCl. Adv. Mater. 1996, 8 (9), 771.

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.inorgchem.9b01050. Bond valence calculations (PDF) Accession Codes

CCDC 1920161 and 1920750 contain the supplementary crystallographic data for this paper. These data can be obtained free of charge via www.ccdc.cam.ac.uk/data_request/cif, or by emailing [email protected], or by contacting The Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax: +44 1223 336033. F

DOI: 10.1021/acs.inorgchem.9b01050 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry (6) Taguchi, Y.; Kitora, A.; Iwasa, Y. Increase in Tc upon reduction of doping in LixZrNCl superconductors. Phys. Rev. Lett. 2006, 97 (10), 107001. (7) Yamanaka, S.; Hotehama, K.; Kawaji, H. Superconductivity at 25.5 K in electron-doped layered hafnium nitride. Nature 1998, 392 (6676), 580−582. (8) Yamanaka, S. Intercalation and superconductivity in ternary layer structured metal nitride halides (MNX: M = Ti, Zr, Hf; X = Cl, Br, I). J. Mater. Chem. 2010, 20 (15), 2922−2933. (9) Fogg, A. M.; Green, V. M.; O’Hare, D. Superconducting metallocene intercalation compounds of beta-ZrNCl. Chem. Mater. 1999, 11 (2), 216−217. (10) Zhang, S.; Tanaka, M.; Zhu, H. K.; Yamanaka, S. Superconductivity of layered beta-HfNCl with varying electron-doping concentrations and interlayer spacings. Supercond. Sci. Technol. 2013, 26 (8), 085015. (11) Ye, G. J.; Ying, J. J.; Yan, Y. J.; Luo, X. G.; Cheng, P.; Xiang, Z. J.; Wang, A. F.; Chen, X. H. Superconductivity in YbxMyHfNCl (M = NH3 and THF). Phys. Rev. B: Condens. Matter Mater. Phys. 2012, 86 (13), 134501. (12) Hotehama, K.; Koiwasaki, T.; Umemoto, K.; Yamanaka, S.; Tou, H. Effect of Swelling on the Superconducting Characteristics in Electron-Doped beta-ZrNCl and HfNCl. J. Phys. Soc. Jpn. 2010, 79 (1), 014707. (13) Tou, H.; Maniwa, Y.; Koiwasaki, T.; Yamanaka, S. Unconventional superconductivity in electron-doped layered Li0.48(THF)yHfNCl. Phys. Rev. Lett. 2001, 86 (25), 5775−5778. (14) Taguchi, Y.; Hisakabe, M.; Iwasa, Y. Specific heat measurement of the layered nitride superconductor LixZrNCl. Phys. Rev. Lett. 2005, 94 (21), 217002. (15) Ekino, T.; Takasaki, T.; Muranaka, T.; Fujii, H.; Akimitsu, J.; Ya ma na ka , S . T un neling spectr oscopy o f M gB 2 an d Li0.5(THF)yHfNCl. Phys. B 2003, 328 (1−2), 23−25. (16) Tou, H.; Maniwa, Y.; Yamanaka, S. Superconducting characteristics in electron-doped layered hafnium nitride: N-15 isotope effect studies. Phys. Rev. B: Condens. Matter Mater. Phys. 2003, 67 (10), 100509. (17) Uemura, Y. J.; Fudamoto, Y.; Gat, I. M.; Larkin, M. I.; Luke, G. M.; Merrin, J.; Kojima, K. M.; Itoh, K.; Yamanaka, S.; Heffner, R. H.; MacLaughlin, D. E. μSR studies of intercalated HfNCl superconductor. Phys. B 2000, 289, 389−392. (18) Stewart, G. R. Heavy-Fermion Systems. Rev. Mod. Phys. 1984, 56 (4), 755−787. (19) Joukovski, E.; Tugushev, V.; Avignon, M. Weak antiferromagnetism in the Kondo lattice. Phys. B 1999, 259−261, 225−226. (20) Pfleiderer, C. Superconducting phases of f-electron compounds. Rev. Mod. Phys. 2009, 81 (4), 1551−1624. (21) Juza, R.; Sievers, R. Nitridhalogenide des thoriums. Z. Anorg. Allg. Chem. 1968, 363, 258−272. (22) Deubner, H. L.; Rudel, S. S.; Kraus, F. A simple access to pure thorium (IV) halides (ThCl4, ThBr4, and ThI4). Z. Anorg. Allg. Chem. 2017, 643 (23), 2005−2010. (23) Larson, A. C.; Von Dreele, R. B. General Structure Analysis System (GSAS). Los Alamos National Laboratory Report LAUR 86− 748; The University of California: 1994. (24) Toby, B. H. EXPGUI, a graphical user interface for GSAS. J. Appl. Crystallogr. 2001, 34, 210−213. (25) Silva, G. W. C.; Yeamans, C. B.; Cerefice, G. S.; Sattelberger, A. P.; Czerwinski, K. R. Synthesis and Nanoscale Characterization of (NH4)4ThF8 and ThNF. Inorg. Chem. 2009, 48 (13), 5736−5746. (26) Shamoto, S.; Kato, T.; Ono, Y.; Miyazaki, Y.; Ohoyama, K.; Ohashi, M.; Yamaguchi, Y.; Kajitani, T. Structures of beta-ZrNCl and superconducting Li0.16ZrNCl: double honeycomb lattice superconductor. Phys. C 1998, 306 (1−2), 7−14. (27) Yamanaka, S.; Yasunaga, T.; Yamaguchi, K.; Tagawa, M. Structure and superconductivity of the intercalation compounds of TiNCl with pyridine and alkali metals as intercalants. J. Mater. Chem. 2009, 19 (17), 2573−2582.

(28) Tauc, J. Optical properties and electronic structure of amorphous Ge and Si. Mater. Res. Bull. 1968, 3 (1), 37−46. (29) Felser, C.; Seshadri, R. Electronic structures and instabilities of ZrNCl and HfNCl: implications for superconductivity in the doped compounds. J. Mater. Chem. 1999, 9 (2), 459−464. (30) Feng, F.; Guo, H.; Li, D.; Wu, C.; Wu, J.; Zhang, W.; Fan, S.; Yang, Y.; Wu, X.; Yang, J.; Ye, B.; Xie, Y. Highly efficient photothermal effect by atomic-thickness confinement in two-dimensional ZrNClnanosheets. ACS Nano 2015, 9 (2), 1683−1691. (31) Wang, X.; Zhang, S.; Fu, H.; Gao, M.; Ren, Z.; Chen, G.-F. Dominant role of processing temperature in electric field induced superconductivity in layered ZrNBr. New J. Phys. 2019, 21, 023002. (32) Matar, S. F.; Pouchard, M. Ab initio investigation of the crystal and electronic structures of the nitride fluoride ThNF. Solid State Sci. 2013, 18, 123−126.

G

DOI: 10.1021/acs.inorgchem.9b01050 Inorg. Chem. XXXX, XXX, XXX−XXX