Electron-deficient copper pnictides: A₂Mg₃Cu₉Pn₇ (A = Sr, Eu; Pn = P, As) and Eu₅Mg_2.39Cu_16.61As₁₂

Abstract

A series of new magnesium copper pnictide intermetallics, A₂Mg₃Cu₉Pn₇ (A = Sr, Eu; Pn = P, As) and Eu₅Mg_2.39Cu_16.61As₁₂, were synthesized through high temperature flux reactions. By the single-crystal X-ray diffraction technique, their structures were accurately determined. Sr₂Mg₃Cu₉P₇, Eu₂Mg₃Cu₉P₇, Sr₂Mg₃Cu₉As₇ and Eu₂Mg₃Cu₉As₇ crystallize in the Zr₂Fe₁₂P₇ structure type (S.G. P6̄, no. 174) with cell parameters: a = 9.7736(19)/9.7417(7)/10.0550(10) Å/10.0410(8) Å, c = 3.9280(16)/3.9008(5)/4.0544(9)/4.0364(6) Å, respectively, while Eu₅Mg_2.39Cu_16.61As₁₂ adopts the Zr₅Co₁₉P₁₂ structure type (S.G. P6̄2m, no. 189) with cell parameters of a = 13.2521(6) Å and c = 4.0910(3) Å. These compounds can be viewed as an extensive branch of the ternary RE−M−Pn system (RE = rare-earth elements; M = transition metals; Pn = pnictogen) with the metal-to-nonmetal ratio being close or equal to 2 : 1. Interestingly, with the divalent alkaline-earth cations being introduced, these metal-rich phases exhibit electron deficiency according to the DFT calculations. Eu₂Mg₃Cu₉As₇ is metallic based on the resistivity measurement, consistent with theoretical prediction, and the magnetic susceptibility data of Eu₂Mg₃Cu₉As₇ confirmed the divalent oxidation state and weak ferromagnetic coupling for the europium cations.

---

Introduction

Zintl phases are polar intermetallics composed of metals with very different electronegativities. Usually, in these compounds valence electrons are assumed to be completely transferred from the cations to anions, which result in a closed-shell configuration for all constituent atoms.^1–4 Thus, Zintl phases are typically semi-conductors or insulators. However, electron deficiency can also exist in polar intermetallics and such an idea has been demonstrated by Sr₃In₅,⁵ Ce_5−xCa_xGe₄,⁶ Ca₅MgAgGe₅,⁷ Rb₁₆Cd_25.39(3)Sb₃₆,⁸ Rb₄Zn₇As₇ and Rb₇Mn₁₂Sb₁₂.⁹ In particular, with the Zintl concept extended and the transition metals included, the crystal and electronic structures of such analogues can be even more complicated, which result in interesting physical properties. For example, superconductivity was reported by hole-doping Ba_1−xK_xFe₂As₂.¹⁰ Similarly, by delicately tuning the hole-concentration of Yb₁₄Mn_1−xAl_xSb₁₁, high thermoelectric performance was proven.¹¹ For the A₉T_4+xPn₉ series (A = divalent rare-earth cations, T = d⁵ or d¹⁰ transition metals, Pn = As, Sb, Bi), it is obvious that the crystal structures and thermoelectric properties are closely related to the electron-deficient levels in the structure.^12–16

The Re–M–Pn ternary phases (RE = rare-earth elements; M = transition metals; Pn = pnictides) belong to a huge intermetallic system, with which interesting physical properties such as catalysis and magnetic properties were frequently investigated.^17,18 Among these compounds, the metal-rich analogues, whose metal-to-nonmetal ratio is close or equal to 2 : 1, represent a large family bearing various derivative structures related to the binary phosphide M₂P. As a typical extensive branch, Zr₂Fe₁₂P₇-type compounds were also frequently probed recently for their possible ordered structures when the Fe sites are occupied by two different transition metals.¹⁹ Although there have been many ternary members of such a structure, to date there are only a few examples of the quaternary pnictide phases, i.e., RE₂M₃M′₉Pn₇ (RE = Gd, Er; M = Cr, W; M′ = Fe, Co; Pn = P, As).²⁰ It was not until recently that the ordered rare-earth variants RE₂Mn₃Cu₉Pn₇ (Pn = P, As) have been reported²¹ and no such alkaline-earth examples have been discovered to date.

In this work, we successfully obtained four structure-ordered quaternary pnictide compounds, A₂Mg₃Cu₉Pn₇ (A = Sr, Eu; Pn = P, As). The substitution of the trivalent rare-earth elements by divalent cations results in electron deficiency in the compound, according to the electronic band structure calculations. With the above success, we continued to extend the synthesis and interestingly, discovered a new compound Eu₅Mg_2.39Cu_16.61As₁₂, derived from the parent structure Ln₅Cu₁₉P₁₂ (Ln = La, Ce) which typically forms with trivalent rare-earth cations.²² These new quaternary phases are very interesting since they probably imply new applications such as thermoelectrics for their tunable structures. Moreover, the replacement of the rare-earth cations by alkaline-earth cations will also provide easy cases for the studies of electronic band structures of these phases, as it is obvious that the handling of rare-earth cations in theoretical calculations is much more complicated. Magnetic and electrical conducting properties of Eu₂Mg₃Cu₉As₇ are reported as well.

Experimental

Synthesis

The reactants strontium ingot (Alfa, 99.5%), europium ingot (Alfa, 99.9%), magnesium slice (Alfa, 99.99%), copper grain (Alfa, 99.999%), phosphorus piece (Alfa, 99.999%), arsenic piece (Alfa, 99.999%) and lead grain (Alfa, 99.99%) were all purchased and used as received. All preparation processes were carried out in an argon-filled glovebox with an oxygen level <1 ppm or under vacuum.

Pb-flux reactions were utilized to synthesize the title compounds. Initially, Eu₂Mg₃Cu₉As₇ was obtained from the reaction mixture with the atomic ratio Eu : Mg : Cu : As : Pb = 0.4 : 0.6 : 1 : 1 : 15, which was loaded in an alumina crucible and then subsequently sealed in a silica tube by using a flame under vacuum. The reaction mixture was first heated to 1173 K at a rate of 200 K h⁻¹ and then kept at this temperature for 24 h; after this homogenization process, the furnace was slowly cooled down to 773 K at a rate of 5 K h⁻¹ and at this temperature the excess molten lead was decanted quickly by centrifugation. After the structures and compositions of these compounds were accurately determined by single-crystal X-ray diffraction, they were all reproduced with the stoichiometric ratio of A₂Mg₃Cu₉Pn₇. However, an optimization of the reaction procedure with a slightly different ratio of Eu : Mg : Cu : As : Pb = 0.2 : 0.3 : 0.8 : 0.7 : 10 also resulted in the discovery of a new phase Eu₅Mg_2.39Cu_16.61As₁₂. Attempts to synthesize other analogues with such a structure type failed. The products of these compounds are typically black needle-shaped crystals. Although various synthetic procedures have been tried, large crystals suitable for property measurements can only be obtained from the compound Eu₂Mg₃Cu₉As₇.

Single-crystal X-ray diffraction and structural refinement

For structure determination, single crystals were mounted on the tip of a glass capillary using epoxy resin AB glue adhesive. The data were collected with a Bruker SMART APEX-II CCD area detector on a D8 goniometer using graphite-monochromated Mo Kα radiation (λ = 0.71073 Å). As all the title compounds are stable in air, this process was performed without any inert gas protection at room temperature. Data reduction and integration, together with global unit cell refinements were performed by the INTEGRATE program of the APEX2 software.²³ Structure refinements were carried out with the use of the SHELX (version 6.12) program package.²⁴ Anisotropic atomic displacement parameters are applied for all atoms.

For compound Eu₂Mg₃Cu₉As₇, a non-centrosymmetric space group P6̄ (no. 174) was suggested by the program and the structure was solved by direct methods. These results are also consistent with the previous structure solution on the rare-earth cation analogues. For compound Eu₅Mg_2.39Cu_16.61As₁₂, a chiral space group P3 was first suggested, and the structure refinements with anisotropic thermal parameters were converged quickly to a small R value with a reasonable Flack parameter as well. However, later checking by using the PLATON program with the WinGx system, ver. 1.15 ²⁵ indicates higher symmetry in a new space group P6̄2m, which turns the structure exactly into the Ln₅Cu₁₉P₁₂ type, but with the Cu4 site statistically occupied by 0.203(8) Cu and 0.797(8) Mg. Note that the structure solution based on the lower symmetric space group P3 also leads to a similar disorder in the structure. The site occupancies of all atoms except for the mixed Mg/Cu were checked with the freed occupancies, which are very close to 100%.

Important information on the data collection and structure refinement of A₂Mg₃Cu₉Pn₇ (A = Sr, Eu; Pn = P, As) and Eu₅Mg_2.39Cu_16.61As₁₂ is summarized in Table 1. Standardized atomic coordinates and isotropic atomic displacement parameters of Eu₂Mg₃Cu₉As₇ and Eu₅Mg_2.39Cu_16.61As₁₂ are collected in Table 2. The atomic coordinates and isotropic displacement parameters of Sr₂Mg₃Cu₉P₇, Sr₂Mg₃Cu₉As₇ and Eu₂Mg₃Cu₉P₇ are provided in the ESI.† Further information in the form of CIF has been deposited with Fachinformationszentrum Karlsruhe, 76344 Eggenstein-Leopoldshafen, Germany, (fax: (49) 7247-808-666; e-mail: crysdata@fiz-karlsruhe.de) – depository CSD-number 431458-431462 for Eu₂Mg₃Cu₉As₇, Eu₂Mg₃Cu₉P₇, Sr₂Mg₃Cu₉As₇, Sr₂Mg₃Cu₉P₇, and Eu₅Mg_2.39Cu_16.61As₇, respectively. Important bonding distances are tabulated in Tables 3 and 4.

Table 1 Selected crystal data and structure refinement parameters for A₂Mg₃Cu₉Pn₇ (A = Sr, Eu; Pn = P, As) and Eu₅Mg_2.39Cu_16.61As₁₂

a R 1 = ∑||Fo| − |Fc||/∑|Fo|; wR2 = [∑[w(Fo^2 − Fc^2)^2]/∑[w(Fo^2)^2]]^1/2, and w = 1/[σ^2Fo^2 + (A·P)^2 + B·P], P = (Fo^2 + 2Fc^2)/3; A and B are weight coefficients.
Formula	Sr2Mg3Cu9P7	Sr2Mg3Cu9As7	Eu2Mg3Cu9P7	Eu2Mg3Cu9As7	Eu5Mg2.39Cu16.61As12
fw/g mol^–1	1036.82	1344.47	1165.50	1473.15	2772.34
T/K			296(2)
Radiation			Mo-Kα, 0.71073 Å
Space group	P6̄ (no. 174)	P6̄ (no. 174)	P6̄ (no. 174)	P6̄ (no. 174)	P6̄2m (no. 189)
Cell dimensions
a/Å	9.7736(19)	10.0550(10)	9.7417(7)	10.0410(8)	13.2521(6)
c/Å	3.9280(16)	4.0544(9)	3.9008(5)	4.0364(6)	4.0910(3)
V/Å^3	324.95(16)	354.99(9)	320.59(5)	352.43(7)	622.20(6)
ρ calc./g cm^−3	5.298	6.289	6.037	6.941	7.399
μ Mo Kα/ cm^−1	1.750	2.262	1.980	4.021	4.556
Final R indices^a [I > 2σ(I)]	R 1 = 0.0410	R 1 = 0.0234	R 1 = 0.0218	R 1 = 0.0258	R 1 = 0.0183
	wR2 = 0.0704	wR2 = 0.0413	wR2 = 0.0439	wR2 = 0.0495	wR2 = 0.0416
Final R indices^a [all data]	R 1 = 0.0590	R 1 = 0.0283	R 1 = 0.0246	R 1 = 0.0306	R 1 = 0.0190
	wR2 = 0.0774	wR2 = 0.0425	wR2 = 0.0461	wR2 = 0.0512	wR2 = 0.0418

---

Table 2 Refined atomic coordinates and isotropic displacement parameters for Eu₂Mg₃Cu₉As₇ and Eu₅Mg_2.39Cu_16.61As₁₂

Atoms	Wyckoff	x	y	z	U eq.^a (Å^2)
a U eq. is defined as one-third of the trace of the orthogonalized U^ijtensor.
Eu2Mg3Cu9As7
Eu1	1f	2/3	1/3	0.5	0.0109(3)
Eu2	1c	1/3	2/3	0	0.0087(3)
Mg	3k	0.2251(6)	0.1097(6)	0.5	0.0125(12)
Cu1	3k	0.3793(2)	0.4310(2)	0.5	0.0136(4)
Cu2	3j	0.1597(2)	0.2860(2)	0	0.0125(4)
Cu3	3j	0.4225(3)	0.0556(3)	0	0.0151(5)
As1	3k	0.1165(2)	0.41123(19)	0.5	0.0086(3)
As2	3j	0.4196(2)	0.30701(19)	0	0.0078(3)
As3	1a	0	0	0	0.0081(5)
Eu5Mg2.39Cu16.61As12
Eu1	3g	0.80974(4)	0	0.5	0.00848(15)
Eu2	2c	1/3	2/3	0	0.00906(18)
Cu1	6k	0.36067(11)	0.48053(11)	0.5	0.0142(3)
Cu2	6j	0.18242(10)	0.37656(11)	0	0.0138(2)
Cu3	3g	0.27944(12)	0	0.5	0.0129(3)
Cu4/Mg4	3f	0.4300(2)	0	0	0.0086(9)
Cu5	1a	0	0	0	0.0102(6)
As1	6k	0.17035(8)	0.48024(8)	0.5	0.0084(2)
As2	3f	0.17749(11)	0	0	0.0082(3)
As3	3f	0.63341(10)	0	0	0.0090(3)

---

Table 3 Important interatomic distances (Å) in A₂Mg₃Cu₉Pn₇ (A = Sr, Eu; Pn = P, As)

Atom pairs			Distances (Å)
		Sr2Mg3Cu9P7	Sr2Mg3Cu9As7	Eu2Mg3Cu9P7	Eu2Mg3Cu9As7
A1–	Pn2 × 6	3.062(4)	3.136(1)	3.010(3)	3.105(1)
A2–	Pn1 × 6	3.069(4)	3.158(1)	3.032(3)	3.132(1)
Mg1–	Pn1	2.583(8)	2.681(4)	2.577(6)	2.690(5)
	Pn2 × 2	2.733(6)	2.827(3)	2.728(4)	2.818(4)
	Pn3 × 2	2.738(4)	2.816(3)	2.724(3)	2.812(4)
Cu1–	Pn1	2.419(6)	2.504(2)	2.404(4)	2.493(3)
	Pn1	2.506(6)	2.554(2)	2.494(4)	2.545(2)
	Pn2 × 2	2.416(4)	2.503(1)	2.428(3)	2.508(2)
Cu2–	Pn1 × 2	2.450(3)	2.535(1)	2.441(2)	2.527(1)
	Pn2	2.404(6)	2.503(2)	2.428(4)	2.510(3)
	Pn3	2.404(2)	2.486(1)	2.403(2)	2.493(2)
Cu3–	Pn1 × 2	2.482(4)	2.546(1)	2.480(3)	2.547(2)
	Pn2	2.403(5)	2.492(2)	2.402(4)	2.480(3)
	Pn2	2.513(5)	2.551(2)	2.464(4)	2.539(3)

---

Table 4 Important interatomic distances (Å) in Eu₅Mg_2.39Cu_16.61As₁₂

Atom pairs		Distances (Å)	Atom pairs		Distances (Å)
Eu1–	As2 × 4	3.185(1)	Eu2–	As1 × 5	3.101(1)
	As3 × 2	3.106(1)		As1	3.101(1)
Cu1–	As1	2.511(2)	Cu3–	As1 × 2	2.484(1)
	As1	2.520(2)		As2 × 2	2.451(1)
Cu2–	As1 × 2	2.513(1)	Cu4/Mg4–	As1 × 4	2.867(1)
	As2	2.535(2)		As3	2.696(3)
	As3	2.509(1)	Cu5–	As2 × 3	2.352(1)

---

Elemental analysis

Energy Dispersive X-ray Spectroscopy (EDS) was taken on a picked single crystal of Eu₅Mg_2.39Cu_16.61As₁₂ with a Hitachi FESEM-4800 field emission microscope equipped with a Horiba EX-450 EDS. The measured composition is very close to the result obtained from the single-crystal X-ray diffraction data (ESI†).

Magnetic measurements

Since the crystals of all compounds are generally very small and the yields are also very poor, the magnetic susceptibility measurements were only carried out for Eu₂Mg₃Cu₉As₇ as the product is relatively easy to accumulate. Field-cooled direct-current (DC) magnetization measurements for a polycrystalline sample of Eu₂Mg₃Cu₉As₇ (8.231 mg) were performed by using a MPMS Quantum Design SQUID magnetometer. The applied magnetic field was 500 Oe, and the measured temperature interval was from 5 to 300 K. The collected data were converted to molar magnetic susceptibility χ(T) = M/H in the emu mol⁻¹ unit.

Powder X-ray diffraction

The powder X-ray diffraction pattern of Eu₂Mg₃Cu₉As₇ was recorded at room temperature with a Bruker AXS X-ray powder diffractometer using Cu Kα radiation. The results were used for phase purity analysis only and the data acquisition was performed by using Bruker software with a step size of 0.04° in the 2θ mode (ESI†).

Electrical resistivity

The electrical resistivity of a single crystal of Eu₂Mg₃Cu₉As₇ (needle, 0.2 × 0.2 × 2 mm³) was measured on a Quantum Design Physical Property Measurement System (PPMS) equipped with an AC transport controller (model 7100). The crystal was mounted on four very thin copper wires glued to the needle using a highly conducting silver epoxy paste for standard four-probe electrical resistivity measurements. The measured temperature range was from 100 to 300 K and a current of 100 μA was used.

Electronic structure calculations

Computation of the total density of states (TDOS) and electronic band structure was carried out by Wien2k²⁶ with the full potential linearized augmented plane wave method (FP-LAPW).^27,28 In order to better understand the structure and bonding of these compounds, the program “LMTO 4.7”²⁹ with the linear muffin-tin orbital (LMTO) method³⁰ was used to calculate the crystal Hamilton orbital population (COHP). Two model compounds, Sr₂Mg₃Cu₉As₇ and Eu₅Cu₁₉As₁₂ were chosen for such a theoretical investigation. In the latter, the Eu 4f electrons were treated as the core states.

Results and discussion

Structure description

The RE₂M₃M′₉Pn₇ series are derived from the Zr₂Fe₁₂P₇ structure and with the substitution of Fe by two kinds of transition metals, new ordered structures may result if these two metals prefer different coordination geometries. Such an idea was demonstrated recently by RE₂Mn₃Cu₉Pn₇ (Pn = P, As), in which the Mn atoms occupy the site coordinated by five Pn atoms in a square-pyramidal geometry and the Cu atoms reside in the center of a tetrahedron composed of four Pn atoms.²¹ The relevant synthesis and discovery of new compounds have been well conducted with large RE cations (La–Nd, Sm, Gd–Dy), however, the inclusion of divalent alkaline-earth atoms in this structure is a first. After all, these two types of cations obviously represent different charges and in such cases structural transformation is usually expected. Surprisingly, the four title compounds reported in this work, Sr₂Mg₃Cu₉P₇, Eu₂Mg₃Cu₉P₇, Sr₂Mg₃Cu₉As₇ and Eu₂Mg₃Cu₉As₇, are actually isostructural with RE₂Mn₃Cu₉Pn₇. This phenomenon is very interesting because it probably suggests an unexpected electronic structure and even mixed-valency for these compounds. Another benefit of these compounds lies in the significantly different atomic numbers between Mg and Cu, for which the problem of possible site mixing can be readily verified through structural refinements.

Eu₅Mg_2.39Cu_16.61As₁₂ was discovered incidentally in the reactions aiming at optimizing the synthetic procedure of Eu₂Mg₃Cu₉As₇. The structure of this compound can be viewed as derived from the Zr₅Co₁₉P₁₂ type.³¹ The flexibility of valence electrons in such ternary intermetallics has been well demonstrated by previous references.^32–36 The components of these compounds are various, for example, the cations can be rare-earth, alkaline-earth or transition metals, whereas the anions may come from the pnictogen or tetrel elements. Typical ternary analogues include but are not limited to RE₅Ir₁₉P₁₂,³² RE₅Ru₁₉P₁₂,³³ and Ca₅T₁₉P₁₂ (RE = rare-earth metals; T = Rh, Ir),³⁴ to name just a few. In particular, compound Sr₅Mg₁₉Ge₁₂ can be described as a Zintl phase with the electrons rationalized precisely, [Sr²⁺]₅[Mg²⁺]₁₉[Ge⁴⁻]₁₂,³⁵ and small homogeneity ranges were also suggested on the alkaline-earth positions for A_5+xMg_18−xTt₁₃ (A = Sr, Ba; Tt = Si, Ge).³⁶ Compared to other transition metals, the RE–Cu–Pn ternary systems were rarely investigated and only two phosphides, Ln₅Cu₁₉P₁₂ (Ln = La, Ce), were reported.²² In addition, the related quaternary derivatives of RE₅M₃M′₁₆Pn₁₂ were also very few, and only the Ln₅Zr₃Ni₁₆As₁₂ (Ln = La and Y) series was probed so far.³⁷ With the above successes in configuring the ordered structure of A₂Mg₃Cu₉Pn₇ (A = Sr, Eu; Pn = P, As), the Eu₅Mg_2.39Cu_16.61As₁₂ is also interesting since in this compound, the mixing between Mg and Cu atoms is significant and also the existence of the non-stoichiometric composition of this compound may indicate a similar electron deficiency in the structure, governed by the Mg/Cu concentrations at the mixed site.

A₂Mg₃Cu₉Pn₇ (A = Sr, Eu; Pn = P, As) and Eu₅Mg_2.39Cu_16.61As₁₂ all crystallize in hexagonal crystal systems, while their space groups are different in that the former belongs to P6̄ (no. 174), the latter belongs to P6̄2m (no. 189), and their crystallographic data are presented in Table 1. They all represent a large family bearing various derivative structures related to the binary phosphide M₂P, whose metal-to-nonmetal ratio is close or equal to 2 : 1. The pnictides A₂Mn₃Cu₉Pn₇ (Pn = P, As) are quaternary derivatives of the hexagonal Zr₂Fe₁₂P₇-type structure, and the copper-rich Eu₅Mg_2.39Cu_16.61As₁₂ adopts the Zr₅Co₁₉P₁₂ structure type. Their structures viewed in projection down the c-axis are shown in Fig. 1, in which CuPn₄ tetrahedrons define the overall framework of the structure. These CuPn₄ tetrahedral units are connected through a corner- or edge-sharing manner. In both structures, tunnels form along the c-axis with the alkaline-earth cations filled into the cavities in the anionic framework. Although these two structures are similar in the anionic fragments, their cation coordination environments are subtly different. For A₂Mg₃Cu₉Pn₇, the Mg and Cu atoms are ordered, and occupy different crystallographic sites and feature very different coordination geometries, whereas for Eu₅Mg_2.39Cu_16.61As₁₂, the Mg cations are partially substituted by Cu atoms and in such a case, this structure can be readily viewed as derived from the Zr₅Co₁₉P₁₂ type³¹ with the Mg/Cu statistically disordered at the Co sites.

Fig. 1 Polyhedral and ball-and-stick view of crystal structure of A₂Mg₃Cu₉Pn₇ (A = Sr, Eu; Pn = P, As) (a) and Eu₅Mg_2.39Cu_16.61As₁₂ (b), viewed in the projection down the c-axis.

A more clear illustration of the coordination geometries of various cations as well as the Cu atoms is presented in Fig. 2. For typical alkaline-earth or rare-earth cations such as Sr and Eu, they all exhibit a trigonal prismatic coordination fashion with CN = 6 in both structures. Interestingly, Mg has a relatively smaller coordination number (CN = 5) in these compounds, which adopts the square pyramidal coordination geometry and is reasonable in consideration of its relatively smaller cationic size compared to Sr or Eu. The Cu atoms generally form strong covalent bonds with the pnictide anions coordinated and they prefer the tetrahedral sites with CN = 4. As a typical transition metal, Cu atoms may also act as cations such as in Eu₅Mg_2.39Cu_16.61As₁₂, for which the square pyramidally coordinated Mg can be partially substituted by Cu. As mentioned above, the atomic numbers between Mg and Cu are significantly different, which makes the structure solution of these phases by X-ray diffraction easily conducted. In addition, the possible Cu/Mg mixing at other tetrahedral or trigonal planar sites can be evidentially ruled out. The composition of Eu₅Mg_2.39Cu_16.61As₁₂ from the crystallographic data coincides with the results of EDX analyses as well.

Fig. 2 Coordination geometries for Cu and Mg atoms in A₂Mg₃Cu₉Pn₇ (A = Sr, Eu; Pn = P, As) (a): Cu (CN4, tetrahedral), Mg (CN5, square pyramidal) atoms, and in Eu₅Mg_2.69Cu_16.31As₁₂ (b): Cu (CN4, tetrahedral), Cu/Mg(CN5, square pyramidal), Cu(CN3, trigonal planar). The coordination geometries for Sr and Eu cations are provided for comparison as well.

All CuPn₄ tetrahedrons in A₂Mg₃Cu₉Pn₇ are slightly distorted and the corresponding Cu–As distances fall into the reasonable single-bond distances: d(Cu1–Pn) = 2.404(4)–2.5536(18) Å, d(Cu2–Pn) = 2.4034(15)–2.5351(11) Å and d(Cu3–Pn) = 2.402(4)–2.551(2) Å, respectively. However, for Eu₅Mg_2.39Cu_16.61As₁₂ the Cu atoms also feature square pyramidal and trigonal planar coordination besides the tetrahedral fashion, as illustrated by Cu4 and Cu5 atoms in Fig. 2b. The Cu–As bonding distances in this compound are also in a relatively broad range of 2.3522(14)–2.8671(8) Å. Compared with the regular interatomic distances of Cu–As pairs in other copper-rich analogues such as 2.411–2.595 Å in BaCu₈As₄,³⁸ 2.222(2) Å in Ba₂Cu_16.33As₁₀,³⁹ these values are still reasonable except for the Cu4–As1 (2.8671(8) Å) and Cu4–As3 (2.696(3) Å). However, taking into account the special square pyramidal coordination geometry of the Cu4 cation, these interactions obviously cannot be simply treated as covalent single bonds. According to references, displacive disorder was also found in this type of structure with the transition metals at the same Wyckoff position.^40,41 In the current case, the Cu4 site in Eu₅Mg_2.39Cu_16.61As₁₂ was statistically occupied by 0.203(8) Cu and 0.797(8) Mg, suggested by the single crystal X-ray diffraction data and EDX analyses.

Electronic structure

Sr₂Mg₃Cu₉As₇ and a hypothetical model of Eu₅Cu₁₉As₁₂ were used to calculate the density of states (DOS), the crystal orbital Hamilton population (COHP) and electronic band structures. The results are compared side-by-side in Fig. 3 and 4, respectively.

Fig. 3 Calculated total density of states (TDOS) and partial density of states (PDOS) for Sr₂Mg₃Cu₉As₇ (a) and the hypothetical compound Eu₅Cu₁₉As₁₂ (b). Crystal orbital Hamilton population (COHP) analyses are also provided for Sr₂Mg₃Cu₉As₇ (c) and Eu₅Cu₁₉As₁₂ (d). The dotted line marks the Fermi level.

Fig. 4 Calculated electronic band structures of the Sr₂Mg₃Cu₉As₇ (a) and Eu₅Cu₁₉As₁₂ (b). Fermi level is chosen as the energy reference.

The calculated total and projected density of states (TDOS and PDOS) are shown in Fig. 3a. Both compounds show a pseudo-gap in the TDOS curves and below the energy gap the valence bands are crossed by the Fermi level, indicating an electron deficient character for these structures. In the vicinity of the Fermi level, all constituent atoms have significant contributions except for Cu. In addition, the d-orbitals of Cu atoms are almost localized in an energy region from −5.0 to −2.0 eV. Such a characteristic indicates that in these compounds, the Cu anion can be readily treated as Cu⁺ and based on such a conclusion, the electron counting of these compounds can be established, which results in electron-deficient rationalization in both cases: [Sr²⁺]₂[Mg²⁺]₃[Cu⁺]₉[As³⁻]₇ and [Eu²⁺]₅[Cu⁺]₁₉[As³⁻]₁₂. The former is only 2-electron deficient, but the latter will be 7-electron deficient according to this hypothetical mode. In such a situation, if the Cu⁺ is substituted by Mg²⁺ as in Eu₅Mg_2.39Cu_16.61As₁₂, the electron-deficiency numbers can be reduced to 4.61. If normalized by the compounds’ formulae, the deficiency numbers are 0.10 and 0.13 per atom for Sr₂Mg₃Cu₉As₇ and Eu₅Mg_2.39Cu_16.61As₁₂, respectively, which are still comparable to each other.

The electron deficiency in these compounds can be more informative with the bonding analysis of these compounds. As illustrated in the corresponding COHP curves (Fig. 3b), all Cu–As bonds are well optimized in both structures, however, for the interactions between the alkaline-earth or rare-earth cations and pnictogen anions, the Fermi level crosses a region with significantly bonding interactions. These results indicate that the electron deficiency mainly involves the bonds of A–As, and the physical properties of these compounds should be predominantly governed by the alkaline-earth and pnictogen atoms in the compounds. The electron deficiency can be further illustrated by the calculated electronic band structures, as shown in Fig. 4. It is very clear that the Fermi level crosses the top of the valence bands with a pocket open at G. According to the band structure, the band gaps are almost close in both compounds, corresponding to the pseudo-gaps appearing in their DOS curves. This means that these compounds may be poor-metallic or semi-conducting. Since for LDA or GGA calculations, the underestimation of the band gap is very common, these discrepancies do not impede the assertion that these phases can be considered as Zintl phases.

Magnetic susceptibility measurement

The magnetic susceptibility data of Eu₂Mg₃Cu₉As₇ is shown in Fig. 5 over a temperature range from 5 and 300 K and under an applied field of 500 Oe. In the low temperature region (below 20 K), a rapid increase in susceptibility indicates that Eu₂Mg₃Cu₉As₇ bears ferromagnetic ordering in the structure. The ferromagnetic ordering can be more clearly seen by plotting the inverse magnetic susceptibilities versus temperature, which is also shown as an inset picture. Above 20 K, the inverse susceptibility, χ⁻¹(T), follows the paramagnetic Curie–Weiss law χ(T) = C/(T − θ_P), where C is the Curie constant and θ_P is the Weiss temperature. The effective magnetic moment μ_eff was calculated from the equation μ_eff = (8C)^1/2μ_B.⁴² As deduced from the linear regression, the θ value for Eu₂Mg₃Cu₉As₇ is 12.97 K. This positive θ value also supports the above argument that the magnetic coupling among the Eu²⁺ ions is ferromagnetic. Similar weak ferromagnetism has been frequently reported in other europium pnictides such as Eu₃Tt₂As₄ (Tt = Si, Ge),⁴³ Eu₂CdAs₂,⁴⁴ Eu₅Sn₂As₆ ⁴⁵ and Eu₃Ga₂P₄.⁴⁶ The calculated effective magnetic moment is 8.05μ_B per Eu atom, which is a little different from, but still in good agreement with the Hund's rule derived value of 7.94μ_B for the free-ion Eu²⁺ with total angular momentum J = 7/2 and the Landé g factor = 2. These results also rule out the possibility that mixed valency may exist for the europium cations in these compounds.

Fig. 5 Temperature dependence of the magnetic susceptibility data of Eu₂Mg₃Cu₉As₇, measured under an applied field of 500 Oe. Inset shows the inverse of susceptibility versus temperature.

Electrical resistivity measurement

The temperature-dependent resistivity data on a single crystal of Eu₂Mg₃Cu₉As₇ are provided in Fig. 6, measured over a region between 100 and 300 K. The resistivity increases slowly with the increasing temperature from 60.94 μΩ cm at 100 K to 125.28 μΩ cm at 300 K. Such behavior indicates that Eu₂Mg₃Cu₉As₇ should be poor metallic, which is reasonable taking into account the electron deficiency state in the structure and is also supported by the theoretical prediction above. Compared to a nominally electron-precise analogue such as Ce₂Mn₃Cu₉P₇ with a resistivity of 1100 μΩ cm at room temperature,²¹ Eu₂Mg₃Cu₉As₇ is obviously much more electrically conducting due to the electron deficiency.

Fig. 6 Temperature-dependent electrical resistivity data measured on a single crystal of Eu₂Mg₃Cu₉As₇ by the four-probe method.

Conclusions

In summary, five new quaternary pnictide Zintl phases, A₂Mg₃Cu₉Pn₇ (A = Sr, Eu; Pn = P, As) and Eu₅Mg_2.39Cu_16.61As₁₂, were synthesized by the Pb-flux reactions. With divalent alkaline-earth or rare-earth cations being introduced into these structures, electron deficiency resulted, which was confirmed by DFT theoretical calculations as well as the corresponding magnetic susceptibility and electrical resistivity measurements. The discovery of these new examples may indicate new applications such as thermoelectrics for such a family-rich intermetallic system due to their tunable electronic structures.

Acknowledgements

The authors acknowledge the financial support from the National Natural Science Foundation of China (Grant No. 51271098, 51321091).

References

1. J. D. Corbett, Inorg. Chem., 2010, 49, 13.
2. R. Nesper, Angew. Chem., Int. Ed. Engl., 1991, 30, 789.
3. Chemistry, Structure and Bonding of Zintl Phases and Ions, ed. S. M. Kauzlarich, VCH Publishers, New York, 1996, and the references therein.
4. A. M. Guloy, Polar Intermetallics and Zintl Phases along the Zintl Border. Inorganic Chemistry in Focus III, Wiley-VCH Verlag GmbH &Co. KGaA, Weinheim, Germany, 2006.
5. D.-K. Seo and J. D. Corbett, J. Am. Chem. Soc., 2001, 123, 4512–4518.
6. L.-M. Wu, S.-H. Kim and D.-K. Seo, J. Am. Chem. Soc., 2005, 127, 15682–15683.
7. S. Ponou, S. Lidin, Y.-M. Zhang and G. J. Miller, Inorg. Chem., 2014, 53, 4724–4732.
8. W.-Z. Zheng, P. Wang, L.-M. Wu, Y. Liu and L. Chen, Inorg. Chem., 2010, 49, 5890–5896.
9. M. Khatun, S. S. Stoyko and A. Mar, Inorg. Chem., 2013, 52, 12682–12690.
10. M. Rotter, M. Tegel and D. Johrendt, Phys. Rev. Lett., 2008, 101, 107006.
11. C. A. Cox, E. S. Toberer, A. A. Levchenko, S. R. Brown, G. J. Snyder, A. Navrotsky and S. M. Kauzlarich, Chem. Mater., 2009, 21, 1354–1360.
12. S. Bobev, Inorg. Chem., 2004, 43, 5044–5052.
13. S.-Q. Xia and S. Bobev, J. Am. Chem. Soc., 2007, 129, 10011–10018.
14. X.-C. Liu, Z. Wu, S.-Q. Xia, X.-T. Tao and S. Bobev, Inorg. Chem., 2015, 54, 947–955.
15. S. K. Bux, A. Zevalkink, O. Janka, D. Uhl, S. Kauzlarich, J. G. Snyder and J.-P. Fleurial, J. Mater. Chem. A, 2014, 2, 215–220.
16. S. Ohno, A. Zevalkink, Y. Takagiwa, S. K. Buxc and G. J. Snyder, J. Mater. Chem. A, 2014, 2, 7478–7483.
17. S. T. Oyama, T. Gott, H. Zhao and Y.-K. Lee, Catal. Today, 2009, 143, 94–107.
18. R. Prins and M. E. Bussell, Catal. Lett., 2012, 142, 1413–1436.
19. W. Jeitschko, U. Meisen and J. Albering, Dalton Trans., 2010, 39, 6067–6073.
20. E. Dhahri, J. Phys.: Condens. Matter, 1996, 8, 4351–4360.
21. S. S. Stoyko, K. K. Ramachandran, C. S. Mullen and A. Mar, Inorg. Chem., 2013, 52, 1040–1046.
22. R. J. Cava, T. Siegrist, S. A. Carter, J. J. Krajewski Jr., W. F. Peck and H. W. Zandbergen, J. Solid State Chem., 1996, 121, 51–55.
23. G. M. Sheldrick, SHELXTL, University of Göttingen, Göttingen, Germany, 2001.
24. Bruker APEX2, Bruker AXS Inc., Madison, WI, 2005.
25. WinGX. A Windows Program for Crystal Structure Analysis: L. J. Farrugia, J. Appl. Crystallogr., 1999, 32, 837.
26. P. Blaha, K. Schwarz, G. K. H. Madsen, D. Kvasnicka and J. Luitz, WIEN2k, An Augmented Plane Wave+Local Orbitals Program for Calculating Crystal Properties, Technische Universität, Vienna, Austria, 2001.
27. G. K. H. Madsen, P. Blaha, K. Schwarz, E. Sjostedt and L. Nordstrom, Phys. Rev. B: Condens. Matter, 2001, 64, 195134.
28. K. Schwarz, P. Blaha and G. K. H. Madsen, Comput. Phys. Commun., 2002, 147, 71–76.
29. O. Jepsen and O. K. Andersen, The Stuttgart TB-LMTO Program, Version 4.7.
30. H. L. Skriver, The LMTO Method, Springer, Berlin, 1984.
31. V. Ghetta, P. Chaudouet, R. Madar, J. P. Senateur and A. B. Lambert, J. Less-Common Met., 1986, 120, 197–201.
32. U. Pfannenschmidt, U. C. Rodewald, R.-D. Hoffmann and R. Pottgen, J. Solid State Chem., 2011, 184, 2731–2737.
33. V. Ghetta, P. Chaudouet, R. Madar, J. P. Senateur and B. Lambert-Andron, J. Less-Common Met., 1989, 146, 299–307.
34. A. Wurth, A. Löhken and A. Mewis, Z. Anorg. Allg. Chem., 2002, 628, 661–666.
35. S. J. Steinwand, W. M. Hurng and J. D. Corbett, J. Solid State Inorg. Chem., 1991, 94, 36–44.
36. R. Nesper, S. Wengert, F. Zürcher and A. Currao, Chem. – Eur. J., 1999, 5, 3382–3389.
37. J. Y. Pivan, R. Guerin, O. Pena, J. Padiou, E. H. Ghadraoui and M. Rafiq, Eur. J. Solid State Inorg. Chem., 1990, 27, 617–631.
38. I. Pilchowski, A. Mewis, M. Wenzel and R. Gruehn, Z. Anorg. Allg. Chem., 1990, 588, 109.
39. S. J. Kim, J. Ireland, C. R. Kannewurf and M. G. Kanatzidis, Chem. Mater., 2000, 12, 3133–3138.
40. V. Ghetta, P. Chaudouet, R. Madar, J. P. Senateur and B.-A. Lambert, J. Less-Common Met., 1986, 120, 197.
41. U.-R. Jakubowski and W. Jeitschko, J. Less-Common Met., 1988, 136, 261.
42. C. J. O'Connor, Magnetochemistry-Advances in Theory and Experimentation, in Progress in Inorganic Chemistry, ed. S. J. Lippard, John Wiley & Sons, Inc., New York, 2007, vol. 29, pp. 203–283.
43. X.-C. Liu, S.-Q. Xia, X.-W. Lei, M.-Y. Pan and X.-T. Tao, Eur. J. Inorg. Chem., 2014, 13, 2248–2253.
44. J. Wang, M. Yang, M.-Y. Pan, S.-Q. Xia, X.-T. Tao, H. He, G. Darone and S. Bobev, Inorg. Chem., 2011, 50, 8020–8027.
45. J. Wang, S.-Q. Xia and X.-T. Tao, Inorg. Chem., 2012, 51, 5771–5778.
46. N. Tsujii, C. A. Uvarov, P. Klavins, T.-H. Yi and S. M. Kauzlarich, Inorg. Chem., 2012, 51, 2860–2866.

---

Footnote

† Electronic supplementary information (ESI) available: EDX analysis on the single crystals of Eu₅Mg_2.39Cu_16.61As₁₂; crystallographic data on Sr₂Mg₃Cu₉P₇, Sr₂Mg₃Cu₉As₇ and Eu₂Mg₃Cu₉P₇; powder X-ray diffraction patterns of Eu₂Mg₃Cu₉As₇. See DOI: 10.1039/c6qi00221h

---