The phosphide oxides La3Cu4P4O2 and La5Cu4 P4O4Cl2 were synthesized from lanthanum, copper(I) oxide, red phosphorus, and lanthanum(III) chloride through a ceramic technique. Single crystals can be grown in a NaCl/KCl flux. Both structures were refined from single crystal X-ray diffractometer data: I4/mmm, a = 403.89(4), c = 2681.7(3) pm, wR2 = 0.0660, 269 F2 values, 19 variables for La3Cu4P4O2 and a = 407.52(5), c = 4056.8(7) pm, wR2 = 0.0905, 426 F2 values, 27 variables for La5Cu4P4O4Cl2. Refinement of the occupancy parameters revealed full occupancy for the oxygen sites in both compounds. The structures are composed of cationic (La2O2)2+ layers and covalently bonded (Cu4P4)5– polyanionic layers with metallic characteristics, and an additional La3+ between two adjacent (Cu4P4)5– layers. The structure of La5Cu4P4O4Cl2 comprises two additional LaOCl slabs per unit cell. Temperature-dependent magnetic susceptibility studies revealed Pauli paramagnetism. The phosphide substructure of La3Cu4P4O2 was studied by 31P solid state NMR spectroscopy. By using a suitable dipolar re-coupling approach the two distinct resonances belonging to the P24– and the P3– units could be identified.
The structural chemistry of pnictide oxides is mainly characterized by the large family of ZrCuSiAs type compounds with the general formula RETPnO (RE = rare earth element, T = divalent transition metal cation, Pn = P3–, As3–, Sb3–). These compounds have layer-like structures with an AB stacking of (REO)+ cationic and (TPn)– anionic layers [1–4]. Partial substitution of the oxygen site by fluorine can induce superconductivity. The so far highest transition temperature of 55 K was observed for SmFeAsO1–xFx . Complete fluorine occupancy within the tetrahedra is only possible with a lower-valent cation, e.g. in BaMnPF  or SrFeAsF (≡(SrF)+(FeAs)–) . Such fluorine-centered layers also occur in (SrF)2Ti2Pn2O (Pn = As, Sb)  and (SrF)2Fe2OQ2 (Q = S, Se) .
An entirely new slab in the family of pnictide oxides was observed in the structure of La5Cu4As4O4Cl2 . This phase was first obtained during systematic salt flux syntheses of samples in the La-Cu-As-O system. In contrast to many other synthesis attempts, the alkali halide flux was not inert and reacted as a so-called reactive flux , delivering the chloride anions for formation of La5Cu4As4O4Cl2. The targeted synthesis was then possible from lanthanum filings, arsenic powder, copper(I) oxide, and lanthanum oxychloride.
Parallel work in the corresponding phosphide system now led to single crystals of La5Cu4P4O4Cl2. Its crystal structure and magnetic properties are reported herein. We have also reinvestigated the structure of La3Cu4P4O2  on the basis of single crystal diffraction data and 31P solid state NMR spectroscopy.
Starting materials for the syntheses of La3Cu4P4O2 and La5Cu4P4O4Cl2 were rods of lanthanum (Smart Elements, >99.9 %), copper(I) oxide (ABCR, >99 %), red phosphorus (ABCR, >99.999 %) and anhydrous lanthanum chloride (ChemPur, >99.9 %). Lanthanum was employed as filings, which were prepared from the rods under dried (sodium wire) paraffin oil, subsequently washed with cyclohexane and stored under argon prior to synthesis. Argon was purified by means of titanium sponge (900 K), silica gel, and molecular sieves. Anhydrous lanthanum chloride was used to prepare LaOCl – a suitable precursor for the synthesis of La5Cu4P4O4Cl2 – which could be obtained by calcination (973 K, 24 h according to LaCl3 + H2O → LaOCl + 2HCl with the water partial pressure in air) and was checked for purity by X-ray powder diffraction.
Polycrystalline samples of La3Cu4P4O2 and La5Cu4 P4O4Cl2 were prepared by weighing filings of lanthanum, copper(I) oxide, phosphorus and LaOCl in the case of the quintenary compound in the ideal stoichiometric ratios of 3:2:4 and 3:2:4:2, respectively. Amounts of 0.5 g of these mixtures were carefully ground, cold pressed to pellets at ca. 100 bar and sealed in evacuated silica tubes. Synthesis was carried out in a resistance furnace. The first annealing sequence for La3Cu4P4O2 was 573 K for 12 h followed by 773 K for 72 h. In the case of La5Cu4P4O4Cl2 the sequence was 773 K (24 h) and 1173 K (72 h). Annealing rates were 50 K h–1, with cooling by shutting off the furnace. After regrinding the obtained samples, they were again pressed to pellets and sealed in evacuated silica tubes. For the La5Cu4P4O4Cl2 sample the pellet was additionally placed in an alumina crucible. The second annealing sequences were carried out at higher temperatures of 1173 K (120 h) for La3Cu4P4O2 and 1473 K (48 h) followed by 1373 K (120 h) for La5Cu4P4O4Cl2. Cooling of the La3Cu4P4O2 sample was again by shutting off the furnace, whereas La5Cu4P4O4Cl2 was quenched in iced water. Single phase La3Cu4P4O2 was obtained as a polycrystalline black powder. Single crystals for structure determination were obtained from a salt flux synthesis. Powder of La3Cu4P4O2 (200–300 mg) and an equimolar NaCl/KCl mixture (~1 g) were sealed in an evacuated silica tube, heated to 1223 K (72 h) and slowly cooled by a rate of 2 K/h to 773 K and finally to ambient temperature by shutting off the furnace. Suitable single crystals could be obtained after dissolving the flux in demineralised water. In contrast, the quenched tablet of La5Cu4P4O4Cl2 showed black color with metallic luster and suitable single crystals could be separated from the crushed sample. Unfortunately, La5Cu4P4O4Cl2 could not be obtained phase pure. LaOCl was always observed as a minor impurity in the X-ray powder pattern. Additionally, 31P MAS NMR spectra (see below) revealed further phosphorus species, mainly from amorphous phosphate impurities.
2.2 X-ray diffraction
The polycrystalline La3Cu4P4O2 and La5Cu4P4O4Cl2 samples were characterized through Guinier powder patterns (Enraf-Nonius camera, type FR 552); imaging plate detector, Fujifilm BAS-1800, CuKα1 radiation and α-quartz (a = 491.30, c = 540.46 pm) as an internal standard. Least-squares refinements led to the tetragonal lattice parameters listed in Table 1. Correct indexing of the patterns was ensured through intensity calculations .
|Formula weight, g mol–1||826.8||1207.5|
|Calculated density, g cm–3||6.28||5.95|
|Absorption coefficient, mm–1||24.5||22.5|
|Detector distance, mm||40||80|
|Exposure time||67 s||15 min|
|ω range; increment, deg||–||0–180; 1.0|
|Integr. param. A; B; EMS||9.0; –8.5; 0.040||10.3; 1.3; 0.013|
|Crystal size, μm3||80 × 15 × 3||45 × 20 × 10|
|Transm. ratio (max/min)||0.93/0.62||0.80/0.54|
|θ range, deg||5–33||2–32|
|Range in hkl||±6, ±6, ±38||±6, ±6, ±60|
|Total no. reflections||1050||2232|
|Reflections with I > 3 σ(I)/Rσ||131/0.0974||283/0.0484|
|Goodness-of-fit on F2||1.08||2.10|
|R1/wR2 for I > 3 σ(I)||0.0355/0.0574||0.0442/0.0879|
|R1/wR2 for all data||0.0905/0.0660||0.0671/0.0905|
|Largest diff. peak/hole, e Å–3||1.54/–2.11||2.53/–3.01|
Irregularly shaped black single crystals of La5Cu4 P4O4Cl2 were selected from the crushed annealed pellet. La3Cu4P4O2 crystals were separated from a salt flux. The crystals were glued to thin quartz fibers using bees wax and first studied on a Buerger camera (using white Mo radiation) to check their quality. Intensity data were collected at ambient temperature on a Stoe IPDS-II diffractometer (graphite monochromated MoKα radiation, λ = 71.073 pm, oscillation mode) for La5Cu4P4O4Cl2 and in the case of La3Cu4P4O2 on a four-circle diffractometer (Stoe StadiVari, μ-source, MoKα radiation, λ = 71.073 pm, oscillation mode) with an open Eulerian cradle setup equipped with a reverse-biased silicon diode array detector (Dectris Pilatus 100 K, resolution: 487 × 195 pixel, pixel size: 0.172 × 0.172 mm2). Numerical absorption corrections (along with scaling for the StadiVari data set) were applied to the data sets. Details about the data collections and the crystallographic parameters are summarized in Table 1.
2.3 Structure refinements
Both data sets showed body-centred tetragonal lattices with high Laue symmetry and no further systematic extinctions. Space groups I4/mmm were found to be correct during the structure refinements in agreement with earlier studies on La3Cu4P4O2  and the arsenide La5Cu4As4O4Cl2 . The starting atomic parameters were deduced using the charge-flipping algorithm of Superflip , and both structures were refined with anisotropic displacement parameters for all atoms with Jana2006 . For better comparison with the literature data, in the final refinement cycles we transformed the positions to the setting given in  and . Refinement of the occupancy parameters in a separate series of least-squares cycles revealed full occupancy for all sites within two standard deviations. The final difference Fourier syntheses revealed no residual peaks. The refined atomic positions, displacement parameters, and interatomic distances are given in Tables 2 and 3.
Further details of the crystal structure investigation may be obtained from Fachinformationszentrum Karlsruhe, 76344 Eggenstein-Leopoldshafen, Germany (fax: +49-7247-808-666; e-mail: email@example.com, http://www.fiz-karlsruhe.de/request_for_deposited_data.html) on quoting the deposition number CSD-430391 (La3Cu4P4O2) and CSD-430390 (La5Cu4P4O4Cl2).
2.4 EDX data
The irregularly shaped La3Cu4P4O2 and La5Cu4P4O4Cl2 crystals studied on the diffractometers were semiquantitatively analyzed by EDX using a Zeiss EVO® MA10 scanning electron microscope in variable pressure mode. LaF3, Cu, GaP, and KCl were used as standards. Except the oxygen content, the experimentally observed compositions were all within ±3 at % close to the ideal ones. No impurity elements – especially chlorine in the case of La3Cu4P4O2 – were detected.
2.5 Physical property measurements
Magnetic susceptibility measurements of La5Cu4P4O4Cl2 were carried out on a Quantum Design Physical Property Measurement System (PPMS) using the VSM option. 31.643 mg of the powdered sample were packed in a polypropylene capsule and attached to the sample holder rod. Measurements were performed in the temperature range of 3–300 K with magnetic flux densities up to 80 kOe (1 kOe = 7.96 × 104 A m–1).
2.6 31P solid state MAS NMR
31P solid state MAS NMR spectra of La3Cu4P4O2 were recorded on a Bruker Avance III 300 MHz spectrometer (B0 = 7.05 T) with spinning speeds of 28 and 30 kHz in conventional cylindrical ZrO2 MAS rotors with 2.5 mm diameter at a resonance frequency of 121.442 MHz. The sample was ground and mixed with powdered quartz under dry cyclohexane to reduce the density and electrical conductivity. The experiments were carried out with 90° pulse lengths of 2.5 μs and delays of 0.5 s. Phosphoric acid (85 %) was used as an external reference. All spectra were recorded with the Bruker Topspin software  and analyzed with the Dmfit software . Chemical shifts are referenced to 85 % H3PO4 solution. The experimental data are listed in Table 4.
3 Results and discussion
3.1 Crystal chemistry
The structures of La3Cu4P4O2  and La5Cu4P4O4Cl2 were refined from single crystal X-ray diffractometer data. Their unit cells are presented in Fig. 1. Large parts of both structures are similar, i.e. the [Cu2P2] anionic and the [LaO] cationic layers. A look at the interatomic distances (Table 3) shows almost similar values for these structural slabs. We start the following crystal chemical discussion with the La3Cu4P4O2 structure.
La3Cu4P4O2 is the first member of a family of few quaternary phosphide oxides which are isopointal [18, 19] with the silicide Zr3Cu4Si6 . The rare earth (RE) compounds RE3Cu4P4O2 (RE = Ce, Pr, Nd, Sm) [12, 21] and La3Ni4P4O2  are isotypic with La3Cu4P4O2. With the higher congener arsenic, the phases RE3T4As4O2–δ with RE = La, Ce, Pr, Nd, Sm, and T = Ni, Cu [10, 23] have recently been reported.
The main point regarding these phases concerns their non-electron-precise description, e.g. (3La3+)9+(4Cu+)4+(P24–)(2P3–)6–(2O2–)4–, leaving an excess negative charge. Partial oxygen occupancy of the oxygen sites, like in Pr3Cu4P4O1.50 and Sm3Cu4P4O1.63  could be an explanation. Nevertheless, one has to keep in mind that La3Cu4P4O2 is a metallic conductor and a black solid with metallic luster for crystalline pieces . The original structure refinement of La3Cu4P4O2 showed an enhanced isotropic displacement parameter for the oxygen site, which might be indicative of a split position or a reduced occupancy parameter. The present structure refinement allowed anisotropic refinement of all sites, and a separate refinement of the occupancy parameters gave no hint for oxygen vacancies. This is in line with the neutron powder diffraction data for La3Ni4P4O2 . Oxygen vacancy formation was suggested for the arsenides RE3T4As4O2–δ , but not really substantiated on the basis of experimental data.
Assuming full occupancy of the oxygen site and the essentially ionic bonding between lanthanum and oxygen, the best approximation for an electron-precise formula splitting would be La3+(La2O2)2+(Cu4P4)5–, leaving the electronic adjustment to the pnictide polyanion. The latter has two degrees of freedom: (i) the P–P bond distance and (ii) the possibility of partial copper oxidation. While we observe a P–P distance of 227 pm in La3Cu4P4O2, close to the single bond distance of 223 pm , a much longer P–P distance of 258 pm occurs in La3Ni4P4O2 , a tendency towards breaking the P–P bond and leaving higher electron density at the phosphorus atoms as compared to La3Cu4P4O2.
This peculiar electronic situation is most likely directly related to the anisotropic coordination of the Cu2P2 layer: on one side by La3+ cations (higher cationic charge) and on the opposite side by the [LaO]+ layer (lower cationic charge). Especially for the series of RE3T4As4O2–δ arsenides , this crystalline anisotropy strongly affects the magnetic and transport properties. In contrast, the pnictide layers of the many ZrCuSiAs phases  have isotropic coordination by [REO]+ layers on both sides (alternating AB stacking with the [TPn]– layers).
The structural anisotropy discussed for La3Cu4P4O2 also occurs in the new quintenary compound La5Cu4 P4O4Cl2 which is isotypic with La5Cu4As4O4Cl2 . The main difference to La3Cu4P4O2 concerns the insertion of two ionic LaOCl slabs in the unit cell. Similar to the quaternary compound, also for La5Cu4P4O4Cl2 the ionic formula splitting La3+(La4O4)4+(Cu4P4)5–(2Cl–)2– is the best approximation. The electronic situation of the (Cu4P4)5– slab is exactly the same in both compounds.
La5Cu4P4O4Cl2 is also a black compound and its metallic behavior is evident from the Pauli susceptibility (vide infra). Electronic structure calculations on isotypic La5Cu4 As4O4Cl2  led to a strongly anisotropic hybrid material with covalently bonded tetrahedral CuAs4/4 layers that are separated by larger ionic insulating oxide and chloride blocks. With respect to chemical bonding we can safely apply a rigid band model to La5Cu4P4O4Cl2 described herein.
3.2 Magnetic properties of La5Cu4P4O4Cl2
The temperature dependence of the magnetic susceptibility of La5Cu4P4O4Cl2 is presented in Fig. 2. The overall susceptibility data (Fig. 2, top) are low with a value of χ = 0.4(1) × 10–4 emu mol–1 at 300 K, classifying La5Cu4P4O4Cl2 as a Pauli paramagnet. The upturn towards lower temperatures is due to a small paramagnetic contribution, most likely an impurity phase. In agreement with the low susceptibility data we also observe tiny moments in the magnetization isotherms at 10 and 100 K (Fig. 2, bottom). Low-field low-temperature measurements gave no hint for superconductivity.
3.3 31P Solid state MAS NMR spectroscopic data of La3Cu4P4O2
The 31P MAS NMR spectrum of La3Cu4P4O2 is shown in Fig. 3. As expected, two signals are obtained, consistent with the two phosphorus sites of equal population (1:1 area ratio). The spin-lattice relaxation times are in the millisecond range, which is uncommonly short for spin-1/2 nuclei in the solid state. In addition, considerable resonance shifts relative to the H3PO4 standard are observed. These spectroscopic features are consistent with the metallic character previously reported for this compound from electrical conductivity and magnetic susceptibility measurements . The large resonance shifts signify unpaired conduction electron spin density at the Fermi edge at the phosphorus sites, causing a Knight shift contribution to the resonance frequencies observed. For assigning the two distinct resonances at 250 and 459 ppm to the phosphorus sites P(1) and P(2) we can exploit the dependence of the homo-nuclear 31P-31P dipole-dipole interactions on the inter-nuclear distances. Owing to the direct P–P bond the P(2) species experience much stronger dipolar couplings than the isolated P(1) species. Experimentally, the dipolar coupling strength can be assessed by measuring the site-resolved spin-spin relaxation times T2 by a rotor-synchronized π/2 – τ – π – τ MAS-spin echo experiment with variable effective evolution times 2n τ with τ = νrot–1 (Fig. 4), where 2n is the number of rotor cycles. The results, summarized in Fig. 4 and Table 4, indicate that the spin-spin relaxation times of the two types of phosphorus species differ significantly and clearly suggest that the signal at 459 ppm must be assigned to the P24– dumbbell species P(2).
This work was supported by the Deutsche Forschungsgemeinschaft. We thank Dr. Thomas Wiegand (ETH Zürich) for his participation in the early stages of this work.
 R. Pöttgen, D. Johrendt, Z. Naturforsch. 2008, 63b, 1135. Search in Google Scholar
 T. C. Ozawa, S. M. Kauzlarich, Sci. Techn. Adv. Mater. 2008, 9, 033003. Search in Google Scholar
 D. C. Johnston, Adv. Phys. 2010, 59, 803. Search in Google Scholar
 D. Johrendt, H. Hosono, R.-D. Hoffmann, R. Pöttgen, Z. Kristallogr. 2011, 226, 435. Search in Google Scholar
 Z.-A. Ren, W. Lu, J. Yang, W. Yi, X.-L. Shen, Z.-C. Li, G.-C. Che, X.-L. Dong, L.-L. Sun, F. Zhou, Z.-X. Zhao, Chin. Phys. Lett.2008, 25, 2215. Search in Google Scholar
 H. Kabbour, L. Cario, F. Boucher, J. Mater. Chem. 2005, 15, 3525. Search in Google Scholar
 M. Tegel, S. Johansson, V. Weiss, I. Schellenberg, W. Hermes, R. Pöttgen, D. Johrendt, Eur. Phys. Lett. 2008, 84, 67007. Search in Google Scholar
 R. H. Liu, Y. A. Song, Q. J. Li, J. J. Ying, Y. J. Yan, Y. He, X. H. Chen, Chem. Mater. 2010, 22, 1503. Search in Google Scholar
 H. Kabbour, E. Janod, B. Corraze, M. Danot, C. Lee, M.-H. Whangbo, L. Cario, J. Am. Chem. Soc. 2008, 130, 8261. Search in Google Scholar
 M. Eul, D. Johrendt, R. Pöttgen, Z. Naturforsch. 2009, 64b, 1353. Search in Google Scholar
 D. E. Bugaris, H.-C. zur Loye, Angew. Chem., Int. Ed. 2012, 51, 3780. Search in Google Scholar
 R. J. Cava, H. W. Zandbergen, J. J. Krajewski, T. Siegrist, H. Y. Hwang, B. Batlogg, J. Solid State Chem. 1997, 129, 250. Search in Google Scholar
 K. Yvon, W. Jeitschko, E. Parthé, J. Appl. Crystallogr.1977, 10, 73. Search in Google Scholar
 L. Palatinus, G. Chapuis, J. Appl. Crystallogr.2007, 40, 786. Search in Google Scholar
 V. Petříček, M. Dušek, L. Palatinus, Z. Kristallogr.2014, 229, 345. Search in Google Scholar
 Bruker Corp., Topspin (version 2.1), Karlsruhe, Germany, 2008. Search in Google Scholar
 D. Massiot, F. Fayon, M. Capron, I. King, S. Le Calvé, B. Alonso, J.-O. Durand, B. Bujoli, Z. Gan, G. Hoatson, Magn. Reson. Chem. 2002, 40, 70. Search in Google Scholar
 L. M. Gelato, E. Parthé, J. Appl. Crystallogr. 1987, 20, 139. Search in Google Scholar
 E. Parthé, L. M. Gelato, Acta Crystallogr. 1984, A40, 169. Search in Google Scholar
 F. Thirion, G. Venturini, B. Malaman, J. Steinmetz, B. Roques, J. Less-Common Met. 1983, 95, 47. Search in Google Scholar
 J. W. Kaiser, W. Jeitschko, Z. Naturforsch. 2002, 57b, 165. Search in Google Scholar
 T. Klimczuk, T. M. McQueen, A. J. Williams, Q. Huang, F. Ronning, E. D. Bauer, J. D. Thompson, M. A. Green, R. J. Cava, Phys. Rev. B2009, 79, 012505. Search in Google Scholar
 J. K. Wang, A. Marcinkova, C.-W. Chen, H. He, M. Aronson, E. Morosan, Phys. Rev. B2014, 89, 094405. Search in Google Scholar
 R. Pöttgen, W. Hönle, H. G. von Schnering, Phosphides: Solid State Chemistry, in Encyclopedia of Inorganic Chemistry, Vol. VII, 2nd ed. (Ed.: R. B. King), Wiley, New York, 2005, 4255. Search in Google Scholar
©2016 by De Gruyter