Crystal structure and physical properties of shchurovskyite-related K₂CaCu₆O₂(PO₄)₄

Abstract

A new oxocuprate phosphate, K₂CaCu₆O₂(PO₄)₄, was obtained as tabular dark-green single crystals through flux-assisted high-temperature synthesis. It crystallizes in the space group P2/c and represents the centrosymmetric modification of the fumarolic arsenate mineral shchurovskyite, K₂CaCu₆O₂(AsO₄)₄. The crystal structure of the compound under study is characterized by the presence of large channels filled with potassium cations. Bond-valence energy landscape calculations based on structural data showed the probability of K⁺ migration through the crystal structure along the b axis at an energy of 0.6 eV. K₂CaCu₆O₂(PO₄)₄ ordered into a canted antiferromagnetic state in two steps at T_N1 = 22 ± 2 K and T_N2 = 10 ± 1 K, as revealed by magnetization and specific heat measurements.

---

Introduction

The eruption of the Tolbachik volcano in 1975 resulted in the appearance of a huge number of new copper-based minerals of fumarolic origin.¹ Most of them are oxocuprate salts incorporating additional oxygen atoms, which are not part of the acid residue. The peculiar magnetic properties of these natural compounds stem from the Jahn–Teller effect, leading to the distortion of Cu²⁺-centered polyhedrons and the formation of various configurations (square, tetrahedral, square-pyramidal, trigonal–bipyramidal, octahedral, and heptahedral) of the magnetically active cation. These distortions correlate with variations in the symmetry of the entire crystal structure and with the pronounced anisotropy of magnetic exchange interactions. Among the numerous fumarolic oxocuprate sulfates, the most promising for investigating physical properties are nabokoite, KCu₆CuTeO₄(SO₄)₅Cl;² atlasovite, KCu₆FeBiO₄(SO₄)₅Cl;³ and elasmochloite, Na₃Cu₆BiO₄(SO₄)₅;⁴ all possessing a square kagome magnetic network. This network attracts attention as a platform for elusive spin-liquid quantum ground state, the Holy Grail of modern condensed matter physics.⁵ Adjacent to this family is fedotovite K₂Cu₃O(SO₄)₃,⁶ in which the edge-shared tetrahedral spin clusters of Cu²⁺ ions form a one-dimensional array. It has been shown both experimentally and theoretically that an effective spin S = 1 cluster-based Haldane state is realized in this compound at low temperatures.^7,8 Urusovite, CuAlAsO₅, is a valence bond crystal with a quantum ground state,⁹ while a short-range magnetic order has been established in ilinskite, KCu₅O₂(SeO₃)₂Cl₃,¹⁰ and a long-range magnetic order sets in chloromenite, Cu₉O₂(SeO₃)₄Cl₆,¹¹ and averievite, Cu₅V₂O₁₀(CsCl).¹²

Shchurovskyite, K₂CaCu₆O₂(AsO₄)₄, is one more fumarolic mineral¹³ that serves as the archetype for a family with a complex Cu-based magnetic subsystem. Recently, a series of new shchurovskyite-related compounds have been synthesized using hydrothermal and chemical vapor transport methods.^14,15 However, the physical properties of shchurovskyite itself and its phosphate analogues have not been studied. Aksenov et al. reported that Rb₂CaCu₆O₂(PO₄)₄ evidenced the ferrimagnetic ground state with residual magnetization equal to 1/3 of saturation magnetization.¹⁴ Herein, we report the flux-assisted high-temperature synthesis of a new potassium calcium oxocuprate phosphate, K₂CaCu₆O₂(PO₄)₄, structurally related to fumarolic shchurovskyite. The crystal structure of this compound was determined using single-crystal X-ray diffraction, its physical properties were experimentally characterized, and the K⁺-ion conductivity was theoretically evaluated.

Crystal structure

The title phase was first obtained as a by-product during an attempt to synthesize KNa₃Cu₇(PO₄)₆ crystals (in press). Initial reagents K₂CO₃ (0.0156 g), Na₂CO₃ (0.0359 g), (NH₄)₂HPO₄ (0.1790 g), and CuCl₂·2H₂O (0.2695 g) were taken in a molar ratio of 1 : 3 : 12 : 14 (total 0.5 g), mixed with the same amount of KCl (0.5 g) used as a flux and then gradually heated up to 690 °C in a porcelain crucible. The new phase was obtained as tabular dark-green crystal concretions (Fig. 1). Isolated mechanically, high-quality crystals were subsequently used for further investigations by SEM and single-crystal X-ray diffraction.

Fig. 1 Photograph (left) and SEM image (right) of the title phase.

To obtain a larger amount of the compound for magnetic measurements, targeted solid-state synthesis was performed. This time, K₂CO₃ (0.2317 g), CaCO₃ (0.1678 g), CuCl₂·2H₂O (1.7149 g), and (NH₄)₂HPO₄ (0.8856 g) were taken in stoichiometric proportions of 2 : 1 : 6 : 4 (total 3 g). The reagents were thoroughly mixed, ground, and gradually heated up to 870 °C, followed by slow cooling in three stages: first to 540 °C for 53 hours (5° per hour), then to 300 °C for 12 hours (12° per hour), and finally, to room temperature. A fine-crystalline dark-green product was obtained. According to the powder X-ray diffraction (PXRD) data, a minor impurity phase, Ca₃Cu₃(PO₄)₄, was detected, as shown in Fig. S1 of the (SI).

Crystals were mounted on a special plate, coated with carbon, and examined for their composition and homogeneity by energy-dispersive spectroscopy (EDS) using a JEOL JSM-IT500 scanning electron microscope (Laboratory of Analytical Techniques of High Spatial Resolution, Department of Petrology, Faculty of Geology, Lomonosov Moscow State University). The crystal morphology obtained by SEM is shown in Fig. 1. Semi-quantitative EDS analysis, performed on an unpolished sample under standard conditions (accelerating voltage = 20 kV, beam current = 10 nA, exposure time = 50 s), revealed the presence of K, Ca, Cu, P, and O in an atomic ratio of 1.6 : 0.9 : 5.9 : 4 : 17. NaCl, potassium feldspar (KAlSi₃O₈), metallic Cu, and GaP were used as reference standards.

A small dark-green crystal, selected using a binocular microscope, was examined by preliminary single-crystal X-ray diffraction. As its quality was good enough, it was used for full structural investigation. The data were collected using an Xcalibur (Oxford Diffraction) diffractometer equipped with Mo Kα radiation and corrected for Lorentz and absorption effects.¹⁶ The crystal structure was solved by direct methods in the monoclinic space group P2/c using the SHELXT program¹⁷ and refined by full-matrix least-squares on F² using SHELXL,¹⁸ both implemented in the WinGX software package.¹⁹ During the refinement, partial cationic disorder was detected for potassium and calcium sites. The K1 and K2 positions were located 0.69(2) Å apart and refined with site occupancies of 0.21(2) and 0.79(2), respectively. Calcium was found to be isomorphically substituted by Cu²⁺ ions, with refined occupancies of 0.876(8) (Ca) and 0.124(8) (Cu). The final refinement converged to R = 0.038, yielding the formula K₂(Ca_0.876Cu_0.124)Cu₆O₂(PO₄)₄, which is consistent with the elemental composition determined by EDS X-ray spectroscopy. Crystallographic data and refinement details are summarized in Table 1, and atomic coordinates with equivalent isotropic displacement parameters are listed in Table S1 of the SI. The crystal structure was visualized using the Diamond software.²⁰

Table 1 Crystal information and details of X-ray data collection and refinement

Crystal data
Chemical formula	K2(Ca0.876Cu0.124)Cu6O2(PO4)4
M r	914.45
Crystal system, space group	Monoclinic, P2/c
Temperature (K)	293
a, b, c (Å)	8.3527 (3), 5.6597 (2), 16.8327 (6)
β (°)	94.605 (4)
V (Å^3)	793.18 (5)
Z	2
Radiation type	Mo Kα
µ (mm^−1)	9.380
Crystal size (mm)	0.07 × 0.04 × 0.03

---

Data collection
Diffractometer	Xcalibur
Absorption correction	Gaussian
T min, Tmax	0.673, 0.775
No. of measured, independent and observed [I > 2σ(I)] reflections	6049, 1822, 1471
R int	0.052
(sin θ/λ)max (Å^−1)	0.650

---

Refinement
R[F^2 > 2σ(F^2)], wR(F^2), S	0.038, 0.066, 1.21
No. of reflections	1822
No. of parameters	147
Δρmax, Δρmin (e Å^−3)	1.12, −1.03

---

The crystal structure of the title compound consists of two symmetrically independent orthophosphate tetrahedra, four distinct copper coordination polyhedra, a calcium octahedron, and a large potassium polyhedron (Fig. 2). Both PO₄ groups exhibit the lowest symmetry (C₁) and typical P–O bond lengths ranging from 1.522(3) to 1.564(3) Å, with an average of 1.54 Å (Table S2 of the SI). Each phosphate tetrahedron shares all its vertices with copper or calcium polyhedra.

Fig. 2 Independent fragment of the crystal structure of title phase in the ellipsoid mode at the 90% probability level. Symmetry codes: ′1 − x, 1 − y, 1 − z; ″1 − x, 2 − y, 1 − z; ‴2 − x, y, 1.5 − z; *x, 1 − y, 0.5 + z; **1 − x, y, 1.5 − z; ***x, 2 − y, 0.5 + z; ′*−1 + x, y, z.

Copper cations occupy four crystallographically independent sites with distinct oxygen coordination environments. Cu1 and Cu2 atoms lie on twofold rotation axes at the 2f Wyckoff site, positioned one above the other along the b axis. The anionic coordination around Cu1 is octahedral, with two pairs of short Cu–O bonds of 1.925(3) and 1.960(3) Å and one pair of elongated bonds of 2.652(3) Å. Cu2O₄ with nearly planar rectangles are characterized by Cu–O distances of 1.896(3) and 1.934(3) Å; the next pair of oxygen atoms is located at 2.983(3) Å.

Cu3 and Cu4 cations occupy general positions (4g Wyckoff sites) and are coordinated by five oxygen atoms at distances ranging from 1.881(4) to 2.251(3) Å. Additional oxygen atoms are found at 3.194(3) and 2.863(3) Å from Cu3 and Cu4, respectively. All Cu–O bond lengths are typical for divalent copper ions exhibiting pronounced Jahn–Teller distortion. The oxidation states of the copper cations were further confirmed by bond-valence sum (BVS) calculations (Table S3).²¹

Calcium atoms occupy the special 2e Wyckoff site located on a twofold rotation axis. Their coordination environment is octahedral, with Ca–O distances ranging from 2.336(4) to 2.489(4) Å. The large monovalent K⁺ ions are eight-coordinated, forming irregular polyhedra with K–O bond lengths between 2.644(5) and 3.534(8) Å.

The framework of the title phase is based on the complex and condensed arrangement of the copper polyhedra. The Cu1O₆ octahedra share O1–O1 edges with square-planar Cu2O₄ groups to form straight columns extending along the b axis (Fig. 3a). These rods are decorated on both sides along the c direction by CuO₅ pyramids (Fig. 3b). Pairs of Cu3O₅ and Cu4O₅ pyramids, which share common edges, are attached to the chains through the same O1 ligands. As a result, each O1 oxygen atom is coordinated exclusively to four copper polyhedra. The Cu3O₅ pyramid shares a single vertex O1 with the polyhedra of the rod, whereas the Cu4O₅ pyramid shares the O1–O7 edge with a Cu1O₆ octahedron (Fig. 3c).

Fig. 3 Structural fragments of the title phase: columns of the Cu1O₆ and Cu2O₄ polyhedra extended along the b axes (a) which are further decorated with the Cu3O₅ and Cu4O₅ pyramids (b); clusters of copper polyhedra, showing the connectivity between copper polyhedra and coordination for the O1 oxygen atom (c); and heteropolyhedral Cu–P layers (d).

These “decorated” copper ribbons are interconnected by sharing edges between neighboring Cu3O₅ pyramids, resulting in the corrugated layers parallel to the bc plane. These layers are reinforced with P1O₄ tetrahedra, which share all the vertices with copper polyhedra (Fig. 3d). Within this layer, adjacent ribbons alternate their orientation along the b axis, directed “up” or “down” as defined by the apical O5 vertex of the Cu4O₅ pyramid.

P2O₄ tetrahedra link Cu–P blocks into a triperiodic heteropolyhedral framework possessing large channels between blocks (Fig. 4). Two channels run parallel to the b axis and one runs along the c axis of the unit cell, and they are occupied by slightly disordered K⁺ cations (K1 and K2 positions) and smaller Ca²⁺ ions (partially substituted by Cu²⁺ cations).

Fig. 4 Crystal framework of the title phase in the ac (a) and ab (b) projections, showing intercrossing channels filled with K⁺ and Ca²⁺ ions.

As mentioned earlier, the O1 oxygen atom is coordinated exclusively by four Cu²⁺ cations, which classifies the new phase as an oxocuprate phosphate. Natural oxocuprate salts, especially arsenates, are widely distributed in fumarolic exhalations, such as at the Tolbachik volcano (Kamchatka Peninsula, Russia). For these structures, the anion-centering approach^22,23 is commonly employed to describe the condensed cationic substructure.

In the crystal structure of K₂CaCu₆O₂(PO₄)₄, oxo-centered [OCu₄] tetrahedra are linked via shared Cu–Cu edges to form [O₂Cu₆]⁸⁺ dimers shaped like a bow-knot (Fig. S2). These dimers associate with the P1O₄ tetrahedra to create quasi-layers of the type {[O₂Cu₆]⁸⁺(PO₄)₂}²⁺, called “oxide floors”.²⁴ Along the a axis, the oxide floors alternate with “salt floors”, |(K₂Ca)(PO₄)₂|²⁻, which consist of large K-centered eightfold polyhedra, CaO₆ octahedra, and P2O₄ tetrahedra (Fig. S2 of the SI).

These oxo-centered dimers [O₂Cu₆]⁸⁺ are the key structural elements in the frameworks of the shchurovskyite structural group, including the title compound, natural oxocuprate arsenates, and synthetic phosphates, as will be discussed further. Similar [O₂M₈] structural units also occur in the family of synthetic oxosalts with the general formula [M₃O]M(TO₄)₂ (M = Zn, Cu, and Ni; T = P, As, and V), as reviewed by Bakakin (2025).²⁴ Furthermore, the euchlorine family of natural fumarolic oxocuprate sulfates, comprising puninite, Na₂[Cu₃O](SO₄)₃, features crystal structures based on the same [O₂Cu₆] building units.²⁵ Comparable motifs are also observed in natural and synthetic selenite-chlorides, such as prewittite, KPb_0.5Cu[PbCu₅O₂]Zn(SeO₃)₂Cl₁₀,²⁶ and K[Cu₃O](SeO₃)₂Cl.²⁷ Thus, the oxo-centered [O₂Cu₆] dimers demonstrate an exceptional structural versatility, capable of integrating with diverse anionic groups, from phosphates and arsenates to sulfates and selenites. to form a remarkable variety of complex crystal architectures.

Physical properties

Magnetic susceptibility as a function of temperature measurements was performed using a “Quantum Design” MPMS-7T (SQUID) system under an applied field of 0.1 T. Crystals of K₂CaCu₆O₂(PO₄)₄ were ground and placed in a gelatin capsule fastened in a plastic straw for immersion into the SQUID system. The specific heat and the isothermal magnetization up to 9 T were measured using a “Quantum Design” PPMS-9T system.

The temperature dependences of dc magnetic susceptibility χ_dc = M/H in K₂CaCu₆O₂(PO₄)₄ taken at µ₀H = 0.1 T in both field-cooled (FC) and zero field-cooled (ZFC) regimes are shown in the main panel of Fig. 5a. These curves practically coincide above the Néel temperature T_N1 = 22 K. At a lower temperature, they slightly diverge and show a rounded hump followed by a sharp increase in the second antiferromagnetic transition at T_N2 = 10 K. The low-temperature phase possesses a weak spontaneous magnetic moment. The inverse dc magnetic susceptibility is shown in Fig. 5b, where the fit by the Curie–Weiss law is shown by the solid line.

Fig. 5 (a) Temperature dependences of magnetic susceptibility χ_dc(T) in K₂CaCu₆O₂(PO₄)₄ under an applied magnetic field of 0.1 T in both FC and ZFC modes. The inset represents both real and imaginary parts of the ac magnetic susceptibility curves at various frequencies. (b) Inverse dc magnetic susceptibility with the fit using the Curie–Weiss law shown by a solid line.

To clarify the features of the two-step transition, the ac magnetic susceptibility χ_ac has been measured at a probing field of 3 Oe in the frequency range f = 5 × 10²–10⁴ Hz, as shown in the inset of Fig. 5. The peak at T_N1 = 20 K is accompanied by the change in the slope of curves at T_N2 = 10 K. The positions of these features are not sensitive to the frequency.

Above T_N1, the magnetic susceptibility χ_dc follows the Curie–Weiss law:

A fit of the χ(T) data in the 200–300 K temperature interval gives a temperature-independent diamagnetic contribution χ₀ = −(2.1 ± 0.9) × 10⁻⁴ emu mol⁻¹, the Weiss temperature Θ = −98 ± 4 K and the Curie constant C = 3.08 ± 0.06 emu K mol⁻¹. The negative value of Weiss temperature points to dominant antiferromagnetic interactions between the Cu²⁺ cations. From the Curie constant, the effective magnetic moment per formula unit is as follows:

µ_eff = (8C)^1/2µ_B, (2)

which is calculated to be µ_eff = 4.96µ_B. This value is higher than the spin-only moment, µ^calc_eff = 4.24μ_B, expected for six (n = 6) Cu²⁺ (S = 1/2) ions per formula unit. The experimentally found g-factor, g = 2.32, stems from the value of the Curie constant through the ratio as follows:

8C = ng²S(S + 1), (3)

This value is typical for Cu²⁺ systems with substantial spin–orbit coupling. K₂CaCu₆O₂(PO₄)₄ stands out by the rich variety of the local environment of copper ions. The CuO₄, CuO₅ and CuO₆ units can be separated in this compound. It means that the spin–orbit coupling and the g-factor of Cu²⁺ ions may vary within wide limits.²⁸

The field dependence of magnetization taken at T = 2 K is shown in Fig. 6. At µ₀H = 7 T, the magnetization reaches a value of 0.5μ_B, which indicates that the system is far from the saturation. The calculated saturation magnetization M_sat is about 7μ_B. The weak hysteresis loop is visible for T < T_N1 (inset of Fig. 6).

Fig. 6 Field dependences of magnetization in K₂CaCu₆O₂(PO₄)₄ at various temperatures. Inset: the enlarged fragment of the M(H) curve.

The presence of two subsequent magnetic transitions was confirmed by specific heat measurement, as shown in Fig. 7. The singularities barely seen in the C_p(T) curve (inset of Fig. 7) are better pronounced in C_p/T vs. T² presentation at T_N1 = 22 K and T_N2 = 11 K, which is consistent with the magnetic susceptibility data. Low magnitudes of these anomalies may be due to the fact that most part of the magnetic entropy is released at much higher temperatures. It should be noted that the experimental data on magnetic susceptibility start deviating from the extrapolation of the Curie–Weiss fit at about 85 K. The C_p/T versus T² data allow identifying the phase transitions at T_N1 and T_N2 because in this representation, the lattice contribution at low temperatures is linear. The deviations from linearity correspond to magnetic contribution.

Fig. 7 C _p/T vs. T² plot for K₂CaCu₆O₂(PO₄)₄. Inset: temperature dependence of specific heat C_p.

The relation between two main thermodynamic properties, i.e., magnetization and specific heat, stems from Boltzmann statistics. It has been shown by a general theoretical argument that the variation in the magnetic specific heat of a ‘simple’ antiferromagnet, in particular the singular behavior in the region of transition, should be closely similar to the behavior of the function ∂(χT)∂T, where χ is the zero-field susceptibility.²⁹ The positions of singularities in Fisher's specific heat, d(χT)/dT, are seen at T_N1 = 22 K and T_N2 = 10 K, as shown in Fig. S3 of the SI.

The transitions at T_N1 and T_N2 are of different natures. At T_N1, the system transforms from paramagnetic into magnetically ordered state with spontaneous magnetization. This transition is well pronounced in both real and imaginary parts of the ac magnetic susceptibility, as shown in the inset of Fig. 5. At T_N2, the transition occurs between two magnetically ordered states. Contrary to the real part of ac susceptibility, the imaginary part does not evidence any anomalies at T_N2. It means that no significant dissipation processes accompany this transition. This is in line with the observation that both phases at T < T_N2 and T_N2 < T < T_N1 possess similar properties, i.e. the presence of a small uncompensated moment and the hysteresis loop, as shown in the inset of Fig. 6.

Crystal chemistry of the shchurovskyite family of compounds

As it was mentioned earlier, the title oxocuprate phosphate, K₂CaCu₆O₂(PO₄)₄, represents a new synthetic modification of the natural fumarolic arsenate shchurovskyite K₂CaCu₆O₂(AsO₄)₄.¹³ Both phases exhibit similar crystal structures, which is reflected in their closely comparable unit-cell parameters and volumes (Table 2). Nevertheless, the title phase crystallizes in the centrosymmetric space group P2/c, in contrast to the polar C2 structure of shchurovskyite, indicating that K₂CaCu₆O₂(PO₄)₄ corresponds to a new structure type.

Table 2 Crystallographic parameters, structural characteristics and synthesis data of shchurovskyite-related salt-inclusion oxocuprates A₂B[M₆O₂](TO₄)₄

Formula	Space group, Z, V (Å^3)	Unit-cell parameters, (Å and °)	Cu-layers characteristics	Genesis/synthesis conditions	Ref.
a Standard setting P2/c was converted to P2/a for convenient comparison between polymorphs. b The authors setting was changed for the same reason.
Shchurovskyite, K2Ca[Cu6O2](AsO4)4	C2	a 17.2856(9)	Columns of CuO4 and CuO6	Sublimates of Arsenatnaya fumarole at temperatures above 380 °C	13
	2	b 5.6705(4), β 92.953(6)	Decoration with CuO5
	839.24(9)	c 8.5734(6)	Vertice connection
Rb2Ca[Cu6O2](PO4)4	C2	a 16.8913(4)	Columns of CuO4 and CuO6	Hydrothermal synthesis at 420 °C	14
	2	b 5.6406(4), β 93.919(3)	Decoration with CuO5
	794.57(4)	c 8.3591(3)	Vertice connection
K2Ca[Cu6O2](PO4)4	P2/a^a	a 16.8327(6)	Columns of CuO4 and CuO6	Flux in KCl at 690 °C	This paper
	2	b 5.6597(2), β 94.605(4)	Decoration with CuO5
	793.18(5)	c 8.3527(3)	Edge connection
Dmisokolovite, K3[Cu5AlO2](AsO4)4	C2/c	a 17.0848(12)	Columns of CuO4 and AlO6	Sublimates of Arsenatnaya fumarole at temperatures above 380 °C	13
	4	b 5.7188(4), β 91.716(6)	Decoration with CuO5
	1614.7(2)	c 16.5332(12)	Vertice connection
K2Cu[Cu6O2](PO4)4	P1̄^b	a 8.2612(4), α 103.239(4)	Columns of CuO5	Gas-phase crystallization at 800 °C	15
	1	b 5.7787(3), β 95.813(4)	Decoration with CuO5 and CuO6
	382.58(3)	c 8.3717(4), γ 96.821(4)	Edge connection
K2.35Cu0.825[Cu6O2](PO4)4	P21/n	a 16.7138(4)	Columns of CuO5	Gas-phase crystallization at 800 °C	15
	8	b 11.2973(3), β 90.775(2)	Decoration with CuO5 and CuO6
	3172.47(13)	c 16.8031(4)	Edge connection

---

Beyond these two compounds, several additional synthetic and natural phases have been assigned to the shchurovskyite-related structural family. Although mostly non-isostructural, all of them conform to the general formula A₂B[M₆O₂](TO₄)₄, where A is monovalent large alkali cation K⁺ or Rb⁺; B is primarily Ca²⁺ but in some cases Cu²⁺ or K⁺; T is P or As; M is Cu, or a combination of Cu and Al.

Within this family, only one isostructural-to-shchurovskyite synthetic phase was obtained – the Rb-substituted phosphate, Rb₂Ca[Cu₆O₂](PO₄)₄.¹⁴ A second mineral in this group, dmisokolovite, K₂K[Cu₅AlO₂](AsO₄)₄, was discovered together with shchurovskyite in Arsenatnaya fumarole sublimates.¹³ Interestingly, its composition is notable for containing only K⁺ cations (rather than both K⁺ and Ca²⁺), with charge balance maintained by the substitution of Cu²⁺ for Al³⁺. The crystal framework of dmisokolovite was shown to be a derivative to that of shchurovskyite.¹³ Additionally, two more shchurovskyite-related compounds have recently been synthesized by gas-phase crystallization under conditions emulating natural fumarolic environments: triclinic K₂Cu[Cu₆O₂](PO₄)₄ and monoclinic K_2.35Cu_0.825[Cu₆O₂](PO₄)₄.¹⁵ The relationships among these phases – considering archetypes and derivative superstructures that correlate with Z and unit-cell volumes – were comprehensively outlined by Kornyakov and Krivovichev.¹⁵

As shown in Table 2, all members of the shchurovskyite family differ primarily in the salt part of their structures, while sharing very similar compositions and cationic substructures within the “oxide floor” fragments {[O₂M₆]⁸⁺(TO₄)₂}. Nevertheless, the anionic environment of the copper polyhedra varies together with the changes in symmetry, which is clearly illustrated by the Cu-layer slices of the frameworks in Fig. 8.

Fig. 8 Layers built by the copper polyhedra in the crystal structures of K₂CaCu₆O₂(PO₄)₄, space group P2/a^a (a); shchurovskyite, K₂CaCu₆O₂(AsO₄)₄, sp. gr. C2 (b); K₂CuCu₆O₂(PO₄)₄, sp. gr. P1̄^b (c) and K_2.35Cu_0.825Cu₆O₂(PO₄)₄, sp. gr. P2₁/n (d). ^a,bSettings as in Table 2.

Both centrosymmetric synthetic K₂CaCu₆O₂(PO₄)₄ (Fig. 8a) and polar natural shchurovskyite K₂CaCu₆O₂(AsO₄)₄ (Fig. 8b) contain the same columns built of alternating CuO₄ planar squares and CuO₆ octahedra in their structures. However, differences emerge in the orientation and connectivity of the copper pyramids that decorate these rods. In the synthetic phosphate, the Cu4O₅ pyramids alternate the orientation of their O5 apical vertices (up/down) along the b axis in adjacent ribbons. In contrast, analogous Cu2O₅ pyramids in shchurovskyite maintain a single orientation, contributing to the polar character of the layer. The same polarity is reinforced by the second type of Cu1O₅ pyramids, which share only one vertex (O1) with each other. In the synthetic modification, this pyramid is slightly twisted, resulting in edge-sharing (O4–O4) connections between neighboring Cu3O₅ pyramids.

In crystal structures of the other two synthetic compounds – triclinic K₂Cu[Cu₆O₂](PO₄)₄ (Fig. 8c) and monoclinic K_2.35Cu_0.825[Cu₆O₂](PO₄)₄ (Fig. 8d) – the columns consist entirely of CuO₅ pyramids (here and below, only Cu–O bonds ≤2.9 Å are considered). Remarkably, the orientation of these pyramids was shown to form different sequences: UDUDUD for K₂Cu[Cu₆O₂](PO₄)₄ and UUDDUU for K_2.35Cu_0.825[Cu₆O₂](PO₄)₄, where U and D denote apical vertices pointing upward or downward relative to the plane of the copper layer.¹⁵ Furthermore, the combination of pyramids and octahedra contributes additional “decoration” to the columns – whereas in all other structures of the shchurovskyite family, only pyramids play this role (Table 2). At the same time, the decorating pyramids in the synthetic phases are linked by edge-sharing, in contrast to the vertex-sharing connectivity typical of the natural phases and the Rb-containing isostructural analogue. This edge-sharing pattern is consistent with centrosymmetric structures, while vertex-sharing is associated with the polar frameworks of minerals (Table 2).

Thus, most monoclinic shchurovskyite-related compounds crystallize in different space groups and structure types. This pronounced structural flexibility stems from the strong tendency of Cu²⁺ polyhedra to undergo Jahn–Teller distortions, which promote structural rearrangements driven by temperature effects and compositional variations.

Large channels accommodating K⁺ ions run along the b axis through the framework of K₂CaCu₆O₂(PO₄)₄ (Fig. 4a). The channels are framed with four copper pyramids and four phosphate tetrahedra alternating with each other. The effective width of the channel³⁰ is 2.5 Å × 5.2 Å (the smallest O8–O8 and the longest O4–O4 distances, respectively), suggesting easy migration of K⁺ ions through the structure. Thus, we made calculations of potential migration paths for K⁺ ions based on the structural topology by means of the ToposPro program³¹ and evaluated the energy barriers using the bond-valence energy landscape algorithm,³² as shown in Fig. 9, using the VESTA software.³³ For the topological calculations, based on the Voronoi-Dirichlet partition, the threshold values were established as follows: r_sd = 1.70 Å and r_chan = 2.38 Å, which derive from the sizes of K⁺ and O²⁻ ions.³⁴ As a result, the 1D paths for K⁺ ions were established running along the b axis through the large channels, described above. The BVEL calculations based on structural data confirmed the probability of K⁺ migration through the crystal structure, which can occur along the b axis at 0.6 eV (Fig. 9).

Fig. 9 Bond-valence energy landscape isosurfaces (isosurface level 15) for potassium atoms in the K₂CaCu₆O₂(PO₄)₄ structure shown in axonometric projection.

Summary

Single crystals of a new shchurovskyite-like oxocuprate phosphate, K₂CaCu₆O₂(PO₄)₄, were obtained via a flux method. Its crystal structure features rods composed of alternating CuO₆ and CuO₄ polyhedra, sharing edges. These rods are decorated by CuO₅ pyramids. In contrast to the vertex-sharing connectivity typical of the shchurovskyite and its Rb-analogue, Rb₂Ca[Cu₆O₂](PO₄)₄, the decorating pyramids show the title-phase edge-sharing. This difference in polyhedral linkage correlates with a significant variation in the copper–copper distances between decorated pyramids of neighboring rods: Cu3–Cu3 = 3.194 Å (K) vs. Cu1–Cu1 = 3.760 Å (Rb). Our study once again demonstrated the flexibility of the shchurovskyite structural archetype. This flexibility is governed by modifications to the “salt” part of the structure, in contrast to the conserved cationic substructure of the {[O₂M₆]⁸⁺(TO₄)₂} “oxide-floor”. The main variations occur principally in the “salt” cation sites, correlating with the changes in geometry and/or the connectivity of Jahn–Teller-active Cu²⁺ polyhedra. Bond-valence energy landscape calculations based on structural data confirmed the assumed probability of K⁺ migration through the crystal structure along the b axis at 0.6 eV.

Recently, it has been shown that Rb₂Ca[Cu₆O₂](PO₄)₄ orders ferrimagnetically at T_C = 25 K and evidences a 1/3 magnetization plateau starting at the zero magnetic field. It is because in the anion-centered [O₂Cu₆] dimer of the Rb-based compound (see Fig. 6a of ref. 14), the collinear UUDDUU spin state (U = up, D = down) is realized due to the competition of purely antiferromagnetic exchange interactions. The density functional theory calculations of super-exchange parameters in Rb₂CaCu₆O₂(PO₄)₄ point to the possible low-dimensional magnetic behavior of this compound, since they confirm weak magnetic coupling between the Cu-based layers. However, the actual character of the low-dimensional magnetism (1D or 2D) remains uncertain, owing to the presence of weak exchange interactions within the layers. Experimentally found value of the most intense component of the g-tensor reach g = 2.31 ± 0.01, which is in good correspondence with the calculated value and the value of g-factor g = 2.32 found in K₂CaCu₆O₂(PO₄)₄. The low-dimensional magnetic behavior of Rb₂CaCu₆O₂(PO₄)₄ is masked by the fierce competition between multiple ferromagnetic and antiferromagnetic exchange interactions.

Contrary to these observations, K₂CaCu₆O₂(PO₄)₄ appears to be canted antiferromagnet with weaker spontaneous magnetization most probably due to the Dzyaloshinskii–Moriya interaction. The K-based compound orders at two steps at T_N1 = 22 ± 2 K and T_N2 = 10 ± 1 K, where T_N1 corresponds to the onset of the long-range magnetic order, while the transition at T_N2 corresponds to some rearrangement of the magnetic structure. Notably, weak spontaneous magnetization is present at both T_N2 < T < T_N1 and T < T_N2 ranges. The magnetic structures in these temperature ranges can be established in the neutron scattering experiments.

Conflicts of interest

There are no conflict to declare.

Data availability

Supplementary information (SI) is available: XRPD (Fig. S1), atomic coordinates (Table S1), interatomic distances (Table S2), bond-valence data (Table S3), the anio-centering projection of the title compound’ crystal structure (Fig. S2), and the temperature dependence of Fisher's specific heat (Fig. S3). See DOI: https://doi.org/10.1039/d6dt00109b.

CCDC 2505942 for K₂CaCu₆O₂(PO₄)₄ contains the supplementary crystallographic data for this paper.³⁵

Acknowledgements

We thank N. V. Zubkova for her assistance in obtaining the X-ray diffraction data set. We are grateful to V. O. Yapaskurt for the X-ray spectral analysis. This work has been supported by Lomonosov Moscow State University (award no. AAA-A16-116033010121-7) and by the Ministry of Science and Higher Education of the Russian Federation within the framework of the Priority-2030 strategic academic leadership program at NUST MISIS. G. K. is grateful for the FMUF-2022-0002 state research project of IEM RAS. Access to the powder diffractometer Tongda-TDM-20 (Geology Faculty) was granted by the Lomonosov Moscow State University Program of Development.

References

1. I. V. Pekov, A. A. Agakhanov, N. V. Zubkova, N. N. Koshlyakova, N. V. Shchipalkina, F. D. Sandalov, V. O. Yapaskurt, A. G. Turchkova and E. G. Sidorov, Russ. Geol. Geophys., 2020, 61(5–6), 675–688.
2. F. Pertlik and J. Zemann, Mineral. Petrol., 1988, 38(4), 291–298.
3. V. R. Popova, V. A. Popov, N. S. Rudashevskiy, S. F. Gravatskikh, V. O. Polyakov and A. F. Bushsmakin, Zap. Vses. Mineral. O-va., 1987, 116, 358–367 . (in Russian).
4. I. V. Pekov, S. N. Britvin, A. A. Agakhanov, M. F. Vigasina and E. G. Sidorov, Eur. J. Mineral., 2019, 31(5–6), 1025–1032.
5. A. Vasiliev, O. Volkova, E. Zvereva and M. Markina, npj Quantum Mater., 2018, 3(1), 18.
6. G. L. Starova, S. K. Filatov, V. S. Fundamensky and L. P. Vergasova, Mineral. Mag., 1991, 55(381), 613–616.
7. V. Y. Pomjakushin, A. Podlesnyak, A. Furrer and E. V. Pomjakushina, Phys. Rev. B, 2024, 109(14), 144409.
8. M. Fujihala, T. Sugimoto, T. Tohyama, S. Mitsuda, R. A. Mole, D. H. Yu, S. Yano, Y. Inagaki, H. Morodomi, T. Kawae, H. Sagayama, R. Kumai, Y. Murakami, K. Tomiyasu, A. Matsuo and K. Kindo, Phys. Rev. Lett., 2018, 120(7), 077201.
9. A. N. Vasiliev, O. S. Volkova, E. A. Zvereva, A. V. Koshelev, V. S. Urusov, D. A. Chareev, V. I. Petkov, M. V. Sukhanov, B. T. Rahaman and B. T. Saha-Dasgupta, Phys. Rev. B:Condens. Matter Mater. Phys., 2015, 91(14), 144406.
10. D. I. Badrtdinov, E. S. Kuznetsova, V. Y. Verchenko, P. S. Berdonosov, V. A. Dolgikh, V. V. Mazurenko and A. A. Tsirlin, Sci. Rep., 2018, 8(1), 2379.
11. H. C. Wu, K. N. Denisova, D. Menzel, D. C. Kakarla, O. V. Maximova, T. W. Kuo, Z. H. Yang, C. H. Lee, W. H. Li, H. Berger, C. W. Wang, C. K. Chang, Y. C. Chuang, J.-Y. Lin, M. Gooch, C. W. Chu, A. N. Vasiliev and H. D. Yang, Phys. Rev. B, 2019, 100(24), 245119.
12. A. S. Botana, H. Zheng, S. H. Lapidus, J. F. Mitchell and M. R. Norman, Phys. Rev. B, 2018, 98(5), 054421.
13. I. V. Pekov, N. V. Zubkova, D. I. Belakovskiy, V. O. Yapaskurt, M. F. Vigasina, E. G. Sidorov and D. Y. Pushcharovsky, Mineral. Mag., 2015, 79, 1737–1753.
14. S. M. Aksenov, E. Y. Borovikova, V. S. Mironov, N. A. Yamnova, A. S. Volkov, D. A. Ksenofontov, O. A. Gurbanova, O. V. Dimitrova, D. V. Deyneko, E. A. Zvereva, O. V. Maximova, S. V. Krivovichev, P. C. Burns and A. N. Vasiliev, Acta Crystallogr., Sect. B:Struct. Sci., Cryst. Eng. Mater., 2019, 75, 903–913.
15. I. V. Kornyakov and S. V. Krivovichev, Crystals, 2021, 11, 807.
16. Agilent Technologies, CrysAlis PRO, Yarnton, UK, 2014.
17. G. M. Sheldrick, Acta Crystallogr., Sect. A:Found. Adv., 2015, 71, 3–8.
18. G. M. Sheldrick, Acta Crystallogr., Sect. C:Struct. Chem., 2015, 71, 3–8.
19. L. J. Farrugia, J. Appl. Crystallogr., 2012, 45, 849–854.
20. K. Brandenburg, DIAMOND, Crystal Impact GbR, Bonn, Germany, 2006.
21. I. D. Brown and D. Altermatt, Acta Crystallogr., Sect. B:Struct. Sci., 1985, 41, 244–247.
22. S. V. Krivovichev and S. K. Filatov, The Crystal Chemistry of Inorganic Compounds and Minerals that Contain Radicals of Anion-Centered Tetrahedra, St. Petersburg University, 2001.
23. S. V. Krivovichev, O. Mentre, O. I. Siidra, M. Colmont and S. K. Filatov, Chem. Rev., 2013, 113(8), 6459–6535.
24. V. V. Bakakin, J. Struct. Chem., 2025, 66, 1491–1526.
25. O. I. Siidra, E. V. Nazarchuk, A. N. Zaitsev, E. A. Lukina, E. Y. Avdontseva, L. P. Vergasova, N. S. Vlasenko, S. K. Filatov, R. Turner and G. A. Karpov, Eur. J. Mineral., 2017, 29, 499–510.
26. R. R. Shuvalov, L. P. Vergasova, T. F. Semenova, S. K. Filatov, S. V. Krivovichev, O. I. Siidra and N. S. Rudashevsky, Am. Mineral., 2013, 98, 463–469.
27. V. M. Kovrugin, M. Colmont, O. Mentré, O. I. Siidra and S. V. Krivovichev, Mineral. Mag., 2016, 80, 227–238.
28. S. A. Al'tshuler and B. M. Kozyrev, Electron paramagnetic resonance, Academic Press, 2013.
29. M. E. Fisher, Philos. Mag., 1962, 7, 1731–1743.
30. L. McCusker, F. Liebau and G. Engelhardt, Microporous Mesoporous Mater., 2003, 58, 3–13.
31. V. A. Blatov, A. P. Shevchenko and D. M. Proserpio, Cryst. Growth Des., 2014, 14, 3576.
32. BVlain Python Package, available at: https://github.com/dembart/BVlain (accessed December 2024).
33. K. Momma and F. Izumi, J. Appl. Crystallogr., 2008, 41, 653–658.
34. V. A. Blatov, Crystallogr. Rev., 2004, 10, 249.
35. CCDC 2505942: Experimental Crystal Structure Determination, 2026,  DOI:10.25505/fiz.icsd.cc2q3mvb.

---