Synthesis and Characterization of Oxide Chloride Sr2VO3Cl, a Layered S = 1 Compound

The mixed-anion compound with composition Sr2VO3Cl has been synthesized for the first time, using the conventional high-temperature solid-state synthesis technique in a closed silica ampule under inert conditions. This compound belongs to the known Sr2TmO3Cl (Tm = Sc, Mn, Fe, Co, Ni) family, but with Tm = V. All homologues within this family can be described with the tetragonal space group P4/nmm (No. 129); from a Rietveld refinement of powder X-ray diffraction data on the Tm = V homologue, the unit cell parameters were determined to a = 3.95974(8) and c = 14.0660(4) Å, and the atomic parameters in the crystal structure could be estimated. The synthesized powder is black, implying that the compound is a semiconductor. The magnetic investigations suggest that Sr2VO3Cl is a paramagnet at high temperatures, exhibiting a μeff = 2.0 μB V–1 and antiferromagnetic (AFM) interactions between the magnetic vanadium spins (θCW = −50 K), in line with the V–O–V advantageous super-exchange paths in the V–O layers. Specific heat capacity studies indicate two small anomalies around 5 and 35 K, which however are not associated with long-range magnetic ordering. 35Cl ss-NMR investigations suggest a slow spin freezing below 4.2 K resulting in a glassy-like spin ground state.


■ INTRODUCTION
Two-dimensional (2d) layered compounds are of interest for possible intercalation 1 and superconductivity 2 applications. A recent chemical design to increase the structural anisotropy, to form layered materials, is that of mixed-anion compounds. 3,4 By using two different anions, even further effects can be obtained, such as polar, heteroleptic coordinations, and charge ordering. One of the most investigated 2d lattices are the Ruddlesden−Popper (RP) phases that usually are pure oxides. By introducing a part chlorine in the oxide lattice, its anisotropy should be altered. This was successful for the compound series Sr 2 TmO 3 Cl (Tm = Sc, 5 Mn, 6 Fe, 7 Co, 8 Ni 9 ), which is structurally related to the 214 RP phases. There are higher homologues of the oxychloride RP phases, such as, Sr 3 Co 2 O 5 Cl 2 and Sr 4 Mn 3 O 8 Cl 2 , indicating the further design of RP phases if a further Tm can be introduced. 6, 8 Yet, a few of the Tms in Sr 2 TmO 3 Cl have not been reported. Herein, we describe the synthesis and fundamental properties of the vanadium homologue, Sr 2 VO 3 Cl. Note, IUPAC recommends to write Sr 2 VClO 3 , because of the alphabetical order of anions, but, for historical reasons, the halogen will be placed last in the formula throughout this report.

■ MATERIALS AND METHODS
In order to obtain an X-ray pure sample of the title compound, SrO, SrCl 2 (Alfa Aesar 99.995%), V (Alfa Aesar 99.5%), and V 2 O 5 (British Drug Houses, pro analysis) in amounts corresponding to the stoichiometry Sr 2 VO 2.875 Cl were used.
SrO was made in-house from SrCO 3 , (Fluka ≥ 98%) by an over-night heating at 1030°C under dynamic vacuum (p < 10 −2 mbar). The obtained SrO was proven to be X-ray pure, and it was only handled without exposure to air. The slight oxygen non-stoichiometry was intentional, because all starting composition with more oxygen resulted in samples with small amounts of V(IV) compounds, such as SrVO 3 . 10 Minor oxide impurities in the vanadium metal is expected to be the cause of this observation. The powder mixture was ground thoroughly in an agate mortar inside a glovebox (O 2 and H 2 O < 2 ppm) and all handling of the precursors was performed therein. The powder mixture was placed in a corundum crucible, which was melt-sealed inside a silica ampule. The solid-state reaction was performed in a muffle furnace at 1100°C for 10 h (heating rate: 5°min −1 ) and the subsequent cooling was done at an ambient rate. Note that no reaction was observed between the sample and crucible after the solid-state reaction.
Powder X-ray Diffraction (pXRD). X-ray diffraction at room temperature was performed using a Bruker D8 Discover with a Bragg−Brentano geometry, a Ge(1 1 1) Johanssen monochromator to select CuKα 1 radiation (λ = 1.5406 Å), and a Lynxeye as the detector. The obtained data were indexed using the software Topas. 11,12 Further analyses, including Rietveld refinement, were done using the Jana2006 software package. 13 In the refinement, the background was fitted by Chebyshev polynomials including 15 terms. First, the two unit cell parameters were refined together with the zero-position in a Le Bail fit, where the Pseudo-Voigt function was used to simulate the peak shapes. In the final refinement, all atomic fractional coordinates, as allowed by symmetry restrictions, as well as isotropic thermal displacement parameters were refined. Expecting anisotropic crystal morphology, preferred orientation along the (1 1 0) direction, according to March−Dollase, was included, significantly improving the overall fit. Within the diffraction range, there were 46 observed intensities that were simulated with 13 free parameters. All specific data are shown in Table 1.
Scanning Electron Microscopy (SEM). SEM images were captured using a Hitachi SU8230 field emission scanning electron microscope (FESEM). A current of 5 μA at an acceleration voltage of 5 kV was used. The elemental composition was determined using a XFlash 6|10 EDX detector. Averaged values of measurements from 10 different crystallites were used to estimate the elemental composition of the sample.
Magnetic and Thermodynamic Measurements. A Quantum Design Physical Property Measurement System (PPMS) was used to measure the magnetic susceptibility at three different magnetic field strengths (μ 0 H = 0.1, 0.5, 1 T) from 4 to 375 K. The polycrystalline sample was packed in a polypropylene sample holder and measured in a VSM insert at 40 Hz with 2 mm amplitude. The specific heat capacity was studied from 2 to 300 K on a non-sintered pellet at zero magnetic field in the same PPMS using the standard nonadiabatic thermal relaxation method, including an addenda subtraction from the absolute signal.
Solid-State Nuclear Magnetic Resonance (ss-NMR). NMR measurements were performed using a Tecmag Apollo spectrometer with a standard probe head and a sweepable Oxford Instruments magnet. Temperatures were controlled using a 4 He temperature insert (VTI). Temperatures below 4.2 K were achieved by pumping at the VTI. 35 Cl NMR spectra were obtained using the frequency sweep method at a fixed magnetic field of 7 T.

■ RESULTS
The resulting powder of Sr 2 VO 3 Cl is completely black, and no larger crystals could be observed after synthesis. During characterization, it was observed that the sample is slightly sensitive in air/to humidity, slowly decomposing with time, which leads to visible changes in its magnetic properties. SEM/EDX. Observation done in SEM shows that the sample has a high degree of crystallinity, see Figure 1. The crystal morphology resembles stacked plates with angles of 90°. The semiquantitative elemental analysis, as determined by EDX on 10 different crystallites, is Sr 2.1(1) V 0.94(5) O x Cl 1.0(2) when assuming all elements but oxygen to add up to four. Despite the facts that the compound is a poor conductor and the polycrystalline nature of the sample can cause topology issues for the EDX detector, these results agree fairly well with the expectations.
Powder X-ray Diffraction and Rietveld Refinement. The synthesized sample exhibits high crystallinity: the peak shapes are all narrow and high intensities, even at high angles, are observed. When comparing the Rietveld refinement with the observed data, the sample seems to be almost X-ray pure. The largest deviation between the observed and calculated intensities occurs at the peaks of highest intensities, indicating minor errors in the peak-shape function. The largest extrinsic intensity (from a secondary phase) is found close to 2θ = 35°, as indicated by an asterisk in Figure 2. However, its origin is unclear, although a SrO reflection can be expected at that angle.
In the title compound, two different vanadium-to-oxygen distances of 1.83 and 2.01 Å are observed. These values are significant shorter compared to the V−O distances in V 2 O 3 of 1.97 and 2.05 Å. 14 The observed distance between vanadium and chlorine is 2.94 Å for Sr 2 VO 3 Cl, much larger than the observed bond length of 2.417 and 2.418 Å for VCl 3 . 15 The calculated bond valence sum (BVS) for vanadium is 2.87 when including chlorine 16   value 3. This suggests that the vanadium coordination is most suitably described as a distorted octahedron with a 5 + 1 coordination. The vanadium atoms form 2d-layers as indicated in Figure 3a. This structure is directly related to a 214 RP phase, with the main difference that one of the oxygen atoms has been replaced by a chlorine. The two strontium atoms are differently coordinated: one of the strontium atoms only coordinates to oxygen while the other coordinates to both oxygen and chlorine, see Figure 3c. From the observed V−O− V angle of 162°, magnetic super-exchange is expected to be important, probably giving rise to AFM interactions. 17,18 Magnetic and Specific Heat Measurements. According to the magnetic susceptibility measurements, the title compound is paramagnetic down to at least 35 K as the signal increases when decreasing the temperature, following the Curie−Weiss law ( Figure 4). However, at low temperatures, the magnetic behavior changes: zero-field cooled (ZFC) and field-cooled cooling (FCC)/field-cooled warming (FCW) data are different. The temperature where the curves separate depends on the field strength: for 0.1 T, ZFC and FCC/FCW curves separate at ∼18 K as compared to ∼12 K for 1 T.
As observed in Figure 5, representing a 1/χ versus T plot, the high-temperature paramagnetic range can be approximated with a straight line. From a Curie−Weiss approximation, the effective magnetic moment μ eff = 2.0 μ B V −1 was determined. A diamagnetic correction was not performed because its influence on the paramagnetic signal is smaller than measurement errors. This is different from the expected magnetic moment for an S = 1 ion, such as the spin-only V 3+ ion: However, if an orbital moment would be active on the vanadium ion, it would oppose the spin moment, thus explaining, in part, the observation. The Curie− Weiss temperature was determined to be θ CW = − 50 K, indicating dominant antiferromagnetic spin−spin interactions. This is in accordance with the almost straight super-exchange paths between vanadium and oxygen.
The specific-heat data (shown in Figure 6a) indicate that the phonon contribution dominate in the high-temperature range. As the measurement was performed on a cool-pressed polycrystalline sample, the coupling to the sample holder was not optimal, but always above roughly 80%. This could in part explain the difference between the C p and the Dulong−Petit limit close to room temperature. Interestingly, extra con-    tributions to the specific heat appear below 5 and around 35 K, as shown in the insets of Figure 6a.
In order to obtain the magnetic contribution to the specific heat, the lattice contribution was estimated by fitting the experimental C p data, excluding points in the ranges T < 8 K and 25 K < T < 50 K, where significant features were observed. The fit was performed using the following Debye model: where R is the universal gas constant, the index i sums over two different Debye temperatures Θ D1 and Θ D2 , each one representing a part of the total unit cell formed by n = 7 atoms (n 1 = 3 and n 2 = 4). The total estimated lattice contribution (shown by the red line in the lower inset of Figure  6a) was obtained by extrapolating the fit function to the temperature range 2 K < T < 70 K. After subtracting the lattice contribution from the experimental C p data, the magnetic contribution to the specific heat was obtained, which is plotted as C mag /T as a function of temperature in Figure 6b (left axis). A clear peak-like feature is observed between 30 and 40 K. The magnetic entropy (S mag ) was evaluated by integrating C mag /T as shown in Figure 6b (right axis). An entropy release of approximately 8.8 J mol −1 K −1 at 70 K is, in principle, close to the expected for a S = 1 ground state, (R ln(2S + 1)) = 9.1 J mol −1 K −1 . It is worth noticing that due to the fit approximations, the extracted magnetic contribution to the specific heat has a semiquantitative character only. Nevertheless, this does not alter the general behavior observed: as indicated by the vertical line in Figure 6b at T = 30 K, most of the entropy is not associated with the peak-like feature around 35 K, ruling it out as the signature of magnetic long-range ordering. Instead, roughly 96% of the entropy release takes place for temperatures T < 30 K, where there is no evidence of a sharp λ peak, but only a small anomaly for T < 5 K as shown in the upper inset of Figure 6a. The possible nature of such behavior is discussed in the following sections.
ss-NMR. 35 Cl NMR is used as a local probe to check for the presence of possible magnetic ordering in Sr 2 VO 3 Cl. 35 Cl is a spin-3/2 nucleus for which one would expect a powder NMR spectrum with a central line (1/2 ↔ −1/2 transition) together with quadrupole singularities on both sides (corresponding to the −3/2 ↔ −1/2 and 3/2 ↔ 1/2 transitions). Figure 7 shows the frequency sweep 35 Cl NMR spectra measured at H = 7 T. At high temperatures, the NMR spectra consist of a single and narrow central line with a broad background. This can be qualitatively described by a powder distribution with a small quadrupole splitting of about 0.02 MHz and a large quadrupole broadening of about 0.6 MHz (see the fit at 50 K). The NMR lines broaden monotonically as the temperature is lowered, and  no evidence of magnetic long-range ordering is observed, while the line broadening below 4.2 K is likely due to spin freezing at low temperatures.

■ DISCUSSION
As all known compositions in the Sr 2 TmO 3 Cl series are isostructural, changes in unit cell parameters can be used to estimate local properties, such as the electronic configuration of the transition metal. The unit cell volume of the Sr 2 TmO 3 Cl members (Figure 8a) scales relatively well with increasing effective nuclear charge going from Sc to Ni. For the unit cell parameter c (Figure 8b), a sudden change in the trend is observed for both Mn and Ni. As V(III) is d 2 , and the coordination is quasi octahedral (Schonflies symmetry, C 4v ), the electrons will occupy B 2g and E g orbitals (corresponding to the t 2g orbitals in an octahedral field). 19 22,23 The single electron in the B 1g orbital could result in Jahn−Teller distortions and therefore explain the suddenly elongated c unit cell vector in comparison with the other homologues.
On further examining the magnetic data of Sr 2 TmO 3 Cl ( Figure 5), a difference between ZFC and FC curves is observed. The usual interpretation of this observation is the existence of magnetic domains. Nevertheless, long-range magnetic ordering is excluded considering that typical signs of second-order transitions are absent in the specific heat. On the other hand, fluctuating short-range AFM correlations, supplied by the advantageous super-exchange vanadium− oxygen paths, cannot explain the ZFC/FC splitting in the magnetization curves. Instead, based on the ss-NMR data, a freezing of spin-clusters (or soft domains) at lower temperatures is a tentative explanation that agrees well with a magnetic field dependent freezing temperature. This scenario is supported by the magnetic entropy calculation, where it is observed that a large part of the entropy is released at low temperature. Alternatively, the anomaly observed in the specific heat below 5 K could be associated to a two-level Schottky anomaly. As already reported, 19 the quasi-octahedral coordination of vanadium has a C 4v symmetry that leads to dorbital splitting where the d xz and d yz are degenerate but different from d xy . These discrete energy levels could then dominate the behavior of the system when thermal excitation is comparable to the energy spacing between d orbitals. The probability of populating the upper level via thermal excitations can manifest itself as a broad peak in the specific heat. Systematic measurements under several applied magnetic fields would be necessary to confirm a probable Schottky character of the anomaly, which is, however, beyond the scope of this work.
Experimentally determined Weiss constants, Neél temperatures, Tm−O−Tm angles, and Tm−O distances observed in the Sr 2 TmO 3 Cl (Tm = Sc, Mn, Fe, Co, Ni) analogues are shown in Table 2.
When comparing the magnetic interaction strengths for the different homologues, it is obvious that the transition metal has a great influence on the magnetic behavior of the RP oxide chlorides: the Weiss constants range from negative to slightly positive. A special case is seen for the Ni analogue where a positive Weiss constant is observed, although its ground state is indeed AFM ordered. 9 The fact that Co and Fe analogues are already AFM ordered at room temperature can be explained by their relatively larger magnetic moment of high-spin d 5 and d 6 ions. However, Mn (HS d 4 ) and Co (HS d 6 ) have the same S moment, but differ in the number of electrons in the B 1g and A 1g orbitals; Co has one electron in all but the doubly occupied B 2g orbital, making it more capable of being involved in relatively stronger super-exchange interactions. As expected, due to the different spin sizes, the homologues with Tm = V 3+ (d 2 ) and Tm = Mn 3+ (d 4 ) have very different mean superexchange interaction strengths. No obvious trends are observed between the Tm−O eq −Tm angles/Tm−O eq distances and the magnetic interactions.  a The angles and distances are between the Tms and the equatorial oxygen atoms.

■ CONCLUSIONS
A new Sr 2 TmO 3 Cl (Tm = Sc, Mn, Fe, Co, Ni) analogue with Tm = V 3+ has been discovered, as observed by X-ray diffraction data and elemental analysis. Thermodynamic and 35 Cl NMR measurements suggest a high-temperature paramagnetic state with spin freezing at low temperatures. A comparison among the structural homologues reveals that the possible transition metal electronic configuration has a strong influence on the unit cell parameters. The choice of Tm also strongly influences the magnetic interaction strengths and the magnetic ground state in a nontrivial way. Obviously, the higher homologues of the RP phases with chlorine replacing oxygen are a promising area for future research.
■ ASSOCIATED CONTENT

* sı Supporting Information
The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.3c01151. CIF file from JANA2006 containing the results from the here-used Rietveld refinement of powder X-ray data (CIF)