Structural and electronic tunability of Ruddlesden–Popper oxyfluorides through nickel–copper substitution in La₂Ni_1−xCu_xO_2.5F₃ (0 ≤ x ≤ 1)

Abstract

In this contribution we report on the synthesis, structure and optical characterization of Ruddlesden–Popper oxyfluorides La₂Ni_1−xCu_xO_2.5F₃ (0 ≤ x ≤ 1) obtained by topochemical low-temperature fluorination of La₂Ni_1−xCu_xO₄ with polyvinylidene fluoride (PVDF). Our study reveals that the anionic ordering in the tetragonal unit cell of La₂NiO_2.5F₃ persists even at high Cu substitution levels (x = 0.9), with minimal change in unit cell volume. This observation is contrary to expectations based on Jahn–Teller induced unit cell distortions, which were previously reported for the oxides La₂Ni_1−xCu_xO₄, as well as for the closely related oxyfluorides La₂Ni_1−xCu_xO₃F₂. The pure copper-containing compound La₂CuO_2.5F₃ crystallizes in a triclinic version of the same structure, and the symmetry lowering is attributed to the enhanced space requirements of the Jahn–Teller elongated CuO₄F₂ octahedra. The structural investigations based on XRD and ND Rietveld refinements are supported by low-field ¹⁹F MAS NMR experiments. We also report the results of diffuse reflectance UV-Vis measurements, which are complemented by DFT calculations. Here, we demonstrate a strong impact of the Cu substitution on the electronic structure of the oxyfluorides, resulting in band gap energies in the range of 3.4 eV to 1.3 eV, spanning the whole visible spectrum. Notably, first photocatalytic water splitting tests reveal a considerable hydrogen evolution activity for x = 0.2, highlighting the potential of Ruddlesden–Popper oxyfluorides for solar energy applications.

---

Introduction

Compounds with the Ruddlesden–Popper (RP) structure type (AX)(ABX₃)_n (A, B are the cations and X are the anions) are a hot topic in solid state materials chemistry. The characteristic arrangement of ABX₃ perovskite layers with thickness n alternatingly interconnected with one halite AX layer results in a high structural flexibility, which gives rise to different electronic, magnetic, and optical properties. Therefore Ruddlesden–Popper compounds are interesting for both fundamental research and industrial applications, for example in the fields of energy storage,^1,2 catalysis,^3,4 and electronics.^5,6 Recently superconductivity was observed for the infinite 2D nickelates like NdNiO₂ ^7,8 and also for the bilayer n = 2 and trilayer n = 3 nickelates La₃Ni₂O₇ and La₄Ni₃O₁₀ when pressurized to 14 GPa⁹ and 69 GPa¹⁰ respectively.

An important part of the compositional flexibility of RP-compounds results from a high flexibility towards deviations from the ideal anion composition. In the case of n = 1 both over-stoichiometric¹¹ and under-stoichiometric¹² compounds are found. As a result, aliovalent anion substitution, for example, of one O²⁻ anion with two F⁻ anions can be used to obtain compounds with modified physical properties. Such fluorination reactions are for example possible by reacting oxide precursors with the fluorinated polymers polyvinylidene fluoride (PVDF, [CH₂CF₂]_n) or poly tetrafluoroethylene (PTFE, [CF₂CF₂]_n) in a non-oxidizing low temperature reaction¹³ yielding the corresponding oxyfluorides. These fluorinations often show highly complex reaction mechanisms involving several less fluorinated intermediates.¹⁴ The resulting oxyfluorides generally show different physical properties than the precursor oxides. Examples are the significantly increased antiferromagnetic transition temperature in Sr_3.5La_0.5Fe₃O_7.5F_2.6 from T_N ≈ 50 K to T_N > 450 K,¹⁵ and the increased optical band gaps (E_g) of 3.4 eV in the 3F-oxyfluoride La₂NiO_2.5F₃ ¹⁶vs. 1.3 eV ¹⁷ for La₂NiO₄ which was obtained from UV-Vis measurements as well as band structure calculations making this compound of potential interest for applications using the high energy range of the solar spectrum. A similar change of E_g from 2.3 eV in LaBaInO₄ to 1.9 and 3.3 eV in LaBaInO₃F₂ was recently reported to result from topochemical fluorination.¹⁸

In our previous work on the substitution series La₂Ni_1−xCu_xO₃F₂ (0.0 ≤ x ≤ 1.0) we reported a strong impact of the Ni/Cu ratio on the structural distortion (orthorhombic → monoclinic → triclinic, with increasing x)^19,20 which is linked to an increased Jahn–Teller distortion of the Ni/CuO₄F₂ octahedra. We also highlighted the resulting change in thermal stability (increase of the decomposition temperature from 460 °C for x = 0.0 and 420 °C for x = 1.0 up to 520 °C with x = 0.6)²¹ as well as the tunability of the antiferromagnetic Néel temperature in the range of 40 K to 260 K.

In this study, we present the even higher fluorinated compounds of the substitution series La₂Ni_1−xCu_xO_2.5F₃ with three F⁻ per formula unit over the complete range (0.0 ≤ x ≤ 1.0). The compounds are obtained by the topochemical low-temperature fluorination with PVDF as fluorination agent. Structural characterization using XRD and ND Rietveld refinements clarify the impact of the Ni/Cu ratio on the changes of the anion-ordered crystal structure. In addition, the Ni/Cu ratio results in strong changes in the optical band gaps which are analysed by UV-Vis measurements. Our results are complemented by DFT calculations and photocatalytic hydrogen evolution tests, which show a significant activity.

Experimental section

Synthesis

The precursor oxides of the La₂Ni_1−xCu_xO₄ solid solution were synthesized in steps of x = 0.1 by a citric acid assisted combustion method as reported before.¹⁹ In short, stoichiometric amounts of La₂O₃ (Merck) (dried at 900 °C for 10 h), Ni powder (Sigma-Aldrich) and copper(II) acetate (98%; Sigma-Aldrich) were dissolved in distilled water containing a few drops of concentrated HNO₃ and citric acid (molar ratio metal ions : citric acid = 1 : 3). The oxides were obtained by evaporating the water on a hot plate at 100 °C, carbonization of the resulting gel at 350 °C and subsequent calcination at 950 °C for 6 h in a box furnace in static air. The oxyfluorides La₂Ni_1−xCu_xO_2.5F₃ were synthesized by mixing the oxides with polyvinylidene fluoride ((PVDF/CH₂CF₂)_n) (Alfa Aesar) in a molar ratio of 1 : 1.5 (oxide : CH₂CF₂) with a small excess of 2% polymer. The mixtures were slowly (2 °C min⁻¹) heated to 340 °C, kept at this temperature for 28 h and afterwards allowed to cool down to room temperature in the box furnace.

Characterization

X-ray diffraction (XRD) patterns were recorded on a Bruker AXS D8-Advance diffractometer operating in Bragg–Brentano geometry. Cu-Kα_1,2 radiation was used and the device was equipped with a 1D silicon strip detector (LYNXEYE). Patterns were recorded in the angular range 2θ = 10–140° with a step size of 0.01° and a time of 3 s per step. Additionally, a STOE STADI MP diffractometer operating in transmission geometry with monochromatic Mo-Kα₁ radiation and equipped with a DECTRIS MYTHEN2 1 K detector was used to record patterns in the angular range of 2θ = 5–75° with a data point resolution of 0.015° and a total acquisition time of 2 h per scan.

Neutron powder diffraction (NPD) data of selected oxyfluorides (x = 0.3, 0.7, 0.9, 1.0) were collected on the high-resolution powder diffractometer D2B at the Institute Laue-Langevin in Grenoble, France. Beamtime was granted for proposal 5-23-769.²² Measurements were performed at 300 K with λ = 1.594 Å for sample amounts of ∼1.5 g (6 mm V-cylinder) and acquisition times of about 3 h per sample. For Rietveld refinements of XRD and NPD data GSAS II²³ was used.

The samples were checked for possible residual fluorination agent as well as other carbon related impurities by infrared spectroscopy on an ATR-FT-IR Bruker Tensor 27 spectrometer in the range 4000–250 cm⁻¹.

UV-Vis spectra were recorded in diffuse reflection mode on a PerkinElmer Lambda 19 spectrometer in praying mantis geometry in the range of λ = 1350–250 nm with 1 nm resolution. BaSO₄ (99.998%, Alfa Aesar) was used as white standard.

The La, Ni, and Cu contents of pelletized powder samples (10 wt% cellulose as binder) were quantified by X-ray fluorescence spectroscopy (XRF) using a Panalytical Epsilon 4 spectrometer. Data analysis was carried out based on the fundamental parameter approach (Omnian mode) based on 6 independent scans with excitation voltages between 5 and 50 kV in combination with filters of different materials and thicknesses (Ag, Cu, Al, and Ti).

Iodometric titration was used for the determination of the average Ni/Cu oxidation states. About 35 mg of the samples were dissolved in concentrated HCl containing an excess of KI. Afterward, 1 g of NaHCO₃ was added to adjust the pH and to create a CO₂ saturated atmosphere preventing the solutions from oxidation. For titration a 0.005 M Na₂S₂O₃ solution was used. The obtained oxidation state was calculated from three averaged titrations per sample.

¹⁹F Magic Angle Spinning (MAS) NMR experiments were performed on a 1.4 T magnet in combination with an Avance II Bruker NMR console running Topspin V2.1, operating at a frequency of 56.4 MHz. Magic angle sample spinning on the 1.4 T magnet was carried out with a home-made conical stator with the sample packed in 3D printed conical rotors, spinning at approximately 19 kHz.²⁴ The chemical shift values refer to CFCl₃, according to the IUPAC list of reference compounds.²⁵ The spectra were analyzed with deconv2Dxy.²⁶ The repetition delay was set as 1 s.

Photocatalytic hydrogen evolution experiments were performed for selected oxyfluorides with x = 0.0, 0.2, and 0.4. 75 mg of the samples were dispersed in 150 mL of ultrapure water with TOC = 2 ppb containing 10% ethanol. The dispersion was continuously stirred at of 20 °C (Lauda thermostat (ECO RE1050G)). The reactor was flushed with Argon (grade 5.0, purity ≥ 99.999%) with a flow rate of 100 mL min⁻¹ (Bronkhorst mass flow controller) to remove all residues of air. Afterwards, the flow rate was reduced to 25 mL min⁻¹ of Argon and the online detection of the evolved gases was performed using a GC2014 gas chromatograph from Shimadzu, equipped with a shin carbon ST column (Restek) and a thermal conductivity detector, using Argon as carrier gas. Dark measurements were performed before starting the irradiation of the sample with a 300 W Xe lamp (Quantum Design) for 2.0 to 2.5 h. Then the lamp was turned off and the evolved gases were detected until a reduced hydrogen evolution was detected. Afterwards, an H₂PtCl₆ solution was added through a rubber sealing without opening the reactor to reach a concentration of 0.5 wt% of Pt. Then, the lamp was turned on again and the sample was irradiated for another 2 h. The lamp was turned off and the detection of the evolved gasses was performed until no hydrogen could be detected anymore. For comparison, a 10% aqueous ethanol solution was also measured without the addition of a sample. The ethanol solution was irradiated for 3.5 h with the same 300 W Xe lamp. Afterwards, the gas evolution was detected until no gas was evolved anymore.

Density functional theory (DFT) calculations were performed within the projector augmented wave (PAW) formalism,²⁷ as implemented in the Vienna ab initio simulation package (VASP).^28,29 A plane-wave basis cutoff of 520 eV was set and total energy convergence was ensured to be within 0.01 meV per unit cell. For geometry optimization, the Perdew–Burke–Ernzerhof exchange–correlation functional for solids (PBEsol)³⁰ was applied and the Γ-centered k-point grids with a density of 2000 k-points per reciprocal atom (kppa) were used to sample the Brillouin zone. An effective on-site Coulomb correction³¹ of 6.2 eV was applied for Ni 3d states and denser (4000 kppa) k-grids were used to obtain the electronic structures. For compositions containing both Ni and Cu various magnetic and structural configurations were generated using the cluster expansion method implemented in the ATAT package.³² To simulate the behavior of disordered Ni/Cu occupations, the band gaps were averaged for multiple ordered configurations with weights according to the Boltzmann distribution of their total energies, using the generalized quasichemical approximation (GQCA).^33,34 Furthermore, the band structure of the mixed cationic configuration with the largest gap at x = 0.25 (75% Ni ratio) was un-folded into the pure La₂NiO_2.5F₃ Brillouin zone using the Python package Easyunfold.³⁵

Results and discussion

Structure and composition of oxyfluorides La₂Ni_1−xCu_xO_2.5F₃

Based on the results of our previous work on La₂NiO_2.5F₃ the oxyfluorides are best prepared by topochemical fluorination of oxides obtained from soft chemistry synthesis. In contrast, fluorination of solid state precursors usually results in partial decomposition of the desired oxyfluoride before complete formation.¹⁶ More reactive soft chemistry precursor oxides were therefore prepared by a citrate-based approach. These precursor oxides were also used for the closely related oxyfluorides La₂Ni_1−xCu_xO₃F₂ which we reported previously.^19,21 For the x = 0 compound (La₂NiO_2.5F₃) a pronounced metastability was found in previous in situ XRD experiments.^14,16 Preliminary temperature-dependent in situ XRD experiments were therefore used to establish suitable reaction conditions for the samples in this work. Based on these investigations bulk synthesis was performed at 340 °C for 28 h. Such reaction optimization experiments were especially needed for the Cu-rich compounds of the substitution series as the stability of the oxyfluorides was found to further decrease with increasing Cu substitution level. All oxyfluorides with the formula La₂Ni_1−xCu_xO_2.5F₃ were obtained phase pure in the full range of 0 ≤ x ≤ 1 applying these conditions. The absence of significant amounts of unreacted PVDF as well as other residual organic compounds in the final products was routinely checked by IR spectroscopy (not shown) and later confirmed by ¹⁹F MAS NMR spectroscopy (vide infra) for the sample with x = 0.0, 0.3, and 0.7. Only for the x = 1 sample a weak PVDF signal was found most probably resulting from the slight excess of PVDF which was used for the synthesis. The La : Ni : Cu ratio of the resulting products was determined via XRF spectroscopy and no major deviations from the expected value of La : (Ni_1−xCu_x) = 2 : 1 was found as the ratios were ranging from 1.9 to 2.1. Additionally, for the Ni : Cu ratio no considerable deviations from the nominal x values were observed either. The fluorination with PVDF is expected to be non-oxidative³⁶ and therefore no deviations of the oxidation state of nickel/copper from +2 is expected. This assumption was checked for two selected samples (x = 0.3, and 0.7) by iodometric titration and average oxidation states of +2 (2.07(2), and 2.02(2)) was found for the B-type cations for both samples. This finding is in good agreement with the results of iodometric titrations, which we previously performed for the starting oxides as well as members of the 2F-substitution series La₂Ni_1−xCu_xO₃F₂, where also no hints for a change of the oxidation states was found.¹⁹ An average oxidation state of +2 and fully occupied anion sites derived from neutron powder diffraction (vide infra) were seen as confirmation of the nominal O_2.5F₃ anion stoichiometry. This is in agreement with the results obtained with an F⁻ ion sensitive electrode that were previously performed for the x = 0.0 compound La₂NiO_2.5F₃.¹⁶

The XRD patterns of all oxyfluorides La₂Ni_1−xCu_xO_2.5F₃ are shown in Fig. 1. Only subtle changes in the reflection positions are observed for all compounds with 0 ≤ x ≤ 0.9 and the diffraction patterns qualitatively resemble the one of La₂NiO_2.5F₃. This observation is highly surprising as both the starting oxides and the 2F-oxyfluorides were previously found to exhibit prominent Ni/Cu-dependent unit cell distortions with a corresponding symmetry lowering from tetragonal to orthorhombic for the oxides (I4/mmm → Bmab)³⁷ and orthorhombic to monoclinic to triclinic for the oxyfluorides (Cccm → C2/c → P1̄).¹⁹ This symmetry lowering is the result of an increased space requirement of the BX₆ octahedra caused by Jahn–Teller elongation due to introduction of Cu²⁺ with d⁹ electron configuration. In the case of the here investigated oxyfluorides a deviating diffraction pattern is only observed for x = 1 where clear shoulders appear at both sides of the (113) peak (indexed and marked with arrows in Fig. 1). These shoulders in combination with a further splitting of several other reflections indicate a symmetry lowering to triclinic. A triclinic unit cell metric was also obtained for the corresponding 2F-oxyfluoride La₂CuO₃F₂.²⁰ It has to be noted though that both compounds (La₂CuO₃F₂ and La₂CuO_2.5F₃) exhibit clearly different diffraction patterns resulting from different structural distortions.

Fig. 1 XRD patterns of all members of the solid solution La₂Ni_1−xCu_xO_2.5F₃ and a detailed view of the (113) and (200) main reflections. The shift of (200) is highlighted by a slightly curved, dotted line as guide for the eye. For x = 1 indices are given for two of the additional reflections resulting from the triclinic unit cell symmetry.

Rietveld refinements were conducted for all oxyfluorides in order to quantify the Ni/Cu-dependent changes in the unit cell parameters as well as to confirm the structural distortions. These refinements were performed as joint refinements based on two independent X-ray datasets (Cu-Kα_1,2, and Mo-Kα₁ radiation). By this approach we were able to achieve higher data redundancy combining the advantages of the increased angular resolution of the Cu-Kα_1,2 patterns with the higher accessible Q-range in the transmission Mo-Kα₁ data. For selected substitution levels (x = 0.3, 0.7, and 0.9) neutron powder diffraction patterns were additionally combined with the XRD data. Both XRD patterns were given the weighting factor 0.5 and the NPD data got the weighting factor 1 in the refinements. The Rietveld plots for the X-ray and neutron diffraction patterns of x = 0.3 and 0.7 are shown in Fig. 2. Rietveld plots of all other refinements and the obtained crystallographic parameters can be found in the supplement (Fig. S1, S2, Tables S1 and S3). Due to the high structural similarity the assignment of the anions to the different positions was performed according to the pure nickel compound (La₂NiO_2.5F₃). Here based on DFT and BVS calculations oxygen was found to be located on the equatorial octahedral positions (O_eq1; 4d, and O_eq2; 4g) as well as one interstitial position (O_int; 2b) and fluorine is located on the apical octahedral positions (F_ap; 8m) as well as on one interstitial site (F_int; 4c) giving an overall ¾ interstitial anion occupation.¹⁶

Fig. 2 Rietveld plots obtained for the joint refinements of X-ray and neutron powder diffraction data of La₂Ni_0.7Cu_0.3O₃F₂ (a and b) and La₂Ni_0.3Cu_0.7O₃F₂ (c and d). Refinements were carried out using the structure model of La₂NiO_2.5F₃ (P4₂/nnm).

From the difference Fourier map obtained for the NPD data missing scattering density was found to be located near the apical anion positions for both x = 0.3 and 0.7. A similar observation was previously made for the neutron diffraction data of the 2F-oxyfluoride La₂Ni_0.2Cu_0.8O₃F₂ for which a split apical position (F_ap) due to local occurrence of Jahn–Teller elongated CuO₄F₂ and non-elongated NiO₄F₂ octahedra was derived.²⁰ Final refinements of the x = 0.3 and 0.7 datasets of this work were therefore performed in the La₂NiO_2.5F₃ structure model including a split apical (8m) anion position (denoted as F_ap1, and F_ap2). With this structure model good fits were obtained for all datasets and two different inter atomic distances are obtained for the apical octahedral positions (x = 0.3: Ni/Cu–F_ap1 = 2.15 Å and Ni/Cu–F_ap2 = 2.37 Å, x = 0.7: Ni/Cu–F_ap1 = 2.07 Å and Ni/Cu–F_ap2 = 2.36 Å) indicative of the Jahn–Teller stretching of the CuO₄F₂ octahedra. These distances are close to the ones observed for La₂Ni_0.2Cu_0.8O₃F₂ (Ni/Cu–F_ap = 2.09 Å and 2.30 Å).²⁰ The refinement of fractional site occupation factors (sof) (assuming a complete total occupation as found in initial refinement runs) gave values of sof_Fap1/sof_Fap2 ≈ 0.7/0.3 for x = 0.3 and 0.3/0.7 for x = 0.7 resembling the nominal Ni/Cu values. The good agreement of the occupation factors with the atomic distances is a clear confirmation that the different octahedral geometries indeed result from the Jahn–Teller distortion of the Cu²⁺-environment and that the shorter distance reflects the Ni–F_ap1 bond while the longer distance corresponds to the Cu–F_ap2 bond. The refined unit cell of x = 0.3 and a detailed view of one (Ni,Cu)O₄F₂ octahedron (for x = 0.3) is shown in Fig. 3. The structure of x = 0.7 is highly similar and therefore not shown here.

Fig. 3 The crystal structure of La₂Ni_1−xCu_xO_2.5F₃ (x = 0.3) obtained from Rietveld refinements. An enlarged view of one Ni/CuO₄F₂ octahedron is additionally shown highlighting the different atomic distances and tilting directions.

In addition to the increased Cu–F_ap2 distance, the CuO₄F₂ octahedron exhibits a significantly increased octahedral tilt compared to the NiO₄F₂ octahedron. Interestingly, no indications for a splitting for the equatorial position (O_eq2; 4g) were observed. This results in stronger deviations of the F_ap–Ni/Cu–O_eq2 angle from the expected 90° angle, i.e. ∠(F_ap1–Ni/Cu–O_eq2) ∼88° and ∠(F_ap2–Ni/Cu–O_eq2) ∼80° for both Ni/Cu ratios. A possible explanation is the contraction of the equatorial plane associated with the Jahn–Teller stretching, which is opposed to a further deflection of the equatorial position (4g) along c.

Structure refinements with the same tetragonal structure model gave good fits for the X-ray diffraction data of all other compounds with x = 0.0 to 0.9. No additional unindexed or systematically extinct peaks, which would point to a different unit cell symmetry, were observed. The split apical position of x = 0.3 and 0.7 was not taken into account in these refinements as the anionic positions are subjected to an increased error in the presence of heavy scatterers like La, Ni, and Cu. The tetragonal structure of the pure nickel compound La₂NiO_2.5F₃ can therefore be confirmed to exist throughout the entire substitution series (except for x = 1; vide infra). The absence of structural transitions shows that the 3F structure has a significantly higher tolerance for additional octahedral stretching than the structural variant of the 2F oxyfluorides, where a symmetry reduction from orthorhombic (Cccm) to monoclinic (C2/c) symmetry already occurs at x = 0.2 due to the need of an additional tilting component of the octahedra.¹⁹

The unit cell parameters and selected atomic distances, which were obtained from the refinements, are shown in Fig. 4 for x = 0.0 to 0.9. For the cell parameters a and c a non-linear dependence on the Cu substitution level is apparent. The substitution series therefore does not follow Vegard's law. While parameter a decreases monotonously with x, c exhibits a non-monotonous dependency, with a minimum at x = 0.1–0.2 and a maximum at x = 0.8. The difference between these two extrema is 0.124 Å. This corresponds to only about 1/3 of the overall change in the long axis (0.370 Å) of the corresponding 2F oxyfluorides.¹⁹ The deviation from the expected linear trend of a, and c might hint to a phase transition that is not resolved by the powder diffraction data. The assumption is in concordance with changes in the NMR spectra of the x = 0.7 sample (vide infra). To check for a possible phase transition, additional refinements in the five tetragonal translationengleiche sub groups of P4₂/nnm (P4̄n2, P4̄2m, P4₂nm, P4₂22, and P4₂/n) were carried out for the x = 0.3, and 0.7 datasets. None of these structure models resulted in a significant decrease of the quality factors R_w or χ² despite the higher number of refinable parameters. We therefore stick to the P4₂/nnm structure model as correct choice for all compounds with x ≤ 0.9. The non-monotonous cell parameter dependence on x also affects the unit cell volume. With increasing Cu content there is first a volume contraction up to x = 0.3, followed by a volume expansion reaching a maximum at x = 0.6, which is then followed by another volume contraction. The relative volume change compared to La₂NiO_2.5F₃ is slightly negative with the strongest change of −0.9% observed for x = 0.3. This finding is a serious difference to the 2F oxyfluorides La₂Ni_1−xCu_xO₃F₂, in which Ni/Cu substitution leads to an almost linear enlargement of the unit cell of ∼5%. The apical Ni/Cu-F_ap atomic distance (shown in Fig. 4d) clearly increases with the Cu content, which is expected due to the Jahn–Teller elongation. For the equatorial distances, a decrease is found for increasing x, which is in concordance with the expectations from the Jahn–Teller distortion. In principle, a corresponding increase in c would also be expected. However, the elongation of the copper-containing octahedra is compensated by a stronger tilting along [110] as described above for x = 0.3 and 0.7.

Fig. 4 Evolution of the unit cell parameters a, and c (a) and cell volume (b) as well as the apical (Ni/Cu−F_ap) and equatorial (Ni/Cu−O_eq) distances (c and d) with the copper content (x) as obtained by Rietveld refinements. The lines serve as guides to the eye. Error bars are not given as they are in the range of the symbol size.

Refinements of the x = 1.0 oxyfluoride La₂CuO_2.5F₃ were not successful when applying the tetragonal structure model due to the presence of additional diffraction peaks. Indexing these peaks in the XRD data can be performed in comparison to the x = 0.9 dataset as depicted in Fig. 5. Using the (113) main signal of the tetragonal structure, for example, it becomes clear that the observed splitting of the signal is not explained by a monoclinic unit cell symmetry, which would result in two signals for (113) and (1̄13). The observation of a more complex splitting in at least three signals is the sign of a triclinic unit cell giving rise to (11̄3̄) and (113̄). This interpretation is further supported by the splitting of the (204) peak into (204̄) and (204), which is also shown in the figure. An analogous splitting into (204) and (024) would have to be linked to clearly different values for a and b, which would, for example, give rise to a clear splitting of the (200) signal. However, this peak is not split, which is why a triclinic distorted version of the P4₂/nnm unit cell with a ≈ b is most likely. Full profile fits with the LeBail method were used to extract the unit cell parameters from the XRD data (P1̄, a = b = 5.7079(1) Å, c = 13.0818(2) Å, α = 89.65(1)°, β = 89.59(1)°, γ = 89.92(3)°).

Fig. 5 Comparison of selected regions of the X-ray diffraction patterns of x = 0.9 and x = 1.0 obtained with Cu-Kα_1,2 radiation. Selected peaks are indexed highlighting the transition from tetragonal to triclinic.

Structure refinements of La₂CuO_2.5F₃ were carried out with this triclinic unit cell. The refinements were performed as joint refinements against one XRD and NPD dataset. The structure model in space group P1̄ was set up in the coordinate system of the tetragonal unit cell of the 3F-oxyfluorides and the Rietveld plots are shown in Fig. 6. The anionic positions were taken from the positions found in the CuO₄F₂ octahedra of the tetragonal structure model for x = 0.9. A good fit to the XRD data is obtained indicating a correct structure model with reliable cation positions (the structural parameters are given in Table S2 in the SI). In the NPD data some unmatched intensity is clearly visible. A reasonable fit to the NPD data can be obtained when freely refining the anion positions but this results in unreasonably strongly distorted CuO₄F₂ octahedra most probably due to too many refinable variables. For this reason, we do not report the results of this refinement here. Additionally, by closely inspecting the XRD data, some mismatched reflections of low intensity can be found at 0.81, 1.72, 1.76 Å⁻¹ and 1.95 Å⁻¹ of which the later three are marked with asterisks in the inset of Fig. 6a. These peaks cannot be described in the current structure model or be assigned to any previously known impurity phase. The same peaks are also present in different La₂CuO_2.5F₃ sample batches and might hint to an even more complex supercell arising from anion ordering. The full structure description of this compound will be addressed in a future article with the aid of high resolution/high intensity synchrotron XRD data, as well as high intensity, lower Q NPD data.

Fig. 6 Rietveld plots obtained for the refinement of La₂CuO_2.5F₃ in space group P1̄ based on X-ray (a) and neutron (b) powder diffraction data. The position of LaOF signals is highlighted by violet diamonds. Asterisks highlight the additional reflections which were not indexable.

¹⁹F MAS NMR experiments were performed to further study the fluorine environment for four selected compounds (x = 0.0, 0.3, 0.7, and 1.0), and the ¹⁹F MAS NMR spectra are shown in Fig. 7. Spinning sidebands are successfully suppressed by the choice of a low field magnet (1.4 T) in combination with the high sample spinning frequency (∼20 kHz). The repetition delay of 1 s was sufficiently long so that the fluorine atoms in the diamagnetic environments can also be measured quantitatively. The broadening caused by hyperfine interactions, dipolar coupling to Cu and F atoms nevertheless limits the resolution of the spectra. Based on the broadness of the peaks, all are found to originate from paramagnetic fluorine environments. One exception is the signal at −89 ppm that is assigned to unreacted PVDF (the spectrum of pure PVDF is additionally included) and its amount has to be below the detection limit of the IR-spectrometer used. The observation of two signals for x = 0.0, 0.3 and 1.0 points to the presence of two different fluorine environments, which is in good agreement with the expectations stemming from the structural refinements, as two different F-sites (namely: apical octahedral (8m) and interstitial sites (4c)) were derived for La₂NiO_2.5F₃ from BVS calculations.¹⁶ The presence of two signals in similar intensity ratios between the compounds is therefore seen as confirmation of the proposed structural similarity of all oxyfluorides including the x = 1.0 compound. The observation of only one signal for x = 0.7 is counter intuitive as the same two anion sites are expected due to the strong structural similarity. A possible explanation is a change in the crystal symmetry, which is in agreement with the nonlinear change in the unit cell parameters (vide supra) and can't be ruled out based on the here discussed powder diffraction data.

Fig. 7 ¹⁹F MAS NMR spectra of La₂Ni_1−xCu_xO₃F₂ with x = 0.0, x = 0.3, x = 0.7, x = 1.0, and PVDF. The samples were spinning at ≈ 20 kHz in a 1.4 T magnetic field. The dashed lines serve as guides to the eyes. The asterisks and hashes denote isotropic peaks.

The observed chemical shifts differ for all compositions and for the most intense peak values in the range of δ = −20 ppm to 15 ppm are obtained without a clear correlation with the Cu content. The chemical shifts of the second signal are in the range of 89 ppm to 98 ppm. Differing chemical shifts were also observed for the ¹⁹F MAS NMR spectra obtained for the 2F oxyfluorides La₂Ni_1−xCu_xO₃F₂ and a possible explanation was seen in the changing La–F interactions due to changes in the F-environment caused by the Jahn–Teller elongation.¹⁹ The same explanation holds for the spectra shown here. In addition, changes in the hyperfine shift, which is related to the distance and orientation of the paramagnetic centers Ni/Cu to F will also add to the different chemical shift values. It is interesting to note that the isotropic peaks of the 3F oxyfluorides presented here are clearly shifted to higher resonance frequencies compared to the 2F oxyfluorides for which δ-values in the region of −44 ppm were found.¹⁹ This highlights the clear presence of different fluorine environments for both substitution series which is expected hence their different crystal structures. For example, in La₂NiO_2.5F₃ decreased F_ap–F_ap distances of neighboring octahedra compared to the 2F-oxyfluroides (d(F_ap–F_ap) ≈ 2.88 Å vs. 2.77 Å La₂NiO₃F₂vs. La₂NiO_2.5F₃ respectively) are accompanied by additional F_ap–F_int interactions over a distance of 2.43 Å.

Optical band gap determination and photocatalytic hydrogen evolution

In our previous study, it was shown that La₂NiO_2.5F₃ has an optical band gap of 3.4 eV ¹⁶ which is unusually large compared to the oxide and the 2F oxyfluoride La₂NiO₃F₂ (both appear black to the eye; La₂NiO₄E_g = 1.3 eV ¹⁷). The influence of Ni/Cu substitution on the optical properties of the oxyfluorides was studied by UV-Vis spectroscopy as well as DFT calculations. Diffuse reflectance UV-Vis spectra were obtained in order to quantify the changes in the optical band gap energy. These spectra consist of several features for the compounds in the range of x = 0–0.6, which indicate different transitions in the band gap region (compare Fig. 8a). Tauc plots of the data converted by the Kubelka–Munk function with exponents of ½ (indirect transition) and 2 (direct transition) were used to determine the band gap energy of these transitions.³⁸ These plots are shown in the supplement (Fig. S3) for x = 0.4 as example. The use of the exponent n = ½ allows the determination of E_g for all observed features e.g. band transitions, while the feature at higher energies is strongly favoured if a direct transition (n = 2) is assumed. This is why the n = ½ exponent seems more reasonable which is also in concordance with the initial reported data of La₂NiO_2.5F₃. The band gap energies (shown in Fig. 8b) were nevertheless determined for all compounds applying both exponents. For n = 2, only one transition could be reasonably analysed in each case and, compared to n = ½, larger values of E_g are determined. Based on the visual impression of the samples with x = 0.0 to 0.6 which are shown in Fig. 9. It can be seen that the change in E_g is accompanied by a change in the colour of the oxyfluoride powder ranging from greyish white (x = 0) via ochre to reddish brown for x = 0.5 (E_g: 1.93/1.31 eV with n = 1/2). Starting from x ≥ 0.6, the samples look dark brown to black. Assuming of a direct band transition, a value of E_g = 1.83 eV is obtained for the x = 0.6 compound corresponding to an absorption edge of 678 nm. When an exponent of ½ is used, a more realistic band gap energy of an indirect band transition in the near IR range of 1.25 eV (992 nm) is obtained. This again supports the assumption of indirect band transitions analogous to x = 0.0.

Fig. 8 (a) Normalized diffuse reflectance UV-Vis spectra for the oxyfluorides La₂Ni_1−xCu_xO_2.5F₃ and (b) band gap energies E_g obtained from Kubelka–Munk transformation assuming an indirect (n = 1/2) as well as a direct (n = 2) band gap transition. The size of the symbols corresponds to the estimated error of the determination method. The step like singularity at 860 nm in (a) is an experimental artifact resulting from switching detectors.

Fig. 9 Optical photographs of La₂Ni_1−xCu_xO_2.5F₃ samples prepared from drying isopropanol/powder slurries on microscopy cover glasses highlighting the influence of Ni/Cu substitution of the colour of the samples.

The band structure of La₂NiO_2.5F₃ which was previously obtained applying the modified Becke–Johnson functional, revealed strong nickel d-orbital contributions to the valence band near the band gap.¹⁶ It was therefore assumed that the insertion of one additional electron in the d-orbitals by substitution of Ni²⁺ (d⁸ electron configuration) by Cu²⁺ (d⁹ electron configuration) together with a strong Jahn–Teller distortion has to result in an increase in the d-orbital contribution to the valence band. To verify this assumption, DFT calculations were performed to investigate the impact of Ni/Cu-substitution on the band structure and to explain the experimentally observed change in E_g. For these calculations, a disordered Ni/Cu occupation was approximated by simulation of ordered occupations of Ni and Cu in a (110) × (1̄10) supercell with 68-atoms (8 disordered sites). The amount of different tested configuration was highest for the low copper containing compounds with x = 0.25 where the strongest changes in E_g were observed in the experimental data. For each configuration the band gap was calculated using the PBEsol functional with an on-site Coulomb correction for Ni-d states. The resulting band gaps for ordered configurations were used to get an estimation of the gap of the disordered system via applying the generalized quasichemical approximation (GQCA).^33,34 The results for different Cu concentrations are plotted in Fig. 10a. It is observed that the general decrease of E_g with increasing x is well presented in the calculated results, although as expected the PBE (+U) method underestimated the gap of these strong correlated systems. Moreover, for most (∼90%) of the configurations, the calculated band gap is indirect, and the smooth change of the calculated gap above x = 0.75 is consistent with the lowest E_g data obtained from the experimental measurements (Fig. 8b). Note that we used an AFM spin configuration for x = 1, which correctly captured the non-zero gap nature of La₂CuO_2.5F₃, while FM or non-magnetic spin-configurations result in metallic (zero gap) ground states. The band structure and DOS of x = 0.25 (depicted in Fig. 10b) was obtained from unfolding the supercell band structure to the unit cell k-path with the easyunfold³⁵ Python-package. Here the appearance of Cu-d and O-p band contribution around 1–1.5 eV within the wide bandgap of La₂NiO_2.5F₃ is confirmed and the decrease of E_g with x therefore indeed results from the additional Cu-d electrons.

Fig. 10 (a) Band gap energy vs. copper content x data obtained from DFT calculations applying the modified Becke–Johnson functional. Data were obtained for several sample structures per x order to simulate a disordered Ni/Cu distribution. (b) Band structure and DOS of the x = 0.25 configuration obtained with the easyunfold Python package.

The experimental E_g values in the region of the visible light makes the oxyfluorides of potential interest for the application as solar photocatalysts. First hydrogen evolution experiments were therefore performed for selected compounds with x = 0.0, 0.2, and 0.4 in order to test for photocatalytic hydrogen evolution reaction activity. Experiments in the presence of ethanol as sacrificial agent were performed in two steps: 1st with the pure oxyfluoride suspension and 2nd after the addition of Pt (0.5 wt%) as co-catalyst which was photo-deposited in situ. The obtained H₂ evolution rates are plotted in Fig. 11 for x = 0.2, and 0.4 as well as the 10% ethanol solution without any catalyst. For the x = 0.0 sample no hydrogen was detected and the data is therefore excluded from the figure to maintain clarity. For the x = 0.2 and 0.4 compounds hydrogen evolution in the range of 0.3–0.45 μmol h⁻¹ is obtained without the addition of Pt as cocatalyst. After the photodeposition of Pt, the hydrogen evolution increases only for the x = 0.2 sample to a hydrogen evolution rate of ∼0.7 μmol h⁻¹. The photodeposition of Pt seems to be unsuccessful as it does not increase the hydrogen evolution rate for x = 0.4 and has little effect for x = 0.2. When comparing the amount of evolved hydrogen of the oxyfluorides with the data of the 10% ethanol solution without catalyst an increased hydrogen evolution rate is only found for the x = 0.2 sample. Here the maximum value of ∼0.7 μmol h⁻¹ after the photodeposition of Pt is slightly higher than the rate found for the pure sacrificial agent itself (∼0.6 μmol l⁻¹ peak activity). This highlights the necessity of pure sacrificial agent test reaction to check for the reliability of the obtained data.

Fig. 11 Photocatalytic hydrogen evolution rate for oxyfluorides with x = 0.2 and 0.4 with 10% ethanol as sacrificial agent before and after addition of 0.5 wt% Pt as co-catalyst. The data of the ethanol solution without catalyst is additionally shown.

The overall low amount of evolved hydrogen might be attributed to high defect concentrations in the oxyfluorides, which is indicated by reflection broadening that is observed during topochemical fluorination of the oxides with sharper diffraction peaks. Such defects might act as traps/recombination centres for the photo generated electron/hole pairs. Our aim therefore is to prepare samples with less defects for further investigations.

Conclusion

In this article, we report on the highly fluorinated Ruddlesden–Popper oxyfluorides La₂Ni_1−xCu_xO_2.5F₃, which were synthesized through a topochemical fluorination reaction of the corresponding oxides obtained from the citrate route with PVDF as fluorination agent. Single-phase samples were obtained in increments of x = 0.1, and a comprehensive structural study was conducted using X-ray and neutron powder diffraction data together with ¹⁹F MAS NMR experiments which allowed to investigate the F environment. Our results show that the compounds up to x = 0.9 adopt the anion-ordered tetragonal (P4₂/nnm) structure of La₂NiO_2.5F₃, characterized by layer-wise alternating tilts of the Ni/CuO₄F₂ octahedra. The stability of the tetragonal structure throughout the (almost) entire substitution range is surprising, given that closely related 2F-oxyfluorides exhibit a strong symmetry reduction upon Ni²⁺ substitution by the Jahn–Teller active Cu²⁺. In contrast, the pure cuprate endmember La₂CuO_2.5F₃ exhibits a clear symmetry reduction to triclinic P1̄. We also characterized the optical properties of the oxyfluorides using UV-Vis measurements in combination with DFT calculations, which revealed that Ni/Cu substitution results in a Cu-d contribution to the valence band. This enables E_g to be modified across a broad range of 3.5 eV to 1.3 eV, covering the entire visible light range. Preliminary tests of the suitability of the oxyfluorides for photocatalytic hydrogen evolution showed a considerable hydrogen evolution rate for x = 0.2. In a follow-up article, we will discuss the impact of Ni/Cu substitution on the thermal stability, formation and decomposition reaction pathways, and magnetic properties of the title oxyfluorides La₂Ni_1−xCu_xO_2.5F₃.

Author contributions

J. J.: conceptualization, data curation, formal analysis, investigation, methodology, visualization, and writing of the original draft. C. R.: collection of neutron powder diffraction data and review. K. X. and J. S. a. d. G.: collection and analysis of ¹⁹F MAS NMR data and review. H.-C. W. and M. A. L. M.: performing and interpretation of theoretical calculations and review. A. H. and R. M.: performing and interpretation of photocatalytic test reactions and review. S. G. E.: resources, review, discussion and editing.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data supporting the findings of this study are available within the article and its SI. The crystallographic data of all oxyfluoride has been deposited at the CCDC under deposition numbers 2471688–2471698, and can be obtained from https://www.ccdc.cam.ac.uk/. The neutron diffraction data are available under the following link https://doi.ill.fr/10.5291/ILL-DATA.5-23-769. Any other information should be requested to the corresponding authors.

Rietveld plots for the refinement of XRD data of La₂Ni_1−xCu_xO_2.5F₃; Exemplarly Kubleka–Munk plots for the bandgap determination of the x = 0.4 compound; crystallographic data of La₂Ni_1−xCu_xO_2.5F₃ obtained from Rietveld refinements. See DOI: https://doi.org/10.1039/d5ta05605e.

Acknowledgements

The neutron diffraction experiments was performed on the D2B diffractometer run by the Institut Laue-Langevin (ILL). Beamtime was granted under the proposal number 5-23-769 (Doi: https://doi.ill.fr/10.5291/ILL-DATA.5-23-769).²² We thank the ILL for the allocated beamtime.

References

1. M. A. Nowroozi, K. Wissel, J. Rohrer, A. R. Munnangi and O. Clemens, LaSrMnO₄: Reversible Electrochemical Intercalation of Fluoride Ions in the Context of Fluoride Ion Batteries, Chem. Mater., 2017, 29(8), 3441–3453,  DOI:10.1021/acs.chemmater.6b05075.
2. M. A. Nowroozi, S. Ivlev, J. Rohrer and O. Clemens, La₂CoO₄: A New Intercalation Based Cathode Material for Fluoride Ion Batteries with Improved Cycling Stability, J. Mater. Chem. A, 2018, 6(11), 4658–4669,  10.1039/c7ta09427b.
3. Y. Huang, Y. Wei, S. Cheng, L. Fan, Y. Li, J. Lin and J. Wu, Photocatalytic Property of Nitrogen-Doped Layered Perovskite K₂La₂Ti₃O₁₀, Sol. Energy Mater. Sol. Cells, 2010, 94(5), 761–766,  DOI:10.1016/j.solmat.2009.12.020.
4. B. Zhao, R. Wang, W. Yao, X. Yang and B. Zhou, The Effect of Copper Substitution on La₂Ni_1-xCu_xO₄ Catalysts Activity for Simultaneous Removal of NO_x and Diesel Soot, Catal. Lett., 2009, 132(1–2), 41–49,  DOI:10.1007/s10562-009-0041-9.
5. X. Q. Liu, J. W. Wu, X. X. Shi, H. J. Zhao, H. Y. Zhou, R. H. Qiu, W. Q. Zhang and X. M. Chen, Hybrid improper ferroelectricity in Ruddlesden-Popper Ca₃(Ti,Mn₎₂O₇ ceramics, Appl. Phys. Lett., 2015, 106(20), 202903,  DOI:10.1063/1.4921624.
6. Q. Zhang, A. Solanki, K. Parida, D. Giovanni, M. Li, T. L. C. Jansen, M. S. Pshenichnikov and T. C. Sum, Tunable Ferroelectricity in Ruddlesden-Popper Halide Perovskites, ACS Appl. Mater. Interfaces, 2019, 11(14), 13523–13532,  DOI:10.1021/acsami.8b21579.
7. D. Li, K. Lee, B. Y. Wang, M. Osada, S. Crossley, H. R. Lee, Y. Cui, Y. Hikita and H. Y. Hwang, Superconductivity in an Infinite-Layer Nickelate, Nature, 2019, 572(7771), 624–627,  DOI:10.1038/s41586-019-1496-5.
8. L. Si, W. Xiao, J. Kaufmann, J. M. Tomczak, Y. Lu, Z. Zhong and K. Held, Topotactic Hydrogen in Nickelate Superconductors and Akin Infinite-Layer Oxides ABO₂, Phys. Rev. Lett., 2020, 124(16), 1–8,  DOI:10.1103/PhysRevLett.124.166402.
9. H. Sun, M. Huo, X. Hu, J. Li, Z. Liu, Y. Han, L. Tang, Z. Mao, P. Yang, B. Wang, J. Cheng, D. X. Yao, G. M. Zhang and M. Wang, Signatures of Superconductivity near 80 K in a Nickelate under High Pressure, Nature, 2023, 621(7979), 493–498,  DOI:10.1038/s41586-023-06408-7.
10. Y. Zhu, D. Peng, E. Zhang, B. Pan, X. Chen, L. Chen, H. Ren, F. Liu, Y. Hao, N. Li, Z. Xing, F. Lan, J. Han, J. Wang, D. Jia, H. Wo, Y. Gu, Y. Gu, L. Ji, W. Wang, H. Gou, Y. Shen, T. Ying, X. Chen, W. Yang, H. Cao, C. Zheng, Q. Zeng, J. G. Guo and J. Zhao, Superconductivity in pressurized trilayer La₄Ni₃O_10−δ single crystals, Nature, 2024, 631(8021), 531–536,  DOI:10.1038/s41586-024-07553-3.
11. R. De Barros, M. Ceretti, W. Schmidt, V. Y. Pomjakushin and W. Paulus, Growth and Oxygen Stoichiometry Control of High-Quality La₂CoO_4+δ Single Crystals (δ = 0.25), Cryst. Growth Des., 2022, 22(9), 5542–5551,  DOI:10.1021/acs.cgd.2c00631.
12. S. Ebbinghaus, M. Fröba and A. Reller, The “Realstruktur” of the System La_2-xSr_xCu_1-yRu_yO_4-δ Studied by Rietveld and Extended X-ray Absorption Fine Structure Spectroscopy, J. Phys. Chem. B, 1997, 101(48), 9909–9915,  DOI:10.1021/jp971551n.
13. P. R. Slater, Poly(vinylidene fluoride) as a reagent for the synthesis of K₂NiF₄-related inorganic oxide fluorides, J. Fluorine Chem., 2002, 117(1), 43–45,  DOI:10.1016/S0022-1139(02)00166-5.
14. J. Jacobs, A. Bivour, V. Sikolenko, H. Kohlmann, T. C. Hansen, J. R. Hester, K. Xu, J. Schmedt auf der Günne and S. G. Ebbinghaus, Unveiling the Fluorination Pathway of Ruddlesden-Popper Oxyfluorides: A Comprehensive in Situ X-Ray and Neutron Diffraction Study, J. Am. Chem. Soc., 2025, 147(11), 9739–9751,  DOI:10.1021/jacs.4c18187.
15. A. Bivour, J. Jacobs, F. Daumann, G. Hörner, B. Weber, C. Ritter and S. G. Ebbinghaus, Structure and Magnetic Properties of the n = 3 Ruddlesden-Popper Oxyfluoride La_0.5Sr_3.5Fe₃O_7.5F_2.6, Inorg. Chem., 2024, 63(43), 20427–20437,  DOI:10.1021/acs.inorgchem.4c02835.
16. J. Jacobs, M. A. L. Marques, H. C. Wang, E. Dieterich and S. G. Ebbinghaus, Structure, Magnetism, and Thermal Stability of La₂NiO_2.5F₃: A Ruddlesden-Popper Oxyfluoride Crystallizing in Space Group P4₂/nnm, Inorg. Chem., 2021, 60(17), 13646–13657,  DOI:10.1021/acs.inorgchem.1c01957.
17. S. Boumaza, R. Brahimi, L. Boudjellal, A. Belhadi and M. Trari, Photoelectrochemical study of La₂NiO₄ synthesized using citrate sol gel method—application for hydrogen photo-production, J. Solid State Electrochem., 2020, 24(2), 329–337,  DOI:10.1007/s10008-019-04470-8.
18. S. Perween, K. Wissel, Z. Dallos, M. Weiss, Y. Ikeda, S. Vasala, S. Strobel, P. Schützendübe, P. M. Jeschenko, U. Kolb, R. Marschall, B. Grabowski and P. Glatzel, Topochemical Fluorination of LaBaInO₄ to LaBaInO₃F₂, Their Optical Characterization, and Photocatalytic Activities for Hydrogen Evolution, Inorg. Chem., 2023, 62(40), 16329–16342,  DOI:10.1021/acs.inorgchem.3c01682.
19. J. Jacobs, H. C. Wang, M. A. L. Marques, K. Xu, J. Schmedt auf der Günne and S. G. Ebbinghaus, Ruddlesden-Popper Oxyfluorides La₂Ni_1-xCu_xO₃F₂ (0 ≤ x ≤ 1): Impact of the Ni/Cu Ratio on the Structure, Inorg. Chem., 2024, 63(13), 6075–6081,  DOI:10.1021/acs.inorgchem.4c00399.
20. J. Jacobs, J. R. Hester and S. G. Ebbinghaus, Cuprate Oxyfluorides La₂Cu_0.8Ni_0.2O₃F₂ and La₂CuO₃F₂ with “Channel-like” Anion Ordering, Inorg. Chem., 2022, 61(43), 17202–17211,  DOI:10.1021/acs.inorgchem.2c02776.
21. J. Jacobs, H. C. Wang, M. A. L. Marques and S. G. Ebbinghaus, Ruddlesden-Popper Oxyfluorides La₂Ni_1-xCu_xO₃F₂ (0 ≤ x ≤ 1): Impact of the Ni/Cu Ratio on the Thermal Stability and Magnetic Properties, Inorg. Chem., 2024, 63(24), 11317–11324,  DOI:10.1021/acs.inorgchem.4c01330.
22. J. Jacobs, A. Bivour, S. G. Ebbinghaus and C. Ritter, Solving the Structure of New n = 1 and n = 3 Ruddlesden–Popper Oxyfluorides with Unique Structural Distortions by High Resolution Neutron Powder Diffraction, 2021,  DOI:10.5291/ILL-DATA.5-23-769.
23. B. H. Toby and R. B. Von Dreele, GSAS-II: The Genesis of a Modern Open-Source All Purpose Crystallography Software Package, J. Appl. Crystallogr., 2013, 46(2), 544–549,  DOI:10.1107/S0021889813003531.
24. K. Xu, F. Aldudak, O. Pecher, M. Braun, A. Neuberger, H. Foysi and J. Schmedt auf der Günne, High resolution solid-state NMR on the desktop, Solid State Nucl. Magn. Reson., 2023, 126, 101884,  DOI:10.1016/j.ssnmr.2023.101884.
25. R. K. Harris, E. D. Becker, S. M. C. De Menezes, R. Goodfellow and P. Granger, NMR Nomenclature: Nuclear Spin Properties and Conventions for Chemical Shifts - IUPAC Recommendations 2001, Solid State Nucl. Magn. Reson., 2002, 22(4), 458–483,  DOI:10.1006/snmr.2002.0063.
26. D. Jardón-Álvarez and J. Schmedt auf der Günne, Reduction of the Temperature Gradients in Laser Assisted High Temperature MAS NMR, Solid State Nucl. Magn. Reson., 2018, 94, 26–30,  DOI:10.1016/j.ssnmr.2018.08.002.
27. D. Joubert, From Ultrasoft Pseudopotentials to the Projector Augmented-Wave Method, Phys. Rev. B:Condens. Matter Mater. Phys., 1999, 59(3), 1758–1775,  DOI:10.1103/PhysRevB.59.1758.
28. G. Kresse and J. Furthmüller, Efficiency of Ab Initio Total Energy Calculations for Metals and Semiconductors Using a Plane-Wave Basis Set, Comput. Mater. Sci., 1996, 6(1), 15–50,  DOI:10.1016/0927-0256(96)00008-0.
29. G. Kresse and J. Furthmüller, Efficient Iterative Schemes for Ab Initio Total-Energy Calculations Using a Plane-Wave Basis Set, Phys. Rev. B, 1996, 54(16), 11169–11186,  DOI:10.1103/PhysRevB.54.11169.
30. J. P. Perdew, K. Burke and M. Ernzerhof, Generalized Gradient Approximation Made Simple, Phys. Rev. Lett., 1996, 77(18), 3865–3868,  DOI:10.1103/PhysRevLett.77.3865.
31. S. Dudarev and G. Botton, Electron-Energy-Loss Spectra and the Structural Stability of Nickel Oxide: An LSDA+U Study, Phys. Rev. B, 1998, 57(3), 1505–1509,  DOI:10.1103/PhysRevB.57.1505.
32. A. van de Walle, Multicomponent multisublattice Alloys, nonconfigurational entropy and other additions to the Alloy Theoretic Automated Toolkit, CALPHAD:Comput. Coupling Phase Diagrams Thermochem., 2009, 33(2), 266–278,  DOI:10.1016/j.calphad.2008.12.005.
33. A. Sher, M. Van Schilfgaarde, A. B. Chen and W. Chen, Quasichemical Approximation in Binary Alloys, Phys. Rev. B, 1987, 36(8), 4279–4295,  DOI:10.1103/PhysRevB.36.4279.
34. L. Teles, J. Furthmüller, L. Scolfaro, J. Leite and F. Bechstedt, First-Principles Calculations of the Thermodynamic and Structural Properties of Strained and Alloys, Phys. Rev. B, 2000, 62(4), 2475–2485,  DOI:10.1103/PhysRevB.62.2475.
35. B. Zhu, S. R. Kavanagh and D. Scanlon, Easyunfold: A Python Package for Unfolding Electronic Band Structures, J. Open Source Softw., 2024, 9(93), 5974,  DOI:10.21105/joss.05974.
36. O. Clemens and P. R. Slater, Topochemical modifications of mixed metal oxide compounds by low-temperature fluorination routes, Rev. Inorg. Chem., 2013, 33(2–3), 105–117,  DOI:10.1515/revic-2013-0002.
37. A. Aguadero, J. A. Alonso, M. J. Escudero and L. Daza, Evaluation of the La₂Ni_1-xCu_xO_4+δ System as SOFC Cathode Material with 8YSZ and LSGM as Electrolytes, Solid State Ionics, 2008, 179(11–12), 393–400,  DOI:10.1016/j.ssi.2008.01.099.
38. P. Makuła, M. Pacia and W. Macyk, How To Correctly Determine the Band Gap Energy of Modified Semiconductor Photocatalysts Based on UV-Vis Spectra, J. Phys. Chem. Lett., 2018, 6, 6814–6817,  DOI:10.1021/acs.jpclett.8b02892.

---