Serendipitous formation and characterization of K₂[Pd(NO₃)₄]·2HNO₃

2.1 Synthesis

The reaction was performed in an evacuated, torch sealed borosilicate glass ampoule (l=300 mm, Ø=16 mm, thickness of the tube wall=1.8 mm). A mixture of 50 mg (0.47 mmol) Pd and 100 mg (0.33 mmol) K₂B₄O₇·4 H₂O in 1 mL fuming nitric acid was heated up to a temperature of 353 K and kept there for 24 h. Afterwards, the ampoule was cooled with a rate of 6 K h⁻¹ down to room temperature. Red platelets were obtained on the ampoule wall. The mother liquor was separated from the crystals by decantation, the acid containing side was cooled with liquid nitrogen and after cracking of the ampoule, several crystals were transferred directly into inert oil. The remaining bulk material was dried under reduced pressure in a Schlenk flask and transferred into the glovebox for further treatment. The handling was carried out strictly under inert conditions, as the compound appears to be sensitive to moisture.

2.2 X-ray structure determination

A STOE Stadi P diffractometer with curved Ge(111)-monochromatized MoKα₁ radiation (λ=0.7093 Å) was used to characterize a polycrystalline sample by X-ray powder diffraction. Diffraction intensities were measured with a Mythen2 1 K detector (Dectris, Switzerland) microstrip detector with 1280 strips.

The Rietveld refinements were accomplished with the software package Diffracplus-Topas^® 4.2 (Bruker AXS GmbH, Karlsruhe, Germany) [12] based on single-crystal data for β-K₂[Pd(NO₃)₄] [11] (collection code: 166766) and β-HBO₂ [13] (collection code: 34640) from the ICSD [14], [15], [16]. The reflection shapes were modelled using modified Thompson-Cox-Hastings pseudo-Voigt profiles [17], [18]. Instrumental contributions on reflection profiles were corrected from the refinement of a standard (LaB₆) [19]. The background was fitted with Chebychev polynomials up to the 6^th order [20].

Under a polarisation microscope, a suitable single crystal of K₂[Pd(NO₃)₄]·2HNO₃ with a diameter of 30 μm was fixed on the tip of a MicroMount™ (MiTeGen, LLC, Ithaca, NY, USA) and immediately placed into a stream of cold N₂ inside the diffractometer. The intensity data was collected with a Bruker D8 Quest diffractometer (Bruker, Karlsruhe, Germany) equipped with a Photon 100 detector system and an Incoatec microfocus source generator (multi-layered optic, monochromatized MoKα radiation, λ=0.71073 Å). The collection strategy, concerning the ω and φ scans, was optimized using the Apex 3 [21] program package. Thus, a complete data set up to high angles with high redundancies was received. For data processing and data reduction, the program Saint [22] was employed. Thereafter, multi-scan absorption corrections were applied with the program Sadabs [23].

In the course of the structure solution and parameter refinement with anisotropic displacement parameters for all non-hydrogen atoms with the Shelxs/l-17 [24], [25], [26], [27] software suite, the monoclinic space group P2₁/c was found to be correct. The hydrogen atom H7 was refined isotropically, while employing a bond restraint of 0.84(2) Å using the DFIX command.

Relevant details of the data collection and evaluation are listed in Table 1, the atomic coordinates and the isotropic displacement parameters in Table 2, and the anisotropic displacement parameters in Table 3. Interatomic distances and bond angles are shown in Table 4 and the hydrogen bond data in Table 5.

CCDC 1897204 contains the supplementary crystallographic data for this paper. These data can be obtained free of charge from The Cambridge Crystallographic Data Centre viawww.ccdc.cam.ac.uk/data_request/cif.

3.1 Crystal structure

According to the single-crystal diffraction data, K₂[Pd(NO₃)₄]·2HNO₃ crystallizes in the monoclinic space group P2₁/c (no. 14) with the cell parameters a=10.1715(4), b=8.9294(3), c=8.8055(3) Å, and β=98.13(1)°. The [Pd(NO₃)₄]²⁻ anion consists of a central palladium atom, which is coordinated monodentately in a square planar fashion by four NO₃ groups, as is depicted in Fig. 2 (left) in comparison to the similar [Pd(NO₃)₄]²⁻ anion found in β-K₂[Pd(NO₃)₄] (right). The mean Pd–O bond length in K₂[Pd(NO₃)₄]·2HNO₃ is 2.01 Å and, as the palladium atom is situated on a centre of inversion, only two crystallographically independent nitrate groups are distinguishable. The deviation of the O–Pd–O angle from the theoretical value for a perfect square planar coordination is 0.17°. The nitrate groups are nearly planar, deviating only 0.2(2)° for the nitrate group of N1 and 0.3(2)° for the nitrate group of N2 from planarity. The terminal N–O distances with 1.213(3) to 1.247(2) Å are noticeably shorter than the palladium coordinating O–N distances possessing values of 1.301(2) Å each. The O–N–O bond angles vary from 116.2(2) to 123.9(2)° and are in good accordance with the theoretical value of 120° for a trigonal planar structure.

Fig. 2:

The [Pd(NO₃)₄]²⁻ anion in K₂[Pd(NO₃)₄]·2HNO₃ (left) is built up from a central palladium atom, which is coordinated in a monodentate fashion by four NO₃⁻ anions. It is very similar to the anion in β-K₂[Pd(NO₃)₄] (right).

Each [Pd(NO₃)₄]²⁻ anion is surrounded by eight potassium cations with Pd···K distances of 4.053(2) to 5.800(1) Å, which is in good agreement with Pd···K distances of 4.3905(4) to 5.5658(4) Å found for β-K₂[Pd(NO₃)₄] [11]. The two crystallographically different potassium sites are not fully occupied. The K1 position reveals a site occupancy factor (s.o.f.) of 0.72 and the K2 site a s.o.f. of 0.28. The initial results were obtained by refining the potassium positions with a free variable. Determining the same positions freely results in similar values of K1 s.o.f. 0.70 and K2 s.o.f. 0.15. The potassium cation K1 is surrounded by ten oxygen atoms, with K–O distances ranging from 2.738(2) to 3.242(2) Å. The potassium cation K2 is surrounded by nine oxygen atoms, with K–O distances ranging from 2.493(8) to 3.354(9) Å. Both potassium sites are depicted in Fig. 3. The potassium cations are surrounded by NO₃⁻ anions and HNO₃ molecules. The distance between the two potassium sites is 0.794(9) Å.

Fig. 3:

Each potassium site is surrounded by NO₃⁻ anions. Left: The K1 site is surrounded by six NO₃⁻ anions and one HNO₃ molecule. Right: The K2 site is surrounded by four NO₃⁻ anions and two HNO₃ molecules.

The [Pd(NO₃)₄]²⁻ anion is surrounded by four HNO₃ molecules. The array allows for hydrogen bonds between hydrogen atoms H7 and the oxygen atoms O2 with an H7···O2 distance of 1.82(4) Å and an O7–H7···O2 bond angle of 172(4)° (see Fig. 4 and for crystallographic details Table 5). The hydrogen atom was fixed at a bond length of 0.84(2) Å to the oxygen atom O7 using the DFIX command. The HNO₃ molecule is nearly planar with a O9–O8–O7–H7 torsion angle of 4(3)° and exhibits two shorter and one longer N–O distances with 1.188(4), 1.194(3), and 1.358(3) Å, respectively. The hydrogen atom is bonded to the oxygen atom belonging to the elongated N–O bond. The O–N–O bond angles comprise two smaller and one larger angle with 115.3(3), 116.2(3), and 128.5(3)°.

Fig. 4:

The moderately strong hydrogen bond O7–H7···O2 is shown as a fragmented line [28].

The general packing motif for the complex anion is determined by two alternating orientations for the square-planar complex anion with a change of 74.73(4)°, as is shown in Fig. 5. K₂[Pd(NO₃)₄]·2HNO₃ can be regarded as a HNO₃ intercalation variant of β-K₂[Pd(NO₃)₄]. The [Pd(NO₃)₄]²⁻ anion in β-K₂[Pd(NO₃)₄] is surrounded by potassium cations and further complex anions. In K₂[Pd(NO₃)₄]·2HNO₃ on the other hand, the complex anion is surrounded by disordered potassium cations and HNO₃ molecules along the a axis. Both compounds are very similar with the main difference being the intercalation of HNO₃ groups.

Fig. 5:

Left: K₂[Pd(NO₃)₄]·2HNO₃ featuring layers along the bc plane and stacks along the a axis. It can be regarded as a HNO₃ intercalation variant of β-K₂[Pd(NO₃)₄] (right). The different potassium sites are shown in purple. The less occupied K2 site is emphasized in form of transparent spheres.

3.2 Powder diffraction data

In a glovebox, the bulk material was ground to a powder in an agate mortar and transferred to a glass capillary (Ø=0.3 mm). Figure 6 shows the experimentally obtained powder diffraction data (black; top) in comparison to the data simulated from the single-crystal diffraction data (red; bottom). It is obvious that the diffraction patterns do not match. Upon further investigation it became clear that the sample already decomposed, probably caused by vacuum treatment, which was necessary for the transfer into the glovebox. The powder pattern can be identified as a mixture of β-K₂[Pd(NO₃)₄] and β-HBO₂ [13], which formed as a side product of the reaction. The Rietveld refinement of the decomposition products is shown in Fig. 7. The experimental data are shown in black, the best fit profile in red, and the difference curve in blue.

Fig. 6:

Observed powder pattern (MoKα₁ radiation, λ=0.7093 Å) (top), compared to the theoretical powder pattern of K₂[Pd(NO₃)₄]·2HNO₃ calculated from the single-crystal data (bottom).

Fig. 7:

Rietveld refinement of the powder X-ray diffraction data. The observed powder pattern can be attributed to β-K₂[Pd(NO₃)₄] and β-HBO₂.