Organoarsonate Functionalization of Heteropolyoxotungstates.

Functionalization of the {P8W48} polyoxotungstate (POT) archetype with aromatic organoarsonates results in the first homometallic {P8W48} derivatives, with the general formula [(RAsVO)4PV8WVI48O184]32- [R = C6H5 (1) or p-(H2N)C6H4 (2)]. Short As-O bonds here induce unusual bending of the otherwise rigid {P8W48} macrocycle, breaking its D4h symmetry. The obtained species also represent the first lacunary POTs functionalized with organoarsonates and can potentially act as polyoxometalate precursors themselves. We elaborate solution stability in different aqueous media using 1H and 31P NMR spectroscopy and possible pathways for subsequent transformations in aqueous solutions of the functionalized polyanions. Recrystallization of the K+/Li+/dimethylammonium salt of 2 from 4 M LiCl solution yielded a further functionalized POT, [(H3NC6H4AsO)3P8W48O184Hx{WO2(H2O)2}0.4](30.2-x)- (3), revealing dissociation of the organoarsonate fragments in slightly acidic aqueous solutions followed by their rearrangement within the inner POT cavity.

Introduction

The combination of structural versatility, redox robustness, and the ability to act as multidentate ligands toward oxophilic heterometals renders polyoxotungstates (POTs), a large family of discrete polynuclear oxo complexes of WVI and WV ions, highly attractive for application in many areas ranging from catalysis, electrochemistry, photochromism, and magnetism to medicine.[1] Further expansion of these clusters’ structural diversity can be achieved by their functionalization with organic moieties, e.g., organosilyl, organophosphonate and -phosphoryl, organotin, and other groups.[2]

Until now, the reactivity of organoarsenic-based reagents toward POTs has received relatively little attention. While a series of organoarsinates [R2AsVMoVI4O15H]2− (R = CH3, C2H5, C6H5)[3] and organoarsonate-functionalized polyoxomolybdates with formulas [(RAsVO3)4MoVI4O10]4−,[4h,4q] [(RAsVO3)2MoVI5O15]4−,[4e] [(MeAsVO3)MoVI6O18(H2O)6]2−,[3b] [(RAsVO3)2MoVI6O18]4−,[4a,4d,4e,4h,4l] [(RAsVO3)2MoVI6O19H](6−,[4b,4f,4g] [(PhAsVO3)MoVI7O22]4−,[4d] and [(RAsVO3)4MoVI12O34]4−,[4c,4i−4m,4o] and several MoV-based and mixed-valent MoV/VI species[4n,4p,4q] [R = CH3, n-C3H7, C6H5, C6H4(CH3), p-C6H4(NH2), p-C6H4(OH), p-C6H4(COOH), p-C6H4(CN), o-C6H4(NO2), C6H3-4-OH-3-NO2, C6H3-4-OH-3-N(H)C(O)CH3, etc.] are known, organoarsonate-containing POTs are restricted to the isopolyoxotungstate derivatives [(RAsVO3)2W6O18]6− and [(RAsVO3)W7O24H]7−.[4k,5] These species were obtained by the direct condensation of tungstate ions with organoarsonates in aqueous media (pH 5−8).

In order to prepare novel hybrid polyanions offering the potential for subsequent integration of magnetic metal ions, we explored the reactivity of organoarsonates with multilacunary POTs. The seminal, highly stable macrocyclic POT [H7P8W48O184]33− ({P8W48})[6] was chosen for these experiments because it is regarded as highly stable and comparably inert toward substitution of its phosphate heterogroups and provides adequate coordination sites between its four {P2W12O48} building blocks. In {P8W48}, the latter are linked by oxo bridges into a wheel-shaped structure with a large inner cavity (approximately 1 nm in diameter), able to accommodate various transition and rare-earth metals as well as their oxo/hydroxo clusters, e.g., in [Cu20X(OH)24(H2O)12P8W48O184]25− (X = Cl, Br, I), [Cu20(N3)6(OH)18P8W48O184]24−, and [K8⊂{P8W48O184}{MoVIO2}4{MoV4O10(H2O)3}2]24−.[7] Moreover, it can act as a secondary building block in numerous fully inorganic coordination polymers, as exemplified by the 3D framework compounds K18Li6[Mn8(H2O)48P8W48O184]·108H2O with huge rigid voids[8] or the recently reported Li9K7W1Co10[H2P8W48O186]·nH2O, which can undergo several crystal-to-crystal transformations, resulting in structures of different connectivities (from 0D up to 3D).[9] An additional advantage of {P8W48} as the precursor for further functionalization is its solubility and stability in aqueous solutions in the unusually broad pH range from 1 to 8.[6]

Very recently, we have reported the first examples of {P8W48} complexes with the main-group metals [K4.5⊂(ClSnII)8P8W48O184]17.5− and [(HOSnII)6P8W48O184]34−.[10] At the same time, to our knowledge, no {P8W48} derivatives with nonmetal elements are known. We here report two [(RAsVO)4PV8WVI48O184]32− polyanions with R = C6H5 (1) and p-(H2N)C6H4 (2) that represent not only the first homometallic {P8W48} derivatives but also the first lacunary POTs functionalized with organoarsonates possessing the potential to act as precursors for organoarsonate-functionalized heterometal polyoxometalate (POM) complexes.

Experimental Section

General Methods and Materials

The starting reagents were used as purchased without further purification. K28Li5[H7P8W48O184]·92H2O (KL-{P) was synthesized according to the published procedure,[6] and its identity and purity were confirmed by IR and 31P NMR spectroscopy. Elemental analysis results [inductively coupled plasma atomic emission spectroscopy (ICP-OES) and C, H, and N] were obtained from the Central Institute for Engineering, Electronics and Analytics (ZEA-3), Forschungszentrum Jülich GmbH (Jülich, Germany). Thermogravimetric/differential thermal (TGA/DTA) analysis measurements were carried out with a Mettler Toledo TGA/SDTA 851 in dry N2 (60 mL min−1) at a heating rate of 5 K min−1. Vibrational spectra were recorded on a Bruker Vertex 70 Fourier transform infrared (FT-IR) spectrometer coupled with a RAM II FT-Raman module (1064 nm Nd:YAG laser) on KBr disks for the FT-IR measurements and a solid material for the Diamond ATR FT-IR and Raman measurements. 1H and 31P NMR spectra were recorded at room temperature in 5 mm tubes using a Bruker Avance 600 MHz spectrometer equipped with a prodigy probe, operating at 600.15 MHz for 1H and 242.95 MHz for 31P, and with a Varian Inova 400 MHz spectrometer equipped with an Auto-X-PFG probe, using a resonance frequency of 161.834 MHz for 31P. Chemical shifts are reported with respect to Si(CH3)4 (1H) and 85% H3PO4 (31P); all chemical shifts downfield of the reference are reported as positive values. The solid-state 31P magic-angle-spinning (MAS) NMR measurements were performed at ambient temperature using a Bruker AVANCE III spectrometer with a 9.4 T magnetic field. The spectrometer was equipped with a Bruker triple resonance probe for rotors of 2.5 mm diameter. The ground sample powder was filled into a ZrO2 rotor for experiments with and without sample spinning. Rotation frequencies of 35 and 15 kHz were applied during MAS. Eight-fold cyclization of the pulse sequences with a cycle delay of 120 s was used. The signal shift is referred to H3PO4.

Synthetic Procedures

Synthesis of K8Li17[(CH3)2NH2]7[(C6H5AsO)4P8W48O184]·130H2O (KLD-1)

A solution of C6H5AsO3H2 (0.0505 g, 0.25 mmol) in 5 mL of 1 M LiCl with pH 2.2 adjusted by glacial acetic acid was added dropwise to a solution of KL-{P (0.37 g, 0.025 mmol) in 15 mL of 1 M LiCl under vigorous stirring. The obtained reaction mixture with pH 3.0 was then placed in an oven and heated at 60 °C for 4 days. After cooling to room temperature, aqueous 1 M [(CH3)2NH2]Cl (20 drops) was added to the resulting solution and left for evaporation (pH 2.97). Transparent colorless blocklike crystals of KLD-1 formed in 2 weeks. The pH of the solution after crystallization remained unchanged at 2.96. The crystals were collected by filtration and dried in air. Yield: 0.134 g (34 % based on {P8W48}). Elem anal. Calcd for C38H336As4K8Li17N7O318P8W48 (found), mass %: C, 2.89 (2.90); H, 2.15 (2.25); As, 1.90 (1.94); K, 1.98 (1.87); Li, 0.75 (0.78); N, 0.62 (0.66); P, 1.57 (1.55); W, 55.91 (53.9). IR (KBr pellets, cm−1): 3438 (s, br), 1633 (s), 1461 (w), 1438 (w), 1135 (m), 1083 (m), 1022 (m), 939 (s), 820 (s), 727 (s), 667 (s), 466 (s). Raman (solid sample, λe = 1064 nm, cm−1): 1003 (w), 969 (s), 886 (m), 637 (m), 388 (w), 214 (m). 1H NMR (D2O): 7.72 (d, 2H), 7.61 (m, 1H), 7.54 (m, 2H).

Synthesis of K10.5Li14[(CH3)2NH2]3.5[(H3NC6H4AsO)4P8W48O184]·92H2O (KLD-2)

A solution of p-(H2N)C6H4AsO3H2 (0.0542 g, 0.25 mmol) in 5 mL of 1 M LiCl with pH 2.2 adjusted by glacial acetic acid was added dropwise to a solution of KL-{P (0.37 g, 0.025 mmol) in 15 mL of 1 M LiCl with vigorous stirring. The obtained reaction mixture with pH 3.0 was then placed in an oven and heated at 60 °C for 4 days. After cooling to room temperature, 8 drops of aqueous 1 M [(CH3)2NH2]Cl was added to the yellowish reaction mixture, and the resulting solution was left for evaporation (pH 3.03). Light-yellow crystals of KLD-2 formed after 1 week. The pH of the solution after crystallization remained virtually unchanged at 2.98. The product was collected by filtration and dried in air. Yield: 0.154 g (41 % based on {P8W48}). Elem anal. Calcd for C31H240As4K10.5Li14N7.5O280P8W48 (found), mass %: C, 2.47 (2.46); H, 1.60 (1.59); As, 1.99 (1.95); K, 2.72 (2.68); Li, 0.64 (0.62); N, 0.70 (0.77); P, 1.64 (1.62); W, 58.52 (58.02). IR (KBr pellets, cm−1): 3422 (s, br), 1629 (s), 1496 (w), 1465 (w), 1413 (w), 1324 (w), 1135 (m), 1085 (m), 1023 (m), 935 (s), 833 (s), 728 (s), 671 (s), 464 (s). Raman (solid sample, λe = 1064 nm, cm−1): 966 (s), 877 (m), 802 (w), 708 (m), 633 (m), 382 (w), 313 (w), 204 (m). 1H NMR (D2O): 7.50 (d, 2H), 6.88 (d, 2H).

Synthesis of K7.2Li23−[(H3NC6H4AsO)3P8W48O184H{WO2(H2O)2}0.4]·nH2O (KL-3)

A sample of KLD-2 (0.040 g, 0.003 mmol) was dissolved in aqueous 4 M LiCl (22 mL) with heating at 70 °C for 40 min. After that, the solution (pH 2.7) was divided into four vials and left for evaporation at room temperature. Needlelike crystals of KL-3 started to form after 1 week. The crystals were collected by filtration and dried in air. Yield: 0.021 g. IR (KBr pellets, cm−1): 3440 (s, br), 1629 (s), 1497 (w), 1416 (w), 1326 (w), 1294 (w), 1135 (m), 1083 (m), 1022 (m), 937 (s), 816 (s), 677 (s), 464 (s). Raman (solid sample, λe = 1064 nm, cm−1): 969 (s), 890 (m), 801 (w), 695 (m), 633 (m), 370 (w), 320 (w), 207 (m).

Synthesis of KL-4

A sample of KLD-2 (0.040 g, 0.003 mmol) was dissolved in aqueous 4 M LiCl (20 mL), followed by the addition of FeCl3·6H2O (0.006 g, 0.021 mmol), while the reaction mixture was stirred at 70−100 °C for 1 h. The resulting solution with pH 2.2 was filtered while hot to remove a small amount of fine precipitate. The filtrate produced a yellow precipitate after cooling, which was collected the next day. Elem anal. (ICP-OES): Fe:As:P:W = 7.7:4:8:48. IR (KBr pellets, cm−1): 3434 (s, br), 1635 (s), 1496 (w), 1417 (w), 1322 (w), 1144 (m),1084 (m), 1020 (m), 935 (s), 916 (s), 796 (s), 723 (s), 470 (s).

Single-crystal diffraction data for KLD-1, KLD-2, and KL-3 were collected on a SuperNova (Agilent Technologies) diffractometer at 120 K with Mo Kα radiation (λ = 0.71073 Å) for KLD-1 and Cu Kα radiation (λ = 1.54184 Å) for KLD-2 and KL-3. The crystals were mounted in a Hampton cryoloop with Paratone-N oil to prevent water loss. Absorption corrections was applied numerically based on Gaussian integration over a multifaceted crystal model.[2] The SHELXTL software package[3] was used to solve and refine the structures. The structures were solved by direct methods and refined by a full-matrix least-squares method against |F|2 with anisotropic thermal parameters for all heavy atoms (As, K, P, and W) with application of ISOR instructions for the refinement of some heavily disordered K cations. The K+ counterions in the structure of KL-3 with the site occupancy factor less than 0.25 were refined in isotropic approximation. The relative site occupancy factors for the disordered positions of the K cations as well as O atoms of the crystal water molecules were first refined in an isotropic approximation with Uiso = 0.05, then fixed at the obtained values, and refined without thermal parameter restrictions. No Li+ positions and H atoms of the crystal water molecules have been located. H atoms of the phenyl rings of the PhAsO32− and p-arsanilate ligands and the amino and methyl groups of p-arsanilates and the located dimethylammonium cations, respectively, were placed in geometrically calculated positions. Because of severe disorder, only 4 DMA+ cations could be located in the structure of KLD-1 and no DMA+ cations in the structure of KLD-2, while 7 and 3.5 DMA+ counterions are present in these structures, respectively, based on elemental analysis. Because of this disorder, there are only 15 highly disordered water molecules in KLD-1 (with O site occupancy from 0.0625 to 0.5) out of 130, the number determined from elemental analysis and TGA for this compound. Similarly, the O positions for only 18.5 out of 92 water molecules were located in the structure of KLD-2. This is consistent with the large solvent-accessible volume remaining in the structures. For the overall consistency, the final formulas in the crystallographic data correspond to the compositions of the bulk materials determined by elemental analysis and TGA. The rather high values of Rint for KLD-1, KLD-2, and KL-3 (0.124, 0.134, and 0.135, respectively), combined with several violations of systematic absences, are consistent with the twinning issue (especially considering the needlelike shape of the crystals). Because we were unable to discern any pattern in the list of reflections for which F(obs) ≫ F(calc), we were not able to systematically remove composite reflections. Additional crystallographic data are summarized in Table .

Table 1

Crystallographic Data and Structure Refinement Details for KLD-1, KLD-2, and KL-3

sample	KLD-1	KLD-2	KL-3
radiation source	Mo Kα	Cu Kα	Cu Kα
empirical formula	C38H336As4K8Li17N7O318P8W48	C31H240As4K10.5Li14N7.5O280P8W48	C18H215.6As3K7.2Li17N3O283.6P8W48.4
fw, g mol−1	15784.16	15079.26	14783.50
cryst syst	tetragonal	triclinic	triclinic
space group	I4̅	P1̅	P1̅
a, Å	25.1943(1)	23.0426(7)	14.4150(6)
b, Å	25.1943(1)	24.3124(7)	21.7712(7)
c, Å	25.6924(3)	29.9784(8)	24.5919(6)
α, deg	90	80.834(2)	69.384(3)
β, deg	90	82.616(2)	78.910(3)
γ, deg	90	14949.4(7)	82.697(3)
volume, Å3	16308.3(2)	14949.4(7)	7073.8(4)
Z	2	2	1
Dcalcd, g cm−3	3.214	3.350	3.470
abs coeff, mm−1	17.519	36.373	38.142
F(000)	14328	13528	6602
cryst size, mm3	0.28 × 0.29 × 0.32	0.06 × 0.08 × 0.24	0.06 × 0.09 × 0.33
θ range for data collection	4.10−25.67	8.88−66.59	8.95−65.09
completeness to θmax, %	99.5	99.5	99.3
index ranges	−30 < h < +30, −30 < k < +30, −31 < l < +31	−27 < h < +27, −28 < k < +28, −35 < l < +35	−16 < h < +16, −24 < k < +25, −28 < l < +28
reflns collected	153179	262401	94589
indep reflns	15399	52558	23947
Rint	0.1239	0.1341	0.1351
obsd [I > 2σ(I)]	13962	32564	15078
abs corrn	numerical based on Gaussian integration over a multifaceted crystal model
Tmin/Tmax	0.0118/0.0844	0.0130/0.2352	0.0120/0.2044
no. of data/restraints/param	15399/32/469	52558/48/1716	23947/0/869
GOF on F2	1.042	1.035	1.031
R1, wR2 [I > 2σ(I)]	0.0487, 0.1280	0.0921, 0.2304	0.0842, 0.2142
R1, wR2 (all data)	0.0562, 0.1360	0.1457, 0.2771	0.1339, 0.2628
largest diff peak/hole, e Å−3	2.486/−1.409	4.387/−2.824	3.308/−2.972

Results and Discussion

Syntheses and Crystal Structures

Polyanions 1 and 2 have been prepared by heating of KL-{P with phenylarsonic (1) or p-arsanilic (2) acid in a 1 M LiCl solution (pH 3.0) at 60 °C. The pH of the reaction mixture was adjusted by the addition of glacial acetic acid to the initial organoarsonate solution. On the one hand, this slightly improves the solubility of the organoarsonic acids in a 1 M LiCl medium. On the other hand, it reduces pH fluctuations while mixing the [P8W48O184]40− and organoarsonate ligand solutions compared to directly mixing the reagents in 1 M aqueous LiCl followed by the addition of acetic acid to the final mixture. 1 and 2 crystallize as the hydrated mixed potassium/lithium/dimethylammonium (DMA+) salts K8Li17[(CH3)2NH2]7[(C6H5AsO)4P8W48O184]·130H2O (KLD-1) and K10.5Li14[(CH3)2NH2]3.5[(H3NC6H4AsO)4P8W48O184]·92H2O (KLD-2) in the tetragonal I4̅ and triclinic P1̅ space groups, respectively. The precise numbers of K+, Li+, and DMA+ countercations as well as crystal water molecules in KLD-1 and KLD-2 were established from elemental and thermogravimetric analyses. 1 and 2 can also be crystallized with Rb+ and Cs+ countercations instead of DMA+; however, 1 and 2 cannot be isolated directly from the reaction mixture without additional counterions.

Single-crystal X-ray diffraction reveals that the structures of 1 and 2 comprise four {RAsVO} units [R = C6H5 (1) and p-(H2N)C6H4 (2)] that are covalently bound to the two inner rims of the {P8W48} wheel through As−O−W bonds (Figures and S1). The AsV ions of the {RAsVO} groups in 1 and 2 occupy four of the eight available vacant sites of {P8W48}, and each coordinates to two O atoms belonging to two adjacent {P2W12} units [As−O = 1.683(12)−1.718(12) Å in 1 and 1.671(19)−1.73(2) Å in 2]. The tetragonal AsV coordination sphere is completed by a terminal oxo ligand [As−O = 1.661(15) Å in 1 and 1.63(2)−1.68(3) Å in 2] and a C atom of the phenyl (1) or p-aminophenyl (2) group [As−C = 1.891(9) Å in 1 and 1.883(16)−1.930(17) Å in 2]. The aromatic rings are directed outward from the inner cavity of the polyanion, minimizing potential steric hindrances (Figures and S1). The bond lengths and angles within the {P8W48} framework and organic moieties are in the usual ranges (see Table S2). Two AsV ions occupy opposite (alternating) coordination sites on one rim of the POT wheel, while the other two are located on the other rim, binding to opposite vacant sites, which are in orthogonal positions to the first two (Figure b, right), resulting in approximately C2 symmetry (if slight rotations of the phenyl and p-aminophenyl groups are not considered).

Figure 1

Structures of polyanion 2 in perpendicular views. WO6 octahedra of the front (lime green) and rear (light blue) inner rims of the {P8W48} wheel defining the coordination sites for the p-arsanilate groups (in a ball-and-stick representation; front-facing H2NPhAsO groups, dark-gray bonds; As, dark green; rear-facing groups, light-gray bonds; As, rose). Dashed lines indicate deviation from the planar regular structure of {P8W48}. Color code: WO6, gray; PO4, yellow polyhedra; C, black; N, blue; O, red spheres. Some H atoms are omitted for clarity. See Figure S1 for the structure of 1.

Relatively short As−O bonds compared to the M−O bonds in the complexes of {P8W48} with heterometals (1.9−2.3 Å) cause slight distortions within the {P8W48} skeleton. Thus, the macrocyclic wheel is somewhat tilted toward two {RAsVO} units (Figure b, right): the O···O distance between the two O ions coordinated to one AsV constitutes 2.65 Å for 1 and is in the range from 2.63 to 2.68 Å for 2, while the O···O distance between the corresponding O centers of the noncoordinated vacant sites is 3.39 Å for 1 and 3.21−3.36 Å for 2. This distortion probably limits the number of incorporated {RAsVO} groups to four (vs eight), even if larger L/POT ratios are used in the reaction mixture (e.g., 20:1 instead of 10:1), and is responsible for the geometry of the complexes because coordination of the {RAsVO} groups at the four specific positions allows minimization of the strain within the POT skeleton. Charge neutrality for KLD-2 requires the presence of four protons. The bond-valence-sum calculations[12] indicate that these protons are not likely associated with any of the O atoms of the polyanion (see Table S3 for details). At the same time, the relatively low pH used in the synthesis of 2 suggests protonation of the amino groups of the p-arsanilate ligands, similar to the situation observed for the p-arsanilate-functionalized polyoxomolybdates.[4k]

Solution Studies

The solution behaviors of 1 and 2 in aqueous media have been examined by 1H and 31P NMR spectroscopy. The 1H NMR spectrum of a KLD-1 solution in D2O (Figure S7) exhibits that expected for the phenyl ring combination of a doublet (7.72 ppm) and two triplets (7.61 and 7.54 ppm) in an approximate 2:1:2 integral ratio. The 1H NMR spectrum of a D2O solution of KLD-2 (Figure S8) shows two doublets (7.565 and 6.97 ppm) corresponding to the two pairs of symmetrically equivalent protons of the phenyl rings of the p-arsanilate ligands, while the signal of the NH2 group protons overlaps with the strong H2O signal. The signals in the spectra of 1 and 2 are significantly shifted compared to those of noncoordinated phenylarsonate [7.765 (d), 7.70 (t), and 7.60 (t) ppm] and p-arsanilate [7.565 (d) and 6.97 (d) ppm] anions (Figures S7 and S8), indicating the stability of 1 and 2 in an aqueous medium over the short time period of the 1H NMR measurement.

The solution stability over longer time periods (1 h to 1 day) has been studied using 31P NMR spectroscopy. On the basis of the crystal structures of KLD-1 and KLD-2, we expect only a singlet for both polyanions in case they are intact in solution. In addition, two overlapping signals at −6.9 and −7.1 ppm as well as a wider signal contribution downfield to these signals were observed in the solid-state 31P MAS NMR spectrum for KLD-1 (at a rotation frequency of 15 kHz; Figure S9). The appearance of several signals is most likely due to disorder of K+ cations bound in the inner cavity of the polyanion to phosphate O atoms that in the crystal lattice renders phosphates bound and not bound to K+ nonequivalent. Solution 31P NMR spectra were measured at room temperature in several media of various basicity and ionic strength, such as H2O (Figures S10 and S11), 1 M LiClaq (Figures S12 and S13), 4 M LiClaq (Figures S14 and S15), 2 M Li2SO4/H2SO4 buffer (pH 3.0; Figures S16 and S17), 0.5 M LiOAc (pH 6.76, Figures S18 and S19), and 2 M LiOAc (pH 7.6; Figures S20 and S21) solutions as well as 0.5 M Tris buffer with pH 7.6 (Figures S22 and S23). While the compounds were well-soluble in water, heating has to be applied to dissolve KLD-1 and KLD-2 in Li-based media and in Tris buffer. Already 1 h after redissolution, the spectra of 1 and 2 in H2O and all Li-based media exhibit two sets of closely spaced signals: one at around −6.5/−7 ppm, and the other one slightly shifted upfield, while the exact positions, number, and relative intensities of these signals are changed after 1 day. These results are evident of at least partial decomposition of the polyanions, which most likely occurs by dissociation of the organoarsonate ligands. The cleanest spectra in these media and, in turn, best stability against hydrolysis were observed in 4 M LiCl solutions after 1 h, which exhibit main signals at −7.1 ppm (for 1) and −6.7 ppm (for 2), along with small upfield peaks at −7.6 ppm (1) and −7.3 ppm (2) (Figures S14 and S15), which compare well with the chemical shifts in the solid state and differ from the chemical shift for the P atoms of the noncoordinated {P8W48} wheels in this medium (−6.9 ppm). However, after 1 day, the spectrum of the same solution shows several overlapping signals, similar to the results obtained in the other tested media.

At the same time, the spectra of KLD-1 and KLD-2 solutions in 0.5 M Tris buffer (pH 7.6) exhibit a singlet at −8.3 and −8.4 ppm, respectively, after 1 h, 4 h, and 1 day. The chemical shifts unambiguously differ from the signal at −8.0 ppm observed for non-functionalized {P8W48} in the same medium appearing along with additional small peaks, which strongly supports at least one-day stability of the organoarsonate {P8W48} derivatives 1 and 2 in this medium. Thus, the above results point to a significant potential of 1 and 2 to act as novel organically functionalized POM precursors in Tris aqueous solutions at pH 7−8.

Recrystallization and Complexation with Heterometal Experiments

For further insight into the hydrolysis processes in the LiCl media, we analyzed crystals isolated from the solution of KLD-2 in 4 M LiCl. Structural analysis revealed the composition K7.2Li23−[(H3NC6H4AsO)3P8W48O184H{WO2(H2O)2}0.4]·nH2O (KL-3). The three AsV centers in the polyanion 3 bind to O atoms of the vacant sites situated on opposite rims of the {P8W48} wheel (Figure ). The two inversion-symmetric AsV positions are fully occupied, while the remaining AsV ion (along with its p-aminophenyl and terminal oxo ligands) is disordered over the two remaining positions with equal occupancy. This means that either (a) only a single [(H3NC6H4AsO)3P8W48O184H{WO2(H2O)2}0.4](30.2− polyanion type makes up the crystals of KL-3 (with three monoprotonated p-arsanilate ligands bound to the {P8W48} macrocycle) or (b) there is an equimolar mixture of species with two and four organoarsonates, [(H3NC6H4AsO)2P8W48O184H{WO2(H2O)2}0.4](33.2− and [(H3NC6H4AsO)4P8W48O184H{WO2(H2O)2}0.4](27.2−, respectively. The additional four vacant sites of {P8W48} in 3 are surprisingly occupied by WVI centers (with a relative site occupancy factor of 0.1), which is related to minor {P8W48} decomposition during heating. This result shows that the {RAsVO} groups in 1 and 2 readily dissociate in slightly acidic aqueous solutions, followed by their recomplexation with {P8W48} POT, indicating complex equilibria in aqueous solutions of KLD-1 and KLD-2, which explains their very complex 31P NMR spectra. This behavior is also similar to that observed previously for SnII complexes of {P8W48}, [K4.5⊂(ClSn)8P8W48O184]17.5−.[10]

Figure 2

Structures of polyanion 3 in perpendicular views. The color code is as in Figure . The four additional underoccupied WVI centers (sof 0.1) are shown as brown spheres, and their terminal O and H2O groups could not be located due to low occupancy.

Interestingly, the coordination mode of AsV centers in 3 does not allow for a significant bending-type distortion of the POT wheel, as observed in 1 and 2 (Figure , right). On the other hand, this results in significant distortion of the arsonate groups, as reflected by the much wider As−O bond length range in KL-3 [As−OPOT = 1.55(2)−1.83(3) Å; As−Oterminal = 1.69(3) and 1.91(4) Å] compared to KLD-1 and KLD-2. Another consequence is “compression” of the macrocylce toward p-arsanilate groups. Thus, distances between O atoms bound by AsV centers constitute 2.74−2.75 Å, while the O···O separation of the remaining four vacant sites is 3.64−3.65 Å. The “compression”-type distortion of {P8W48} is not unprecedented and is commonly observed in its heterometal complexes, which do not possess C4 symmetry.[7c,14]

To test the ability of using polyanions 2 as precursors for magnetic organoarsonate-functionalized {P8W48} derivatives, we added various heterometals in solutions of KLD-2 in an aqueous 4 M LiCl medium. The microcrystalline product KL-4 isolated from the reactions with FeIII exhibit {Fe8(H3NC6H4AsO)4P8W48} stoichiometry based on elemental analysis results. Unfortunately, the small crystal dimensions thus far have precluded single-crystal XRD measurements; however, the IR spectrum for KL-4 (Figure S3) strongly supports coordination of the FeIII ions to the p-arsanilate-functionalized polyanion (see the Supporting Information for details). This implies that KLD-1 and KLD-2 can act as possible novel POM precursors for organofunctionalized heterometal {P8W48} derivatives.

Conclusions

In summary, we have successfully isolated the first examples of the main-group, nonmetal-element-decorated {P8W48} derivatives with the general formula [(RAsO)4P8W48O184]32− [R = C6H5 or p-(H2N)C6H4], also representing the first heteropolyoxotungstates functionalized with organoarsonates. In these species, four {RAsVO} moieties are connected to the POT wheel via As−O−W bonds, resulting in C2-symmetric complexes. Because of the relatively short As−O bond length compared to the M−O bonds of transition or rare-earth and SnII metal ions, the formation of [(RAsO)4P8W48O184]32− leads to thus-far-unprecedented distortions of the {P8W48} wheel, which is slightly flexed toward coordinated {RAsVO} groups in 1 and 2. Our solution studies revealed dissociation of the organoarsonate fragments from the title polyanions in a slightly acidic aqueous medium followed by their rearrangement within the inner POT cavity. Polyanions of the composition [{H3NC6H4AsO}3P8W48O184H{WO2(H2O)2}0.4](30.2− (3) have been crystallized as the hydrated mixed K+/Li+ salt as the product of these structural rearrangements. At the same time, the title polyanions appear to be stable in the time frame from 1 h to at least 1 day (depending on the medium) in slightly basic solvents (pH 7.6).

The presence of vacant sites that remain noncoordinated by AsV centers in the inner cavity of the {P8W48} macrocycle in 1−3 and their feasible steric accessibility open perspectives for subsequent functionalization of the organoarsonate−POM hybrids by magnetic metal centers, an avenue that we are currently investigating.