Polyoxoniobates and Polyoxotantalates as Ligands—Revisited

Abstract

This short review summarizes our contribution to the coordination chemistry of noble metals (organometallic fragments of Rh, Ir, Ru and hydroxo Pt(IV)) and polyoxocomplexes of niobium and tantalum.

1. Introduction

The group 5 (V, Nb, Ta) polyoxometalates (commonly defined as anionic polynuclear oxocomplexes) form two large but apparently distinct families, which, despite certain resemblance, show almost no overlap [1,2,3,4,5,6,7,8,9]. Vanadium(V) prefers aggregation into decaniobate [V₁₀O₂₈]⁶⁻ and rearrangement of this very stable structure (search in CBSD yields about 150 structurally characterized salts of this anion with various organic cations) can be realized only in the presence of PO₄³⁻ as template at low pH, producing a “bicapped” Keggin-type [PV₁₄O₄₂]⁹⁻ anion [10,11,12,13,14,15]. Vanadium (IV) (but not Nb(IV) or Ta(IV)) can aggregate around differently-shaped anionic templates (like Cl⁻, NCS⁻ etc.) producing a family of reduced polyoxovanadates built from tetragonal pyramids, with fascinating magnetic properties [16,17]. The chemistry of polyoxoniobates (PONb) and tantalates is much less studied and for a long time it was confined to preparations of alkali metal salts of hexaniobate and hexatantalate, [Nb₆O₁₉]⁸⁻ and [Ta₆O₁₉]⁸⁻. The beginning of the modern era in this chemistry can be dated back to the works of Spinner, Pope and Stucky in the late 1960s–1980s [18,19,20,21,22,23,24], who obtained evidence for condensation of [Nb₆O₁₉]⁸⁻ into larger entities, and discovered its ability to coordinate certain transition metals. This chemistry is quite different from polyvanadates and is dominated by Lindqvist hexametalates [M₆O₁₉]⁸⁻, in spite of the existence of the decaniobate, [Nb₁₀O₂₈]⁶⁻ [25,26] and recently reported decatantalate [Ta₁₀O₂₈]⁶⁻ [27]. By contrary, free [V₆O₁₉]⁸⁻ is unknown, and the reason for this is perhaps that, if existed, it would have an extraordinarily high charge density (defined as the anionic charge of the POM divided by the number of non-hydrogen atoms of the POM). Among the known POM, the highest charge density (0.32) is achieved precisely for [M₆O₁₉]⁸⁻ [9]. As V is smaller than Nb, [V₆O₁₉]⁸⁻ would have the highest charge density per volume ratio. It can be stabilized only by coordination of four positively charged {Cp*Rh}²⁺ units [28,29].

The high charge density means that [M₆O₁₉]⁸⁻ anions are expected to act as efficient ligands towards various metals. This is indeed so, and a few examples of coordination of [Nb₆O₁₉]⁸⁻ to Mn(IV) and Ni(IV) were reported, as early as in 1967–1969 [18,19,20,21,22,23,24], and Co(III), reported only in 2011 [30], giving complexes of the type [M(Nb₆O₁₉)₂]ⁿ⁻, where two hexaniobate units (each acting as tridentate ligand) sandwich an octahedrally coordinated central metal ion. There is also a strong association of hexaniobate with alkali metal cations, even in water, where the existence of neutral [A₈Nb₆O₁₉] (A = Rb, Cs) species was established by Small Angle X-ray Scattering (SAXS) [31]. Curiously enough, no tantalate analogues of [M(Nb₆O₁₉)₂]ⁿ⁻ could be obtained [32]. It is doubtless highly desirable to expand this still rather modest list of existing complexes. However, strongly basic conditions, necessary to prevent hydrolysis of hexaniobate (and precipitation of niobium oxide), severely limit the choice of suitable metal precursors (they would simply give insoluble hydroxides). This drawback can be obviated in two different ways. First, in order to limit the hydrolysis and ensuing condensation, a part of the coordinating positions of a transition metal can be occupied by a polydentate (or polyhaptic) ligand (or simply by a strongly coordinated unidentate ligand, such as CO). This can be achieved for Co(III), Cr(III), Cu(II) and Ni(II) in the presence of polyamines [33,34,35,36,37]. Particularly attractive for this purpose are the isolobal fragments d⁶-fac-{ML₃}, where ML₃ stands for {M(CO)₃}⁺ (M = Mn, Re), {(arene)Ru}²⁺, {Cp*Rh}²⁺, {Cp*Ir}²⁺. In fact, a decade ago Pope et al. used [M₆O₁₉]⁸⁻ (M = Nb, Ta) as robust inorganic tridentate ligands in reactions with [Mn(CO)₃(CH₃CN)₃]ClO₄ or [Re(CO)₅Br], which gave rise to trans-[(M'(CO)₃)₂M₆O₁₉]⁶⁻ (M’ = Mn, Re; M = Nb, Ta) [38]. Proust et al. reported the syntheses of a series of complexes combining {(arene)Ru}²⁺ and [Nb₆O₁₉]⁸⁻ with Ru/POM stoichiometries ranging from 1:1–4:1 [39]. Coordination of organometallic fragments, such as {(p-cymene)Ru}²⁺ or {Cp*Rh}²⁺, to POM adds extra possibilities for the synthesis of hybrid (i.e., combining chemically very different moieties within the same structure) complexes. This development reflects not only academic interest in such compounds, being a logical development of POM chemistry in general, but also is driven by the well-known catalytic properties of many POM complexes of transition metals, in particular, with noble metals [40].

The second way is to use a metal which forms stable well-defined and water soluble hydroxocomplexes, such as Pt(IV), for which both [Pt(OH)₄(H₂O)₂] and [Pt(OH)₆]²⁻ exist as individual species [41]. In this case, a particular attraction for coordinating Pt(IV) to [Nb₆O₁₉]⁸⁻ comes from the reports about use of PONb in photocatalysis. In this case, a photoactivator is necessary, which role is typically fulfilled by [PtCl₆]²⁻ [42]. From this point of view, direct coordination of Pt(IV) to PONb is very desirable. Pt(IV), being octahedral and sufficiently oxophilic, could either substitute a Nb atom in the hexaniobate structure, similar to [H₂Pt^IVV₉O₂₈]⁵⁻ which is simply a Pt(IV)-monosubstituted classical decavanadate [43,44], or be “grafted” as a capping atom. As we will see, the latter is the case.

In this short review, we focus on the recent advances in the field of group V (in particular, Nb and Ta) POM complexes, concentrating on the synthesis of Lindqvist-based POM with incorporated noble metal centers and their potential applications as electrocatalysts for water oxidation. The second part addresses related studies with heteropolyniobiate homologues coordinating noble metal fragments. The third part deals with the current use of analytical tools to unravel the integrity and reaction dynamics in the solution and selected examples are discussed.

2. Reactivity of [M₆O₁₉]⁸⁻ towards Noble Metal Complexes

In this section, studies of reactivity of the Lindqvist [M₆O₁₉]⁸⁻ (M = Ta and Nb) POM towards a noble metal complex are discussed, with an emphasis on finding suitable experimental conditions for noble metal grafting, the main structural features of the resulting solids, as well as their behavior in aqueous solution.

2.1. Reaction [M₆O₁₉]⁸⁻ with {Cp*Rh}²⁺ and {Cp*Ir}²⁺

Heating of [Cp*RhCl₂]₂ with K₇H[Nb₆O₁₉]·13H₂O or Cs₈[Ta₆O₁₉]·14H₂O in water at 80–90 °C in 2:1 molar ratio yields yellow-orange solutions, from which K₄[{Cp*Rh}₂Nb₆O₁₉]·20H₂O (1) and Cs₄[{Cp*Rh}₂Ta₆O₁₉]·18H₂O (2) can be readily isolated in high yields (80% for Nb and 70% for Ta) [45]. Crystal structures of both complexes contain hybrid bi-capped organometallic-POM anions trans-[{Cp*Rh}₂M₆O₁₉]⁴⁻ (Figure 1) with the typical Lindqvist-type oxide metal core (metal octahedron with one μ₆-, twelve μ₂- and six terminal oxygen ligands) decorated at an opposite pair of the triangular {M₃O₃} faces with two capping organometallic fragments {Cp*Rh}²⁺. In both cases, exclusively trans-isomers crystallize. No cis-isomers were observed in solution or in solid. Both 1 and 2 produce basic solutions in water; 3 mM solutions have the pH value of 8.8, which is much lower in comparison to that produced by alkali metal salts of parent [M₆O₁₉]⁸⁻ of comparable concentration (10.5–11.0). The ESI(-) mass spectra of aqueous solutions of 1 and 2 display prominent peaks assigned to the [{Cp*Rh}₂M₆O₁₉ + 2H]²⁻ dianions, accompanied by minor signals from the singly-charged [{Cp*Rh}₂M₆O₁₉ + 3H]⁻ anions. However, the protonation degree observed by ESI-MS does not necessarily mirrors that in the solution because of the influence of the ESI process itself on the protonation [46]. For example, the most efficient ionization mechanism observed by ESI-MS for the decaniobate [Nb₁₀O₂₈]⁶⁻ ion was the uptake of up to three protons, whereas solution studies indicated no protonation [26,47].

Figure 1. View of trans-[{Cp*Rh}₂M₆O₁₉]⁴⁻ in the crystal structure of K₄[(Cp*Rh)₂Nb₆O₁₉]·20H₂O (1) and Cs₄[(Cp*Rh)₂Ta₆O₁₉]·18H₂O (2) (ball and stick model).

Figure 1. View of trans-[{Cp*Rh}₂M₆O₁₉]⁴⁻ in the crystal structure of K₄[(Cp*Rh)₂Nb₆O₁₉]·20H₂O (1) and Cs₄[(Cp*Rh)₂Ta₆O₁₉]·18H₂O (2) (ball and stick model).

The reactions of [M₆O₁₉]⁸⁻ with [Cp*RhCl₂]₂ in 1:1 molar ratio under the same conditions gave solutions, from which only poorly diffracting yellow solids could be isolated. These solids can be extracted with methanol yielding yellow solutions. The ESI mass spectra of the methanolic extracts were consistent with the presence of POM/Rh complexes featuring 1:1 stoichiometry of general formula ([{Cp*Rh}M₆O₁₉]⁶⁻ (3, M = Nb; 4, M = Ta).

Grafting of {Cp*Ir}²⁺ fragments onto the [M₆O₁₉]⁸⁻ (M = Nb, Ta) proceeds essentially in the same manner with [Cp*IrCl₂]₂ as the source of the {Cp*Ir}²⁺ fragment. Both 2:1 [{Cp*Ir}₂Nb₆O₁₉]⁴⁻ (5, M = Nb; 6, M = Ta) and 1:1 [{Cp*Ir}Nb₆O₁₉]⁶⁻ (7, M = Nb; 8, M = Ta) complexes can be produced in high yields by varying the reagent ratio. In this case, we succeeded in isolation and X-ray characterization of potassium (M = Nb) and sodium (M = Nb, Ta) salts of 2:1 trans-[{Cp*Ir}₂M₆O₁₉]⁴⁻, and sodium salts of the single-capped [{Cp*Ir}M₆O₁₉]⁶⁻ complexes (both for Nb and Ta) (Figure 2).

Compound 5⁴⁻ was tested as electrocatalyst for water oxidation. The cyclic voltammogram of the trans-[{Cp*Ir}₂Nb₆O₁₉]⁴⁻ (5⁴⁻) in 1 M Na₂SO₄ exhibited a strong anodic current due to water oxidation already at 0.8 V versus Ag/AgCl. The current (296 μA) at 1.6 V in the presence of the complex was 5.1 times higher than that (57.7 μA) for the blank solution.

Figure 2. Ball and stick model of single-capped [{Cp*Ir}M₆O₁₉]⁶⁻ hybrid anions (M = Nb, Ta).

Figure 2. Ball and stick model of single-capped [{Cp*Ir}M₆O₁₉]⁶⁻ hybrid anions (M = Nb, Ta).

2.2. Complexation of [Ta₆O₁₉]⁸⁻ with {(C₆H₆)Ru}²⁺

Na₈[Ta₆O₁₉]∙24.5H₂O reacts with [(C₆H₆)RuCl₂]₂ (POM/Ru ratio 1:1) in water, yielding a yellow solution which contains single-capped [(C₆H₆)RuTa₆O₁₉]⁶⁻ (9⁶⁻) complex as the main product [48]. The progress of the reaction was monitored by ESI-MS as shown in Figure 3.

Figure 3. ESI(−) mass spectrum of aqueous solutions recorded at Uc = 10 V of the products of reaction between Na₈[Ta₆O₁₉]·24.5H₂O and [(C₆H₆)RuCl₂]₂ with POM / Ru ratio 1:1.

Figure 3. ESI(−) mass spectrum of aqueous solutions recorded at Uc = 10 V of the products of reaction between Na₈[Ta₆O₁₉]·24.5H₂O and [(C₆H₆)RuCl₂]₂ with POM / Ru ratio 1:1.

The ESI(−) mass spectrum of this reaction features the [(C₆H₆)RuTa₆O₁₉ + 5H]⁻ anion (m/z 1574.5) as the base peak, accompanied by a minor signal due to the doubly charged [(C₆H₆)RuTa₆O₁₉ + 4H]²⁻ species (m/z 786.7). A minor peak at m/z 1751.5 was attributed to the double-capped [{(C₆H₆)Ru}₂Ta₆O₁₉ + 3H]⁻ anion with two coordinated {(C₆H₆)Ru}²⁺ units. Evaporation of the reaction solution gave yellow plates of Na₁₀[{(C₆H₆)RuTa₆O₁₈}₂(μ-O)]·39.4H₂O (10) in 80% yield, with unexpected [{(C₆H₆)RuTa₆O₁₈}₂(μ-O)]¹⁰⁻ (10¹⁰⁻) anion (see Scheme 1) built from two fragments {(C₆H₆)RuTa₆O₁₈} joined via a common linear µ₂-O bridge. The formation of [{(C₆H₆)RuTa₆O₁₈}₂(μ-O)]¹⁰⁻ can be viewed as coordination-induced condensation of two hexatantalates according to the equilibrium:

[{(C₆H₆)Ru}Ta₆O₁₉]⁶⁻ + H₂O = [{(C₆H₆)Ru}Ta₆O₁₉H]⁵⁻ + OH⁻
2[{(C₆H₆)Ru}Ta₆O₁₉H]⁵⁻ = [{(C₆H₆)RuTa₆O₁₈}₂(μ-O)]¹⁰⁻ + H₂O

(1)

It is worth stressing that such condensation is unknown for free, non-coordinated hexatantalate. Perhaps, the reduced charge of the 1:1 complex [{(C₆H₆)Ru}Ta₆O₁₉]⁶⁻ facilitates interaction between two negatively charged species; coordination of electron-accepting fragment {(C₆H₆)Ru}²⁺ may affect electronic density on the oxygen atoms of the POM part as well.

Decreasing the POM/Ru ratio to 1:2 leads to the bis-capped complex trans-[{(C₆H₆)Ru}₂Ta₆O₁₉]⁴⁻ (11), isolated as Na₄[{(C₆H₆)Ru}₂Ta₆O₁₉]∙20H₂O (Na₄([11]∙20H₂O) in high yield. The ESI(−) mass spectrum displays the signal from the triply protonated [{(C₆H₆)Ru}₂Ta₆O₁₉ + 3H]⁻ anion (m/z 1751.5) as the main peak, accompanied by a minor signal from the doubly protonated [{(C₆H₆)Ru}₂Ta₆O₁₉ + 2H]²⁻dianion (m/z 875.3). A small amount of monocapped [{(C₆H₆)Ru}Ta₆O₁₉ + 5H]⁻ species coexists with the [{(C₆H₆)Ru}₂Ta₆O₁₉]⁴⁻ anion, thus suggesting a plausible dynamic equilibrium between mono- and bis-capped species. The relationship between the Ru/Ta species is summarized in Scheme 1:

Scheme 1. Formation of 1:1 and 1:2 complexes from [Ta₆O₁₉]⁸⁻ and [(C₆H₆)RuCl₂]₂.

Scheme 1. Formation of 1:1 and 1:2 complexes from [Ta₆O₁₉]⁸⁻ and [(C₆H₆)RuCl₂]₂.

2.3. Reaction [Nb₆O₁₉]⁸⁻ and Pt⁴⁺

Hydrothermal reaction (150 °C) of [Pt(OH)₄(H₂O)₂] with K₇H[Nb₆O₁₉]·13H₂O in 1:1 molar ratio gives a yellow solution, which shows only one signal in ¹⁹⁵Pt NMR spectrum at 3189 ppm. Evaporation of this solution yields yellow crystals of Cs₂K₁₀[Nb₆O₁₉{Pt(OH)₂}]₂·13H₂O (Cs₂K₁₀-12). Compound Cs₂K₁₀-12 was characterized by X-ray structural analysis (see Figure 4 left) [49].

Platinum niobate 12 is a unique dimeric complex, unprecedented in the PONb chemistry, and consists of two Pt(IV) centers coordinated each by three oxygen atoms from a {Nb₃O₃} face of a hexaniobate (d(Pt-O)_av 2.028(13), 2.010(14), and 2.124(14) Å), a terminal oxygen of the adjacent [Nb₆O₁₉]⁸⁻ anion (d(Pt-O) 1.998(11) Å), and bearing two terminal OH groups (d(Pt-O) = 2.015(16) and 1.994(14) Å). Freshly prepared aqueous solution of 12 gives ¹⁹⁵Pt NMR spectrum with only one signal at 3189 ppm, the same as is observed in the mother liquor. The negative ESI mass spectrum also strongly suggests that the integrity of the dimeric [Nb₆O₁₉{Pt(OH)₂}]₂¹²⁻ polyanion is preserved in solution.

Figure 4. Structure of [Nb₆O₁₉{Pt(OH)₂}]₂¹²⁻ (12¹²⁻) (left) and structure of [Pt(Nb₆O₁₉)₂]¹²⁻(13¹²⁻) (right), ball and stick models.

Figure 4. Structure of [Nb₆O₁₉{Pt(OH)₂}]₂¹²⁻ (12¹²⁻) (left) and structure of [Pt(Nb₆O₁₉)₂]¹²⁻(13¹²⁻) (right), ball and stick models.

Reaction of [Pt(OH)₄(H₂O)₂] and K₇H[Nb₆O₁₉]·13H₂O in the 1:2 molar ratio at 160–190 °C gives a yellow solution, which has two ¹⁹⁵Pt NMR signals at 3189 and 3422 ppm with relative intensities 0.75/0.25. Slow evaporation of this solution in a glass beaker produces light-yellow crystals of Na₂K₁₀[Pt(Nb₆O₁₉)₂]·18H₂O (Na₂K₁₀-13) in 20% yield (Na⁺ comes from glass leaching). The sandwich-type anion [Pt(Nb₆O₁₉)₂]¹²⁻ (13¹²⁻) is a new (fourth) member of the well-known [M(Nb₆O₁₉)₂]ⁿ⁺ (M = Mn, Ni, Co) family (Figure 4 right). All the Pt-O distances are equivalent, being 2.042(1) Å. The salt Na₂K₁₀-13 is isotypic with K₆Na₂[Co^IIIH₅(Nb₆O₁₉)₂]·26.5H₂O [30].

Further evaporation of the solution gives Cs₂K₁₀-12. A freshly prepared solution of Na₂K₁₀-13 demonstrates only one, slightly broadened, signal at 3422 ppm. Aged solutions give rise to increasingly intense signal at 3189 ppm, corresponding to the formation of [Nb₆O₁₉{Pt(OH)₂}]₂¹²⁻ (12). This observation implies a dynamics in solution, where [Pt(Nb₆O₁₉)₂]¹²⁻ (13) slowly converts into [Nb₆O₁₉{Pt(OH)₂}]₂¹²⁻ (12). Taking advantage of the versatility and high sensitivity of the ESI-MS technique for tracing solution speciation in the POM chemistry [47,50], we recorded ESI-MS spectrum of aqueous solution of Cs₂K₁₀-12 equilibrated for one week. ESI-MS results suggest coexistence of at least four different species in solution. Beside 12 and 13, the ESI-MS data suggest the presence of free Lindqvist [Nb₆O₁₉]⁸⁻ polyanion and a species that corresponds to the [Nb₆O₁₉Pt(OH)₃]⁷⁻ anion whose structure might putatively correspond to capping of [Nb₆O₁₉]⁸⁻ with a {Pt(OH)₃}⁺ group. Remarkably, the platinum niobate Cs₂K₁₀-12 exhibits electrocatalytic activity for water oxidation in 0.1 M Na₂SO₄. Oxidation process generates a significant rise in the current for a catalytic process mediated by the complex. The current in the presence of the platinum niobate is several times higher than for the blank solution. This electrochemical behavior is similar to the reported for the nonatungstate complex with a tetrarhodium-oxo core [51].

3. Reactivity of Heteropolyniobates towards Noble Metal Complexes

In 1979, S.K. Ray reported that reaction of K₇H[Nb₆O₁₉]·13H₂O with H₆TeO₆ gave a heteropolyniobate formulated as K₈H₂[TeNb₁₂O₃₈]·13H₂O [52]. This formula was deduced uniquely from elemental analysis. Such stoichiometry could imply that this compound would be analogous with the known heteropolyniobates [(Nb₆O₁₉)₂M]ⁿ⁻ (M = Co, Ni, Mn, Pt) with octahedral Te^VI sandwiched between two hexaniobates. We were curious if it were really so and repeated these experiments, only to find out that the purported “dodecaniobate” was in fact [(OH)TeNb₅O₁₈]⁶⁻, which can be regarded simply as a substituted hexaniobate with one {NbO}³⁺ vertex replaced with a {TeOH}⁵⁺ vertex. In fact, this complex had been already identified by W. Casey and co-workers, who prepared it from Nb₂O₅·xH₂O, Me₄NOH and H₆TeO₆ in methanolic solution [53]. Perhaps, the authors of [52] dealt with an approximately stoichiometric mixture of [(OH)TeNb₅O₁₈]⁶⁻ and [Nb₆O₁₉]⁸⁻, which led them to the wrong conclusion about the composition of the product. Surface charge density of telluroniobate is lower than in the hexaniobate, which means that their reactivity might be different. In order to check the significance of this difference, reactivity studies towards noble metal complexes have been extended in our group to the telluroniobate [(OH)TeNb₅O₁₉]⁶⁻.

3.1. Reaction [(OH)TeNb₅O₁₉]⁶⁻ with {Cp*Rh}²⁺ and {Cp*Ir}²⁺

We investigated reaction of [(OH)TeNb₅O₁₈]⁶⁻ with [Cp*MCl₂]₂ as sources of {Cp*M}²⁺ (M = Rh, Ir). Despite the difference in the surface charge density between [(OH)TeNb₅O₁₈]⁶⁻ and [Nb₆O₁₉]⁸⁻, formation of double-capped complexes takes place as easily as with hexaniobate. Crystal structure of Na₂[{Cp*Rh}₂Te(OH)Nb₅O₁₈]·24H₂O (14) shows coordination of two {Cp*Rh}²⁺ fragments, with the Te atom localized in the central M₄ plane of Lindqvist-type anion as illustrated in Figure 5. Two other metal positions are fully occupied by niobium. However, this disorder corresponds to four different orientations of a single isomer which is the only possible structure for the trans complex in the solid state.

Figure 5. Structure of [{Cp*Rh}₂Te(OH)Nb₅O₁₉]²⁻ anion (14), positions of Te atom with different occupancies are colored in green and purple (ball and stick model).

Figure 5. Structure of [{Cp*Rh}₂Te(OH)Nb₅O₁₉]²⁻ anion (14), positions of Te atom with different occupancies are colored in green and purple (ball and stick model).

3.2. Reaction [(OH)TeNb₅O₁₉]⁶⁻ with {(C₆H₆)Ru}²⁺

Reactivity studies on [(OH)TeNb₅O₁₈]⁶⁻ were also extended to [(C₆H₆)RuCl₂]₂ as source of the {(C₆H₆)Ru}²⁺. ESI-MS can be conveniently used to monitor the advance of the reaction. Grafting of one and two {(C₆H₆)Ru}²⁺ fragments can be immediately inferred from the characteristic m/z shifts to higher values, as illustrated in Figure 6. Moreover, the uptake of one or two Ru-containing fragments confers a new characteristic isotopic pattern due to multiple isotopes of the Ru at the natural abundance. Formation of single and double capped complexes [(OH)TeNb₅O₁₈{(C₆H₆)Ru}]⁴⁻ (15) and [(OH)TeNb₅O₁₈{(C₆H₆)Ru}₂]²⁻ (16) takes place as smoothly as with hexaniobate. Characteristic peaks at m/z 539.1 and 628.1 correspond to formulations [(OH)TeNb₅O₁₈{(C₆H₆)Ru} + 2H]²⁻ and [(OH)TeNb₅O₁₈{(C₆H₆)Ru}₂]²⁻, respectively. Other species based upon 15 and 16 associated Na⁺ or K⁺ adducts are also evident.

Figure 6. ESI mass spectrum of the reaction mixture between [(OH)Nb₅O₁₈]⁶⁻ and [(C₆H₆)RuCl₂]₂ with POM / Ru ratio 1:1.

Figure 6. ESI mass spectrum of the reaction mixture between [(OH)Nb₅O₁₈]⁶⁻ and [(C₆H₆)RuCl₂]₂ with POM / Ru ratio 1:1.

3.3. Reaction [SiNb₁₂O₄₀]¹⁶⁻ with {(C₆H₆)Ru}²⁺

Studies on the complexation with the siliconiobate [SiNb₁₂O₄₀]¹⁶⁻ [54] are particularly relevant. Because of its high negative charge, it is expected to be much more “sticky” than well-known Group 6 Keggin complexes [EM₁₂O₄₀]ⁿ⁻ (n = 2–6) (M = Mo, W) which are often regarded as non-coordinating [55]. It can be anticipated that quite a large number of metal fragments could be grafted to [SiNb₁₂O₄₀]¹⁶⁻. Following this rationale, we undertook preliminary exploration of the reaction of [SiNb₁₂O₄₀]¹⁶⁻ with [(C₆H₆)RuCl₂]₂, and ESI-MS results indicate that up to four {(C₆H₆)Ru}²⁺ fragments can be grafted to [SiNb₁₂O₄₀]¹⁶⁻, leading to the species formulated as [{(C₆H₆)Ru}₄SiNb₁₂O₄₀]⁸⁻ (17).

4. Analytical Tools to Study Solution Speciation in Nb and Ta POM Chemistry

One of the general features of the reactivity of POMs is their lability in aqueous solutions up to the point that such solutions can be regarded as virtual libraries of different building blocks, which equilibrate together and can form different structures depending on counter cations, pH and etc. The simplest case is protonation. The hexametalate anions [M₆O₁₉]⁸⁻ (M = Nb, Ta) are very basic. For example, hexaniobate in water exists as [H₃Nb₆O₁₉]⁵⁻ (рН = 8), [H₂Nb₆O₁₉]⁶⁻ (рН = 10), [HNb₆O₁₉]⁷⁻ (рН = 12), and [Nb₆O₁₉]⁸⁻ (рН = 14) [56]. Similarly, for hexatantalate three protonated species, [HTa₆O₁₉]⁷⁻, [H₂Ta₆O₁₉]⁶⁻, and [H₃Ta₆O₁₉]⁵⁻ were found [57,58,59]. Even in aqueous NBu₄OH hexatantalate exists as [H₂Ta₆O₁₉]⁶⁻ [60]. Small-angle X-ray scattering (SAXS) studies of [Ta₆O₁₉]⁸⁻ and [Nb₆O₁₉]⁸⁻ in water have been reported and important differences in the solution behavior have been observed due to the distinctive propensity to aggregation with counteractions for M = Ta and Nb [61]. Uv-vis spectroscopy complemented with DFT calculations have been also reported [62]. More complicated processes connected with rearrangement of POM backbone can take place [63,64,65]. It is obvious that proving the identity between species existing in solid state and solution is a very important part of this chemistry. However, the study of solution behavior of polyniobates and polytantalates is a challenging task. These species are colorless, diamagnetic, show no redox behavior and are, under most conditions, in a dynamic equilibrium. In this section, useful analytical tools that can provide complementary signatures for the Nb and Ta POM characterization in solution are discussed. In particular, the usefulness of ¹H NMR, ¹H NMR Diffusion Ordered Spectroscopy (DOSY), capillary electrophoresis and ESI-MS in concert is underlined. An illustrative example is given to demonstrate the complementarily of all techniques to determine stoichiometry and solution dynamics.

The possibilities of NMR are restricted to the use of ¹⁷O, which unfortunately has very low natural abundance and sensitivity. It was used, for example, to follow the conversion of several niobates in water [66,67]. However, coordination of organometallic fragments to niobates and tantalates provides new opportunities for use of NMR for solution studies. In particular, structural information can be gained from ¹H NMR signals of the organic ligands attached to the grafted heterometal, which would tell us about the overall symmetry of the compound. Moreover, it is possible to estimate diffusion coefficients from ¹H NMR DOSY of the organic part that can be subsequently used to calculate molecular weight of a particular species. This ¹H DOSY NMR technique is becoming a powerful tool for the determination of the size (hydrodynamic radius) and the molecular weight of inorganic clusters [68,69,70]. This technique has not been virtually used in the chemistry of Nb and Ta POM prior to our work.

Another useful tool for solution studies is capillary electrophoresis (CE). It is rarely used in the chemistry of POMs [71,72]. The possibilities opened up by the use of CE for POMs mixtures were studied in a few cases [73]. Although in these works CE revealed itself as a promising analytical tool, its potential for POM chemistry is far from being fully exploited. Combination of CE, HPLC and other methods (in particular ESI-MS and NMR) can provide sufficient information about quantitative and qualitative composition of solutions, numbers of POM species, and their stability and evolution with time.

Accurate determination of the formulae and structure of new compounds remains an important issue in POM chemistry. In this respect, use of soft ionization mass spectrometric techniques, and in particular, ESI techniques, is becoming widespread. Excellent reviews on the use of soft ionization MS techniques can be found elsewhere; for example, Cronin et al. illustrated the versatility of ESI and cryospray ionization techniques not only for characterization of final POM assemblies, but also to prove mechanisms of self-assembly of supramolecular POM-based architectures in solution [47,50]. Ohlin et al. reported an important work on this topic highlighting different approaches to obtain reliable information on the dynamic solution of POM on the basis of ESI-MS [47]. However, there are several potential limitations to using ESI-MS alone for the investigation of POM. This is the case when dynamic speciation involving hydroxy-bridged POM species, the pH value, ionic strength or the counteraction plays an important role. In addition, during the ESI process itself, the samples undergo dramatic changes in pH and concentration, which may perturb the speciation equilibrium. Moreover, due to the high charge state and molecular weight, the use of ESI-MS spectrometers equipped with high resolution analysers is important to facilitate unambiguous molecular compositions’ determination. In this part, we will discuss a case study, where we used ESI-MS, ¹H DOSY NMR and CE techniques, combined together, in order to elucidate the solution behavior of the {(C₆H₆)Ru}²⁺/[Ta₆O₁₉]⁸⁻ system to afford compounds 9 and 10 [48].

4.1. {(C₆H₆)Ru}²⁺/[Ta₆O₁₉]⁸⁻ System—Case Study

The ¹H DOSY NMR experiment performed on a sample of Na₁₀[10]·39.4H₂O in D₂O (see Figure 7) yielded the self-diffusion coefficients D = 315 ± 10 µm²·s⁻¹ and D = 325 ± 10 µm²·s⁻¹, respectively, for the ¹H NMR signals located at 5.85 ppm and 5.94 ppm. The Stokes-Einstein relation $D=\frac{kT}{6\pi \pi R_H}$ (which describes the diffusion coefficient D for spherical objects, where k is the Boltzmann constant, T the temperature taken equal to 300 K in this study, η of the solvent viscosity taken equal to that of water at 300 K (1.002 Pa.s) and R_H the hydrodynamic radius) gave hydrodynamic radii of 6.75–6.96 Å which are compatible with the radii expected for solvated monomeric species [{(C₆H₆)Ru}Ta₆O₁₉]⁶⁻ and [{(C₆H₆)Ru}₂Ta₆O₁₉]⁴⁻ but not with the dimeric [{(C₆H₆)RuTa₆O₁₈}₂(μ-O)]¹⁰⁻.

Figure 7. ¹H DOSY NMR spectra of Na₁₀-10 in D₂O (left) and ¹H DOSY NMR spectra of Na₄-9 in D₂O (right).

Figure 7. ¹H DOSY NMR spectra of Na₁₀-10 in D₂O (left) and ¹H DOSY NMR spectra of Na₄-9 in D₂O (right).

The ¹³C NMR spectrum of Na₄[{(C₆H₆)Ru}₂Ta₆O₁₉]∙20H₂O (Na₄-9) in D₂O exhibits two signals at 81.95 ppm and 81.82 ppm, with the complementary set of signals in the ¹H NMR spectrum at 5.97 (6%) ppm and 5.92 (93%) ppm and 5.85 (≤1%). This pattern corresponds to a mixture of cis (6%) and trans (93%) isomers of [{(C₆H₆)Ru}₂Ta₆O₁₉]⁴⁻, similarly to what was observed for [{(p-cymene)Ru}₂Nb₆O₁₉]⁴⁻⁶, while the minor peak at 5.85 ppm is to be attributed to the 1:1 complex [{(C₆H₆)Ru}Ta₆O₁₉]⁶⁻. The ¹H DOSY NMR experiment gives the self-diffusion coefficients D = 315 ± 10 µm² s⁻¹ and D = 325 ± 10 µm² s⁻¹ for the signals at 5.92 and 5.97 ppm, respectively, whereas the low intensity of the minor peak has prevented accurate determination of the D value. These values correspond to hydrodynamic radii in the 0.675–0.696 nm range. This is identical to that obtained for the solutions of Na₁₀[{(C₆H₆)RuTa₆O₁₈}₂(μ-O)]·39.4H₂O and agree well with the attribution of ¹H NMR peaks to cis-[{(C₆H₆)Ru}₂Ta₆O₁₉]⁴⁻ (5.97 ppm), trans-[{(C₆H₆)Ru}₂Ta₆O₁₉]⁴⁻ (5.92 ppm), and [{(C₆H₆)Ru}Ta₆O₁₉]⁶⁻ (5.85 ppm). These results necessarily imply two equilibria. The first corresponds to cis-trans isomerization process, strongly displaced towards the trans isomer. The latter corresponds to the de-coordination of a {(C₆H₆)Ru}²⁺ moiety. This hypothesis is supported by the observation of a small broad peak at 5.4 ppm, attributed to [(C₆H₆)₂Ru₂(OH)₃]⁺ (see Equation 2).

trans-[{(C₆H₆)Ru}₂Ta₆O₁₉]⁴⁻ = cis-[{(C₆H₆)Ru}₂Ta₆O₁₉]⁴⁻
2[{(C₆H₆)Ru}₂Ta₆O₁₉]⁴⁻ + 3OH⁻ + = 2[{(C₆H₆)Ru}Ta₆O₁₉]⁶⁻ + [(C₆H₆)₂Ru₂(μ-OH)₃]⁺

(2)

ESI(−) mass analysis of Na₁₀-10 in water reveals only the presence of the monomeric 9^6- species, thus indicating that the dimerization depicted in Eq. 1 is a reversible process ongoing from the solution to the solid state and vice versa. The electrophorograms recorded for aqueous solutions of Na₁₀-10 and Na₄-9 are presented in Figure 8. The solutions of Na₁₀-10, to which some amount of Na₄-9 has been added as reference, give two well-resolved peaks (Figure 7). The second peak corresponds to 9, which was confirmed by UV-spectroscopy and by the standard addition method. The first peak must therefore correspond to the monomeric [{(C₆H₆)Ru}Ta₆O₁₉]⁶⁻ species (or its dominant protonated form), because the signal from the dimer with its larger size and charge, should appear after the signal of 10, and with a longer retention time. Thus, combination of ESI-MS and CE gives coherent evidence that in aqueous solutions [{(C₆H₆)RuTa₆O₁₈}₂(μ-O)]¹°⁻ dissociates into the monomeric [{(C₆H₆)Ru}Ta₆O₁₉]⁶⁻.

Figure 8. The electrophorogram of complex Na₄-9 (up). The electrophorogram of complex Na₁₀-10 solution with addition of complex Na₄-9 as reference (down). Fused-silica capillary 50µm i.d. × 75 sm; sample buffer: 7.5·10⁻³М borate (pH 9.18); running buffer: 7,5·10⁻² M borate (pH 9.18).

Figure 8. The electrophorogram of complex Na₄-9 (up). The electrophorogram of complex Na₁₀-10 solution with addition of complex Na₄-9 as reference (down). Fused-silica capillary 50µm i.d. × 75 sm; sample buffer: 7.5·10⁻³М borate (pH 9.18); running buffer: 7,5·10⁻² M borate (pH 9.18).

5. Conclusions

The chemistry of niobium and tantalum polyoxoanions still belongs to a little researched area of inorganic chemistry. Many unusual complexes still remain hidden, and isolation of principally new [(Ta₆O₁₈)₂(µ-O)]¹⁴⁻ backbone is a tantalyzing hint at this potential richness. In this short review, we summarized our recent research focused on the coordination of noble metals to hexametalates. Grafting of {Cp*M}²⁺ (M = Rh, Ir) on the surface of Lindqvist-type hexametalates (including [(OH)TeNb₅O₁₉]⁶⁻) was studied with full characterization of the hybrid complexes in solid state and solution. Coordination of {(C₆H₆)Ru}²⁺ to [Ta₆O₁₉]⁸⁻ gives unprecedented [({(C₆H₆)Ru}Ta₆O₁₈)₂(µ-O)]¹⁰⁻. Complexation of Pt(IV) with [Nb₆O₁₉]⁸⁻ was studied with different techniques. New dimeric complex [(Nb₆O₁₉{Pt(OH)₂})₂]¹²⁻ is obtained when hexaniobate reacts with Pt(IV) in 1:1 molar ratio, while increasing the Pt/Nb₆ ratio to 2:1 gives sandwich-type complex [Pt(Nb₆O₁₉)₂]¹²⁻. We used different, both traditional and non-traditional, analytical tools for studying the solution behavior of individual reaction products and reaction mixtures, a challenging task for the systems where few analytical methods are readily available. The ¹H, ¹³C and DOSY NMR techniques combined with ESI-MS and CE highlight are adequate tools for characterization of complex solutions. An illustrative example dealing with the {(C₆H₆)Ru}²⁺/[Ta₆O₁₉]⁸⁻ system demonstrates complementarity of all these techniques for determination of stoichiometry and solution dynamics, an approach both promising and highly recommendable to trace unambiguously the identity of large inorganic systems [74].