Scheme 1: Proposed synthetic routes for the preparation of 1 .
The first synthetic route worked very well for some “traditional” divalent Ln ions (Sm, Tm, Yb), but these methods have so far not proved transferable to deliver the analogous Dy(II) complex.[20] The second route (Scheme 1, Route B) relies upon salt metathesis methodologies [i.e.MX3 + n M′L M(L)3-n (X)n + n M′X; where M = Ln, M′ = alkali metal, X = halide, L = ligand, n = 1-3] with Dy(III) precursors to afford [Dy{N(SiiPr3)2}2(X)], and have thus far been unsuccessful, presumably due to facile deprotonation of alkyl silyl groups. However, whilst studying silylamide systems, the Mills Group used an anion abstraction strategy commonly employed elsewhere in the periodic table with the silylium reagent [(Et3Si)2H][B(C6F5)4].[20] In principle this strategy could be transferred to any Ln(L)2X species, where X is a halide and L a sterically demanding ligand, though very few examples like this exist.[21] Fortunately, a family of heteroleptic dysprosium metallocene complexes, [Dy(Cpttt)2(X)] (Cpttt = C5H2-1,2,4-tBu3; X = Br, I, BH4) reported by Nief and Jaroshick in 2007 were perfectly suited to the new synthetic strategy;[22] additionally, the [Dy(Cp*)2]+ (Cp* = C5Me5) cation had recently been postulated to be an excellent SMM.[23] Simply substituting DyCl3 instead of DyI3 in the Nief/Jaroschik protocol, and follow-up reaction with the silylium reagent delivered the cationic Dy(III) complex [Dy(Cpttt)2][B(C6F5)4] (2 , Figure 1), which contained the first example of an isolated dysprosocenium cation.[14a] The synthetic team delivered the new molecule to the theoreticians, who suggested that Cpttt did not fulfil the requirements of a “point-charge like ligand” and “would likely be a poor SMM because even the most axial SMMs lose magnetization at zero field”, but that they would have a look anyway. On this occasion, the hunch of the synthetic chemists was correct, and 2 was found to possess a very high U eff (1760 K) and retained magnetization at temperatures as high as 60 K (T H, the maximum temperature at which hysteresis loops remain open at zero applied field). This was a significant improvement compared to the best previous SMMs, such asT H = 14 K in the radical bridged terbium complex [K(18-crown-6)(THF)2][{Tb(N{SiMe3}2)2(THF)}2(μ-η22-N2)].[24]