AccountTuning seleniumiodine contacts: from secondary soft–soft interactions to covalent bonds
Introduction
For a long time, uncharged covalent selenium iodides have been regarded as non-existent [1], [2]. Intermediate SeI compounds had been postulated as early as 1953, when the impact of iodine on the rate of conversion of amorphous into crystalline selenium had been studied in context with early industrial applications of selenium [3]. It was also recognized that organic selenides may act as donors towards molecular iodine and certain iodine compounds leading to ‘charge–transfer complexes’ which are generally more stable than those of the corresponding sulfur ligands [4], [5], [6]. A major breakthrough was the isolation of stable solid SeI3+AsF6− [7]. ‘Stabilization by crystal lattice energy’ [2], [7] has subsequently allowed the preparation of a large number of stable cationic SeI (and various related cationic SI) compounds [2]. At present, fully characterized uncharged covalent SeI compounds with 2c–2e bonds are still extremely rare, but basic conditions for the existence of SeI compounds are now well-understood [2], [8], [9]. The observation that the in vivo deiodination of the prohormone thyroxine (T4) to the biologically active 3,5,3′-triiodothyronine (T3) implies a selenoenzyme, and the postulate of some kind of SeI interaction in the enzyme–substrate complex [10], [11] has stimulated further interest to improve understanding of the interactions of selenium nucleophiles with molecular iodine [12] or iodine compounds.
In this overview, the state of knowledge about different kinds of selenium–iodine interactions will be reviewed by using a series of selected compounds with different SeI distances representing a ‘continuum of SeI interactions’ ranging from about 390 pm (slightly less than the sum of Se and I van der Waals radii) to 244 pm (shorter than a single bond). Between these extremes, all kind of ‘transitions’ are known (for the selected examples see Table 1, Scheme 1). Many of the SeI distances that are longer than those of covalent two center–two electron (2c–2e) bonds (about 252 pm) can be regarded as implying ‘valence shell expansion’ at Se or at I leading to an increase in the number of electron pairs and of the coordination number at Se (typically 10-Se-3) and/or at I (10-I-2) [13].
Although ‘sharp lines’ between different extents of SeI overlap cannot be drawn, we will use the following idealized cases as a rough concept for the classification of Se–I interactions:
- 1.
‘Secondary bonds’, i.e. weak interactions that involve SeI distances which are significantly longer than those of typical 3c–4e SeI systems (about 265–300 pm) but shorter than the sum of the Se and I van der Waals distances (≤about 390 pm). These Se⋯I secondary bonds play a very important role in determining the crystal packing of the various products. Whenever the case, we will point out the importance of these interactions also during the discussion of compounds belonging to the classes II and III.
- 2.
Typical ‘three center–four electron’ bonds, i.e. triiodide-like bonding, that involves a ‘hypervalent’ central atom with, in a symmetrical case, two bonds of bond order 0.5 each. Typical SeI bond lengths in compounds involving the SeI bond order 0.5 are close to 276 pm (about 265–295 pm).
- 3.
Predominantly ‘covalent’ bonds, i.e. bonds that can be described satisfactorily by two center–two electron overlap leading to SeI bond orders close to 1; typical SeI single bond distances are close to 250 pm.
Section snippets
Weak Se⋯I interactions [secondary bonds, d(SeI)>300 pm]
The ‘nonexistence’ of solid binary SeI compounds has been associated with the very similar electronegativity of these two elements, i.e. the lack of ion-covalence resonance energy in covalent SeI bonds. Weak ‘soft–soft interactions’ between selenium and iodine, related to those in both elements (between selenium chains in grey selenium [344 pm] and between iodine molecules within layers of solid iodine: [348 pm]), do not depend on differences in electronegativity. Yet polar contributions like
The diphenyldiselenide–iodine adduct (8) [21]
Addition of one equivalent of iodine to a mixture of diphenyldiselenide and di-p-tolyldiselenide leads to a very broad 77Se signal covering the complete range from 468 to 489 ppm, whereas the products from the single diselenides 1:1 with iodine would give resonance at 468 ppm (Ph2Se2I2) and 488 ppm (p-Tol2Se2I2) [14]. It appears that in the mixture of diselenides, iodine will be coordinated to symmetric and to unsymmetric diselenides, but iodine will also catalyze diselenide scrambling. Ph2Se2
Uncharged compounds with covalent SeI single bonds
The lack of polarity, i.e. of ion-covalence resonance energy, in covalent SeI bonds may explain, that there is no intrinsic tendency to make covalent SeI bonds from SeSe and II bonds. This argument provides, however, not an intrinsic reason why in solution SeSe bonds and II bonds should avoid reacting with each other in a kind of statistic dismutation providing certain amounts of SeI bonded species in equilibria. If a particular SeSe bond would be higher in energy than an average SeSe
Summary
Uncharged covalent SeI compounds with SeI bond orders close to one will have to rely on stabilization methods that prevent their dismutation into diselenides and iodine or into diselenideiodine adducts. Useful concepts will be steric destabilization of the diselenides, (like trans-Tsi2Se2 [72]), or bowl-shaped environments (like that of stable RSeOH and RSI compounds [78]). Stabilization by coordination is an alternative which implies a loss of SeI bond order due to the presence of
References (80)
- et al.
J. Biol. Chem.
(1991) - et al.
Tetrahedron: Asymmetry
(1999) - et al.
Inorg. Chim. Acta
(1997) - et al.
Coord. Chem. Rev.
(1999) - et al.
Polyhedron
(1997) - et al.
Coord. Chem. Rev.
(1999) - (a) T. Klapötke, J. Passmore, Acc. Chem. Res. 22 (1989) 234. (b) J. Passmore, in: R. Steudel (Eds.), Studies in...
Angew. Chem.
(1953)- (a) G.Y. Chao, J.D. McCullough, Acta Crystallogr. 14 (1961) 940. (b) H. Hope, J.D. McCullough, Acta Crystallogr. 15...