MeSi(CH2SnRO)3 (R=Ph, Me3SiCH2): Building Blocks for Triangular‐Shaped Diorganotin Oxide Macrocycles

Abstract The syntheses of the novel silicon‐bridged tris(tetraorganotin) compounds MeSi(CH2SnPh2R)3 (2, R=Ph; 5, R=Me3SiCH2) and their halogen‐substituted derivatives MeSi(CH2SnPh(3−n)In)3 (3, n=1; 4, n=2) and MeSi(CH2SnI2R)3 (6, R=Me3SiCH2) are reported. The reaction of compound 4 with di‐t‐butyltin oxide (t‐Bu2SnO)3 gives the oktokaideka‐nuclear (18‐nuclear) molecular diorganotin oxide [MeSi(CH2SnPhO)3]6 (7) while the reaction of 6 with sodium hydroxide, NaOH, provides the trikonta‐nuclear (30‐nuclear) molecular diorganotin oxide [MeSi(CH2SnRO)3]10 (8, R=Me3SiCH2). Both 7 and 8 show belt‐like ladder‐type macrocyclic structures and are by far the biggest molecular diorganotin oxides reported to date. The compounds have been characterized by elemental analyses, electrospray mass spectrometry (ESI‐MS), NMR spectroscopy, 1H DOSY NMR spectroscopy (7), IR spectroscopy (7, 8), and single‐crystal X‐ray diffraction analysis (2, 7, 8).


Introduction
Metal chalcogenides with the empirical formula (M x E y ) n (M = transition or main group metal;E= O, S, Se; n = infinite) are classical ionic compounds found as ores in nature. Their reactivity and structures in the solid state have been extensively studied since the beginning of chemical research and chemistry textbooks describe essential topics of these compounds. [1a-e] Foralong time,m etal chalcogenides mainly served as raw materials for the metallurgy.However,over the years,c hemists learned also about some intriguing physical properties of these compounds,s uch as semiconductivity, [1a,f] nonlinear optoelectronic behaviour, [1g] and thermochromism, [1a-e] just to mention three out of many.Such properties are of utmost importance for high-tech applications.A cademic curiosity and even more the need for ab etter understanding of structure-property relationships motivated chemists trying to cut out molecular entities from the three-dimensional polymeric metal chalcogenides.This was achieved by formally replacing metal-chalcogen bonds by agreat variety of metalligand bonds,w ith the ligands being inorganic as well as organic moieties.A saresult, the concept of polynuclear metal chalcogenido clusters was established and the achievements made over the years for both main group-and transition metal-containing such clusters were regularly reviewed. [2] This type of chemistry is also well established for the element tin. Randomly selected representatives are (RSn) 4 E 6 (R = organic substituent with or without additional functionality,E = O, [3] S, [4] Se [4] )s howing adamantane-or double decker-type structures,d odecanuclear tinoxo clusters {(RSn) 12 O 14 (OH) 6 }( R= i-Pr, n-Bu, Me 3 SiCH 2 ,f errocenyl), [5] Sn 12 O 8 (OH) 4 (OEt) 28 (HOEt) 4 , [6a] [(BuSn) 12 (m 3 -O) 14 (m 2 -OH) 6 ]-(L 1 ) 2 ·2 EtOH, [6b] [NaO 4 (BuSn) 12 12 ], [6d] tetraorganodistannoxanes {R 2 (X)SnOSn(X)R 2 } 2 (R = organic substituent, X = electronegative substituent), [7] [{R(X)Sn-(CH 2 ) n Sn(X)R}O] 4 (R = organic substituent;X= halide,h ydroxide,c arboxylate; n = 1, [8a] 3-8, 10, 12 [8b] ), [(R 2 SnO) 3 (R 2 SnOH) 2 CO 3 ] 2 (R = organic substituent), [9] and [(2,4,6-i-Pr 3 C 6 H 2 Sn) 8 (m 4 -O) 2 (m 3 -O) 8 -(m 2 -O) 4 (m 2 OH) 8 {Sn-(OH)} 4 ]. [10] Recently,this well-established chemistry got new momentum by the spectacular findings reported by S. Dehnen et al. about laser-induced white light-emitting ability of the simple styryltin silsesquisulphide [(StySn) 4 S 6 ], [11] but also by extending the Sn-nuclearity of tinoxo clusters to the impressive number 34 in [(n-BuSn) 34 Na 2 (OH) 14 O 40 (PA) 8 ]·2-(PA)·8 H 2 O( PA = propionic acetate), as published by L. Zhang et al. [12] One aspect from these studies is that the steric bulk and identity of the substituents bound to the tin centre control the nuclearity of the clusters.F or organic substituents R, the general trend is that, as result of reactions with water or hydroxide,m onoorganotin precursors RSnX 3 (X = halogen, alkoxide,c arboxylate) tend to give clusters of higher nuclearity than diorganotin precursors R 2 SnX 2 do.O nt he other hand, triorganotin compounds R 3 SnX can only give distannoxanes R 3 SnOSnR 3 as result of such reactions.However,in combination with appropriately designed ligands and the concept of self-assembly,t hey as well as diorganotin precursors R 2 SnX 2 may also serve for the formation of largemembered polynuclear rings. [13] It is common knowledge that the complete replacement of the electronegative substituents Xindiorganotin compounds of type R 2 SnX 2 (X = halogen, alkoxide,c arboxylate) with oxide dianion gives the corresponding diorganotin oxides (R 2 SnO) n .Depending on the identity of the organic substituents R, these oxides can either be polymeric (type I, n = 1), [14] trimeric (type II, n = 3) [15][16][17][18][19][20][21][22] or even dimeric (type III, n = 2) (Scheme 1). [23][24][25] Sterically less demanding organic substituents such as n-alkyl or phenyl give polymers.T hese, because of intermolecular O!Sn interactions making the tin atoms five-or even six-coordinate,a re almost insoluble in common organic solvents.I ncreasing the steric bulk of the organic substituents enables the formation of six-or even four-membered rings in which the tin atoms are fourcoordinate.T he same principle holds for the formation of the molecular diorganotin oxides of types IIa (in which two parallel six-membered Sn 3 O 3 rings are linked to each other by three organic spacers), [26] IV (adamantane-type structure), [27] and V (the only crystallographically characterized diorganotin oxide containing an eight-membered Sn 4 O 4 ring). [27] More recently,i ntramolecular N!Sn or P=O!Sn coordination proved to be alternatives to steric bulk for the stabilization of type III diorganotin oxides. [24,25] Considering what is stated above,w ep ose the question whether diorganotin oxides can be synthesized the structures of which are in between the polymers of type I on the one hand and the eight-, six-, and four-membered rings of types II, III,a nd V on the other hand. Herein, we report that diorganotin diiodide precursors containing sterically lesscrowded substituents (Ph, Me 3 SiCH 2 ), but having aparticular tripod architecture give,b yr eaction with appropriate oxide, respectively hydroxide anion releasing reagents,c yclic polynuclear molecular diorganotin oxides (Scheme 1, type VI)of unprecedented large sizes.I nt hese,t he tin atoms adopt the coordination number five.
Finally,a ne lectrospray ionization mass spectrum (ESI MS;S upporting Information, Figures S66-S73 Thereaction of the diorganotin diiodide derivative 6 with sodium hydroxide,N aOH, in am ixture of dichloromethane, methanol, and water (Scheme 2) gave ac rude reaction mixture a 119 Sn NMR spectrum of which was rather complex and showed both broad and sharp resonances between À125 and À170 ppm (see Supporting Information, Figure S82). After the work-up procedure,amicrocrystalline material was obtained. From this,a ne xtremely small crystal was picked and identified by single crystal X-ray diffraction analysis as the molecular diorganotin oxide solvate 8.A lthough the elemental analysis of the bulk crystalline material, obtained from the reaction between 6 and NaOH (Scheme 2), perfectly matches with the empirical formula [Me-Si(CH 2 SnCH 2 SiMe 3 O) 3 ] n ,o ne cannot be sure whether it exclusively consists of the trikonta-nuclear species 8 (n = 10). Given the insufficient amount of material, no powder X-ray diffraction analysis of the bulk material was performed. Figure 4s hows its simplified molecular structure and the Figure caption contains selected interatomic distances and angles.T he compound crystallizes in the triclinic space group P1 with two molecules in the unit cell.
Like in the oktokaideka-nuclear diorganotin oxide 7,a ll tin centres in 8 are five-coordinated and, except Sn(1), show distorted trigonal bipyramidal environments.F or the Sn(2)-Sn(4), Sn(7)-Sn(13), Sn(16)-Sn(25), Sn (29), and Sn(30) atoms,f or each case two carbon atoms (the Me 3 SiCH 2 and the MeSiCH 2 methylene carbon atoms) and one oxygen atom occupy the equatorial positions.T he other two oxygen atoms take the axial positions.T he corresponding C eq -Sn-C eq angles   Figure S81) of the same sample showed am ajor intense broad unsymmetrically shaped signal at d 0.9 ppm and as harp resonance at dÀ21.9 ppm. A 1 HNMR spectrum (Supporting Information, Figure S79) revealed signals for the SiCH 3 ,S iCH 2 Sn, SnCH 2 SiMe 3 ,and Si(CH 3 ) 3 protons with correct integral ratio of 3:6:6:27. Attempts at obtaining 1 HDOSY NMR spectrum failed as the sample became gel-like over time.F rom the NMR data at hand, we conclude that the identity of 8 is not retained in solution. Apparently,t he solution contains am ixture of different species.W ith caution and in analogy to 7,weassign the three sharp 119 Figure S75 in the SupportingInformation shows an image of the complete molecular structure including numbering of the atoms. Although no detailed mechanistic studies have been performed, the formation of 7 and 8 can formally be seen as as tepwise process as shown in Scheme 3. Molecular diorganotin oxides A with adamantane-type structure undergo ringopening dimerization via the intermediate B giving the hexanuclear product C.I nc ase of R = Ph, three C-moieties assemble giving the oktokaideka-nuclear diorganotin oxide 7.
In case of R = Me 3 SiCH 2 ,h owever, C-moieties combine with A-and B-moieties giving,asone product out of probably several, the trikonta-nuclear molecular diorganotin oxide 8. This view gets support from the ESI mass spectrometric studies revealing mass clusters that are in line with the presence of A-a nd C-moieties (vide supra).

Conclusion
In conclusion we have shown that simple tripod-type diorganotin halides such as MeSi(CH 2 SnRI 2 ) 3 (R = Ph, Me 3 SiCH 2 )s erve as precursors for the synthesis of novel belt-shaped molecular diorganotin oxides [MeSi-(CH 2 SnRO) 3 ] n of unprecedented oktokaideka (n = 18) and trikonta (n = 30) nuclearity.T he results obtained fit well into the ongoing interest in large-sized metaloxido clusters in general [2k] and tinoxido clusters of high nuclearity in partic-ular. [5g, 6b,12, 31, 32] Thec oncept shown herein holds great potential for future work and we encourage interested readers to step into the field. Just to mention afew options out of many, the variation of the substituents Ra nd/or replacing the CH 3 group with other substituents,t he variation of the spacing between the silicon and tin centres as well as replacement of the MeSi bridgehead moiety with MeGe or with isoelectronic P, P = E( E= O, S, Se) or PM (M = transition metal moiety such as W(CO) 5 and others) might give ap lethora of novel diorganotin oxides showing polynuclear structures.Moreover, replacing the organic substituent Ri nR oeskys( RSn) 4 O 6 ( R = (Me 3 Si) 2 CH) [3] by aR ' of slightly reduced steric bulk could give well defined oligomers [(R'Sn) 4 O 6 ] n similar to 7 and 8.