J. CHEM. SOC. DALTON TRANS. 1994 2705Preparation of N(SeCI),+X- (X = Sbcl, or FeCI,),F,CCSeN SecC F, + s bc16-, F,CCSeN SeCC F,, F,CCSeSeCC F,and F,CCSeSeC( CF,)C( CF,)SeSeCCF,. Electron Diffraction IStudy of F,CCSeSeCCF, and Crystal Structure of the 1 Eight-membered Heterocycle F,CCSeSeC(CF,)C(CF,)SeSeCCF, tKonstantin B. Borisenko,B Matthias Broschag,b istvan Hargittai,"sa Thomas M . Klapotke,**bDetlef Schroder,= Axel S c h u l ~ , ~ Helmut Schwarzc lnis C. Tornieporth-Oetting andPeter S. White * r da Institute of General and Analytical Chemistry, Budapest Technical University and Hungarian Academyof Sciences, H- 152 1 Budapest, Hungarylnstitut fur Anorganische und Analytische Chemie, Technische Universitat Berlin, Strasse des 1 7 Juni135, D - 10623 Berlin, Germanylnstitut fur Organische Chemie, Technische Universitat Berlin, Strasse des 17 Juni 135, 0- 10623 Berlin,GermanyDepartment of Chemistry, University of North Carolina, Chapel Hill, North Carolina 27599, USAThe salts N(SeCI),+SbCI,- 1 and N(SeCI),+FeCI,- 2 were synthesized by reaction of SeCI,+X-(X = SbCI, or FeCI,) with N(SiMe,),; 1 was also formed by reaction of Se,NCI, with SbCI,.Reaction of 1with SnCI, and F,CCCCF, led to the formation of F,CCSeNSeCCF,+SbCI,- 3. In this reaction theSe,N+ cation is a likely intermediate because SnCI, seems t o be essential for chloride abstraction inthe first reaction step t o generate Se,N+ in situ which then adds F,CCCCF, t o yield 3. Compound 3is a useful building block t o generate selenium compounds such as F,CCSeNSeCCF, 4,F,CCSeSeCCF, 5 and F,CCSeSeC(CF,)C(CF,)SeSekCF, 6.The heterocycle 5 was shown byelectron diffraction to have an approximately planar four-membered ring structure. The structure ofcompound 6 was determined by X-ray crystallo raphy: orthorhombic, space group Pbca, a =the cation F,CCSeNSeCCF,+, ab initio calculations were made on model compounds in which theCF, groups were replaced by a fluorine atom (i-e. FCSeSeCF for 5 and FCSeNSeCF+ for the cationin 3). In addition, mass spectrometric experiments were performed in order to examine the structuresand stabilities of the unligated cation F,CCSeNSeCCF,+ as well as its neutral counterpart. The existenceof the neutral radical 4 was established by means of neutralization-reionization mass spectrometry.-- -10.1920(21), b = 13.0615(20) and c = 22.050(5) 1 .In order to rationalize the structures of 5 and - - - -During the last few years significant advances have been madein the area of Se-N chemistry.'-3 The objective has been thepreparation of polymeric (SeN), which may exhibit moreunusual properties than the superconductor (SN),. SeveralSe-N chlorides which are potential building blocks, becausethey are sources of the SeNSe unit, have been synthesizedrecently. The first examples of ternary Se-N-CI cations wereN(SeCl,),+ and N(SeCl),+. The former was prepared byreaction of SeCl, 'ASF, - with N(SiMe,), [equation (l)]."6SeCl,+AsF,- + SN(SiMe,), - 9SiMe3F + 3AsF, +N, + 6SiMe3C1 + 3N(SeCI,), (1)The N(SeCl), + cation was prepared by different reactionpathways as its GaC1,- or its SbC1,- salt.'-' The salt~~~t Supplementary data available: Further details concerning the crystalstructure determination may be obtained from Fachinformations-zentrum Karlsruhe, Gesellschaft fur wissenschaftlich-technischeInformation mbH, D-76012 Eggenstein-Leopoldshafen, Germany, byquoting reference CSD-57857.Non-SI units employed: Eh z 4.36 x lo-'* J, eV z 1.60 x J,cal = 4.184 J .N(SeCl), 'GaC1,- was obtained from the reaction of Se,NC13with the Lewis acid GaCl, [equation (2)], whereas N(SeCl),+-SbC1,- 1 was prepared by the reaction of SeCl,+SbCl,- withN(SiMe,), [equation (3)].The solid-state structure of theSe,NCl, + GaCI, - N(SeCl),+GaCl,- (2)2SeCI3 + SbCl, - + N(SiMe,), --+N(SeCl),+SbCl,- + SbCl, + C1, + 3SiMe3C1 (3)cation N(SeCl),+ depends on its counter anion.It exists in thecrystalline state as either the u isomer (GaC1,- salt) or the sisomer (SbCl, - salt), both isomers being essentially identical intheir total energy CAE(MP2) = 1 kcal mol-'1, as shown by abinitio computations.Not only is the N(SeCI), + cation of general interest in termsof structure and bonding, it is also a very usefulbuilding block in the synthesis of heterocycles like F3C-CSeNSeCCF, +SbCl, - , F,CCSeNSeCCF,, F,CCSeSeCCF,and F,CCSeSeC(CF,)C(CF,)SeSeCCF,. The synthesis ofthe first by reaction of 1 with F,CC=CCF, and SnCl, isvery interesting, because the formation of the still- - -I 2706 J. CHEM. SOC. DALTON TRANS.1994unknown Se,N+ cation as an intermediate is more than likely.In this paper we summarize the syntheses of the compounds, theresults of some structural investigations (electron diffraction,X-ray) and of ab initio molecular orbital (MO) calculations forFCSeNSeCF+ and FCSeSeCF. We also present the results ofneutralization-reionization mass spectrometry (NRMS) experi-ments on compound F,CCSeNSeCCF, +SbCl, - which showthe existence of F,CCSeNSeCCF, and neutral Se,N as stablecompounds in the gas phase. The experiments presented aresummarized in Scheme 1.- - - -Results and DiscussionThe salt N(SeCl),+SbC16- 1 was obtained by reaction ofSeCl,+SbC1,- and N(SiMe,), in CCl, solution at roomtemperature. It was shown that the N(SeCl),+ cation has adifferent solid-state structure (s isomer, C,) from that in thepreviously reported N(SeCl), 'GaC1,- salt (u isomer, C2v).697At the MP2 level the s-shaped isomer is slightly higher in energythan the u-shaped cation (AE = 1 kcal mol-') and it was shownexperimentally that only marginal differences in cation-anioninteraction can favour either of these species.The compoundN(SeCl), +FeCl,- 2 was synthesized in SO% yield from SeCl,,FeCl, and N(SiMe,), [equations (4) and ( 5 ) ] . The infrared2SeC1, + 2FeC1, + 2SeCl,+FeCl,- (4)2SeC13+FeC1,- + N(SiMe,), ---+N(SeCl),+FeCl,- + FeC1, + Cl, + 3SiMe3C1 (5)spectrum of 2 shows two strong bands at 933 [v,,,,(SeNSe)]and at 432 cm-' [v,,,(SeNSe)] which indicate that theN(SeCl), + cation exists in the u-shaped C,, form.ThecompoundN(SeCl), 'GaC1,- reported by Dehnicke and co-workers wasprepared by a different reaction route from Se,NCl, and GaCl,[equation (2)]. In order to apply this reaction to the synthesis of1, we prepared Se,NCl, from SeC1, and N(SiMe,),.6*8 Thesynthesis of N(SeCl),+X- (X = SbC1,- 1 or FeC1,- 2) ispossible either by reaction of Se,NCl, with Lewis acids likeSbCl, or FeCl, or by the reaction of SeCl, +X- with N(SiMe,),.The latter method is more convenient because the startingSeCl3+FeCl+- + N(SiMe& * N(SeCI)*+FeC4' + .....2SeCI3+SbCIe- --I 1- Se2NCI3t tN (SeCI),'SbCI6-13 45 6Scheme 1 ( i ) N(SiMe,),; (ii) SbCl,; (iii) excess of SnCl,, F,CCCCF,;(iu) NRMS; (u) Na,S,O,; (ui) MS; (uii) 0 "C, 7 dmaterial SeCl, +X- can easily be synthesized without side-products in nearly quantitative yield whereas the yield in thepreparation of Se,NCl, is only 60%.,Compound 1 is a very useful building block in heterocyclicselenium chemistry.Its reaction with hexafluorobutyne andSnCI, led to the formation of the five-membered heterocycleF,CCSeNSeCCF, + SbC1, - 3 [equation (6)]. The cation in -3 is the selenium analogue of the well known 4,5-bis(trifluoro-methyl)-l,3,2-dithiazolylium cation.' It is very likely that thestill unknown Se,N+ cation is formed during the reaction as anunstable intermediate due to the presence of SnCl,, which actsas a chloride acceptor. Subsequently, it is trapped by cyclo-addition to F,CC=CCF,, which leads to the formation of 3.This statement is supported by the fact that 1 did not react togive 3 in the absence of SnCl,.Another possible reactionpathway in which F,C~Se(C1)NSe(Cl)~CF3+SbC1,- mighthave been formed in the first step can be excluded because itshould be sufficiently stable to have been isolated. Moreover, noreaction was observed without SnCl,. In addition, the reactionof 1 with SnCl, (in the absence of F,CCCCF,) in order toobtain the Se,N+ cation (and SnC1,) led to unidentifieddecomposition products. In contrast, the reaction of theanalogous sulfur cation in N(SCl), +AlCl,- is a convenientmethod to synthesize S,N+AlCl,- in good yield." This meansthat SnC1, is necessary to abstract the chlorine atoms via adductformation (cf. ref. 10) and that hexafluorobutyne is essential tostabilize the intermediate for formation of 3.Unfortunately, thecrystals of 3 were not suitable for a structure determination, butthe structure of a model compound in which the CF, groups arereplaced by F atoms has been computed by ab initio methods.The geometry was fully optimized at the Hartree-Fock (HF)and correlated MP2 level. The corresponding structuresincluding bond lengths and angles are shown in Fig. 1. Not onlyat the MP2 level but also at the HF level both the Se-N and theSe-C bond distances are slightly shorter than the sum of thecovalent radii for a single bond (Se-N 1.92, Se-C 1.94 A)''which indicates an additional n-bond interaction between thering atoms (6n-electron aromatic ring).The F,CCSNSCCF, + cation can be chemically reduced withan excess of finely ground sodium dithionite in SO, at roomtemperature to give the 7n-electron radical F,CCSNSCCF, asa sensitive black-green liquid beneath a blue gas in high yield(88%).'2 The same reaction with the selenium cation in salt 3did not yield the 771 radical species, but led under dinitrogenelimination to the formation of F,CCSeSeCCF, 5.Compound5 was obtained in 40% yield as a red liquid (m.p. - 15 "C), whichhas to be stored at liquid-nitrogen temperature in order toprevent dimerization. The compound was previously obtained- --Se SeFig. 1 Computed structures of the cation FCSeNSeeF' (bondlengths in A, angles in ") (for basis set and computational details seeExperimental section): (a) MP2 level, E = 347.9 E,,; (6) HF level, E =346.9 EJ.CHEM. SOC. DALTON TRANS. 1994 2707by refluxing elemental selenium under a flow of N, andF,CCzCCF, [equation (7)] (yield 25%), but the structure wasnot determined.2Se + F,CC=CCF, --+ F,CCSeSeCCF, (7)5The gas-phase structure of compound 5 was investigated byelectron diffraction. The experimental and theoretical molecularintensities and radial distributions are shown in Fig. 2, and themolecular structure including the numbering of the atoms ispresented in Fig. 3. The main results of the least-squaresprocedure are presented in Table 1, and the final molecularparameters with estimated total errors in Table 2. The Se-Cbond length is well determined as its contribution at 1.9 8, is wellseparated from those of the other distances on the radialdistribution curve. As to the C=C, C-C and C-F bonds, theirlengths obtained using different initial values of A(CC) andA(CF) were also stable and the differences referring torefinements from different initial values did not exceed theexperimental errors.The C-C bond lengths in compound 5 and the sulfuranalogue F,CCSSCCF, are in good agreement, taking intoaccount the large experimental errors in the structurenI 50 cm0 10 20 30s A-1I------ c... . . . . . .:. . . . . . . . . :. . . . . . . .~-+. . . . . .-:. . . . . . . . . I0 2 4 6 r / AFig. 2 Experimental (* 0 ) and theoretical (-) molecular intensities.sM(s) and radial distributions f ( r ) . The difference curves are shownbelowFig. 3Se(3) n-W "F(8)- F(7) O F ( 10)Molecular structure of compound 5determination of F,CCSSCCF,.l4 Structural changes in therest of the molecule as compared with simpler systems suggestconsiderable electron-density redistribution. This seems to beconsistent with the overall structure presented by Scheme 2.According to these resonance structures the Se-Se bond is notexpected to be considerably different from a single bond. Thelength of the C-Se bonds is intermediate between the singleC-Se bond in MeSeSeMe" and the double C=Se bond inF,C=Se. l 6Concerning the vibrational amplitudes, they were groupedtogether according to the appearance of the respective distanceson the radial distribution curve. This grouping is indicated inTable 1. The influence of alternative choices of the fixeddifferences between vibrational amplitudes belonging to thesame group on the other parameters was carefully examined.Itwas found that the choice of these differences, in a reasonablerange, did not influence the structural results beyondexperimental error.The conformation of the molecular skeleton was found to beapproximately planar, the S e a - S e and C - M - C dihedralangles being 11.3 2 2 and 8.6 k 3.4', respectively. Thedifference from zero dihedral angles obtained in the electrondiffraction analysis may be a consequence of torsional motion,and, accordingly, a planar equilibrium conformation cannot beexcluded. However, the C-Se and C-C bonds originating fromthe same end of the C==C bond turn in the same direction.Thismay indicate that the deviation of the molecular skeleton fromplanarity is a real structural feature and not merely aconsequence of torsional vibrations.Torsion of the CF, groups around the C-C bonds was alsoexamined. Refinements were made with several initial values ofthe C==C-C-F(7) torsional angle between 0 and 60". Thesmallest R factor was obtained at 27.0'. This might indicate thatrotation of the CF, groups is considerably restricted.In addition, the observed tilt decreases the closest contactbetween Se and F, 3.033(4) A, which is about 0.32 8, shorterthan the sum of the respective van der Waals radii, 3.35 A. Thispoints to an interaction between Se and F of the CF, group as aconsequence of electron-density redistribution in accordancewith the resonance structures of Scheme 2, where the Se atomhas a partial positive charge.Thus electrostatic interactionbetween the positively charged Se and electron-withdrawingfluorine could make the observed conformation the preferredone. This is facilitated by the electron-density redistribution inthe molecular skeleton provided by resonance.The structure of FCSeSeCF as a model compound for 5 wascalculated by ab initio methods (HF, MP2). The results areshown in Fig. 4. The calculated data nicely support theexperimentally obtained (electron diffraction) structural datafor 5.-Se' -\\Scheme 2 Resonance structures for compound 5(a ) ( b )n Fig. 4 Geometry-optimized structure of FCSeSeCF (bond lengths inA, angles in ") (for basis set and computational details see Experimentalsection): (a) MP2 level, E = 293.6 E,,; (b) HF level, E = 292.8 E2708 J.CHEM. SOC. DALTON TRANS. 19941Table 1 Results of least-squares electron diffraction refinement of F , C e C F , ; distances in A, angles in OParameter r a I Group Parameter r a I GroupIndependentC(2)=c(5) 1.363(5) 0.047( 1) (i) C( l)-C(2)=c(5)-C(6) 8.6(21)C(2)-Se(3) 1.882( 1) 0.057(2) (ii) F(7)-C( 1 )-C(2)=C(5) 27 .O( 1 2)C(2)=C(5)-Se(4) 105.1 (2) Tilt * 3.4(4)C( 1 FC(2)=C(5) 130.9(2) A(CC)' 0.125(4)C-C-F (mean) 1 1 1.6( 1) A(CF) ' 0.023( 5)Se(3)-C(2)=C(5 jSe(4) 11.3(14)DependentC( 1 W ( 2 )C-F (mean)Se(3 jSe(4)C(1) * * - C(5)C( 1) - - C(6)C( 1) * * F( 10)C(1) F(11)C ( l ) - - * F(12)C(2) F(7)C(2) - F(8)C(2) F(9)C(2)..* F(10)C(2)***F(11)C(2) - - F(12)C( 1) Se(3)C(l) - Se(4)C(2) .- Se(4)Se(3) - - F(7)1.488(2)1.340(1)2.367( 1)2.593(4)3.3 16(5)3.168(9)3.846( 5)4.5 1 O( 5)2.3 8 5( 6)2.3 1 6( 3)2.3 1 6( 3)2.996( 10)3.297(8)3.590( 5)2.971(4)4.079( 3)2.594(2)4.057( 3)0.05 10.0430.047( 1)0.068( 2)0.1 18(2)0.0980.072(2)0.155(3)0.0820.0820.0820.1080.1080.130(3)0.0880.0880.0580.089Se(3) - . F(8)Se(3) - - - F(9)Se(3) - - F( 10)Se(3) - F( 1 1)Se(3)-.. F(12)F( 7) - F(8)F(7) F( 10)F(7)*-.F(11)F(7) * * F( 12)F(8) * * a F(10)F(8)*--F(11)F(8) - * F( 12)F(9) - F( 10)F(9) * F(11)F(9)s.s F(12)C(2FSe(3)-Se(4)F-C-F (mean)3.03 3 (4)3.665(6)4.757( 10)4.698( 13)4.758( 4)2.160( 1)2.811(12)3.333(10)4.48 l(9)4.48 l(9)4.866(4)5.71 l(7)3.3 33( 1 0)4.649(8)4.866(4)74.27( 5)107.3( 1)0.1030.1400.1060.1060.1060.06 10.1030.1030.1860.1860.1860.135(25)0.1030.1860.186w .. (Vlll)(vii)(vii)(vii)(iii)(v)(v)(vii)(vii)(vii)(ix)(v)(vii)(vii)a Least-squares standard deviations in the last digit are given in parentheses. Angle between the C-C bond and direction of the C3 symmetry axisof the trifluoromethyl group. A(CC) = r[C(l)-C(2)] - r[C(2)S(5)], A(CF) = r[C(2)=c(5)] - r(C-F)mean.Table 2 Bond lengths ( r g in A), bond angles and torsion angles (")with estimated total errors for F,C-CCF,G== 1.364 f 0.007 Se-Se 2.368 f 0.005 c-c 1.489 f 0.004 C-F (mean) 1.342 f 0.003C-Se 1.884 f 0.004C-Se-Se 74.3 f 0.1 C-C-F (mean) 11 1.6 ? 0.3F-C-F 107.3 f 0.2 C - M - C 8.6 f 3.4G==-Se 105.1 f 0.3 Se-C=C-Se 11.3 f 2.0C-C== 130.9 f 0.4 M - C - F 27.0 f 1.9Wavenumberl cm-'Fig. 5 Raman spectrum of compound 5 (647 nm, 50 mW, 20 "C)The four-membered heterocycle 5 was also characterized byits Raman spectrum (Fig.5 ) , which is in agreement with thecorresponding infrared spectrum reported in the literature.Compound 5 is unstable at 0 "C with respect to dimerization.Within 1 week a sample was completely converted into the eight-membered heterocycle 6, which was isolated in the form oforange crystals, which were neither air nor moisture sensitive.The structure of 6 was determined by X-ray crystallography.Two different views are presented in Fig.6. The atomicFig. 6 Two different views"," of the molecular structure ofcompound 6parameters are listed in Table 3, selected bond lengths andangles in Table 4. The molecular structure is a twisted eight-membered ring with four exocyclic CF, groups. All bonddistances compare well with those expected from a simpleo,n-bonding Lewis represen tation.The sulfur analogues of compounds 5 and 6 have been knownsince the 1960s. 14*19 Monomeric F,CCwCCF, is synthesized byrefluxing sulfur in the presence ofhexafluorobutyne andcan easilybe converted into the dimer F,CCSSC(CF,)C(CF,)SSCCF, bykeeping the monomer at 25 "C for 2 months." The dimer is thethermodynamically more stable at low temperatures, but atelevated temperatures (1 80-220 "C) it decomposes yielding theI J.CHEM. SOC. DALTON TRANS. 1994 2709monomer. In contrast to the sulfur species, the seleniumcompound 6 cannot be converted into 5, but decomposes to giveelemental selenium.In addition, we performed mass spectrometric experimentsin order to examine the unimolecular and collision-inducedfragmentations of the F,CCSeNSeCCF,+ cation in 3 as well asthe stability of its neutral analogue in the gas phase. The NRMStechnique has been shown to be a valuable tool for elucidation-_ _ _ _ ~ ~ ~ ~ ~ ~Table 3deviations in parenthesesAtomic coordinates of compound 6 with estimated standardX0.443 19(18)0.255 73( 18)0.453 62( 18)0.264 33( 18)0.314 2(16)0.400 5( 16)0.313 5(17)0.391 5(16)0.256 9( 17)0.459 4( 18)0.258 l(19)0.450 6( 19)0.138 8( 10)0.332 3(10)0.239 5( 10)0.486 9(9)0.384 7(10)0.574 l(10)0.141 7(10)0.332 l(11)0.231 6(10)0.480 5( 10)0.370 4(11)0.561 9(11)Y0.021 75(12)0.023 95( 12)0.260 29( 12)0.267 37( 12)0.096 2(11)0.178 O(11)0.192 2(11)0.105 7(11)0.046 4( 12)0.220 8( 13)0.243 6(13)0.063 O( 13)0.006 3(7)-0.031 3(6)0.1 10 3(7)0.146 9(7)0.289 2(7)0.265 8(8)0.285 8(7)0.316 8(7)0.176 O(7)0.137 8(7)0.01 3 6(8)- 0.002 6(7)Z0.160 44(8)0.101 57(8)0.097 30(8)0.154 69(8)0.031 7(8)0.029 6(7)0.225 l(7)0.226 4(7)- 0.024 6(8)-0.027 3(8)0.281 5(9)0.285 7(8)-0.01 1 7(4)-0.045 l(4)- 0.070 9(4)- 0.067 6(4)-0.054 2(5)-0.015 2(5)0.266 6(4)0.303 O(4)0.325 7(4)0.325 l(4)0.313 2(4)0.273 O(5)Table 4 Selected bond lengths (A) and angles (") of compound 62.3 1 O( 3)1 .896( 1 5)1.901 (1 6)2.309(3)1.918( 15)1.905(16)1.385(22)1.519(23)1.499(24)1.38 l(22)1.52(3)1.543(24)101.2(5)101.7(5)102.3(5)101.9(5)127.6(12)109.3( 10)123.1 ( 14)1 25.8( 1 2)109.2(11)124.8(14)1 26.1 ( 12)109.9(11)123.9( 1 5)128.1(12)109.4(11)122.5(14)109.7( 14)1 12.1 (1 3)114.2(12)106.7( 12)1.343(19)1.350( 19)1.329( 19)1.342(19)1.315(21)1.336(21)1.349(22)1.307(21)1.342(21)1.343(20)1.330(21)1.335(21)106.7( 13)107.0( 1 3)11 1.7(13)11 3.5( 14)110.4(15)108.1 (1 5)105.4( 13)107.3(14)108 .O( 1 5)113.9( 15)112.2(13)107.2( 13)105.6(14)109.5( 16)1 1 1.9( 13)1 12.3( 15)109.2(14)108.3( 14)107.1 (1 5)1 07.8( 13)of the structures of ions and neutral compounds formed in thegas phase.20-2The metastable ion (MI) mass spectrum of F,CCSeNSeC-CF,' (Fig. 7) is dominated by the loss of hexafluorobutyne(loo%), corresponding to the formal cycloreversion product ofthe [2 + 31 addition ofthe SeNSe' dipole to the alkyne. The onlyother unimolecular fragmentation exhibits very low intensityand corresponds to the loss of a neutral SeN unit (1 %). The CA(collisional activation) mass spectrum of F,CCSeNSeCCF, + (Fig. 8) is similar to the MI spectrum; additional peaks are dueto the losses of a N or a F atom, the elimination of a CF, group,and fragments originating from further decomposition ofSe,N'+, i.e.Se," and SeN'. The simple dissociation patternand the low intensities of the radical losses are completely in linewith the aromatic character of cationic F,CCSeNSeCCF, + .The most interesting feature of the NR mass spectrum ofF,CCSeNSeCCF, + (Fig. 9) corresponds to the peak at m/z =336 (for the 80Se, isotopomer) which corresponds to re-ionizedneutral F,CCSeNSeCCF, 4, thus establishing the intrinsicstability of neutral F,CCSeNSeCCF,' in the gas phase. Ascompared to the MI and CA mass spectra, the NR spectrumexhibits a much richer fragmentation pattern. This is not onlydue to the multicollision event in the NR experiment, but alsoreflects the decreased stability of the neutral species.However,the fragments in the NR spectrum are in line with the cyclicstructure of F,CCSeNSeCCF, +, as evidenced by the signalscorresponding to the heteroatom backbone, i.e. Se,N+, Se, +,SeN+ and Se', and those of the hexafluorobutyne moiety, i.e.C4F4'+, C4F3+, C3F3+, C,F,*+, CF3+ and C3F+, respectively.--- - - --Se2N+L F,CCSeN SeCCF,' + I -SeNiFig. 7 The MI mass spectrum of F,CbeNSeCCF,+Se2N+mfz -mfz -Fig. 8 The CA mass spectrum of F,CCSeNSekCF, 2710 J. CHEM. SOC. DALTON TRANS. 1994mlz + - Fig. 9 The NR mass spectrum of F,CCSeNSeCCF,+According to the analysis of the isotopomers of F,CCSeNSeC-CF, ' , the signal for the re-ionized hexafluorobutyne moleculeis almost negligible.We note in passing that the Se,N+ cation,which is also formed upon electron impact, does also exhibit anintense recovery signal in its NR spectrum, and thus the elusiveneutral Se,N' molecule is stable within the ps time-scale of themass spectrometric experiment.*ExperimentalThe reactions were carried out in glass vessels in an inert-gasatmosphere (N,, dry-box). The salt SeCl, 'SbC1,- was directlyprepared from SeC1, and SbCl, and (SeCl),N+SbCl,- wassynthesized as described in the literature and purified byrecrystallization from SO,.7 The compounds N(SiMe,),, SnCl,,SbC1, and FeCl, (all from Aldrich) were used as received,Na,S,O, (Merck) was dried over P4OI0 and F,CCCCF,(Fluorochem) was used after distillation; CCl, (Merck, driedover P4OlO), SO, (Messer Griesheim, stored over CaH,),CH,Cl, (Merck, stored over P4OI0) and CFC1, (Merck, storedover P4OlO) were used after distillation.Infrared spectra were recorded using a Perkin-Elmer 580B or aNicolet Magna spectrometer, Raman spectra using a Jobin YvonRamanor U 1000 spectrometer equipped with a Spectra Physicskrypton-ion laser (647.09 nm), 13C NMR spectra using a BrukerWH 270 instrument (67.9 MHz) and referred to SiMe, and 77SeNMR spectra with a Bruker ARX 400 spectrometer (76.3 1 MHz)and referred to Me,Se.Routine mass spectra were recorded on aVarian MAT 3 1 1 A instrument (electron impact, 70 eV), MI, CIand NR spectra with a modified VG ZAB/HF/AMD large-scalefour-sector spectrometer of BEBE configuration (B = magneticsector, E = electric s e ~ t o r ) .~ ~ , ~ ~ The carbon and nitrogenanalyses were performed by the TU Berlin service.Syntheses.-N(SeCl), + SbCl, - 1 from Se,NCl,. The com-pound SbC1, (0.392 g, 1.31 1 mmol) was added to a solution ofSe,NCl, (0.365 g, 1.31 1 mmol) in SO, and stirred for 2 h atroom temperature. After pumping off the solvent, orangecrystals (0.72 g, 95%) were obtained which were spectro-scopically and analytically identified to be the recentlydescribed N(SeCl), + SbCl, - .N(SeCl),+FeCl,- 2. The compounds SeCl, (1.000 g, 4.53mmol) and FeCl, (0.735 g, 4.53 mmol) were stirred in CCl, (150cm3) for 2 h and N(SiMe,), was added dropwise to the reactionmixture and stirred for another 2 h. The precipitated orangesolid was filtered off, washed with CCl, and dried in vacuum.After recrystallization from SO, 2 was obtained as red needles(0.80 g, 80%) (Found: N, 3.00.Cl,FeNSe, requires N, 3.15%).IR (pure compound between KBr plates): v,,, 933s [vasym-* Note added at proot Condensation of a microwave discharge excitedstream of argon-nitrogen-selenium gave SeN, Se,N and Se,N+, whichwere characterised by IR spectroscopy.26(SeNSe)], 800s, 48 Is, 432vs [v,,,(SeNSe)], 428s [v(SeCl)], 418scm-' [v(SeCl)]. Raman (20mW, - 50 "C): 434(2) [v,,,(SeNSe)],397(10) [v(SeCl)], 381(1) [v(SeCl)], 373(1) [v,(FeCl,-)],332(4) [v,(FeCl,-)], 314(1), 128(5) [y(SeCl)], 113(4) [6(SeC1)].El mass spectrum: m/z 254 (15.9, Fe2C14'), 230 (12, Se2C1,+),193 (13, Se,Cl+), 161 (14, FeCl,'), 150 (84, SeCl,'), 126 (15,FeCl,+), 115 (100, SeCl'), 91 (1 1, FeCl') and 70 (45%, C12+).Se,NCl, (cf.ref. 6). A solution of N(SiMe,), (2.200 g, 9.40mmol) in CH,Cl, (20 cm3) was added dropwise to a suspensionof SeCl, (4.100 g, 18.70 mmol) in CH,Cl, (30 cm3). The reactionmixture was refluxed for 12 h and the solution was filtered offfrom unreacted SeCl, and concentrated to 10 cm3. Cooling to5 "C yielded after 2 d a first fraction of crystals which were notused for further experiments. The second and third fractions ofcrystals of Se,NCl, (1.30 g, 50%) were used for furtherexperiments.F,CCSeNSeCCF, 'SbC1,- 3. The compound F,CC=CCF,(0.926 g, 5.72 mmol) was condensed at - 196 "C onto a frozensuspension of 1 (2.750 g, 4.76 mmol) and SnCl, (0.903 g, 4.76mmol) in SO2 (7 cm3).After warming to room temperature themixture was stirred for 5 d. It changed from red to light brown.After separation of insoluble by-products the solvent andvolatile products were pumped off yielding a white solid whichgave, after recrystallization, fine, colourless needles ofcompound3 (2.13 g, 67%), decomp. > 155 "C (Found: C, 7.15; N,2.25. C,Cl6F,NSbSe, requires c , 7.20; N, 2.10%). IR (KBrpellet): v,,, 1595w, 1530w, 1260s, 1185vs [v(CF)], 1025m,890vw, 718m, 688m, 639w, 600w, 370m, 3355, 304w cm-'[v,(SbCl,)]. Raman (50 mW, 20 "C): 885(1), 714(2), 636(1),mass spectrum: m/z 336 (32, F,CCSeNSeCCF,), 228 (51,SbCl,'), 193 (100, SbCl,'), 174 (35, Se,N'), and 158 (13%,SbCl'). 13C NMR (CD,CN): 6 121.9 [2 C, q, 'J(CF) 278, 4CF,] and 184.9 [2 C, q, 2J(CF) 44 Hz, 2CCF,].F,CCSeNSeCCF, 4.Compound 4 was generated andidentified from 3 by means of NRMS. Briefly, 3 was admitted tothe ion source via the solid probe inlet system (probe tiptemperature 160 "C) and subsequently ionized by a beam ofelectrons having 70 eV kinetic energy in an EI source (repellervoltage ca. 30 V). The ions of interest were accelerated to 8 keVkinetic energy and mass-selected by means of B(l)/E(l) at aresolution of m/Am x 3000. Unimolecular fragmentationsoccurring in the field-free region preceding the second magnetwere recorded by scanning B(2); the spectra so-obtained will bereferred to as MI spectra. For collisional activation the 8 keVions were collided with helium (80% transmission); the fourthsector was not used in this study.For NRMS experiments theprecursor ions were mass-selected by means of B(l) andsubjected to a double-collision experiment. Neutralization wasperformed by collision with xenon (85% transmission); theremaining ions were deflected by a 1000 V potential, the beam offast neutral compounds was subsequently reionized by collisionwith oxygen (85% transmission), and the so-formed ions wereanalysed by scanning E(1). Due to the relatively low vapourpressure of 3, the NRMS experiments were performed with thefirst two sectors only, and the resolution was reduced tom/Am x 2000; both these modifications were necessary toobtain a sensitivity sufficient for the two-stage collisionexperiments.The signals of F,CCSeNSeCCF, + for the 78Se,"Se and ',Se isotopomers were slightly contaminated byisobaric SbH,Cl,+ (n = 0-2) ions, stemming from thermolysisof 3 and its subsequent reactions with background water presentin the mass spectrometer. These contaminations amount to lessthan 1% of the ion flux, evaluated on the basis of thecharacteristic losses of C1' in the spectra of authentic SbH,,Cl6+formed in the ionization of SbCl,. According to this analysisthe contribution of neutral SbH,C16 to the recovery signal inthe NR spectrum amounts to less than 2%. Unfortunately,the thermal stabilities of F,CCSeNSeCCF, ' salts with othercounter ions, e.g. AlC1,- and A@,-, were not sufficient for-486(1), 401(1); 334(10), 288(2), 171(4) cm -' [Vl(SbC16-)].EI-IJ. CHEM. SOC. DALTON TRANS. 1994 271 1NRMS experiments. All spectra were accumulated andprocessed on-line with either the VG 250/11 or the AMD-Intectra data system; five to 20 scans were averaged to improvethe signal-to-noise ratio.F,CCSeSeCCF, 5. A solution of compound 3 (1.920 g, 2.88mmol) and Na2S20, (2.510 g, 14.42 mmol) in SO2 (7 cm3) wasstirred for 2 d at room temperature. The solvent was pumped offat - 50 "C and the orange-red residue, which melts at - 15 "Cto give a red liquid, was condensed into a vessel held at - 20 "C.The product (0.37 g, 40%) was stored at - 196 "C to preventdimerization, m.p. - 15 "C. Raman (50 mW, 20 "C): 1616(1)[v(C=C)], 1285(1), 1259(1), 1144(1) [v(C=F)], 901(1), 751(1),712(3), 651(1), 329(1), 297(10), 270(2), 250(1), 233(1) and 124(1)cm-'.El mass spectrum: m/z 322 ( M + , loo), 303 (18, M -F),160 (32, M - F,CCCCF,), 143 (16, F,CCCCF,+), 93 (10,F,CCC+) and 69 ( l l x , CF,'). NMR (CDCl,): 13C, 6 117.9[2 C, q, 'J(CF) 272, 2 CF,], 124.4 [2 C, q, 2J(CF) 21 Hz,2CCF,]; 77Se, 6 682.2 [2 Se, q, ,J(SeF) 2.3 Hz].F,CCSeSeC(CF,)C(CF,)SeSeCCF, 6. Within 1 week com-pound 5 (0.30 g, 0.94 mmol) slowly dimerized at 0°C andquantitatively yielded the air- and moisture-stable compound 6(0.30 g, 100%). Single crystals suitable for X-ray diffractionstudy were grown by recrystallization from CFCl,, m.p. 142 "C(Found: C, 15.00. C8F12Se4 requires C, 15.00%). IR (KBrpellet): 1562s [v(C=C)], 1230vs, 1179s, 1148 (sh), 1130s, 969m,850m, 809m, 681s, 671 (sh), 595m, 554w, 528m, 460w, 358w,331m and 270m.Raman (50 mW, 20 "C): 1559(3) [v(C==C)],1243(1), 1170(1), 1137(1), 966(1), 843(1), 801(1), 678(3), 667(1),591(1), 553(1), 518(1), 456(1), 329(2), 310(9), 264(10), 214(3),182(5), 152(7), 128(10) and 108(7) cm-'. NMR (CDCl,): 13C,6 120.6 [4 C, q, 'J(CF) 278 Hz, 4 CF,] and 131.4 (4 C, m, 4F,CC);77Se, NMR: 6 378 (4 Se, m).-I I- Electron DiJjCraction Structure Determination of F,CCSeSeC-CF, .-The electron diffraction photographs of the sample wererecorded with two nozzle-to-plate distances, 19 and 50 cm, in amodified EG-1 OOA apparatus 2 9 7 3 0 with a membrane-nozzlesystem at room temperature. The experimental molecularintensity range was 2.000-14.000 k' with step 0.125 A-1 for the50 cm nozzle-to-plate distance, and 9.00-35.75 A-' with step0.25 A-1 for the 19 cm nozzle-to-plate distance.The analysis was carried out by applying the least-squaresmethod to the molecular inten~ities.,~ Electron scatteringfactors were taken from available corn pi la ti on^.^^.^^ Apreliminary model of C, symmetry was chosen on the basis ofthe reported structure of 3,4-bis(trifluoromethyl)dithiete.l 4 TheC, symmetry axis bisects both the C==C and Se-Se bonds in thismodel. Models having C2, or C, symmetry, or no symmetry atall, were also tested. The best agreement with the experimentaldata was achieved for the C, symmetry model. The CF, groupshad local C3 symmetry. The following independent parametersdescribed the molecular geometry: C-Se and G=C bond lengths;A(CC) (difference between C-C and C=C bond lengths); A(CF)(difference between the C=C and mean C-F bond lengths);C=C-Se, C-C=C and mean C-C-F bond angles.Torsionalangles were also refined as independent parameters: Se-C=C-Se, C-C=C-C and C=C-C-F(7) with 0" corresponding to synorientation. In the final stages of the analysis a tilt of theC, symmetry axes from the direction of the C-C bonds wasintroduced.Crystal Structure Determination of Compound 6.-Orangecrystals were obtained from a solution of compound 6 in CFCl,at room temperature (see above).Crystal data. C,F12Se,, M = 639.91, orthorhombic, spacegroupPbca,a = 10.1920(21),b = 13.0615(20),~ = 22.050(5)81,U = 2935.4(10) pi3, 2 = 8, D, = 2.896 g ern-,, F(OO0) =2334.1 1.Orange crystal, dimensions 0.30 x 0.38 x 0.50 mm,p(Mo-Kor) = 10.02 mm-', h(M0-Ka) = 0.7093 A.Data collection andprocessing. Rigaku AFC 6s diffractometerusing the routine DIFRAC,35 -28 scan mode (28,,, = 45.9"),T = - 170 OC, graphite-monochromated Mo-Ka radiation;2757 unique reflections measured, 1779 with Inet > 2.50(Ine,)used for calculations, absorption corrections made usingDIFABS 35936 (maximum, minimum transmission factors =0.096 921,0.021 194).Structure analysis and refinement. The structure was solved bydirect methods and refined by full-matrix least-squares analysisto residuals of R = 0.061, R' = 0.068 for 1312 reflections[ I 2 2.50(Inet)] (2043 total) and 21 8 parameters. All atoms wereassigned anisotropic thermal parameters.Scattering factorswere taken from ref. 33 and corrected for anomalous dispersion.Additional material available from the Cambridge Crystallo-graphic Data Centre comprises thermal parameters andremaining bond lengths and angles.Computational Methods.-The computations were carriedout with the Gaussian 92 37 program using the 6-3 1 + G* basisset for nitrogen, fluorine and carbon. For selenium a(ECP28MWB) 38 quasi-relativistic pseudo-potential and thefollowing basis set was used: Se, (5s5p)/[3s3p]-DZ + p(extended with a single d function, dexp = 0.338).39 Thegeometries for the FCSeNSeCF+ cation and for the neutralFCSeSeCF were fully optimized at the HF level employingstandard procedure^.^' In addition, the geometries were fullyoptimized at the electron-correlated second-order Merller-Plesset level [MP2(full)].- -AcknowledgementsWe thank Maria Kolonits for electron diffraction experimentalwork.Continuous financial support by the DeutscheForschungsgemeinschaft and the Fonds der ChemischenIndustrie is gratefully acknowledged (T. M. K., H. S.). T. M. K.is also indebted to the Technical University of Berlin (FIP 5/15)and to the Bundesminister fur Bildung und Wissenschaft(Graduiertenkolleg) for support. We also thank the co-operative exchange program between the Technical Universityof Berlin (T. M. K.) and the Budapest Technical University(I. H.) as well as the Hungarian Science Research Foundation(I. H.; OTKA, No. 2103) and the J. Varga Foundation of theBudapest Technical University (K.B. B.) for support, andNATO for support ofthex-ray work (T. M. K., P. S. W.; NATOCRG-920034).References1 T. Chivers, Main Group Chem. News, 1993,1, 6.2 . T. M. Klapotke, in The Chemistry of Inorganic Ring Systems, ed.R. Steudel, Elsevier, Amsterdam, 1992, vol. 14, ch. 20, p. 409.3 M. Bjorgvinsson and H. W. Roesky, Polyhedron, 1991, 10,2353.4 M. Broschag, T. M. Klapotke, I. C. Tornieporth-Oetting and5 R. Wollert, Dissertation, Universitat Marburg, Marburg, 1993.6 R. Wollert, A. Hollwarth, G. Frenking, D. Fenske, H. GoesmannandK. Dehnicke, Angew. Chem., 1992,104, 1216.7 M. Broschag, T. M. Klapotke, A. Schulz and P. S. White, Inorg.Chem., 1993,32,5734.8 M. Broschag, T. M. Klapotke, A. Schulz and I. C. Tornieporth-Oetting, ADUC Chemiedozententagung, A 27, Siegen, 1994.9 G.K. MacLean, J. Passmore, M. N. S. Rao, M. J. Schriver,P. S. White, D. Bethell, R. S. Pilkington and L. H. Sutcliffe, J. Chem.SOC., Dalton Trans., 1985, 1405.10 B. Ayres, A. J. Banister, P. D. Coates, M. I. Hansford, J. M. Rawson,C. E. F. Rickard, M. B. Hursthouse, K. M. Abdul Malik andM. Motevalli, J. Chem. Soc., Dalton Trans., 1992, 3097.11 J. E. Huheey, Anorganische Chemie, Walter de Gruyter, Berlin,1988.12 E. G. Awere, N. Burford, C. Mailer, J. Passmore, M. J. Schriver,P. S. White, A. J. Banister, H. Oberhammer and L. H. Sutcliffe,J. Chem. Soc., Chem. Commun., 1987,66.P. S. White, J. Chem. SOC., Chem. Commun., 1992, 1390.13 A. Davison and E. T. Shawl, Inorg. Chem., 1970,9, 1820.14 J.L. Hencher, Q. Shen and D. G. Tuck, J. Am. Chem. SOC., 1976,98,899271215 D. Christen, H. Oberhammer, W. Zeil, A. Haas and A. Darmadi,16 P. DAntonio, C. George, A. H. Lowrey and J. Karle, J. Chem. Phys.,17 S . Motherwell, University Chemical Laboratory, Cambridge, 1978.18 C. K. Johnson, ORTEP, A Fortran Thermal Ellipsoid Plot Program,Technical Report ORNL-5 138, Oak Ridge National Laboratory,Oak Ridge, TN, 1976.J. Mol. Struct., 1980,66, 203.1976,80,618.19 C. G. Krespan, J. Am. Chem. Soc., 1961,83,3434.20 F. W. McLafferty, Science, 1990,247,925.21 H. Schwarz, Pure Appl. Chem., 1989,61,685.22 J. L. Holmes, Ado. Mass Spectrom., 1989, 11, 53.23 J. L. Holmes, Mass Spectrom. Reu., 1989,8,513.24 J. K. Terlouw, Adv. Mass Spectrom., 1989,11,984.25 D. Schroder, J. Hrusak, I. C. Tornieporth-Oetting, T. M. Klapotkeand H. Schwarz, Angew. Chem., 1994,106,223.26 L. Andrews and P. Hassanzadeh, J. Chem. Soc., Chem. Commun.,1994, 1523.27 R. Srinivas, D. Siilzle, T. Weiske and H. Schwarz, Spectrom. IonProcesses, 1991,107,368.28 R. Srinivas, D. Siilzle, W. Koch, C. H. DePuy and H. Schwarz,J. Am. Chem. Soc., 1991,113,5970.29 I. Hargittai, J. Hernadi and M. Kolonits, Prib. Tekh. Eksp., 1972,239.30 I. Hargittai, J. Tremmel and M. Kolonits, Hung. Sci. Instrum., 1980,50, 31.31 I. Hargittai, J. Hernadi, M. Kolonits and G. Schultz, Rev. Sci.Instrum., 1971,42, 546.J. CHEM. SOC. DALTON TRANS. 199432 B. Andersen, H. M. Seip, T. G. Strand and R. Stolevik, Acta Chem.33 R. A. Bonham and L. Schafer, International Tables for X-Ray34 C. Tavard, D. Nicolas and M. Rouault, J. Chim. Phys. Phys.-Chim.35 E. J. Gabe, Y. Le Page, J.-P. Charland, F. L. Lee and P. S. White,36 Y. Le Page, J. Appl. Crystallogr., 1988,21,983.37 M. J. Frisch, G. W. Trucks, M. Head-Gordon, P. M. W. Gill,M. W. Wong, J. B. Foresman, B. G. Johnson, H. B. Schlegel,M. A. Robb, E. S. Replogle, R. Gomperts, K. Andres, K.Raghavachari, J. S. Binkley, C. Gonzales, R. L. Martin, D. J. Fox,D. J. DeFrees, J. Baker, J. J. P. Stewart and J. A. Pople, GAUSSIAN92, Revision B, Gaussian Inc., Pittsburgh, 1992.38 A. Bergner, M. Dolg, W. Keuchle, H. Stoll and H. Preuss, Mol.Phys., 1993,80,1431.39 J. Andzelm, L. Klobukowski, L. Radzip-Andzelm, Y. Sakai andH. Tatewaki, Gaussian Basis Sets for Molecular Calculations,ed. S . Huzinaga, Elsevier, New York, 1984.40 W. J. Hehre, L. Radom, P. v. R. Schleyer and J. A. Pople, Ab InitioMolecular Orbital Theory, Wiley, New York, 1986.Scand., 1969,23,3224.Crystallography, Kynoch Press, Birmingham, 1974, vol. 4.Biol., 1967, 64, 540.J. Appl. Crystallogr., 1989,22,384.Received 17th March 1994; Paper 4/0 1609