Te(SPh)2 [2.406(2) Å],29 Te(SCPh3)2 [2.379(2) Å],30 Me2Te-
(S2CNR2)2, Me2TeX(S2CNR2), Me2TeX[S2P(OR)2], and Me2-
Te(S2COMe)2 (X = Cl, Br or I) [2.484–2.649 Å, the shortest
Te–S for each ligand],31–33 Ph{2,6-(PhS)2C6H3}TeX2,34 and
[Cu(S6)2(TeS3)2]6Ϫ, [Mn(TeS3)2]2Ϫ, [Cu(TeS3)]Ϫ, [Ag(TeS3)]Ϫ, and
[Au(TeS3)2]2Ϫ [2.327–2.395 Å].35–38 The Te–S bonds of 3d are
apparently longer than the reported ones, due to the three-
center four-electron bond character of the X–Te–X bonds in
Ar2TeX2. The X–Te–X (X = Cl, Br, I, or SCN) bond angles are
very close to 180Њ. It is generally accepted that the d-orbital
contribution in these elements is negligible. The Te–halogen
bonds in Ar2TeX2 also cannot be dp-hybrid orbital bonds.
The C–Te–C bond angle in compounds 3a–3d is narrower
than the ideal of 120Њ for trigonal bipyramidal, as expected
from VSEPR theory, but it is much wider [107.6(2)–104.37(9)Њ]
than those reported for Ar2TeX2 [101.1–94.2Њ],7–17 reflecting the
steric requirements of two R-groups (Fig. 1).
We have been interested in the roles of 2,6-methoxy groups
in R-derivatives of a variety of elements. The crystal structures
of R3PSؒH2O and R3PSeؒH2O showed that all the P ؒ ؒ ؒ O
interatomic distances [2.794–3.190 Å] were shorter than the
sum of the van der Waals radii [3.32 Å], implying a direct
electron-donative interaction from the methoxy oxygen to the
phosphorus atom.39 This concept has been proposed by Wood
et al. for the crystal structure of benzyl(2-methoxyphenyl)-
diphenylphosphonium bromide, and was supported by energy
minimization calculations.40 The oxygen–tellurium distances of
3a–3d [2.880–3.323 Å] are much longer than a covalent Te–O
bond observed for (Ph2TeN3)2O [1.972(3) Å],41 for 2,5-diphenyl-
1,6-dioxa-6a-tellurapentalene [2.142(4) and 2.130(4) Å],42 and
for its dibromide [2.135(9) and 2.171(9) Å],42 but they are sig-
nificantly shorter than the sum of van der Waals radii [3.60 Å]
of O and Te.43 Accordingly, it is expected that there is a weak
co-ordinative interaction between the tellurium atom and 2,6-
methoxy oxygens. We believe the interaction is attractive since
the Te–C bond distances in 3a–3d are somewhat shorter than
those reported for phenyl derivatives.7–17 If so, the geometry
of 3a–3d with such Te ؒ ؒ ؒ O secondary intramolecular co-
ordination is pentagonal bipyramidal. This may also explain
why no Te ؒ ؒ ؒ X intermolecular secondary bonds are observed.
DMSO-d6 showed only one resonance possibly due to partial
dissociation of thiocyanate ions. The coalescence temperature
(Tc) of 3d could not be observed in DMSO-d6, but was observed
in CDCl3 (Fig. 2) and in CD3CN.
The activation energies, ∆G‡, for free rotation were calcu-
lated from these temperature-dependent 1H NMR spectra, and
decreased in the order of 3a (90 kJ molϪ1 in DMSO-d6) > 3b
(80 kJ molϪ1 in DMSO-d6) > 3d (≥65 kJ molϪ1 in CDCl3) > 3c
(60 kJ molϪ1 in CDCl3) ≥ 3d (59 kJ molϪ1 in CD3CN). It is of
interest that the activation energy decreased as the halogen
became heavier. This finding is understood by assuming that the
bond distance increased with increasing radius of the halogen
atom, resulting in a smaller rotation barrier. Another possibility
is that there is a co-ordinative interaction between the methoxy
oxygen and TeIV, upon which the positive character of TeIV must
decrease in the order of electronegativity of the halogen atom,
Cl > Br > I, giving a smaller rotational barrier for the R-group.
Consistent with the bond length order of Te–X (Table 3), ∆G‡
of 3d was smaller than those of 3a and 3b. It is noted that the Tc
of 3d varies with the solvent polarity: CDCl3 (≥60 ЊC) > CD3CN
(26 ЊC) > DMSO-d6 (not observed). Although we could not
measure ∆G‡ for 3a and 3b in CDCl3 because of the low boiling
point of the solvent, they must be larger in CDCl3 than
in DMSO-d6. These solvent and halogen effects can best be
understood by the dissociation of the halide ions.
The 13C NMR spectra of compounds 1a,1c,1d,1e showed five
resonances attributable to R-group carbons (Table 2). While 3c
showed analogous resonances, 3a,3b,3d showed six resonances
for the phenyl carbons and two for the methoxy carbons. The
chemical shifts of 1a,1c,1d,1e are almost identical reflecting
the cationic character of these compounds, while considerable
influences of the anionic groups are observed among 3a–3d
reflecting the covalent character. These results are consistent
with the 1H NMR spectra as mentioned above.
The 125Te NMR spectra of some [R3Te]X compounds have
been reported.44 Although a considerable effect of the counter
ions on the chemical shift has been observed for [(alkyl)3Te]X,44
1a,1c,1d,1e showed almost identical chemical shifts at δ 553
1
(Table 2). The chemical shifts of 2 and 3a were observed at
δ 236 and 789, respectively, which may be compared with those
of Ph2Te (δ 680) and Ph2TeCl2 (δ 917).28 The upfield shifts
of those phenyl derivatives are consistent with the electron-
donating character (or co-ordination) of ortho-methoxy groups
in R group. The chemical shifts of 3a,3b,3d fall in the range
of δ 713–789. We could not observe any resonance of 3c due to
the poor solubility.
1H, 13C, and 125Te NMR spectra
1
The H NMR spectra of R2E, REER, R2EO, [MeR2E]ϩ, and
[R2EORЉ]ϩ (E = S, Se or Te) show a triplet due to the 4-proton,
a doublet due to the 3,5-protons, and a very sharp singlet due
to 2,6-dimethoxy protons on the R group, respectively,1 as
observed for compound 1a (Table 1). Identical spectra were
observed for 1aЈ and 1c–1e in accord with the ionic character
of these compounds.
Experimental
General
In contrast, the 1H NMR spectrum of compound 3a in
CDCl3 showed two singlets due to the 2,6-methoxy protons as
well as two doublets due to the 3,5-protons at 25 ЊC (Table 1).
The result is consistent with the structure of 3a in the solid
state, in which each R group is non-symmetrical. It further
indicates that the rotation of R–Te bonds is restricted in solu-
tion. The rotational barrier must occur between the R group
and chlorine atoms rather than between two R groups, since the
spectrum of 3b showed two broad singlets due to 2,6-dimethoxy
protons and that of 3c a normal sharp singlet at 25 ЊC.
The spectrum of 3c showed two signals at Ϫ50 ЊC as shown
in Fig. 2 (c) (coalescence temperature, Ϫ29 ЊC). Analogous
spectra were obtained for DMSO-d6 solutions of both 3a and
3b at 30 ЊC, and were also temperature-dependent (Fig. 2 (a)
and (b)) (coalescence temperatures, 87 and 51 ЊC, respectively).
Interestingly, the spectrum of 3a in D2O showed only one
singlet due to the 2,6-methoxy groups even at 25 ЊC. It is
expected that the chloride ions in 3a dissociate partially in
D2O. The spectrum of 3d measured in CDCl3 or CD3CN
showed two methoxy proton resonances, while that in
1H NMR spectra were recorded for solutions using a JEOL
model JNM-GX270 spectrometer. 1H chemical shifts were
referenced to internal TMS (δ 0.00) in CDCl3 or DMSO-d6
(δ 2.49) or DMSO (δ 2.71) in D2O. 13C and 125Te NMR spectra
were recorded for solutions in CDCl3 using a JEOL model
JNM-ECP500 spectrometer. 13C NMR chemical shifts were
referenced to internal TMS (δ 0.00), 125Te to external diphenyl
ditelluride (δ 450). IR spectra were recorded for Nujol
mulls using a Shimadzu FTIR-8300 spectrophotometer. The
preparations of R2Te, R2TeOؒ(0.67)H2O, and RTeTeR have
been reported elsewhere.1
Preparations
Tris(2,6-dimethoxyphenyl)telluronium chloride hydrate 1a. A
suspension of RLi was prepared from 1,3-dimethoxybenzene
(6.0 cm3, 46 mmol), a 15% hexane solution of n-butyllithium
(26 cm3, 42 mmol), and a catalytic amount of N,N,NЈ,NЈ-
tetramethylethylenediamine (TMEDA) (0.15 cm3) under
argon.1 To this was added dropwise a solution of TeCl4 (2.70 g,
3496
J. Chem. Soc., Dalton Trans., 2000, 3493–3499
Asahara, Masahiro
