[Tri(tbutyl)phosphine]gold(I) iodide has been shown to add
no iodine. Neither a gold(III) complex nor a gold(I) polyiodide
could be isolated.1 The corresponding gold(I) bromide complex,
[(tBu)3P]AuBr, appears to undergo reaction with equimolar
quantities of Br2 in dichloromethane, but no product could be
[Tri(mesityl)phosphine]gold(I) bromide was found to re-
act with excess bromine in dichloromethane to give
a
product of Au–P cleavage, which was identified as bromo-
tri(mesityl)phosphonium tetrabromoaurate(III). Red crystals
were isolated in 87% yield (mp 217 ◦C with decomposition).
1
crystallized. The H NMR spectrum of the reaction mixture
[(Mes)3P]AuBr + 2 Br2 → [(Mes)3PBr]+[AuBr4]−
shows a signal which has a chemical shift similar to that of
the starting material, but its multiplicity has changed from a
simple sharp doublet (A part of an AnX spin systems) to a
“filled in” doublet generally observed for AnXXꢀAꢀn spin systems.
This result suggested a ligand redistribution which leads to
The [(Mes)3PBr]+ cation has been detected as the parent ion
in the mass spectrum of the product. The compound is stable
in air and readily soluble in polar solvents. The 1H, 13C and 31
P
NMR signals appear in the expected shift regions. The rotation
of the mesityl groups about their P–C bonds is hindered (in
dichloromethane at 25 ◦C) rendering the 2,6-methyl groups
and 3,5-hydrogen atoms inequivalent. This pattern suggests a
propeller or paddlewheel structure.15
+
the formation of {[(tBu)3P]2Au} cations. And indeed, with
an excess of bromine a crystalline solid of the composition
+
{[(tBu)3P]2Au} (Br3)−·(Br2) could be isolated and structurally
characterized (below). The by-products could not be identified,
but it is very likely that the anions [AuBr2]− and [AuBr4]−
required by the mass balance are among the counterions
It appears that in the tri(mesityl)phosphine complex—
with three 2,4,6-trimethyl-phenyl groups instead of three 2-
methylphenyl groups in the tri(o-tolyl)phosphine complex—the
gold(I) center is perfectly protected against oxidative addition of
bromine. The situation resembles that of the tri(tbutyl)phosphine
complex, but the reaction takes a different course. The cation
+
which remain in solution. The {[(tBu)3P]2Au} cation was also
observed in the mass spectra of the products.
+
2 [(tBu)3P]AuBr + excess Br2 → {[(tBu)3P]2Au} (Br3)−·(Br2) +
by-products
+
[(Mes)3P]2Au} features a particularly severe steric conflict ow-
ing to the close proximity of two tri(mesityl)phosphine ligands.16
Therefore, any ligand redistribution giving this cation offers no
relief for a crowded system, and in the presence of a strong
oxidizing agent P–Au cleavage is thus the preferred alternative.
This result shows that with the particularly bulky P(tBu)3
ligand even the smaller and more reactive Br2 molecules (as com-
pared to I2 molecules) have no access to the gold(I) center of the
complex and are hence unable to convert it into a gold(III) center.
[Tri(ortho-tolyl)phosphine]gold(I) bromide was found to
readily undergo oxidative addition of a bromine molecule to
give high yields of an orange gold(III) complex of high stability
(mp 150 ◦C with decomposition).
Crystal and molecular structures
(Trimethylphosphine)gold(III) triiodide, (Me3P)AuI3, has re-
cently been structurally characterized. Its gold atom is in a
square planar environment with only minor differences in the
details of bonding for the cis- and trans iodine atoms.1
[(o-Tol)3P]AuBr + Br2 → [(o-Tol)3P]AuBr3
The same structure has now been found for the components
of the new iodine adduct [(Me3P)AuI3]2·(I2). Crystals of this
phase are monoclinic, space group P21/c. The asymmetric unit
contains two inequivalent (Me3P)AuI3 complexes and one I2
molecule. The geometry of the two complexes is very similar, and
in excellent agreement with that of molecules in the phase which
contains no extra iodine. The molecules centered by Au2 have no
contact with the I2 molecule, and their conformation approaches
quite closely the maximum attainable mirror symmetry: The
dihedral angle I21–Au2–P2–C23 is only −1.3◦ (Fig. 2a).
Although the reaction occurs readily and almost
quantitatively, there is severe steric hindrance clearly discernible
in the product: While the three o-tolyl groups have uninhibited
rotational motion in the gold(I) complex, as indicated by their
solution NMR-equivalence (in dichloromethane at 25 ◦C),
these groups are NMR-inequivalent in the gold(III) complex
(Fig. 1). Three separate signals are observed for the three methyl
groups, and a temperature of ca. 80 ◦C is required (in a higher
boiling solvent like chlorobenzene) to induce coalescence. Two
of the three signals have a larger line-width than the third one
already at room temperature, and as the temperature rises, these
two are subject to coalescence first, before the third one also
becomes involved. It is thus obvious that the rotation of the tolyl
groups about the P–C bonds, and possibly even the rotation
of the phosphine ligand about the Au–P bond are restricted
by the cis-bromo substituents of the square planar gold(III)
complex. The ground state conformation was determined by
the crystal structure analysis (below). It shows that only a minor
rotational barrier has to be overcome in order to make two
of the three o-tolyl groups equivalent, the molecule attaining
mirror symmetry. The third substituent is more strongly affected
in its movement and therefore rotates freely only at higher
temperature leading to virtual C3 symmetry.
A pair of molecules, each centered by Au1 is connected by two
I2 molecules as shown in Fig. 2b. Counting all I–I contacts, a 10-
membered ring is formed around a crystallographic center of in-
version. The I–I–I–I bridges are not linear [I1–I2–I13 134.7(1)◦,
I2–I1–I12ꢀ 177.0(1)◦], and the angles at the iodine◦atoms I12
and I13 are strongly different [Au1–I12–I1ꢀ 128.6(1) , Au1–I2–
◦
˚
I13 98.1(1) ]. The I1–I2 distance [2.729(1) A] is significantly
˚
longer than in orthorhombic crystals of iodine [2.667(1) A], and
ꢀ
˚
the distances I1–I12 = 3.360(1) and I2–I13 = 3.570(1) A are
17
˚
shorter than the sum of two van der Waals radii [3.80 A].
The conformation of the molecules centered by Au1 deviates
strongly from mirror symmetry [dihedral angle I13–Au1–P1–
C13 −19.8(5)◦], but it is interesting to note that the Au–I bond
lengths are not affected by the contacts with the I2 molecules
as witnessed by the distances Au1–I11 2.620(1) and Au1–I13
˚
2.623(3) A.
(Trimethylphosphine)gold(III) tribromide, (Me3P)AuBr3, has
already been structurally characterized.7 Crystals have now
been grown for the 1 : 1 bromine adduct [(Me3P)AuBr3]·(Br2)
(orthorhombic, space group Pnma, Z = 4), in which the complex
molecules are linked via bromine molecules into chains (Fig. 3).
Mainly the two cis-bromine atoms (Br2–Br4ꢀ, Br3–Br5) of the
complexes are engaged in this aggregation, but there is also a
sub-van der Waals contact to the trans-bromine atom (Br1–
Br4ꢀ). Surprisingly, all atoms of a chain except for the C1/C1A
atoms (and ignoring the hydrogen atoms) are accommodated on
a crystallographically imposed mirror plane which is bisecting
Fig. 1 1H NMR spectrum of [(o-Tol)3P]AuBr3 (CD2Cl2, 25 ◦C).
D a l t o n T r a n s . , 2 0 0 5 , 1 9 4 0 – 1 9 4 7
1 9 4 1