Aurolysis of α-C-deprotonated group 6 aminocarbene and thiocarbene complexes with Ph3PAu+

Article abstract of DOI:10.1039/b607613k

Deprotonated Fischer-type aminocarbene complexes, (CO) ₅MC(NR₂)CH₃ (M = Cr or W; R = Me or propyl), react with Ph₃PAu⁺ by metal group substitution - (CO) ₅M for Ph₃PAu⁺ - and attachment of the extricated M(CO)₅ to the deprotonated methyl group. (The products may also be seen as aminovinylgold compounds coordinated to M(CO)₅ moieties.) DFT calculations at the B3LYP level of theory using model compounds indicate a clear preference of the gold unit for central C to terminal coordination in the ligand [NMe₂CCH₂]^-, whereas the Cr(CO)₅ has a 7 kcal mol^-1 preference for C(vinyl) coordination compared to N-coordination. In related thiocarbenes, the sulfur donor atom should be the preferred point of attachment for the metal carbonyl unit. The latter prediction is borne out in practice, and in the three products isolated, including Ph₃PAu{C(CH₂)SPh}Cr(CO)₅ in a mixed crystal with [Ph₃PAuSPh]Cr(CO)₅, precisely this coordination mode is present. The latter component of the mixed crystal has also been prepared independently of the vinyl one. The Royal Society of Chemistry 2006.

Full text of DOI:10.1039/b607613k

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
Aurolysis of a-C-deprotonated group 6 aminocarbene and thiocarbene
complexes with Ph3PAu+†
Helgard G. Raubenheimer,*a Matthias W. Esterhuysen,a Gernot Frenking,*b Alexey Y. Timoshkin,c
Catharine Esterhuysena and Ulrike E. I. Horvatha
Received 31st May 2006, Accepted 4th July 2006
First published as an Advance Article on the web 21st July 2006
DOI: 10.1039/b607613k
=
Deprotonated Fischer-type aminocarbene complexes, (CO)5M C(NR2)CH3 (M = Cr or W; R = Me or
propyl), react with Ph3PAu+ by metal group substitution – (CO)5M for Ph3PAu+ – and attachment of
the extricated M(CO)5 to the deprotonated methyl group. (The products may also be seen as
aminovinylgold compounds coordinated to M(CO)5 moieties.) DFT calculations at the B3LYP level of
theory using model compounds indicate a clear preference of the gold unit for central C to terminal
coordination in the ligand [NMe2CCH2]−, whereas the Cr(CO)5 has a 7 kcal mol−1 preference for
C(vinyl) coordination compared to N-coordination. In related thiocarbenes, the sulfur donor atom
should be the preferred point of attachment for the metal carbonyl unit. The latter prediction is borne
=
out in practice, and in the three products isolated, including Ph3PAu{C( CH2)SPh}Cr(CO)5 in a mixed
crystal with [Ph3PAuSPh]Cr(CO)5, precisely this coordination mode is present. The latter component of
the mixed crystal has also been prepared independently of the vinyl one.
We particularly focussed on a-deprotonated, Fischer-type
Introduction
aminocarbene complexes, since their advantages in preference
to alkoxycarbenes in organic synthesis are well documented.3
Thiocarbene complexes, due to their greater lability, are less well
studied.4 To our knowledge, no reactions of nucleophiles derived
from amino- or thiocarbene complexes with any electrophilic
transition metal fragments are known. The potential wealth of
unknown organometallic products that can be obtained by em-
ploying such nucleophiles as organic synthons in organometallic
conversions, therefore, remains unexplored.
The results of our investigation are presented below and clearly
indicate the important role played by the carbene substituents
in the novel isolobal metal substitution at the carbene carbon
atom and subsequent coordination of the group 6 metal carbonyl
moiety.
We have previously shown that anionic Fischer-type carbene
1
−
=
complexes, [(CO)5M C(X)R ] , in which the negative charge
formally resides on X (X = O or NR2, R1 = Ph, R2 = H or
Me; M = Cr or W), react with the fragment Ph3PAu+ to furnish
unusual acyl or imidoyl complexes of gold(I). Since this metal
exchange is related to the hydrolysis of such Fischer-type carbene
complexes through the isolobal analogy, we termed it ‘aurolysis’.
Furthermore, when starting with N-deprotonated aminocarbene
complexes, the expelled (CO)5M unit is firmly N-coordinated.1
On the other hand, a-deprotonated alkoxycarbene complexes,
=
[(CO)5M C(OR)CH2]Li, undergo a metal substitution (group 6
metal for gold) with the same gold phosphine cation to furnish
=
stable vinylether gold complexes, Ph3PAuC(OR) CH2 (R = Me
or Et,) that are unsymmetrically p-coordinated to (CO)5M.2 In
an attempt to further extend the utilisation of anionic carbene
complexes as useful preparative synthons towards gold(I), the
following questions were addressed: (i) would similar aurolysis
reactions occur with a-deprotonated aminocarbene complexes
and how would the carbonyl fragment then be accommodated?
(ii) Could the formation of the product be substantiated by
quantum mechanical calculations? (iii) What predictions are made
for thiocarbene complexes and how are these borne out in practice?
Results and discussion
A. Reactions with aminocarbene complexes
The expected bimetallic, group 6-coordinated aminovinyl gold
complexes, {Ph3PAuC(NR2)CH2}M(CO)5 (1–4, Scheme 1), were
prepared in a similar fashion to the previously reported p-
bonded vinyl ether compounds of gold, i.e. by deprotonation and
transmetallation of Fischer-type carbene complexes.2
aDepartment of Chemistry and Polymer Science, Stellenbosch University,
Private Bag X1, Matieland, 7602, Stellenbosch, South Africa. E-mail: hgr@
sun.ac.za; Fax: +27 21 808 3849; Tel: +27 21 808 3850
Product separation of oily, dark-yellow reaction mixtures (52–
68% yields after separation) was carried out by low temperature
(−15 ◦C) silica gel column chromatography with pentane–diethyl
ether (10 : 1) as eluent, followed by crystallisation from a diethyl
ether solution layered with pentane at −20 ◦C. The same ‘hy-
drolysis mechanism’ of Casey5 and Bernasconi6 now translating
as aurolysis, could again be invoked here to understand the
metal exchange occurring at the carbene carbon. However, further
investigation into the structures of 1–4 indicated a product in
which the (CO)5M moiety is bonded exclusively to the carbon a to
bFachbereich Chemie, Philipps-Universita¨t Marburg, Hans-Meerwein-Str.,
D-35032, Marburg, Germany. E-mail: frenking@chemie.uni-marburg.de;
Fax: +49 6421 282 5566; Tel: +49 6421 282 5563
cSt. Petersburg State University, Department of Chemistry, Inorganic
Chemistry Group, University pr. 26, Old Peterhoff, 198504, St. Petersburg,
Russia
† Electronic supplementary information (ESI) available: Figure showing
disorder in 7/7ꢀ and preparation, physical data and crystallographic results
for pure 7ꢀ. See DOI: 10.1039/b607613k
4580 | Dalton Trans., 2006, 4580–4589
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
Scheme 1
the original C(carbene) and not to a formed vinyl double bond as
in the alkoxyvinyl compounds of gold. This essentially alkyl-type
coordination to both pentacarbonyls affords remarkably stable
products that could withstand column chromatography on SiO2.
Spectroscopic characterisation
Infrared spectroscopy. Pentane solutions of compounds 1–
4 exhibit the typical four absorption band m(CO) spectra of
M(CO)5L compounds. Compared to the corresponding IR ab-
sorptions for the already mentioned alkoxyvinyl complexes, the
bands are now in general shifted approximately 15 cm−1 to lower
frequencies. Such a shift is symptomatic of more back-donation
from the metal atom to the p*-orbitals of its CO ligands indicating
a greater partial negative charge on the M(CO)5 fragments, greater
polarisation of the vinyl-group 6 metal coordination in these
Fig. 1 ORTEP view of 1 (hydrogen atoms on phenyl rings omitted for
clarity) showing the numbering scheme. Ellipsoids are shown at the 50%
probability level.
2
complexes and less g -p-bond interaction by the metal. The broad
degenerate E-vibration in the new complexes, compared to a
lifted degeneracy in the alkoxycarbene complexes, could indicate
a decreased steric influence by the remainder of the gold complex
as M(CO)5 settles in a more terminal position.
the 13C chemical shifts for the terminal carbons in 1–4 appear
in fact quite close to the aliphatic region of the 13C spectrum
(d 36.5, 40.1, 36.0 and 42.1). These upfield shifts both in the
1H and 13C spectra suggest that the terminal ‘vinyl’ CH2-carbon
Mass spectrometry. Molecular ion peaks with intensities ap-
proximately 50% relative to base peaks were seen for each of the
complexes 1–4 (m/z 721, 750, 852 and 881). The homoleptic
rearrangement product, (Ph3P)2Au+, the cation Ph3PAu+, and
the uncoordinated (dialkylaminovinyl) triphenylphosphinegold
moieties (m/z 529 for 1 and 3, and 558 for 2 and 4) were clearly
identified for each complex and the latter formed the base peaks
for 1 and 4.
1
atoms display less s-character in the new compounds. The H
and 13C NMR resonances are, however, still situated somewhat
downfield compared to literature values for methylene groups
directly bonded to negatively charged W(CO)5 fragments.7 The
1
13C{ H} resonances for the gold-coordinated carbon atoms in 1–4
could be easily assigned due to their characteristically large 2JC–P
coupling constants (117.6–134.2 Hz). They appear between d 207.9
and 218.3, indicating carbene-like character.
1
1
NMR spectroscopy. The 1H, 13C{ H} and 31P{ H} NMR spec-
troscopic data for complexes 1–4 are reported in the Experimental
section. As with previously characterised alkoxyvinyl complexes
of this class,2 the 1H NMR spectra are characterised, in particular,
by two distinctive signals, each integrating for a single proton and
representing the two diastereotopic terminal CH2 protons. The
exact origin of the chirality around the CH2 moiety is unknown.
It could be as a result of steric hindrance or even by a weak Cr–
C(7) (compare Fig. 1) interaction. Resonances for one of these
appear, due to weak long range cis-4JP–H coupling, as broadened
singlet or narrow doublet resonances at d 1.93, 1.97, 2.51 and
2.50, whereas those for the other proton are distinct doublets, due
Owing to restricted rotation about the C(coordinated)–N bond
(vide infra) the proton resonances for the NCH3 groups in each of
1 and 3 appear as two clearly separated signals at d 2.79 and 3.45,
and d 2.79 and 3.42, respectively. In the 13C NMR spectra a single
broadened resonance (d 29.2 and 29.4) represents both carbon
atoms in each component. In 2 and 4 every methylene proton of
the ethyl groups is separately represented as a quartet of doublets
in the 1H NMR spectrum. The attached CH3 groups resonate as
doublets of doublets (appearing as pseudo-triplets) at d 1.12, 1.13
and d 1.20, 1.29, respectively. In the 13C spectra of 2 and 4 the
signals for the cis and trans methylene carbon atoms, like those
for the NCH3 groups in 1 and 3, appear as broad resonances (d
25.2 and 27.5) whereas two clearly separated signals each for the
terminal CH3 groups are observed (d 12.5 and 13.6 in 2, d 12.9 and
14.0 in 4).
4
to more efficient JP–H coupling, at d-values of 2.55, 2.52, 3.00
and 2.91. Compared to the related chemical shifts of comparable
protons in previously characterised alkoxyvinyl complexes (in
which the vinyl double bond is more apparent2), these signals
are shifted approximately 0.2 and 1.2 ppm upfield. Similarly,
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 4580–4589 | 4581
©

View Article Online
◦
˚
Table 1 Selected bond lengths (A) and angles ( ) with s.u.s in parenthesis
for complex 1
Au–C(7)
Cr–C(1)
Cr–C(7)
Au–P
C(1)–C(7)
C(7)–N
2.067(5)
2.337(6)
2.947(5)
2.295(1)
1.406(8)
1.348(7)
Cr–C(1)–C(7)
C(7)–Au–P
Au–C(7)–C(1)
Au–C(7)–N
C(1)–C(7)–N
C(7)–N–C(8)
100.9(4)
174.9(1)
118.3(4)
118.7(4)
122.6(5)
122.3(5)
1
The 31P{ H} NMR spectra of 1–4 exhibit singlet peaks, with
chemical shifts between 38.0 and 41.9 ppm for the PPh3 moiety
coordinated to the gold atom.
X-Ray crystal structure determination of complex 1
The low-temperature (178(2) K) crystal and molecular structures
of complex 1 (Fig. 1) were determined by X-ray diffraction
techniques. The molecular structure clearly shows a formal
triphenylphosphine gold(I) moiety coordinated in an unusual
fashion to a square pyramidal Cr(CO)5 fragment. Selected bond
lengths and angles are listed in Table 1.
Of greatest interest in the structure of 1 is the coordination of the
pentacarbonylchromium unit to the gold complex. In contrast to
a related product, referred to above, in which an alkoxyvinylgold
complex becomes p-coordinated to a Cr(CO)5 fragment, a distance
2
˚
˚
of 2.947(5) A (previously already relatively long at 2.629(7) A) be-
tween Cr and C(7) now completely rules out such a bonding mode.
Furthermore, a relatively short C(coordinated)–N separation of
˚
˚
1.348(7) A, a relatively long C(1)–C(7) bond length (1.406(8) A),
tetrahedral coordination around C(1) (hydrogen positions deter-
mined from the difference Fourier map) and planarity around
˚
C(7) (greatest deviation from planarity 0.041(4) A for C(7)),
demonstrate the important contribution made by the zwitterionic
resonance structure in Scheme 2. Such a structure also corresponds
to the results of the NMR investigation discussed above.
ꢀ
˚
Fig. 2 Calculated geometries of 1M and 1M . Distances in A, angles
in ◦
.
The vinyl amine ligand might also bind to the M(CO)5 moiety
via the nitrogen atom of the amine group. Therefore, we also
calculated the isomer of 1M wherein the ligand is bonded to
chromium through its nitrogen atom rather than the vinyl group
(1Mꢀ, Fig. 2(b)). The calculations predict that this isomer is 7.2 kcal
mol−1 higher in energy (electronic) than 1M. The terminal carbon
atom is softer than nitrogen and thus a better donor towards the
relatively soft Cr(0) in Cr(CO)5.
Scheme 2
Quantum chemical calculations
In order to gain insight into the bonding situation in 1, we carried
out DFT calculations of the model compound 1M in which the
PPh3 ligand of 1 is replaced by PH3. Fig. 2(a) shows the optimised
geometry and the most important bond lengths and angles of 1M.
The calculated values of 1M are very similar to the experimental
data of 1. The theoretical value for the C(1)–C(7) distance is
1.408 A, which is in excellent agreement with the experimental
value of 1.406(8) A. In accord with the experimental structure,
the calculations also give rather different distances for Cr–C(1)
We performed further calculations in order to establish whether
the deprotonated alkyl(amino)carbene (or aminovinyl anion,
Scheme 3) ligand binds more strongly through the terminal or
through the central carbon to the metal fragments Cr(CO)5 and
+
Au(PH)3 . The geometry optimisations of the two isomeric forms
˚
of the gold complex yielded only one minimum energy structure,
[(PH3)Au–C(NMe2)(CH2)], with the gold atom bonded to the
central carbon atom. Two isomers of the chromium complex were
found as energy minima, i.e. for both [(CO)5Cr{C(NMe2)CH2}]−
˚
˚
˚
(2.449 A) and Cr–C(7) (3.065 A), indicating much stronger
interaction of chromium with C(1) than with C(7). We conclude
that the bonding interaction between the coordinated vinyl ligand
1
and chromium in 1M and 1 is best described as g -bonding of
a terminal carbanion rather than g -bonding of an olefin. This
agrees with calculations for aminocarbenes coordinated to iron.8
2
Scheme 3
4582 | Dalton Trans., 2006, 4580–4589
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
and [(CO)5Cr{CH2 C(NMe2)}]−. The former species, where the
metal atom is bonded to the central carbon, is 19.3 kcal mol−1
lower in energy than the latter form where the ligand is terminally
bonded. It follows that the aminovinyl ligand clearly binds more
strongly through the central carbon atom. The fact that the
chromium unit is prepared to be terminally coordinated, combined
with gold’s preference for the internal carbon atom, leads to the
observed products.
The results above suggest that the relative stability of the isomers
M and Mꢀ could change if the amino group in 1 is replaced by a
better donor, such as a thiol group. In order to test this hypothesis
the related isomers of the thiocarbene model complexes 5M and
5Mꢀ were also calculated. The DFT calculations indeed predict
that the vinyl thiocarbene ligand now binds more strongly through
the sulfur atom (isomer 5Mꢀ, Fig. 3(b)) than through the vinyl
group (isomer 5M, Fig. 3(a)). The former isomer is 6.6 kcal mol−1
lower in energy than the latter. This means that the relative stability
of the isomers 5M and 5Mꢀ is opposite to that of 1M and 1Mꢀ.
B. Reactions involving thiocarbene complexes
To establish whether Cr(CO)5 bonding to sulfur occurs after
aurolysis as theoretically predicted, similar reactions to those
above were carried out using deprotonated thiocarbene com-
pounds (Scheme 4). Similarly to the only other structurally known
7
=
alkyl-thiocarbene complex, (CO)5Cr C(SPh)Me, the methyl and
propyl groups in the thiocarbene precursor to 69 also appear in
the syn configuration. The comparable bond distances in the two
compounds are almost identical.
Since only compound 6 and a modified version of 7 (vide infra)
could be isolated in pure form, we are thus not presenting a
complete synthetic protocol here. Nevertheless, physical charac-
terisation (including a crystal and molecular structure determi-
nation involving 7) unambiguously indicate that the theoretical
predictions have been borne out in practice.
After reaction with PPh3AuCl and work-up as described before
for the aminocarbene complexes, column chromatography yielded
product 5 as a yellow oil and 6 as a microcrystalline solid. Yellow–
brown platelet crystals containing 7 were later obtained from a
concentrated diethyl ether solution of the viscous product layered
with pentane at −20 ◦C. As discussed later, these crystals were
subjected to an X-ray investigation and shown to consist of a
combination of two complexes, 7 and 7ꢀ, crystallising together
(Scheme 5).
Scheme 5
Unexpectedly, in each of the attempted syntheses of 5–7 another
product identified as Ph3PAuBu, 8, was formed. Its formation
in significant amounts in these reactions (5–15% yield based on
Au) is probably related to the fact that the thiocarbene starting
materials used here are the most unstable carbene complexes
employed in this study. Some measure of decomposition of the
ꢀ
˚
Fig. 3 Calculated geometries of 5M and 5M . Distances in A, angles
in ◦
.
Scheme 4
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 4580–4589 | 4583
©

View Article Online
thiocarbene starting complexes before and during the synthe-
sis of 5–7 was frequently observed. The thiocarbene starting
complexes are also difficult to isolate in analytically pure form
atom. This coupling constant is more than double the size of the
one-bond JP–C coupling (44.7 Hz) observed for the Cipso atoms in
the PPh3 residue.
=
and (CO)5Cr C(SPh)Me, for example, was used in situ as
Infrared spectroscopy. The IR spectra of 5 and 7/7ꢀ (pentane)
prepared, thus also explaining the concomitant formation of 7ꢀ,
Ph3PAuS(Ph)Cr(CO)5. Metal-containing sulfides coordinated to
group 6 metal carbonyls are rare10,11 and we thus also prepared
and characterised 7ꢀ independently by reacting Ph3PAuSPh12 with
(CO)5Cr(thf) (obtained by UV irradiation of Cr(CO)6 in thf),
eqn (1).
(2)
display only the three usual absorption bands (A1(1), E, A1
)
associated with complexes of the type M(CO)5L, in which the
degeneracy of the E vibrational mode has not been lifted in the
carbonyl region. No differentiation between 7 and 7ꢀ is possible in
the spectrum of crystals of 7/7ꢀ and the absorptions approximately
match those of pure 7ꢀ.
As discussed for the IR spectra of 1–4, the most informative IR
absorption mode for complexes of the type Cr(CO)5L is the A1
(2)
(1)
peak assigned to the symmetric stretching of the single terminal
CO ligand trans to the ligand (L) coordination. In the IR spectra of
5–7/7ꢀ these peaks occurring between 1917 and 1921 cm−1 appear
blue shifted (7–27 cm−1) compared to the same COtrans vibrations
in complexes 1–4, indicating that the ligands in 5–7/7ꢀ have smaller
rdonor/pacceptor ratios than the C-coordinated aminovinylgold groups
in 1–4.
Of note is that the remaining m(CO) absorption bands, in
particular the sharp and well defined A1(1) and B1 bands measured
for 5–7/7ꢀ, are also significantly shifted to higher energy compared
to their positions in the IR spectra of the CH2-coordinated
compounds 1–4. We conclude therefore that in 5–7/7ꢀ much less
electron density is transferred to the Cr(CO)5 unit than in 1–4,
indicating coordination through the sulfur rather than the terminal
vinyl group.
In contrast to most other known neutral aryl sulfide complexes
of pentacarbonylchromium,13 7ꢀ is relatively stable and can even
be handled in air.
During the first step of the preparation of 5–7, a slight excess
of BuLi remaining after the deprotonation of the thiocarbene
complexes, is easily foreseeable. This reacts with Ph3PAuCl, added
in the second step of the synthesis, to form 8 through a simple
transmetallation reaction. Since 8, although known,14 has not been
well characterised before, its spectroscopic and physical data are
reported here.
Spectroscopic characterisation of products 5–8
1
1
NMR spectroscopy. The 1H, 13C{ H} and 31P{ H} NMR
spectroscopic data for the isolated products 5–8 are reported in the
Experimental section. The 1H NMR spectra exhibit two distinctive
signals, each integrating for a single proton, representing two
chemically non-equivalent vinyl protons. The cis and trans-vinyl
proton resonances in 5–7 appear in the usual region for protons
in uncoordinated vinyl groups, at d 5.38–5.64 (cis-Hvinyl) and
The fact that clearly-resolved, non-degenerate vibrational
modes originating from the E vibrational mode are only observed
in the m(CO) absorption bands of 6, is a testimony to the greater
steric bulk exhibited by the nPrS group in this molecule compared
to 5, 7 and 7ꢀ, which contain MeS and PhS groups and hence
display a single, two-fold degenerate E vibrational mode.
1
5.76–6.07 (trans-Hvinyl). Similarly, the 13C{ H} resonances for the
Mass spectrometry. The positive-ion FAB mass spectra of
complexes 5–7/7ꢀ were recorded in an m-nitrobenzyl alcohol
matrix. The spectra mainly show the characteristic matrix peaks
(m/z 154 and 307) or the Ph3PAu+ (m/z 459) or (Ph3P)2Au+
(m/z 721) decomposition and fragment ions as peaks of highest
intensity. No molecular ion, or other sensible fragmentation peaks
were observed for 5–7/7ꢀ. Pure 7ꢀ, on the other hand, fragments
beautifully from the molecular ion at 760, with the consecutive
loss of five COs and then Cr and SPh. The conventional EI-mass
spectrum of 8 is illustrated in Scheme 6 (relative intensities in
parentheses).
terminal CH2-vinyl carbon atoms in 5–7 have chemical shifts at
d 121.7, 125.4 and 125.7, in the aromatic region of the 13C NMR
spectrum. Resonances of the carbons coordinated to gold in 5–
7 could be easily identified on account of their characteristically
large 2JC–P coupling constants (125.2–129.4 Hz) at d 181.9, 182.4
and 180.8, respectively.
The positions of the chemical shifts reported here for the vinyl
groups in 5–7 (the latter clearly visible in the spectrum of 7/7ꢀ)
strongly suggest that the vinyl thioether moieties are not coor-
dinated to Cr(CO)5 fragments through their vinyl functionality,
but rather through their sulfur atoms. Further evidence for this
assumption is presented in the discussion of the infrared spectra of
5–7/7ꢀ and the X-ray crystallographic investigation of 7/7ꢀ below.
The resonances for the phenyl groups in 7ꢀ could be assigned
completely (ESI†). Of particular interest in the 13C NMR spectrum
of 8 is the extremely efficient two-bond JP–C coupling (95.4 Hz) of
the gold-coordinated CH2 group with the 31P atom across the gold
X-Ray crystal structure determination of complex 7/7ꢀ and 7ꢀ
The low-temperature (177 K) crystal structure of the co-
crystallised complex mixture 7/7ꢀ, was determined by X-ray
diffraction techniques. The molecular structure of 7 (Fig. 4) clearly
shows, as proposed by the NMR and IR spectroscopic data,
Scheme 6
4584 | Dalton Trans., 2006, 4580–4589
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
◦
˚
Table 2 Selected bond lengths (A) and angles ( ) with s.u.s in parenthesis
for 7 in 7/7ꢀ
Au(2)–C(11)
C(11)–C(12)
C(11)–S(2ꢀ)
Au(2)–P(2)
S(2ꢀ)–Cr(2ꢀ)
S(2ꢀ)–C(241)
Au(1)–Au(2)
2.03(2)
1.32(3)
1.81(2)
2.275(2)
2.457(5)
1.79(2)
3.197(1)
C(11)–Au(2)–P(2)
Au(2)–C(11)–C(12)
Au(2)–C(11)–S(2ꢀ)
Cr(2ꢀ)–S(2ꢀ)–C(11)
C(11)–S(2ꢀ)–C(241)
168.7(7)
128.0(18)
112.7(13)
118.0(8)
102.1(9)
◦
˚
Table 3 Selected bond lengths (A) and angles ( ) with s.u.s in parenthesis
for 7ꢀ in 7/7ꢀ
Au–S
S–C
S–Cr
Au–P
2.326(2), 2.373(6)
1.787(8), 1.77(4)
2.502(2), 2.490(8)
2.258(2), 2.275(2)
S–Au–P
Au–S–C
Au–S–Cr
Cr–S–C
171.11(1), 170.3(2)
100.2(2), 107.6(10)
100.9(1), 108.3(2)
109.4(3), 109.1(10)
Fig. 4 ORTEP view of 7 (hydrogen atoms omitted for clarity) showing
the numbering scheme. Ellipsoids are shown at the 50% probability level.
bond lengths and angles for each of the crystallographically unique
fragments in the crystal structure are listed in Tables 2 and 3.
Despite the rather inaccurate bond lengths and angles obtained
for 7 as a result of the partial site occupancy found for this part
of the structure, the connectivity in 7 could be established beyond
any doubt. The poor accuracy of the bond lengths and angles in
the vinyl group of 7 does, however, not strictly allow comparison
of these to other literature examples of uncoordinated vinyl
ethers/amines/thioethers. The accurate bond lengths determined
˚
for the Au(2)–P(2) bond [2.275(2) A] as well as the bond lengths
and angles in the PPh3 moiety of 7 (and 7ꢀ) are normal.15
Although present in only small quantities in the crystal, it
appears that 7 is stabilised by the formation of an intramolecular
face-to-edge phenyl interaction between the PhS unit and one of
˚
the triphenylphosphine phenyl rings [H(226) · · · C(241) 3.047 A;
˚
H(226) · · · C(242) 3.048 A]. A further effect of this interaction
could be the large Cr(2A)–S(2A)–C(11) angle [118.0(8)◦], although
this could also be related to steric repulsion between the Cr(CO)5
fragment and the vinyl group.
Fig. 5 ORTEP view of the full site-occupancy molecule of 7ꢀ (hydrogen
atoms omitted for clarity) showing the numbering scheme. Ellipsoids are
shown at the 50% probability level.
A C(11)–Au(2)–P(2) angle of 168.7(7)◦ [bent towards Au(1)],
although poorly determined, describes the distortion of the usually
linear gold(I) coordination mode as a result of the relativistically
enhanced16 Au–Au interaction between Au(1) and Au(2). A
similar, and more certain, description of the distortion of Au(1)
toward Au(2) is given by the S–Au–P angles [171.1◦1(1) and
170.3(2)◦] in 7ꢀ deviating significantly from the ideal 180 .
We also embarked on a crystallographic investigation of the
pure 7ꢀ prepared independently. All the bond distances and angles
(compare ESI†) are identical (within the limits of error) to those
in 7ꢀ in the structure discussed above.
The structure of the uncoordinated complex Ph3PAuSPh, has
been reported in the literature no less than four times.12,17 All four
reported structures are isostructural and consist of dimers of the
compound that are linked, in a similar fashion to the structure
reported above, by weak aurophilic interactions [average Au–Au
coordination of a phenylthiolatevinyl(triphenylphospine)gold(I)
moiety to a Cr(CO)5 fragment through the sulfur atom of the
thioether group, rather than through the carbon as found for the
aminovinylgold(I) complex in 1. Complex 7ꢀ (Fig. 5), meanwhile,
is a bimetallic complex consisting of a Ph3PAuSPh complex
coordinated to the Cr(CO)5 fragment through the sulfur atom
of the sulfide group. The final crystallographically refined ratio
of 7/7ꢀ is 1 : 2.87, indicating that there is in actual fact nearly
three times more of 7ꢀ present in the crystal than of 7. The
asymmetric unit of the total structure consists of one molecule
of 7ꢀ linked by a weak aurophilic Au–Au interaction [Au(1)–Au(2)
˚
3.197(1) A] to a molecule of 7 (at 51.7% site occupancy) which
shares that site with a second molecule of 7ꢀ, present at 48.3% site
occupancy (see Fig. S1 in ESI†).
Furthermore, the molecules of 7 and 7ꢀ occupying the same site
share the atom positions in the AuPPh3 moiety and only differ
in the positions of their respective substituted vinyl (only for 7)
and Cr(CO)5 fragments. Within the structure of 7, disorder is
observed for the oxygen atom of the CO ligand trans to the sulfur
coordination[O(7)]. Two positions couldbe identified for this atom
on the difference Fourier map and refined with a final refined
site occupancy ratio of 0.552 : 0.448 [O(71) : O(72)]. Selected
˚
distance 3.144(8) A]. The bond lengths and angles observed in
these structures compare very well to those found for 7ꢀ in both
crystal structures. The greatest differences between the structure
of the Au–Au linked dimers of 7ꢀ and the Ph3PAuSPh structures
in the literature are the lengths of the Au–S bonds, due to
weakening of the bond as a result of the coordination of the S
to the Cr(CO)5 unit. The literature Ph3PAuSPh structures have an
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 4580–4589 | 4585
©

View Article Online
˚
average Au–S bond distance of 2.305(3) A, which is slightly shorter
than the values found for 7ꢀ in the pure and co-crystallised forms
˚
[2.326(2), 2.372(6), 2.328(1) and 2.318(2) A, respectively].
Calculated structural parameters of 5Mꢀ are in good agreement
with the experimentally derived data for 7 (compare Table 2 and
Fig. 3).
Conclusions
In this work we were able to show thata-deprotonated Fischer-type
aminocarbene and thiocarbene complexes of group 6 metal (Cr
and W) carbonyls act as synthons for a-deprotonated aminovinyl
and thiovinyl ligands towards gold(I). Due to the coordination
preferences of the extruded M(CO)5 units, the products in the
two situations differ significantly and selectively. The propensity
of the N-containing functionality towards p-bond formation
effects essentially alkyl coordination to the chosen metal carbonyl
fragment, whereas S-coordination is preferred in the second type
of conversion affording real (M(CO)5-coordinated) thiolatovinyl
complexes of gold.
{Dimethylaminovinyl(triphenylphosphine)gold(I)-
C}pentacarbonylchromium(0), 1
To a solution that contains 0.21 g (0.80 ml) {dimethylamino-
(methyl)}carbene (pentacarbonyl)chromium(0) dissolved in 15 ml
cooled (−78 ◦C) thf was added slowly 0.55 ml 1.6 M BuLi
(0.88 mmol, 1.1 mol equiv.). The mixture was stirred at that
temperature for 30 min. after which 0.40 g (0.80 mmol) Ph3PAuCl
was added and the mixture allowed to reach room temperature
over a period of 3 h. Removal of the solvent in vacuo resulted
in a dark yellow oily residue containing a mixture of products
(TLC). The desired product from broad yellow bands was
isolated by means of low-temperature (−15 ◦C) silica gel column
chromatography using pentane–diethyl ether (10 : 1), as initial
eluent and increasing the polarity to a pentane–diethyl ether ratio
of ∼2 : 1 once the non-polar impurities and unreacted starting
material have been eluted from the column. The product obtained
Experimental
All solvents were dried and purified by conventional methods
and freshly distilled under nitrogen shortly before use. Unless
otherwise stated, all common reagents were used as obtained
from commercial suppliers without further purification. BuLi
was standardised before use according to the procedure reported
by Winkle et al.18 Ph3PAuCl and the group 6 metal Fischer-
type dialkylamino(methyl)carbene starting complexes as well
as methyl(thiocarbene) complexes were prepared according to
procedures described in the literature.19 The preparation and
characterisation of (CO)5CrC(SPr)Me (a precursor for 6) together
with its crystal and molecular structure are reported as ESI.†
All reactions and manipulations involving organometallic
reagents were carried out under a dry nitrogen atmosphere using
standard Schlenk and vacuum-line techniques. NMR spectra were
recorded on Varian INOVA 600 (600 MHz for 1H, 151 MHz
◦
(0.31 g) was a yellow crystalline material. Yield: 54%. Mp 89 C
(decomp.) Anal. Calc. for C27H23NO5PAuCr: C, 45.0; H, 3.2; N,
1.9. Found: C, 45.3; H, 3.3; N, 2.1%. IR (pentane solution): m/cm−1
1896, 1915, 1960, 2044. dH (600 MHz, CD2Cl2): 1.93 (1 H, br, H1),
ꢀ
4
2.55 (1 H, d, Jtrans
= 8.5 Hz, H2), 2.79 (3 H, s, H5,5 ), 3.45
P–H
ꢀ
(3 H, s, ꢀH5,5 ), 7.4–7.6 (15 H, m, Ph); dC (151 MHz, CD2Cl2): 29.2
(br, C5,5 ), 36.5 (br, C3), 129.5 (d, JP–C = 10.9 Hz, Cortho), 130.2
2
1
1
for 13C{ H} and 243 MHz for 31P{ H}) or Varian VXR 300
(d, 1JP–C = 52.1 Hz, Cipso), 131.8 (Cpara), 134.5 (d, 3JP–C = 13.2 Hz,
Cmeta), 216.7 (d, 2JP–C = 134.2 Hz, C4), 220.9 (COcis), 226.5 (COtrans);
dP (243 MHz, CD2Cl2): 38.2. m/z (FAB) 721 (M+, 20%), 529 (100),
459 (70).
(300 MHz for H, 75.4 MHz for 13C{ H} and 121.5 MHz for
1
1
31P{ H}) NMR spectrometers. 1H and 13C chemical shifts are
1
reported in ppm relative to the 1H and 13C residue of the deuterated
solvents while 31P chemical shifts are reported relative to an 85%
H3PO4 external standard solution. IR spectra (4000–400 cm−1,
resolution 4 cm−1) were recorded on a Perkin-Elmer 1600 series
FTIR spectrometer. FAB-MS were recorded on a Micromass DG
70/70E mass spectrometer at the University of Potchefstroom,
South Africa, using xenon gas for bombardment atoms and m-
nitrobenzyl alcohol as matrix. Flash column chromatography
was performed with “flash grade” silica (SDS 230–400 mesh).
Melting/decomposition points were determined on a standard-
ised Bu¨chi 535 melting point apparatus. Crystal structure data
collection and correction procedures were carried out on a Nonius
Kappa CCD diffractometer by Dr J. Bacsa and Ms Hong Su of
the University of Cape Town structural chemistry research group.
The following numbering schemes were utilised for the assign-
ment of chemical shifts:
{Diethylvinyl(triphenylphosphine)gold(I)-
C}pentacarbonylchromium(0), 2
The synthesis followed the same procedure as for 1, starting
from {diethylamine(methyl)carbene}pentacarbonylchromium(0)
(0.26 g, 0.80 mmol). The product was obtained as dark yellow
crystals (0.60 g). Yield: 59%. Mp 82 ◦C (decomp.). Anal. Calc.
for C29H27NO5PAuCr: C, 46.5; H, 3.6; N, 1.9. Found: C, 46.8;
H, 3.8; N, 2.0%. IR (pentane solution): m/cm−1 1893, 1915, 1958,
3
2043. dH (600 MHz, CD2Cl2): 1.12 (3 H, t, JH–H = 7.3 Hz, 10,
ꢀ
ꢀ
3
H6,6 ), 1.13 (3 H, dd, JH–H = 7.3 Hz, 10, H6,6 ), 1.97 (1 H, br,
H1), 2.52 (1 H, br d, Jtrans P–H = 8.1 Hz, H2), 2.73 (1 H, dq,
4
ꢀ
ꢀ
2JH–H = 10.2 Hz, 3JH–H = 7.3 Hz, H5a,5 a,5b,5 b), 3.21 (1 H, dq, 2JH–H
=
ꢀ
ꢀ
10.2 Hz, 3JH–H = 7.3 Hz, H5a,5 a,5b,5 b), 3.46 (1 H, dq, 2JH–H = 10.2 Hz,
4586 | Dalton Trans., 2006, 4580–4589
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
ꢀ
ꢀ
3JH–H = 7.3 Hz, H5a,5 a,5b,5 b), 3.60 (1 H, dq, 2JH–H = 10.2 Hz, 3JH–H
=
{Propylthiolatovinyl(triphenylphospine)gold(I)-
S}pentacarbonylchromium(0), 6
ꢀ
ꢀ
7.3 Hz, H5a,5 a,5b,5 b), 7.4–7.6 (15 H, m, Ph); dC (151 MHz, CD2Cl2):
12.5 (C6), 13.6 (C6), 25.2 (br, C5), 40.1 (br, C3), 129.3 (d, JP–C
=
2
The same preparative method as for 1 but using 0.24 g, 0.80 mmol
of the new methyl(propylthiolato)carbeneCr(CO)5 complex was
employed. The product obtained was a microcrystalline solid.
Yield: ca. 36%. IR (pentane solution): m/cm−1 1917, 1932, 1937,
10.7 Hz, Cortho), 130.2 (d, 1JP–C = 50.8 Hz, Cipso), 132.0 (Cpara), 134.7
(d, 3JP–C = 13.7 Hz, Cmeta), 218.3 (d, 2JP–C = 131.3 Hz, C4), 220.4
(COcis), 226.2 (COtrans); dP (121.5 MHz, CD2Cl2): 38.0. m/z (FAB)
750 (M+, 35%), 721 (70), 558 (70), 459 (100).
3
2063. dH (600 MHz, CDCl3): 0.92 (3 H, t, JH–H = 6.4 Hz, H7),
3
1.73 (2 H, m, H6), 2.80 (2 H, t, JH–H = 7.1 Hz, H5), 5.38 (1 H,
{Dimethylaminovinyl(triphenylphosphine)gold(I)-
C}pentacarbonyltungsten(0), 3
4
4
d, Jcis P–H = 2.7 Hz, H1), 5.88 (1 H, d, Jtrans P–H = 12.5 Hz, H2),
7.3–7.7 (15 H, m, Ph); dC (151 MHz, CDCl3): 13.2 (C7), 23.0 (C6),
46.3 (C5), 125.4 (br, C3), 129.2 (d, JP–C = 9.3 Hz, Cortho), 129.8
2
The synthesis followed the same procedure as for 1, starting
from the tungsten dimethylaminocarbene complex (0.32 g, 0.80
mmol). The product obtained was a yellow crystalline material
(0.58 g). Yield: 68%. Mp 95 ◦C (decomp.). Anal. Calc. for
C27H23NO5PAuW: C, 38.0; H, 2.7; N, 1.6. Found: C, 38.2; H, 2.9;
N, 1.7%. IR (pentane solution): m/cm−1 1900, 1915, 1957, 2052.
(d, 1JP–C = 50.2 Hz, Cipso), 131.5 (Cpara), 134.2 (d, 3JP–C = 14.2 Hz,
Cmeta), 182.4 (d, 2JP–C = 125.2 Hz, C4), 215.9 (COcis), 222.7 (COtrans);
dP (243 MHz, CDCl3): 42.5. m/z (FAB) 721 (65%), 459 (100).
Phenylthiolatovinyl(triphenylphospine)gold(I)-S, 7 and
{phenylthiolato(triphenyl)phosphinegold(I)-
S}pentacarbonylchromium(0), 7ꢀ
ꢀ
dH (600 MHz, CDCl3): 2.51 (1 H, br, H1), 2.79 (3 H, s, H5,5 ),
3.00 (1 H, br, H2), 3.42 (3 H, s, H5,5 ), 7.4–7.6 (15 H, m, Ph); dC
ꢀ
ꢀ
(151 MHz, CD2Cl3): 29.4 (br, C5,5 ), 36.0 (d, JP–C = 3.6 Hz, C3),
129.2 (d, 2JP–C = 11.3 Hz, Cortho), 129.8 (d, 1JP–C = 54.2 Hz, Cipso),
131.5 (Cpara), 134.2 (d, 3JP–C = 13.5 Hz, Cmeta), 201.0 (s, COcis), 203.3
3
Again, a similar synthetic procedure was used but now starting
with 0.26 g (0.80 ml) methyl(phenylthiolato)carbeneCr(CO)5. The
product was obtained in the form of a yellow viscous oil. Yield:
ca. 15%. IR (pentane solution): m/cm−1 1921, 1935, 1979, 2060.
(COtrans), 207.9 (d, JP–C = 118.8 Hz, C4); dP (243 MHz, CDCl3):
2
41.3. m/z (FAB) 852 (M+, 50%), 721 (65), 529 (100), 459 (100).
4
dH (600 MHz, CDCl3): 5.64 (d, Jcis P–H = 5.1 Hz, H1), 6.07 (d,
4Jtrans P–H = 12.0 Hz, H2), 7.1–7.6 (m, Ph, SPh); dC (151 MHz,
{Diethylaminovinyl(triphenylphosphine)gold(I)-
C}pentacarbonyltungsten(0), 4
CDCl3): 125.7 (C3), 129.9–131.8 (PPh, SPh), 180.8 (d, JP–C
=
2
126.9 Hz, C4), 215.4 (COcis), 223.1 (COtrans); dP (243 MHz, CDCl3):
41.9. m/z (FAB) 721 (70%), 459 (100).
The synthesis followed the same procedure as for 1, starting
with the diethylaminocarbene complex of tungsten pentacarbonyl
(0.36 g, 0.80 mmol). The product obtained (0.46 g) was a yellow
crystalline material. Yield: 52%. Mp 85 ◦C (decomp.). Anal. Calc.
for C29H27NO5PAuW: C, 39.5; H, 3.1; N, 1.6. Found: C, 39.7; H,
3.3; N, 1.75%. IR (pentane solution): m/cm−1 1894, 1912, 1958,
Butyl(triphenylphosphine)gold(I), 8
The by-product was obtained in all of the syntheses described
above. After column chromatographic separation (first eluted
product), stripping of solvent in vacuo furnished 8 as a colourless
crystalline material in yields of 5–15% (based on Ph3PAuCl). Mp
106 ◦C (decomp.), Anal. Calc. for C22H24PAu: C, 51.2; H, 4.7.
ꢀ
2053. dH (600 MHz, CD2Cl2): 1.20 (3 H, t, 3JH–H = 7.2 Hz, H6,6 ),
ꢀ
1.29 (3 H, t, 3JH–H = 7.2 Hz, H6,6 ), 2.50 (1 H, d, 4Jcis P–H = 2.7 Hz,
H1), 2.91 (1 H, d, 4Jtrans P–H = 9.0 Hz, H2), 2.96 (dq, 2JH–H = 10.0 Hz,
ꢀ
ꢀ
Found: C, 51.0; H, 4.5%. dH (300 MHz, CDCl3): 0.94 (3 H, 3JH–H
=
3JH–H = 7.2 Hz, H5a,5 a,5b,5 b), 3.45 (1 H, dq, 2JH–H = 10.0 Hz, 3JH–H
=
ꢀ
ꢀ
7.4 Hz, H1), 1.4–1.6 (4 H, m, H2,3), 1.86 (2 H, dt, 3JP–H = 4.3 Hz, 14.7
7.2 Hz, H5a,5 a,5b,5 b), 3.64 (1 H, dq, 2JH–H = 10.0 Hz, 3JH–H = 7.2 Hz,
ꢀ
ꢀ
3
(C1), JH–H = 7.2 Hz, H4), 7.4–7.6 (15 H, m, Ph); dC (75.4 MHz,
H5a,5 a,5b,5 b), 3.83 (1 H, dq, JH–H = 10.0 Hz, JH–H = 7.2 Hz,
2
3
ꢀ
ꢀ
3
2
CDCl3) 34.7 (C2), 30.0 (d, JP–C = 4.3 Hz, C3), 31.3 (d, JP–C
=
=
=
H5a,5 a,5b,5 b), 7.4–7.6 (15 H, m, Ph); dC (151 MHz, CD2Cl2): 12.9
ꢀ
ꢀ
ꢀ
1
2
95.4 Hz, C4), 132.7 (d, JP–C = 44.7 Hz, Cipso), 129.7 (d, JP–C
10.5 Hz, Cortho), 134.6 (d, 3JP–C = 13.3 Hz, Cmeta), 131.6 (d, 4JP–C
(C6,6 ), 14.0 (C6,6 ), 27.5 (br, C5,5 ), 42.1 (d, JP–C = 2.4 Hz, C3),
129.5 (d, 2JP–C = 12.6 Hz, Cortho), 130.2 (d, 1JP–C = 53.5 Hz, Cipso),
131.8 (Cpara), 134.5 (d, 3JP–C = 12.7 Hz, Cmeta), 201.9 (COcis), 203.8
(COtrans), 210.1 (d, 2JP–C = 117.6 Hz, C4); dP (243 MHz, CD2Cl2): d
41.9. m/z (FAB) 881 (M+, 65%), 721 (75), 558 (100), 459 (65).
3
2.4 Hz, Cpara); dP (121.5 MHz, CDCl3): 46.4. m/z (EI) 516 (4%),
488 (8), 459 (35), 262 (100).
Crystal structure determination of complexes 1, and 7/7ꢀ
{Methylthiolatovinyl(triphenylphospine)gold(I)-
S}pentacarbonylchromium(0), 5
The crystallographic data collection and refinement details for
complex 1 are summarised in Table 4. X-Ray quality yellow
platelet single crystals of 1 were obtained by crystallisation from
a concentrated diethyl ether solution layered with pentane at
−20 ◦C. Low temperature (−95 ◦C) data were collected on an
Enraf-Nonius KappaCCD diffractometer20 or Bruker SMART
Apex CCD diffractometer21 using graphite-monochromated Mo-
The same preparative method as for 1 was used with the
methyl(methylthiolato)carbeneCr(CO)5 complex (0.21 g, 0.80
mmol) as reactant. The product obtained was a dark-yellow oil.
Yield: ca. 20%. IR (pentane solution): m/cm−1 1918, 1943, 2063.
dH (600 MHz, CDCl3): 2.47 (3 H, br s, H5), 5.43 (1 H, br, H1),
5.76 (1 H, br, H1), 7.2–7.7 (15 H, m, Ph); dC (151 MHz, CD2Cl3):
˚
Ka radiation (k = 0.71073 A) and scaled and reduced using
1
39.7 (C5), 121.7 (br, C3), 129.5 (d, JP–C = 59.6 Hz, Cipso), 129.8
DENZO-SMN22 and Bruker SAINT23 respectively. Empirical
corrections were performed using SCALEPACK22 and SMART
data were treated with SADABS.24 The structure of 1 was solved
by the heavy atom method and refined anisotropically for all the
(d, 2JP–C = 9.9 Hz, Cortho), 132.1 (Cpara), 134.9 (d, 3JP–C = 13.9 Hz,
Cmeta), 181.9 (d, 2JP–C = 129.4 Hz, C4), 216.4 (COcis), 223.2 (COtrans);
dP (243 MHz, CDCl3): 38.2. m/z (FAB) 721 (70%), 459 (100).
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 4580–4589 | 4587
©

View Article Online
Table 4 Crystallographic data for 1 and 7/7ꢀ
1
7/7ꢀ
Chemical formula
Mr/g mol−1
C27H23O5NPAuCr
721.4
Monoclinic
P21/c
6.6251(2)
23.8365(8)
17.0238(5)
97.573(1)
2664.9(1)
4
1.798
178(2)
6.003
25.00
1.48(C29H20O5SPCrAu)·0.52(C31H22O5SPCrAu)
1534.43
Crystal system
Space group
Monoclinic
P21/c
12.4049(2)
20.7908(3)
23.8896(3)
91.437(1)
6159.38(15)
4
1.655
177(2)
5.265
26.00
˚
a/A
˚
b/A
˚
c/A
b/◦
3
˚
V/A
Z
Dc/g cm−3
T/K
l(Mo-Ka)/cm−1
◦
2hmax
/
Crystal size/mm
Index range, hkl
Reflections collected
Independent reflections (Rint
Observed reflections
Parameters
0.10 × 0.22 × 0.33
0.25 × 0.30 × 0.35
−7 to 4, −28 to 28, −20 to 20
−15 to 14, −24 to 25, −25 to 29
8192
4422 (0.0494)
3729
33837
12099 (0.0328)
10327
)
335
868
R1 (Fo > 2rFo)
wR2 (all data)
0.0333
0.0890
0.0476
0.1087
non-hydrogen atoms by full-matrix least-squares calculations on
F2. All hydrogen atoms in 1 were placed in calculated positions,
except those on the terminal CH2-group coordinated to Cr(CO)5,
which were found on the difference Fourier map and refined
isotropically. The structure of 7/7ꢀ was solved by direct methods.
Peaks of electron density on the difference Fourier map indicated
the presence of disorder, which was subsequently identified as
molecules of 7 and 7ꢀ in the ratio 1 : 2.866. The sum of the
site occupancies of the partial molecular fragments of 7 and 7ꢀ
that share the same site in the crystal lattice was restrained to
1 after which all the non-hydrogen atoms, except O(7)ꢀ, which
exhibits further disorder, were allowed anisotropic displacement
parameters. Bond lengths in the C(24X) and C(25X) phenyl rings
were restrained to sensible values and the carbon atoms in both
rings were subjected to rigid body (DELU), flat geometry (FLAT)
and similar Uij (SIMU) restraints to improve their geometry. Two
O(7)ꢀ atomic positions [O(7)ꢀ1 and O(7)ꢀ2] could be found for
the disorder observed in this CO ligand. The sum of the site
occupancies of the two O(7) atomic sites was restrained to the
refined site occupancy of the disordered fragment they belong
to. All hydrogen atoms in 7 were placed in calculated positions.
The isotropic displacement parameters of the calculated hydrogen
atoms in both structures were fixed at 1.2 (aromatic hydrogens) and
1.5 (methyl hydrogens) times the equivalent isotropic displacement
parameters of their parent atoms. All calculations were performed
using SHELX-9725 under the X-Seed26 environment. Figures were
prepared using ORTEP-3 for Windows, with the ellipsoids at the
50% probability level.27
ndash potential suggested by Becke28 in conjunction with the
LYP29 exchange potential (B3LYP).30 A quasi-relativistic small-
core ECP with a (441/2111/N1) valence basis set for the metal
atoms (N = 4 for Cr and N = 2 for Au)31 and 6-31G(d) all-
electron basis sets32 for the other atoms were employed in the
geometry optimisations. This is our standard basis set II.33 The
nature of the stationary points was examined by calculating the
Hessian matrix at B3LYP/II. All structures are energy minima
on the potential energy surface. The calculations were performed
with the program package Gaussian 98.34
Acknowledgements
From Stellenbosch we thank the NRF (Pretoria) and the Claude
Harris Leon Foundation (UEIH) for financial support. The work
at Marburg was financially supported by the DFG and the Fonds
der Chemischen Industrie. AYT is grateful to the Alexander-von-
Humboldt Foundation for a short-term fellowship.
References
1 H. G. Raubenheimer, M. W. Esterhuysen and C. Esterhuysen, (Special
edition dedicated to H. Schmidbaur), Inorg. Chim. Acta, 2005, 358,
4217.
2 H. G. Raubenheimer, M. W. Esterhuysen, A. Timoshkin, Y. Chen and
G. Frenking, Organometallics, 2002, 21, 3173.
3 (a) K. H. Do¨tz and J. Pfeifer, in Transition Metals for Organic Synthesis,
ed. M. Beller and C. Bolm, Wiley-VCH, Weinheim, 1998, vol. 1, p. 335;
(b) L. S. Hegedus, Transition Metals in the Synthesis of Complex Organic
Molecules, University Science Books, California, 1994, p. 151.
4 (a) C. F. Bernasconi and S. Bhattacharya, Organometallics, 2003, 22,
1310; (b) C. F. Bernasconi, K. W. Kittredge and F. X. Flores, J. Am.
Chem. Soc., 1999, 121, 6630.
CCDC reference numbers 609579–609581.
For crystallographic data in CIF or other electronic format see
DOI: 10.1039/b607613k
5 C. P. Casey, in Transition Metal Organometallics in Organic Synthesis,
ed. H. Alper, Academic Press, New York, 1976, vol. 1, p. 190.
6 C. F. Bernasconi, Chem. Soc. Rev., 1997, 26, 299.
Computational methods
7 R. J. Hoare and O. S. Mills, J. Chem. Soc., Dalton Trans., 1972, 653.
8 L. A. Watson, B. Franzman, J. C. Bollinger and K. G. Caulton,
New J. Chem., 2003, 27, 1769.
The geometries of the molecules were optimised at the gradient-
corrected DFT level using the three-parameter fit of the exchange–
4588 | Dalton Trans., 2006, 4580–4589
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
9 M. W. Esterhuysen, J. Even, C. Esterhuysen, U. E. I. Horvath, E.
Stander and H. G. Raubenheimer, Acta Crystallogr., Sect. E, 2006, 62,
m1625.
21 SMART Data Collection Software, Version 5.629, Bruker AXS Inc.,
Madison, WI, 2003.
22 Z. Otwinowski and W. Minor, Methods Enzymol., 1997, 276, 307.
23 SAINT, Data Reduction Software, Version 6.45, Bruker AXS Inc.,
Madison, WI, 2003.
10 D. Sellmann, R. Weiss, K. Knoch and M. Moll, J. Organomet. Chem.,
1990, 391, 327.
11 M. M. Hossain, H.-M. Lin and S.-G. Shyu, Eur. J. Inorg. Chem., 2001,
2655.
24 SADABS, Version 2.05, Bruker AXS Inc., Madison, WI, 2003.
25 G. M. Sheldrick, SHELX-97 (SHELXS-97 and SHELXL-97), Pro-
grams for crystal structure determination and refinement, University of
Go¨ttingen, Germany, 1997.
12 M. Nakamoto, W. Hiller and H. Schmidbaur, Chem. Ber., 1993, 126,
605.
13 (a) J. P. Marino, in Topics in Sulfur Chemistry, ed. E. Jennings, Georg
Thieme Publishers, London, 1976, p. 33; (b) M. Herberhold and G.
Su¨ss, J. Chem. Res. (M), 1977, 2720; (c) S. Lotz, PhD thesis, Rand
Afrikaans University, Johannesburg, 1978, p. 11.
14 (a) A. Tamaki and J. K. Kochi, J. Organomet. Chem., 1973, 61, 441;
(b) A. N. Nesmeyanov, E. G. Perevalova, D. A. Lemenovskii, A. N.
Kosina and K. I. Grandberg, Izv. Akad. Nauk. SSSR. Ser. Khim., 1969,
9, 2030; (c) K. I. Grandberg, E. Smyslova and A. N. Kosina, Izv. Akad.
Nauk Ser. Khim., 1973, 12, 2787.
26 L. J. Barbour, J. Supramol. Chem., 2001, 1, 189.
27 L. J. Farrugia, J. Appl. Crystallogr., 1997, 30, 565.
28 A. D. Becke, J. Chem. Phys., 1993, 98, 5648.
29 C. Lee, W. Yang and R. G. Parr, Phys. Rev. B, 1988, 37, 785.
30 P. J. Stevens, F. J. Devlin, C. F. Chablowski and M. J. Frisch, J. Phys.
Chem., 1994, 98, 11623.
31 P. J. Hay and R. W. Wadt, J. Chem. Phys., 1985, 82, 299.
32 (a) R. Ditchfield, W. J. Hehre and J. A. Pople, J. Chem. Phys., 1971,
54, 724; (b) W. J. Hehre, R. Ditchfield and J. A. Pople, J. Chem. Phys.,
1972, 56, 2257.
33 G. Frenking, I. Antes, M. Bo¨hme, S. Dapprich, A. W. Ehlers, V.
Jonas, A. Neuhaus, M. Otto, R. Stegmann, A. Veldkamp and S. F.
Vyboishchikov, Rev. Comput. Chem., 1996, 8, 63.
34 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A.
Robb, J. R. Cheeseman, V. G. Zakrzewski, J. A. Montgomery, Jr., R. E.
Stratmann, J. C. Burant, S. Dapprich, J. M. Millam, A. D. Daniels,
K. N. Kudin, M. C. Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi, R.
Cammi, B. Mennucci, C. Pomelli, C. Adamo, S. Clifford, J. Ochterski,
G. A. Petersson, P. Y. Ayala, Q. Cui, K. Morokuma, D. K. Malick,
A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. Cioslowski, J. V.
Ortiz, A. G. Baboul, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I.
Komaromi, R. Gomperts, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-
Laham, C. Y. Peng, A. Nanayakkara, C. Gonzalez, M. Challacombe,
P. M. W. Gill, B. G. Johnson, W. Chen, M. W. Wong, J. L. Andres, M.
Head-Gordon, E. S. Replogle and J. A. Pople, GAUSSIAN 98 (Revision
A.8), Gaussian, Inc., Pittsburgh, PA, 1998.
15 H. Schmidbaur, J. R. Mandel, J. M. Bassett, G. Blaschke and B.
Zimmer-Gasser, Chem. Ber., 1981, 114, 433.
16 P. Pyykko¨, Angew. Chem., Int. Ed., 2004, 43, 4412.
17 (a) J. P. Fackler, R. J. Staples, A. Elduque and T. Grant, Acta.
Crystallogr., Sect. C, 1994, 50, 520; (b) L. G. Kuz’mina, T. V. Baukova,
A. V. Churakov, V. S. Kuz’min and N. V. Dvortsova, Koord. Khim.,
1997, 23, 301; (c) S. Onaka, Y. Katsukawa and M. Yamashita, Chem.
Lett., 1998, 525.
18 M. R. Winkle, J. M. Lansinger and R. C. Roland, J. Chem. Soc., Chem.
Commun., 1980, 87.
19 (a) F. A. Kreissl, in Synthetic Methods of Organometallic and Inorganic
Chemistry (Herrmann/Brauer), ed. W. A. Herrmann, Thieme Verlag,
Stuttgart, 1997, vol. 7, p. 126; (b) Fischer and M. Leupold, Chem.
Ber., 1972, 105, 599; (c) C. T. Lam, C. V. Senoff and J. E. H. Ward,
J. Organomet. Chem., 1974, 70, 273.
20 COLLECT, Data collection software, Nonius B.V., Delft, The Nether-
lands, 1999.
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 4580–4589 | 4589
©