Organometallic gold(III) and gold(I) complexes as catalysts for the 1,3-dipolar cycloaddition to nitrones: Synthesis of novel gold-nitrone derivatives

Article abstract of DOI:10.1016/j.jorganchem.2004.03.001

Gold(III) and gold(I) anionic salts mediate the 1,3-dipolar cycloaddition of N-benzyl-C(2-pyridyl)nitrone (2-PyBN) (1) and methyl acrylate (2) (gold 5-10 mol% with respect to the nitrone) decreasing the reaction time and favouring the formation of the exo

Full text of DOI:10.1016/j.jorganchem.2004.03.001

Journal of Organometallic Chemistry 689 (2004) 1788–1795
www.elsevier.com/locate/jorganchem
Organometallic gold(III) and gold(I) complexes as catalysts for
the 1,3-dipolar cycloaddition to nitrones: synthesis of novel
gold–nitrone derivatives
a
a
a
a,
b
Ana Ade , Elena Cerrada , Marıa Contel , Mariano Laguna *, Pedro Merino ,
ꢀ
ꢀ
b
ꢀ
Tomas Tejero
a
ꢀ
ꢀ
ꢀ
Departamento de Quımica Inorganica, Instituto de Ciencia de Materiales de Aragon, Universidad de Zaragoza-C.S.I.C., E-50009 Zaragoza, Spain
b
Laboratorio de Sꢀıntesis Asimꢀetrica, Departamento de Quꢀımica Orgꢀanica, Instituto de Ciencia de Materiales de Aragꢀon,
Universidad de Zaragoza-C.S.I.C., E-50009 Zaragoza, Spain
Received 11 September 2003; accepted 1 March 2004
Abstract
Gold(III) and gold(I) anionic salts mediate the 1,3-dipolar cycloaddition of N-benzyl-C(2-pyridyl)nitrone (2-PyBN) (1) and
methyl acrylate (2) (gold 5–10 mol% with respect to the nitrone) decreasing the reaction time and favouring the formation of the exo
(cis) isomer. The best catalyst found was Na[AuCl4] (7) able to perform the addition reaction in 56 h (instead of the 96 h required for
the control experiment) and giving an endo/exo relation between isomers of 44/56 (as opposed to 73/27, blank reaction). The cat-
alytic activity of several organometallic gold complexes with the radicals pentafluorophenyl (C6F5) or mesityl (2,4,6-(CH3)3C6H2)
has been also investigated. In some cases the activity is very similar to that obtained with inorganic salts. With the aim of identifying
possible metallic intermediates in the cycloaddition reaction, novel gold(III) and gold(I) nitrone derivatives such as [Au(C6F5)Cl2(2-
PyBN)] (21), [Au(C6F5)2Cl(2-PyBN)] (22) and [Au(C6F5)(2-PyBN)] (23) have been prepared and characterized. The reaction be-
tween [AuCl3(tht)] and 2-PyBN unexpectedly affords the ionic compound [2-PyBN–H][AuCl4] (5) which also displays catalytic
activity and moderate regioselectivity and whose crystal structure has been confirmed by X-ray studies.
Ó 2004 Elsevier B.V. All rights reserved.
Keywords: Dipolar cycloadditions; Nitrones; Gold; Organometallic; Crystal structure
1. Introduction
zaldehydes and alkynes or cyclization of alkenyl furans.
It is noteworthy that some of these reactions [2a] can be
also catalyzed by silver salts but with much lower yields.
Our research group has recently studied the 1,3-dipolar
cycloaddition between nitrones and alkenes mediated by
Lewis acids [3,4]. Despite the extensive study performed
in this field by several laboratories [5] there are not that
many examples in which a metallic complex with a nit-
rone as ligand catalyzes the dipolar cycloaddition. Most
of these reactions require the presence of a metallic
complex and the nitrone in a 1:1 molar ratio. We have
shown, in previous reports, that stoichiometric reactions
of the N-benzyl-C(2-pyridyl)nitrone with different di-
polarophiles in the presence of zinc and silver com-
pounds afforded an increase of both the reaction rate
and the selectivity, favouring the formation of one of the
possible isomers. The existence of zinc and silver-nitrone
The use of Lewis-acid catalysts in modern organic
synthesis has been expanding uninterruptedly during the
last decade [1]. Nowadays the research is targeting more
versatile, more selective, and more reactive catalysts.
Carbon–carbon bond formation is one of the most im-
portant organic processes that can be Lewis-acid pro-
moted. It has been found recently that anhydrous AuCl3
catalyzes the formation of C–C and C–O bonds be-
having as an effective Lewis-acid catalytic system [2].
Some examples include Michael type additions, selective
cross cycloisomerization/dimerization of propargyl or
allenyl ketones, benzannulation between o-alkynylben-
* Corresponding author. Tel.: +34-976-761181; fax: +34-976-761187.
E-mail address: mlaguna@posta.unizar.es (M. Laguna).
0022-328X/$ - see front matter Ó 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2004.03.001

A. Adꢀe et al. / Journal of Organometallic Chemistry 689 (2004) 1788–1795
1789
complexes during the reaction, which behave as inter-
mediates in the cycloaddition process, was confirmed by
different structural techniques [3,4].
We studied the effect of the addition of gold(III) salts
in catalytic amounts (5 or 10 mol%) to the cycloaddition
reaction. Thus, we started our study with the coordi-
nation compound [AuCl3(tht)] (4) that is quite similar to
anhydrous AuCl3 but much more stable in solid state
and solution, and easier to prepare and handle. Com-
pound 4 has got labile tetrahydrothiophene (tht) as one
of the ligands. The addition of 10 mol% of [AuCl3(tht)]
to the reaction mixture reduces the reaction time from 4
to 3 days modifying the regioselectivity of the process
affording an endo/exo mixture of isomers in a 55/45 ra-
tio, respectively (Table 1, entries 1 and 2).
In order to explain the catalytic activity of 4 we
carried out stoichiometric reactions between compound
4 and nitrone 1 and between this compound and the
dipolarophile 2. However, we only could observe reac-
tion in the first case (Eq. (2)). Nevertheless, instead of
the expected [AuCl3(2-PyBN)] gold(III) complex, due to
the plausible displacement of labile tht and a subsequent
coordination of the nitrone 1 to the metallic centre, the
ionic complex [2-PyBN–H][AuCl4] (5) was obtained.
Complex 5 is always the product of the reaction inde-
pendently on the reaction conditions used (different
solvents, including dried toluene, dichloromethane, or
tetrahydrofurane, and working under argon atmo-
sphere).
We have also recently reported on the Lewis acidic
behaviour exhibited by organometallic gold(III) com-
plexes in the addition of nucleophiles to triple bonds [6].
In view of the results obtained with silver derivatives in
the 1,3-dipolar cycloaddition of nitrones to alkenes [5]
and in order to test the potential Lewis acidic character
of gold(III) and gold(I) complexes, we decided to study
their behaviour in cycloaddition reactions. Interestingly,
we have found that they can be used not only as stoi-
chiometric reagents but they can also mediate these
cycloadditions in catalytic amounts (5–10 mol% with
respect to the nitrone). Here we present the results on
the 1,3-dipolar cycloaddition reaction of the N-benzyl-
C(2-pyridyl)nitrone (2-PyBN), 1, and methyl acrylate, 2,
(Eq. (1)) catalyzed by gold(III) and gold(I) derivatives.
Novel gold(III) and gold(I)-nitrone derivatives will be
described.
2. Results and discussion
In the absence of catalyst, the reaction shown in Eq.
(1) takes place in refluxing acetone after four days giving
the two 3,5-disubstituted regioisomers (endo 3b and exo
3c) as major products (73% and 17%, respectively). Of
the two possible 3,4 regioisomers (endo 3d and exo 3a)
only the exo isomer 3a could be detected in the reaction
mixture (ca. 10%).
H
-
+
O
N
O
N
N
AuCl (tht)
+
[AuCl4]
3
Ph
N
Ph
(5)
(1)
(4)
ð2Þ
OMe
The 1H NMR spectrum at room temperature of
complex 5 displays the corresponding N–H signal at
16.5 ppm as a broad singlet as well as signals downfield
displaced from the usual nitrone resonances, except the
signal due to the proton belonging to the carbon in a
position, which is displaced from 8.60 ppm in the free
nitrone to 8.42 ppm in 5.
-
O
N
+
+
N
Ph
O
2
1
COOMe
+
COOMe
N
N
N
N
COOMe
COOMe
++
N
O
N
O
N
O
N O
Ph
3a
Ph
Ph
Ph
The ionic nature of complex 5 and the presence of
such N–H bond was further elucidated by X-ray dif-
fraction studies. A drawing of the molecule (anion and
3b
3c
3d
ð1Þ
Table 1
Catalytic activity of gold(III) and gold(I) salts and complexes in the 1,3-cycloaddition reaction of 1 and 2
Entry
Compound
Catalyst (mol%)
Isomers (%) endo/exo
Time (h)
Conversion (%)a
1
2
3
4
5
6
7
8
9
None
–
73/27
55/45
54/46
63/37
44/56
59/41
57/43
54/46
56/44
96
72
62
72
54
56
46
56
76
100
100
100
100
100
100
100
100
100
[AuCl3(tht)] (4)
[2-PyBN–H][AuCl4] (5)
[2-PyBN–H]Tfo (6)
Na[AuCl4] (7)
10
5
5
5
PPN[AuCl4] (8)
PPN[AuCl4] (8)
PPN[AuCl2] (9)
[AuCl(tht)] (10)
5
10
10
10
a Conversion and selectivity measured by 1H NMR (CDCl3).

1790
A. Adꢀe et al. / Journal of Organometallic Chemistry 689 (2004) 1788–1795
Table 2
cation) is shown in Fig. 1, while selected bond and an-
gles are presented in Table 2. In this case, and as pre-
viously shown for the nitrone 2-PyBN, 1 [3], the system
C–CH@Nþ(Oꢀ)C of the nitrone and the heterocyclic
ring are in the same plane. However, and as opposed to
the free nitrone, the pyridinic nitrogen is syn periplanar
to the oxygen of the nitrone group. This could be due to
an interaction between the oxygen atom of the nitrone
group and the hydrogen that belongs to the N–H bond.
The distance found between the O and the H is 1.95(4)
ꢁ
Bond lengths (A) and angles (°) for [2-PyBN–H][AuCl4] (5)
Bond lengths
Au(1)–Cl(1)
Au(1)–Cl(4)
Au(1)–Cl(3)
Au(1)–Cl(2)
C(15)–N(1)
N(1)–C(8)
2.2696(8)
2.2756(8)
2.2776(8)
2.2870(8)
1.329(4)
1.353(4)
1.293(3)
1.298(5)
1.481(4)
1.440(5)
N(2)–O(1)
N(2)–C(13)
N(2)–C(11)
C(8)–C(13)
ꢁ
ꢁ
A, which is 0.65 A shorter than the sum of the van der
ꢁ
ꢁ
Waals radii (rvdW(H) ¼ 1.20 A and rvdW(O) ¼ 1.40 A) [7].
Bond lengths and angles in the nitrone molecule are
similar to those found in the free nitrone. Thus, the
distances in the N–O and N–C units in 5 are 1.293(3)
Bond angles
Cl(1)–Au(1)–Cl(4)
Cl(1)–Au(1)–Cl(3)
Cl(4)–Au(1)–Cl(3)
Cl(1)–Au(1)–Cl(2)
Cl(4)–Au(1)–Cl(2)
Cl(3)–Au(1)–Cl(2)
N(1)–C(15)–C(16)
C(2)–C(3)–C(4)
90.59(3)
179.41(3)
89.18(3)
89.08(3)
178.96(2)
91.16(3)
120.2(3)
119.9(3)
122.9(3)
124.4(3)
115.2(2)
ꢁ
ꢁ
and 1.298(5) A, respectively and 1.300(3) and 1.299(3) A
in 1. Anion [AuCl4]ꢀ displays a square-planar rear-
rangement for the metallic centre. Similar Au–Cl dis-
tances and angles have been found recently in oxonium
ion compounds of different crown ethers with tetra-
chloroaurate as the counterion [8].
C(15)–N(1)–C(8)
O(1)–N(2)–C(13)
O(1)–N(2)–C(11)
The protonation of the nitrone can be achieved di-
rectly by equimolecular reaction of 1 with triflic acid
(Eq. (3)) affording [2-PyBN–H]Tfo, 6, which shows al-
most identical spectroscopic data to those obtained for
5.
catalytic activity and the moderated regioselectivity
observed. This fact was subsequently confirmed by the
catalytic activity and regioselectivity exhibited by other
gold(III) anionic salts (entries 5–7) such as Na[AuCl4], 7
or the less hygroscopic salt with the bulkier cation PPNþ
(bis(triphenylphosphine)iminium), PPN[AuCl4], 8.
The addition of 5 mol% of 7 produces a great change
in the regioselectivity (from a ratio endo/exo of 73/27 in
the absence of catalyst (entry 1) to an endo/exo ratio of
44/56 (entry 5)). Besides, the addition of gold salts de-
creases the reaction time in all cases and, specifically,
the addition of 10 mol% of 8 decreases this time to less
than half the value (entry 7) for the control experiment
(entry 1).
ð3Þ
2-PyBN + HTfo
[2-PyBN–H]Tfo (6)
We carried out the same catalytic experiment (de-
scribed in Eq. (1)) with [2-PyBN–H][AuCl4] (5) and [2-
PyBN–H]Tfo (6) as catalysts, in order to establish if the
catalytic activity was due to the protonation of the nit-
rone or to the effect of the [AuCl4]ꢀ anion (Table 1,
entries 3 and 4). The obtained results led us to consider
that the gold(III) anion was partly responsible for the
To the best of our knowledge the addition of gold(I)
derivatives in C–C and C–O coupling reactions have
shown to be ineffective [2]. Nevertheless, we decided to
study if gold(I) salts could display catalytic activity in
the cycloaddition reaction object of our investigations.
Unexpectedly, the addition of 10 mol% of PPN[AuCl2],
9 or [AuCl(tht)], 10 to the reaction (Eq. (1)) reduces
considerably the reaction time (from 96 to 56 h in the
first case) and, reasonably in the second case (Table 1,
entries 8 and 9, respectively). In addition, the regiose-
lectivity is similar to that found in the addition of the
gold(III) derivatives.
Recently, our research group has shown that anionic
or neutral organometallic gold(III) complexes where one
or more chlorides have been replaced by an organic
group, like C6F5 or 2,4,6-(CH3)3C6H2 (mes), behave in a
similar way to Na[AuCl4], being efficient catalysts for
Fig. 1. View of the molecular structure of 5 (ellipsoids are drawn at
50% probability level). H atoms have been omitted for clarity.

A. Adꢀe et al. / Journal of Organometallic Chemistry 689 (2004) 1788–1795
1791
the addition of water and alcohols to terminal alkynes
[6]. These complexes have the general formula Q[Au-
RCl3] (Q ¼ BzPPh3, PPN; R ¼ mes, C6F5), [Au(C6F5)
Cl2]2, trans-NBu4[Au(C6F5)X2], X ¼ Cl, Br and [Au-
(C6F5)2Cl]2. The organic groups attached to the gold(III)
centres, pentafluorophenyl (C6F5) and mesityl (2,4,
6-(CH3)3C6H2) were chosen because they are very dif-
ferent in their esteric and electronic properties. We con-
sidered that the evaluation of organometallic gold(III)
compounds in several catalytic reactions could be of
help in order to identify possible intermediates in some
of these reactions. The organic groups could help not
only to stabilize the plausible gold(III) complexes gen-
erated but also, they can be of aid in the identification
of these metallic intermediates by ‘‘labelling’’ the
gold(III) centres for structural techniques such as NMR
spectroscopy.
The catalytic results for several organometallic
gold(III) compounds (numbered in the tables and
throughout the text) tested in the 1,3-cycloaddition re-
action of 1 to 2 (Eq. (1)) are collected in Table 3. There
is always an increase of the reaction rate (with respect to
the control experiment) quite significant, especially in
the case of compound 15 (entry 5). There is also a
general trend to increase the yield of exo (cis) isomers in
favour of the endo (trans) isomers. This is similar to
what it was observed for the gold(III) complexes de-
picted in Table 1. Nevertheless, and as it occurred in the
control experiment, isomer 3d is not obtained with any
of the gold(III) catalysts tested.
It appears that the active catalytic species could be
gold(III) anions of the type [AuCl4]ꢀ, anionic species
such as [AuRCl3]ꢀ or [AuR2Cl2]ꢀ (R ¼ mes, C6F5) and
the neutral species [Au(C6F5)Cl2(tht)] (analogous to
[Au(C6F5)Cl2]2 but isolable). All these species could
coordinate, in the reaction conditions, nitrone (1) or
methyl acrylate (2) molecules to the gold(III) centre.
Organometallic gold(I) complexes with the C6F5 or
mes radicals, can be used in the cycloaddition reaction.
Considering that one requirement for this catalysis is the
presence of an Au–Cl bond, we have chosen ionic de-
rivatives of the type PPN[AuRCl] with R ¼ mes, 18,
C6F5, 19. In these cases (Table 3) only a reduction of the
time reaction was found, while the relation endo/exo is
practically unchanged.
In order to determine the possible intermediates of
the cycloaddition reaction, as confirmed in the case of
zinc and silver derivatives, [4] we performed the reaction
between the nitrone 1 and some of the organometallic
gold(III) and gold(I) complexes (Scheme 1).
As starting gold(III) derivatives with the pentafluor-
ophenyl radical, we used complexes [Au(C6F5)Cl2(tht)]
(15) and [Au(C6F5)2Cl]2 (16), which had the labile tht
ligand or that were dimers that could react easily with 1.
In the first case labile tht is displaced by the nitrone 1
giving rise to complex 21, whose spectroscopic data
point to a coordination of 1 to the metallic centre
through the oxygen atom. Thus, the 1H NMR spectrum
displays a downfield displacement of the nitrone reso-
nances (mainly for d (HC@N) and d (NCH2Ph) reso-
nances) as occurred in silver and zinc complexes
described previously by us [3,4]. The 19F NMR spectrum
exhibits a characteristic pattern for a square-planar
gold(III) complex [6].
In the case of compound [Au(C6F5)2Cl]2 (16), the
isolated complex 22, shows a similar 1H NMR spectrum
than that of 21; however, the 19F NMR spectrum at
room temperature displays six resonances, as expected
for a cis-pentafluorophenyl complex. Some of these
signals are broad and not totally resolved, indicating
that one of the C6F5 ring does not have a free rotation.
The spectrum at )60 °C confirms this assumption and
shows how the broad signals that appear at )124.0 ppm
[Au(C6F5)Cl2(2-PyBN)] (21)
[Au(C6F5)2Cl(2-PyBN)] (22)
i)
-
+
O
N
ii)
N
Ph
(1)
iii)
[Au(C6F5)(2-PyBN)] (23)
i) [Au(C6F5)Cl2(tht)], ii)[Au(C6F5)2Cl]2 , iii) [Au(C6F5)(tht)]
Scheme 1.
Table 3
Catalytic activity of organometallic gold(III) and gold(I) compounds and complexes in the 1,3-cycloaddition reaction of 1 and 2
Entry
Compound
Catalyst (mol%)
Isomers (%) endo/exo
Time (h)
Conversion (%)a
1
2
3
4
5
6
7
8
9
PPN[Au(mes)Cl3] (11)
PPN[Au(C6F5)Cl3] (12)
cis-PPN[Au(mes)2Cl2] (13)
trans-NBu4 [Au(C6F5)2Cl2] (14)
[Au(C6F5)Cl2(tht)] (15)
[Au(C6F5)2Cl]2 (16)
10
10
10
10
10
10
10
10
10
58/42
55/45
55/45
60/40
53/47
77/23
56/44
66/34
66/34
84
60
92
78
48
90
72
48
72
100
100
100
100
100
100
100
100
100
[Au(mes)2Cl]2 (17)
PPN[Au(mes)Cl] (18)
PPN[Au(C6F5)Cl] (19)
a Conversion and selectivity measured by 1H NMR (CDCl3).

1792
A. Adꢀe et al. / Journal of Organometallic Chemistry 689 (2004) 1788–1795
at room temperature (and that correspond to the For-
tho) split in two resonances at )123.1 and )125.6 ppm.
Besides, there is a better resolution for the signals that
correspond to the Fmeta and Fpara (see Section 3).
Reactions of nitrone 1 with gold(III) complexes
containing the organic ligand mesityl [Au(mes)2Cl]2 af-
forded mixtures of unidentified and more complicated
compounds that are currently under investigation.
Complex [Au(C6F5)(tht)] (20) was chosen as a gold(I)
organometallic starting material. The reaction with the
nitrone, 1 (Scheme 1, iii) afforded the stable complex 23,
as a result of the displacement of tht by the nitrone. All
spectroscopic data (see Section 3) point to the coordi-
nation of the nitrone to the metallic centre (through the
O-atom) giving rise to a gold(I) linear compound.
Once the nitrone-complexes (21–23) were synthesized
we carried out the stoichiometric 1,3-cycloaddition re-
action with methyl acrylate (2), to study their behaviour
as possible metallic intermediates. However, the above
mentioned reaction (same reaction conditions as that of
analogous zinc and silver complexes [4]) gave no cyclo-
addition products. Decomposition to metallic gold as a
consequence of reduction from gold(III) and gold(I) to
gold(0) was observed instead in all cases.
Compounds 21–23 were also tested in the cycloaddi-
tion reaction between 2-PyBN (1) and methyl acrylate
(2), Eq. (1), in catalytic amounts (10 mol%). The results
are collected in Table 4. There is only a significant in-
crease of the reaction rate in the case of compound 21
[Au(C6F5)Cl2(2-PyBN)] which has got two chloride at-
oms coordinated to the metallic centre (Table 4, entry
1). In the case of the gold(III) compound [Au(C6F5)2Cl
(2-PyBN)], 22, with two coordinated pentafluorophenyl
groups or the gold(I) complex [Au(C6F5)(2-PyBN)], 23,
the increase is not very high or else, the reaction rate is
decreased (entry 2) in comparison to the control exper-
iment. Nevertheless, a moderate regioselectivity is ob-
served with these novel compounds. The latter results
point to the fact that the coordination of the nitrone to
the gold(III) centre may not be the first step in the 1,3-
cycloaddition reaction. Plausibly the first step should be
the coordination of methyl acrylate (2) to the gold(III)
or gold(I) centres (after the Au–Cl bond break-down
under reaction conditions). However, this coordination
may give rise to very reactive vinyl complexes that
coordinate subsequently the nitrone (1) and give the
1,3-cycloaddition reaction as we have not been able to
isolate stable gold-alkene intermediates. Some gold(I)-
alkene derivatives have been reported on the past [9] as
well as gold(III) and gold(I) vinyl complexes [6,10].
However, the reactions of gold(III) and gold(I) com-
pounds with unsaturated reagents lead usually to quite
unstable species [6].
To summarize we have shown that gold(III) and
gold(I) coordination compounds as well as organome-
tallic derivatives are able to catalyze the 1,3-dipolar
cycloaddition reaction of N-benzyl-C(2-pyridyl) nitrone
and methyl acrylate decreasing the reaction time and
favouring the formation of the exo (cis) isomer. It seems
that the first step in this reaction could be the cleavage of
a Au–Cl or Au–C6F5 bond and subsequent coordination
of the methyl acrylate, giving rise to reactive interme-
diates towards the 2-PyBN nitrone to afford the cyclo-
addition products. The reaction of [AuCl3(tht)] with
2-PyBN unexpectedly affords the ionic compound [2-
PyBN–H][AuCl4] (5) whose crystal structure has been
confirmed by X-ray studies. The catalytic activity and
moderate regioselectivity displayed by this compound
has helped us to find analogous (and more active)
gold(III) and gold(I) inorganic and organometallic cat-
alysts, although the action mechanism of the later could
be different of the former (5). Novel nitrone gold(III)
and gold(I) complexes have been prepared and fully
characterized. From their lack of reaction with methyl
acrylate to give cycloaddition products and their low
catalytic activity, we conclude that they are not plausible
intermediates in the 1,3-dipolar cycloaddition reaction
(as opposed to what it was found for analogous Ag(I)
and Zn(II) nitrone derivatives in our previous investi-
gations). Reactions of gold complexes with nitrones and
cycloaddition reactions of nitrones with different dipol-
arophiles (catalysed by gold complexes) are currently
under investigation by our research group.
3. Experimental
3.1. General procedures
IR spectra were recorded on a Perkin–Elmer 883
spectrophotometer, over the range 4000–200 cmꢀ1, us-
1
ing Nujol mulls between polyethylene sheets. H-, 13C-
and 19F NMR spectra were recorded on a Varian
UNITY 200 or 300 or BRUKER 300 in CDCl3 or
Table 4
Catalytic activity of novel organometallic nitrone gold(III) and gold(I) compounds in the 1,3-cycloaddition reaction of 1 and 2
Entry
Compound
Catalyst (mol%)
Isomers (%) endo/exo
Time (h)
Conversion (%)a
1
2
3
[Au(C6F5)Cl2(2-PyBN)] (21)
[Au(C6F5)2Cl(2-PyBN)] (22)
[Au(C6F5)(2-PyBN)] (23)
10
10
10
61/39
65/35
60/40
72
120
84
100
100
100
a Conversion and selectivity measured by 1H NMR (CDCl3).

A. Adꢀe et al. / Journal of Organometallic Chemistry 689 (2004) 1788–1795
1793
(CD3)2CO solutions; chemical shifts are quoted relative
to SiMe4 (1H, 13C), and CFCl3 (external 19F). The C, H,
N analyses were performed with a Perkin–Elmer 2400
microanalyser. 2-PyBN (1) [11]; PPN[AuCl2] (9) [12];
[AuCl(tht)] (10) [13]; PPN[Au(mes)Cl3] (11) [14]; PPN-
[Au(C6F5)Cl3] (12) [15]; cis-PPN[Au(mes)2Cl2] (13) [6,16];
trans-NBu4[Au(C6F5)2Cl2] (14) [17]; [Au(C6F5)Cl2(tht)]
(15) [18]; [Au(C6F5)2Cl]2 (16) [19]; [Au(mes)2Cl]2 (17)
[16]; PPN[Au(mes)Cl] (18) [14], PPN[Au(C6F5)Cl] (19)
[20]; and [Au(C6F5)(tht)] (20) [21] were prepared by es-
tablished procedures. Methyl acrylate (2) was purchased
from Aldrich whereas Na[AuCl4] (7) was a commercial
3.2.3. Synthesis of [Au(C6F5)Cl2(2-PyBN)] (21)
(Scheme 1)
To a solution of [Au(C6F5)Cl2(tht)] (15) (0.052 g, 0.1
mmol) in dichloromethane (20 ml) was added 2-PyBN
(1) (0.021 g, 0.1 mmol). The mixture was stirred at room
temperature for 1.5 h. Removal of the solvent to ca. 5 ml
and addition of n-hexane (20 ml) afforded 21 as a white
solid that was subsequently filtered off and vacuum
dried. Yield (%): 77. Anal. Calc. for C19H12N2O
AuCl2F5: C, 35.26; H, 1.87; N 4.33%. Found for 21: C,
35.36; H, 2.03; N, 3.92%. IR (cmꢀ1, Nujol): m
1
(C@N) ¼ 1562 cmꢀ1. H NMR (CDCl3): d 9.43 (d, 1H,
~
product from Kemplate Espana S.A. [AuCl3(tht)] (4),
H4), 8.75 (d, 1H, H1), 7.63 (s, br, 1H, H3), 8.11 (t, 1H,
H2) 7.92 (s, 1H, HC@N), 7.50 (m, 2H, H5, H9), 7.40 (m,
3H, H6, H7, H8), 5.22 (s, 2H, N–CH2–Ph). 19F NMR
(CDCl3): d: )125.22 (dd, br, 2F, Fo), )155.31 (t, 1F, Fp),
)161.10 (t, 2F, Fm). 13C NMR (CDCl3): d 149.7, 137.3,
135.8, 133.0, 124.7, 124.1, 72.1. LSIMSþ: m=z (%) ¼ 611
(78%, [M–Cl]þ, 576 (46%, [M–2Cl]þ), 409 (100%, [Au–
2PyBN]þ), 364 (88%, [AuC6F5]).
and PPN[AuCl4] (8) were prepared by oxidative addition
of [AuCl(tht)] (10) or PPN[AuCl2] (9) with PhCl2I (in a
1:1 molar ratio).
3.2. Synthesis
3.2.1. Synthesis of [2-PyBN–H][AuCl4] (5) (Eq. (2))
To a solution of [AuCl3(tht)] (4) (0.039 g, 0.1 mmol)
in dichloromethane (20 ml) was added 2-PyBN (1)
(0.021 g, 0.1 mmol) and the mixture was stirred at room
temperature for 1.5 h. The solution was concentrated in
vacuum to 3 ml. leading to the precipitation of 5 as a
yellow solid which was washed with diethyl ether and
dried in vacuum. Yield (%): 76. Anal. Calc. for
C13H13N2OAuCl4: C, 26.10; H, 2.37; N: 5.09. Found for
5: C, 26.49; H, 2.36; N, 4.91%. IR (cmꢀ1, Nujol):
3.2.4. Synthesis of [Au(C6F5)2Cl(2-PyBN)] (22)
(Scheme 1)
To a solution of [Au(C6F5)2Cl]2 (16) (0.056 g, 0.1
mmol) in dichloromethane (20 ml) was added 2-PyBN
(1) (0.021 g, 0.1 mmol). The mixture was stirred at room
temperature for 30 min and concentrated to ca. 5 ml.
Addition of n-hexane (20 ml) led to the isolation of 22 as
a
white solid. Yield (%): 58. Anal. Calc. for
C25H12N2OAuClF10: C, 38.55; H, 1.55; N 3.59%. Found
1
m(C@N) ¼ 1594 cmꢀ1. H NMR ((CD3)2CO): d 16.5 (s,
for 22: C, 38.70; H, 1.67; N, 3.70%. IR (cmꢀ1, Nujol): m
br, 1H, N–H), 9.22 (d, 1H, H4, J ¼ 5:7 Hz), 9.08 (s, 1H,
HC@N), 8.96 (td, 1H, H2, J ¼ 7:8 Hz, J ¼ 1:2 Hz), 8.45
(d, 1H, H3, J ¼ 7:8 Hz), 8.42 (t, 1H, H1, J ¼ 1:2 Hz), 7.5
(m, 2H, H5, H9), 7.57 (m, 3H, H6, H7, H8), 5.61 (s, 2H,
NCH2–Ph). 13C NMR((CD3)2CO): d 147,0; 140.6;
133.0; 130.1; 129.7; 129.1; 128.8; 128.7; 127.8; 71.7.
(C@N) ¼ 1557 cmꢀ1. H NMR (CDCl3): d 9.43 (d,1H,
1
H4), 8.77 (s, 1H, HC@N), 8.71 (d, 1H, H1), 8.06 (t, 1H,
H2), 7.60 (s, br, 1H, H3), 7.57 (m, 2H, H5, H9), 7.44 (m,
3H, H6, H7, H8), 5.22 (s, 2H, N–CH2–Ph). 19F NMR
(CDCl3, 25 °C): d: )123.1 (m, 2F, Fo) )124.5 (m, vbr,
2F, Fo), )153.6 (t, 1F, Fp), )155.3 (t, 1F, Fp), )159.2
(m, 2F, Fm), )160.8 (m, br, 2F, Fm). 19F NMR (CDCl3,
)60 °C): d: )123.17 (s, 1F, Fo), )123.47 (s, 2F, Fo),
)125.67 (s, 1F, Fo) )152.65 (t, 2F, Fp), )154.31 (t, 2F,
Fp), )158.40 (dt, 4F, Fm), )159.94 (dt, 4F, Fm). 13C
NMR (CDCl3): d 149.6, 148.2, 141.9, 131.9, 130.1,
129.9, 129.6, 129.1, 127.2, 127.0, 74.3. LSIMSþ: m=z
(%) ¼ 743 (100%, [M–Cl]þ), 409 (47%, [Au–(2PyBN)]þ).
3.2.2. Synthesis of [2-PyBN–H][TfO] (6) (Eq. (3))
To a solution of 2-PyBN (1) (0.042 g, 0.2 mmol) in 20
ml of dichloromethane, 18 ll of triflic acid (HCF3SO3,
HTfO) was added. The mixture was stirred at room
temperature during 2 h. Subsequent removal of the
solvent (to ca. 5 ml) and addition of a mixture of diethyl
ether and n-hexane (20 ml) led to the precipitation of 6
as a white solid which was filtered off and vacuum dried.
Yield (%): 45. Anal. Calc. for C14H13O4F3N2S: C, 46.41;
H, 3.62; N 7.73%. Found for 6: C, 46.01; H, 3.38; N,
3.2.5. Synthesisof[Au(C6F5)(2-PyBN)] (23) (Scheme 1)
To a solution of [Au(C6F5)(tht)] (20) (0.045 g, 0.1
mmol) in dichloromethane (20 ml) was added a solution
of 2-PyBN (1) (0.021 g, 0.1 mmol). After 1.5 h stirring at
room temperature the reaction mixture was concen-
trated to ca. 3 ml. The addition of diethyl ether (10 ml)
afforded 23 as a white solid that was further filtered off
and vacuum dried. Yield (%): 52. Anal. Calc. for
C19H12N2OAuF5: C, 39.60; H, 2.09; N 4.86%. Found
for 23: C, 40.05; H, 1.85; N, 4.48%. IR (cmꢀ1, Nujol): m
7.58%. IR (cmꢀ1, Nujol): m (C@N) ¼ 1563 cmꢀ1
,
(SO3CFꢀ3 ), 1281 (vs, br), 1225 (s), 1154 (s). H NMR
((CD3)2CO): d 16.4 (s, br, 1H, N-H), 9.05 (d, 1H, H4,
J ¼ 5:1 Hz), 9.01 (s, 1H, HC@N), 8.79 (td, 1H, H2,
J ¼ 8:4 Hz, J ¼ 1:5 Hz), 8.29 (d, 1H, H3, J ¼ 8:7 Hz),
8.27 (td, 1H, H1, J ¼ 8:3, J ¼ 4.8 Hz), 7.60 (m, 2H, H5,
H9), 7.40 (m, 3H, H6, H7, H8), 5.42 (s, 2H, NCH2–Ph).
1

1794
A. Adꢀe et al. / Journal of Organometallic Chemistry 689 (2004) 1788–1795
1
(C@N) ¼ 1596 cmꢀ1. H NMR (CDCl3): d 9.43 (d,1H,
H4), 8.96 (s, 1H, HC@N), 8.70 (d, 1H, H1), 8.07 (t, 1H,
H2), 7.54 (s, br, 1H, H3), 7.57 (m, 2H, H5, H9), 8.47 (m,
3H, H6, H7, H8), 5.14 (s, 2H, N–CH2–Ph). 19F NMR
(CDCl3): )118.25 (dd, br, 2F, Fo), )161.12 (t, 1F, Fp),
)169.94 (t, 2F, Fm). 13C NMR (CDCl3): d 151.46,
149.05, 140.00, 131.77, 131,14, 129.63, 129,63, 129,19,
125.68, 73.19. LSIMSþ: m=z (%) ¼ 576 (97%, [M]þ), 409
(93%, [M–C6F5]þ), 743 (33%, {Au[(C6F5)(2-Py-BN)]2})
1153 (17%, [Au(C6F5)(2-PyBN)]2).
be established unambiguously by NMR although the
crude material can be purified by PCAR-TCL using a
Chromatotron (2 mm layer thickness) as described
previously [3].
4. Supplementary material
Tables of thermal parameters and observed and cal-
culated structure factors have been deposited at the
Cambridge Crystallographic Data Centre. Any request
for this material should quote a full literature citation
and the reference number CCDC 218856 and may be
obtained from The Director, CCDC, 12 Union Road,
Cambridge CB2 1EZ, UK (fax: +44-1233-336033;
e-mail: deposit@ccdc.cam.ac.uk or www: http://www.
ccdc.cam.ac.uk).
3.3. Crystal structure determination of compound 5
Single crystals were grown by diffusing diethyl ether
into an acetone solution of complex [2-PyBN–H][AuCl4]
(5) mounted on a glass fibre using the inert oil technique.
Crystals of 5 were mounted on a glass fibre on a Bruker–
Siemens Smart CCD diffractometer using graphite
monochromated Mo Ka radiation ðk ¼ 0:71073) with a
nominal crystal to detector distance of 4.0 cm.
Acknowledgements
3.3.1. Crystal data and data collection parameters
M ¼ 552:02,
ꢀ
The authors thank the Direccion General de En-
~
5C13H13AuCl4N2O,
a ¼ 11:280(3),
Monoclinic,
senanza Superior for financial support (BQU2002-
04090-C02-01 and BQU2001-2428), and the Centro de
Excelencia BRUKER-I.C.M.A for supporting the X-ray
facilities. M.C. would like to thank the Ministerio de
ꢀ
Ciencia y Tecnologıa – Universidad de Zaragoza for her
Research Contract ‘‘Ramon y Cajal’’.
ꢁ
A,
b ¼ 10:304(2),
c ¼ 14:238(3)
3
ꢁ
b ¼ 99:236(3)°. V ¼ 1633:4(6) A , T ¼ 150 K, space
group P2ð1Þ=n, graphite monochromated Mo Ka radi-
ation, k ¼ 0:71069 A. Z ¼ 4, Dcal ¼ 2:245 mg mꢀ3, F
ꢁ
(0 0 0) ¼ 1040, colourless prism with dimensions
0.39 ꢁ 0.14 ꢁ 0.079 mm, l ¼ 9:659 mmꢀ1, h range for
data collection 2.14–27.02°, ꢀ10 6 h 6 14, ꢀ13 6 k 6 13,
ꢀ17 6 l 6 18; 10016 reflections collected, 3546 indepen-
dent (Rint ¼ 0:0291).
ꢀ
References
[1] H. Yamamoto, in: Lewis Acids in Organic Synthesis, vol. 1,
Wiley-VCH, New York, 2002.
3.3.2. Structure solution and refinement
The structure was solved by direct methods
[2] (a) A.S.K. Hashmi, L. Schwarz, J.-H. Choi, T.M. Frost, Angew.
Chem. Int. Ed. 39 (2000) 2285;
(
SHELXS97) [21] and refined by full-matrix least-squares
on F2, using the program SHELXL97 [22]. All data were
corrected using the program SADABS [23]. The non-hy-
drogen atoms were refined with anisotropic thermal
parameters. All hydrogen atoms were included in idea-
lised positions. Refinement proceeded to: R ¼ 0:0227,
wR ¼ 0:0652 for 195 parameters, and R ¼ 0:0244,
(b) A.S.K. Hashmi, T.M. Frost, J.W. Bats, J. Am. Chem. Soc.
122 (2000) 11553;
(c) N. Asao, K. Takahashi, S. Lee, T. Kasahara, Y. Yamamoto, J.
Am. Chem. Soc. 124 (2002) 12650.
[3] P. Merino, S. Anoro, E. Cerrada, M. Laguna, A. Moreno, T.
Tejero, Molecules 6 (2001) 208.
[4] P. Merino, T. Tejero, M. Laguna, E. Cerrada, A. Moreno, J.A.
ꢀ
Lopez, Org. Biomol. Chem. 1 (2003) 2336.
wR ¼ 0:0660 for all data, electron density fluctuates in
[5] (a) S. Kanemasa, N. Ueno, H. Shirahase, Tetrahedron Lett. 43
(2002) 657, and references therein;
ꢁꢀ3
the range 1.254 and )1.150 e A
.
(b) For reviews see: K.V. Gothelf, K.A. Jørgensen, Chem.
Commun. (2000) 1449;
3.4. Standard catalytic procedure
(c) K.V. Gothelf, K.A. Jørgensen, Chem. Rev. 98 (1998) 863.
[6] R. Casado, M. Contel, M. Laguna, P. Romero, S. Sanz, J. Am.
Chem. Soc. 125 (2003) 11925.
To a solution of 2-PyBN (1) (0.025 g, 0.1125 mmol) in
acetone (20 ml) methylacrilate (2) (0.106 ml, 1.125
mmol) and the amount of gold catalyst specified in
Tables 1 and 3 were added. The reaction mixture was
refluxed and aliquots were taken at different reaction
times (after cooling down the mixture at room temper-
ature). The solvent from the aliquots was completely
removed and the crude residue was analysed by 1H
NMR (CDCl3). The products ratio 3a:3b:3c could
[7] (a) A. Bondi, J. Phys. Chem. 68 (1994) 441;
ꢀ
~
(b) M.A. Esteruelas, A. Lledos, M. Olivan, E. Onate, M.A.
Tajada, G. Ujaque, Organometallics 22 (2003) 3753.
[8] M. Calleja, K. Jonson, W.J. Belcher, J.W. Steed, Inorg. Chem. 40
(2001) 4978, and references therein.
€
[9] (a) For example: R. Huttel, H. Reinheimer, K. Nowak, Tetrahe-
dron Lett. (1967) 1019;
€
(b) R. Huttel, H. Forkl, Chem. Ber. 105 (1972) 2914;
(c) S. Komiya, J.K. Kochi, J. Organomet. Chem. 135 (1977).

A. Adꢀe et al. / Journal of Organometallic Chemistry 689 (2004) 1788–1795
1795
[10] (a) C.M. Mitchell, F.G.A. Stone, J. Chem. Soc., Dalton Trans.
(1972) 102;
[15] Prepared in a similar way to PPN[Au(mes)Cl3] [13].
[16] M. Contel, A.J. Edwards, J. Garrido, M.B. Hursthouse, M.
Laguna, R. Terroba, J. Organomet. Chem. 697 (2000) 129.
(b) J.A. Jarvis, A. Johnson, R.J. Puddephatt, J. Chem. Soc.,
Dalton Trans. (1973) 373;
ꢀ
ꢀ
[17] R. Uson, A. Laguna, L. Garcıa, M. Laguna, Inorg. Chim. Acta 37
(c) A. Johnson, R.J. Puddephatt, J. Chem. Soc., Dalton Trans.
(1977) 1384;
(1979) 201.
ꢀ
[18] R. Uson, A. Laguna, B. Bergareche, J. Organomet. Chem. 184
(1980) 411.
(d) A. Johnson, R.J. Puddephatt, J. Chem. Soc., Dalton Trans.
(1978) 980.
ꢀ
[19] R. Uson, A. Laguna, M. Laguna, A. Abad, J. Organomet. Chem.
249 (1983) 437.
ꢀ
[11] A. Dondoni, S. Franco, F. Junquera, F.L. Merchan, P. Merino, T.
Tejero, Synth. Commun. 24 (1994) 2537.
ꢀ
[20] R. Uson, A. Laguna, Organomet. Synth. 3 (1985) 325.
[21] G.M. Sheldrick, Acta Crystallogr. Sect. A 46 (1990) 467.
[12] P. Braunstein, R.J.H. Clark, J. Chem. Soc., Dalton Trans. (1973)
1845.
[22] G.M. Sheldrick, Program for Crystal Structure Refinement,
€
ꢀ
[13] R. Uson, A. Laguna, M. Laguna, Inorg. Synth. 26 (1989) 85.
[14] M. Contel, J. Jimenez, P.G. Jones, A. Laguna, M. Laguna, J.
University of Gottingen, Germany, 1997.
[23] M. Sheldrick, SADABS empirical absorption program, University
€
of Gottingen, Germany, 1996.
ꢀ
Chem. Soc., Dalton Trans. (1994) 2515.