Comparison of a range of rhodium-based catalysts for the hydroformylation of selected alkenes

Article abstract of DOI:10.1016/S1381-1169(99)00228-9

To determine whether the activity and selectivity of bimetallic catalysts improve during the hydroformylation of alkenes compared to their monometallic counterparts, the hydroformylation of styrene, 1-hexene, and a variety of amino- and phosphino-alkenes using four Rh-based catalyst systems ((OC)₄W(μ-PPh₂)₂RhH(CO)(PPh₃), (OC)₂RhMo(CO)₃(C₅H₄PPh₂), (OC)₂RhW(CO)₃(C₅H₄PPh₂), and [Rh(OAc)₂]₂/PPh₃) was investigated. The Rh-catalyzed reactions of unsaturated amines with H₂/CO differed in chemo- and regioselectivity depending on the catalyst and the reaction conditions. The reactions of unsaturated amides differed in product ratios with varying catalyst systems.

Full text of DOI:10.1016/S1381-1169(99)00228-9

Ž
.
Journal of Molecular Catalysis A: Chemical 150 1999 133–146
www.elsevier.comrlocatermolcata
Comparison of a range of rhodium-based catalysts for the
hydroformylation of selected alkenes
a
a
a
Ron S. Dickson a,), Jenna Bowen , Eva M. Campi , W. Roy Jackson ,
a
a
a
a
Catrin A.M. Jonasson , Felicity J. McGrath , Daniel J. Paslow , Anastasios Polas ,
a
b
Paul Renton , Serafino Gladiali
Department of Chemistry, Monash UniÕersity, Clayton, 3168, Vic., Australia
Departimento di Chimica, UniÕersita di Sassari, Õia Vienna 2, 07100, Sassari, Italy
a
b
`
Received 25 January 1999; received in revised form 15 March 1999; accepted 30 April 1999
Abstract
Four rhodium-based catalyst systems
1,
2,
w
Ž
. x
2
3, and Rh OAc 2rPPh3 have been used in the hydroformylation of 1-hexene,
styrene and some phosphino-, amino- and amido-alkenes. In general the catalysts showed very similar reactivity and
selectivity. q 1999 Elsevier Science B.V. All rights reserved.
Keywords: Hydroformylation; Rhodium catalysts; Bimetallic catalysts
1. Introduction
Hydroformylation of alkenes has been carried out for many years mainly using mononuclear,
w
x
homogeneous catalysts 1,2 . The use of bimetallic catalysts, both homo- and heterobimetallic, has
w x
been studied more recently 3 .
A good deal of work has been focused on the development of heterobimetallic catalysts in the hope
that the presence of two different adjacent metal centres might lead to increased catalytic activity with
w
x
enhanced chemo- and regioselectivity 4–7 . A general problem associated with the use of bi- or
polymetallic catalysts is their tendency to degrade to monometallic species under the reaction
w x
conditions 8 . Various strategies have been adopted to overcome this problem including the use of
w x
w x
bridging dithiolato 9 or bridging diphosphido ligands 10 . Strong evidence for the retention of the
bimetallic system during the catalytic cycle has been provided, at least for the phosphido-bridged
w
x
heterobimetallic rhodium catalysts 11 . Although there have been some claims as to the enhanced
)
Corresponding author.
1381-1169r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved.
Ž
.
PII: S1381-1169 99 00228-9

(
)
134
R.S. Dickson et al.rJournal of Molecular Catalysis A: Chemical 150 1999 133–146
w
x
activity and selectivity of bimetallic catalysts versus their monometallic counterparts 12–18 , no
systematic comparison using a variety of substrates appears to have been carried out.
In this paper we describe a comparison of the results obtained on hydroformylation of 1-hexene,
styrene and a range of amino- and phosphino-alkenes using three heterobimetallic catalysts and a
w
Ž
. x
rhodium only catalyst system, Rh OAc 2 2rPPh3 with a PrRh ratio of 2r1 which we have used
w
x
previously, e.g., Ref. 10 . It has been shown that the rhodium-catalysed reactions of unsaturated
amines with H2rCO can exhibit significant differences in both chemo- and regioselectivity depending
on the catalyst and the reaction conditions 19–21 . Reactions of the phosphinoalkenes can also show
chemoselectivity in that the initially formed aldehydes are further reduced to alcohols when some
w
x
w
x
catalyst systems are used 22 . These substrates were thus considered good candidates for the
comparison of catalytic systems. In addition, it has been shown that reactions of unsaturated amides
show significant differences in product ratios with varying catalyst systems and thus reactions of an
w
x
amide, N-methyl-N-benzylpent-4-enamide 23,24 were also investigated using the above catalyst
systems.
2. Results and discussion
2.1. Catalyst preparation
The heterobimetallic rhodium compounds chosen for study in these hydroformylation reactions
w
x
were the established phosphido-bridged tungsten-rhodium compound 1 25 and the cyclopentadi-
enylphosphine complexes 2 and 3. The latter compounds were made by the method described by
w
x
Casey et al. 26 but their potential catalytic activity does not appear to have been explored. The
molybdenum compound 2 was prepared in 44% overall yield and the tungsten compound 3 in 37%
Ž
.
Ž
.
Ž
.
w
Ž
.
x
yield from the sequential reaction of CH3CN M CO with Li C5H4PPh2 and Rh CO Cl . It
3
3
2
Ž
.
2 Ž
.
was observed that, in contrast to the 4 h required for preparation of the reagent CH3CN Mo CO ,
the tungsten analogue required prolonged reflux 120 h .
3
3
Ž
.
2.2. Catalytic reactions with styrene and 1-hexene
Ž
.
Reactions of styrene with H2rCO 1:1, 400 psi, 2.76 MPa for 20 h at 508C were carried out using
w
Ž
.
x
Ž
.Ž
.
Rh OAc 2 2rPPh3, HRh CO PPh3 3, and the bimetallic catalysts 1, 2 and 3. All reactions gave
complete conversion to a mixture of the branched and linear aldehydes in a ratio 96:4 "2 . These
Ž
.

(
)
R.S. Dickson et al.rJournal of Molecular Catalysis A: Chemical 150 1999 133–146
135
ratios are identical to those reported previously for reactions of styrene under similar conditions using
w
x
w x
mono 27 and bimetallic 28 catalysts. It was thought that the steric requirements around the Rh
Ž
.Ž
.
atom in the phosphido-bridged complex 1 may be less than in HRh CO PPh3 3 due to the ligand
Ž
.
phosphorus atoms being pulled back into the bridge ring system, but no increase in the percentage of
branched aldehyde was observed. The cyclopentadienylphosphine complexes 2 and 3 also appear to
Ž
.Ž
.
have reduced steric requirements relative to HRh CO PPh3 3 but again no significant change in
regioselectivity was observed.
w x
The hydroformylation of unsubstituted 1-alkenes generally favours linear aldehyde formation 1 .
Ž
.
w
Ž
. x
Reactions of 1-hexene with H2rCO 1:1, 400 psi, 2.76 MPa for 20 h at 608C using Rh OAc 2 2rPPh3
and the bimetallic catalysts 2 and 3 resulted in complete conversion to a mixture of linear and
branched aldehydes. There was only a small variation in regioselectivity from 60:40 for
w
Ž
. x
Rh OAc 2 2rPPh3 to 70:30 for complex 3.
The possibility that these reactions involved breakdown of the heterobimetallic complexes and
catalysis by a monomeric rhodium species thus had to be considered. Evidence for retention of the
w
x
metal–metal bond in the phosphido-bridged species 1 has been discussed previously 11 . A reaction
of 1-hexene using the cyclopentadienylphosphine complex 3 as a catalyst was carried out in which the
1
ratio of substrate to catalyst was 4:1. After removal of C7-aldehydes the residue was analysed by H
and 31P NMR and IR spectroscopy. The 31P NMR was similar to that recorded for 3 and showed the
satellites associated with W–Rh–P coupling. Full identification of the compound after reaction was
1
not possible. Although the H as well as the 31P NMR spectrum was very similar to that of 3, the IR
spectrum showed interesting changes in the region associated with Rh–CO stretching frequencies.
2.3. Catalytic reactions of phosphinoalkenes
The phosphinoalkenes 4; ns1, 2 and 3 were reacted with H2rCO at 808C for 20 h in the presence
of the cyclopentadienylphosphine compounds 2 and 3 and the results compared with published data
w
x
w
Ž
.
x w
x
Ž
.
for the bimetallic complex 1 10,11 and Rh OAc
22 . Table 1 .
2 2
Reactions of the phosphinoalkene 4; ns2 were regiospecific in all cases. Reactions at 808C led to
further reduction of the initially formed aldehyde 5; ns2 to the alcohol 6; ns2. The reduction was
complete in all cases except for the reaction using the W–Rh cyclopentadienylphosphine complex 3,
w
Ž
.2 x2
which gave almost equal amounts of aldehyde and alcohol. Reactions using Rh OAc at lower
w
x
temperatures also gave varying amounts of aldehyde 22 .

(
)
136
R.S. Dickson et al.rJournal of Molecular Catalysis A: Chemical 150 1999 133–146
Table 1
Reactions of phosphinoalkenes 4a
Catalyst system
Reactant 4
Product ratios
5:
6:
7:
8
w
Ž
. x
Rh OAc 2 rPPh3
ns1
ns1
ns1
ns1
–
–
–
13
30
36
28
–
–
–
–
18
20b
18c
30d
2
1
2
3
e
–
w
Ž
. x
Ž
.
.
.
Rh OAc 2 rPPh3
ns2
ns2
ns2
ns2
–
–
–
57
100 86%
–
–
–
–
–
–
–
–
2
Ž
1
2
3
100 77%
Ž
100 79%
43
f
w
Ž
. x
Rh OAc 2 rPPh3
ns3
ns3
ns3
ns3
2
1
2
3
94
100 89%
100
17
–
–
6
–
–
–
–
–
Ž
.
a
Reactions with H2 rCO, 1:1 400 psi, 2.76 MPa at 808C for 20 h with a 1:100 Rh:substrate ratio. Product ratios determined from 1H NMR
Ž
.
spectra; isolated yields of products in parenthesis. Each reaction has been carried out at least twice. Variation in yields and product ratios
w
Ž
. x
were F3% except for reaction of 4; ns3 with Rh OAc 2 rPPh3.
2
b50% starting material.
c46% starting material.
d42% starting material.
e69% starting material.
f
Ž
.
13–50% of Ph2 P CH2 4CH3 formed. Branched and linear products each a mixture of aldehydes and alcohols, branched:linear ratio 85:15
Ž
.
"5 .
The reaction of the phosphinoalkene 4; ns3 using the Mo–Rh cyclopentadienylphosphine 2 was
highly chemo- and regioselective giving only the branched-chain aldehyde 5; ns3, a pure sample of
Ž
.
which was obtained in high yield 89% . This selectivity for branched aldehyde formation was also
displayed by the W–Rh cyclopentadienylphosphine complex 3. These catalysts are thus superior in
w
Ž
. x
that the Rh OAc 2 2rPPh3 system gave significant amounts of double bond saturation and the
tungsten–rhodium phosphido-bridged catalyst 1 was not regiospecific leading to the formation of
w
x
about 6% linear aldehyde 10 .
Reactions of the allylic phosphine 4; ns1 have been shown previously to be significantly slower
w
x
than their longer chain homologues 11,22 . This trend was again confirmed in reactions of the allylic
phosphine using the Mo–Rh cyclopentadienylphosphine 2 and the related W–Rh compound 3 as
catalyst which gave only 60 and 30% conversions, respectively, under the standard conditions. This
reaction also showed poor regioselectivity giving an approximately 50:50 ratio of the alcohols 6;
w
x
ns1 and 8; ns1 for the Mo–Rh compound possibly for reasons suggested previously 11 . The
corresponding W–Rh compound 3 surprisingly gave the aldehydes 5; ns1 and 7; ns1 again with
poor regioselectivity.
(
)
2.4. Catalytic reactions of N-benzyl and N- 1-phenylethyl pent-4-enylamine 9
Zhang and Ojima have demonstrated that the regioselectivity of reactions of N-benzylpent-4-en-
ylamine 9a is sensitive to reaction conditions. Very high regioselectivities for formation of either the
Ž
.Ž
.
3-methylpiperidine 14a or the azepane 15a can be achieved using HRh CO PPh3 3 with varying
w
x
amounts of added phosphine ligand 21 . Formation of 14 and 15 involves cyclisation of the initially

(
)
R.S. Dickson et al.rJournal of Molecular Catalysis A: Chemical 150 1999 133–146
137
Table 2
Reactions of N-benzylpent-4-enamine 9a with H2 rCOa
Product ratiosb
Ž
.
Entry no.
Catalyst system
T
8C
12:
13:
14:
15:
17
w
w
1
1
2
3
Ž
. x
2
1
2
3
4
5
6
Rh OAc 2 rPPh3
60
60
60
40
60
60
28
–
29
27
20
19
19
40
–
16
17
16
10
–
12
–
–
–
22
20
59
21
–
21
40
–
.2 x
c
Ž
Rh OAc 2 rBIPHEPHOS
15d
15e
12f
13
a
Ž
.
Reactions for 20 h with H2 rCO 400 psi, 2.76 MPa .
b Based on integrals from 1H NMR spectra of the total product.
c Implies -5% of compound detected by 1H NMR or GLC.
d 21% starting material.
e48% starting material.
f39% starting material.
formed branched 10 and linear 11 aldehydes to form the cyclic enamines 12 and 13 which are
subsequently hydrogenated under the relatively severe reaction conditions of 1008C and 1800 psi,
12.40 MPa of H2rCO. In order to allow for maximum discrimination, we have reacted both the
Ž
.
benzyl and N- 1-phenylethyl pentenylamines 9 with H2rCO in the presence of the four catalyst
w
Ž
. x
systems Rh OAc 2 2rPPh3, 1, 2, and 3 under milder conditions. Reactions of the N-benzyl
compound 9a are summarised in Table 2 and of the N- 1-phenylethyl compound 9b in Table 3.
Ž
.
1
The products were identified by H NMR spectroscopy and GLC involving comparisons with
literature data andror with authentic samples except for the enamine 13 and its dimerisation product
17. The structures of 13 and 17 were inferred from spectral data and by hydrogenation to the saturated
1
heterocycles 15 and 18. Product ratios are based on integration of H NMR spectra as it proved very
Ž
.
difficult to achieve GLC separation of the regioisomers 12r13 and 14r15 Scheme 1 .
Table 3
Reactions of N-1-phenylethylpent-4-enamine 9b with H2 rCO
Entry no.
Catalyst system
T
Pa psi
Product ratiosb
Ž
8C
.
Ž .
Time h
Ž
.
12:
13:
14:
15:
16:
17
c
w
. x
7
8
9
10
11
12
13
14
15
16
Rh OAc 2 rPPh3
80
80
80
80
80
80
60
80
80
80
20
20
20
60
60
20
20
20
20
20
1200
1500
400
1500
1500
1200
400
400
400
400
39
35
32
28
27
14
5
39
34
20
34
27
–
–
–
38
49
–
18
–
–
–
–
–
–
–
12
–
–
–
–
–
–
27
26
31
w
.2 x
Rh OAc 2 rPPh3
12
–
w
.2 x
Rh OAc 2 rPPh3
37
40
63
15
–
40
18
63
2
d
Ž
Ž
.Ž
.Ž
.
.
HRh CO PPh3
–
–
–
–
–
3
HRh CO PPh3
10
33
46
21
17
3
w
w
1
2
3
Ž
. x
Rh OAc 2 rBIPHEPHOS
Ž
.2 x
Rh OAc 2 rBIPHEPHOS
2
–
13
7
e
–
a400, 1200 and 1500 psi are equivalent to 2.76, 8.28 and 10.34 MPa.
b Based on integrals from 1H NMR spectra of the total product.
c Implies -5% of compound detected by 1H NMR or GLC.
d 20% of the linear aldehyde 11b was observed.
e10% starting material.

(
)
138
R.S. Dickson et al.rJournal of Molecular Catalysis A: Chemical 150 1999 133–146
Scheme 1.
2.4.1. Reactions of N-benzylpent-4-enylamine 9a
Reactions of 9a with H2rCO at 608C gave good conversions to products with the exception of the
cyclopentadienylphosphine complexes 2 and 3 which gave 40–50% of recovered starting material.
The products in all cases, except reaction 2 using the bulky BIPHEPHOS ligand, were a mixture of
the 6- and 7-membered cyclic enamines 12a and 13a, their saturated analogues 14a and 15a and the
dimeric material 17a arising from self-condensation of the enamine 13a. Only the dimer arising from
the 7-ring enamine 13a was detected. Specificity in enamine dimerisation has been previously noted
w
x
where mixtures of 5- and 6-membered enamines only led to cross-coupled products 20 . The
w
x
reactivity of these cyclic enamines has been well established 29,30 . Although the fates of the
Ž
.
intermediate enamines varied with catalyst cf. reactions 1, 3, 5 and 6 the overall regioselectivity for
the branched to linear products 10 and 11 varied little. Thus, when the percentages of 12aq14a vs.
13aq15aq2=17a were compared, the regioselectivities varied only from 40:60 for reaction 3 to
Ž
.
Ž
.
30:70 for reactions 1, 4, 5 and 6. These ratios, which are typical of those obtained from reactions of
1-alkenes, show that no chelation control of selectivity by the amine nitrogen is occurring and suggest
that all catalyst centres are of comparable size. The ability of the different catalysts to hydrogenate the
intermediate enamines differed markedly. No hydrogenation occurred with the cyclopentadienylphos-
Ž
.
phine compound 2 reaction 5 and in contrast a majority of the product was saturated when the
Ž
.
phosphido-bridged compound 1 was used reaction 3 . Other catalysts gave intermediate degrees of
enamine hydrogenations in the range 18% reaction 6 to 26% reaction 1 . The results tentatively
Ž
.
Ž
.
suggest that bimetallic W–Rh complexes, e.g., 1 and 3, are more likely to promote enamine

(
)
R.S. Dickson et al.rJournal of Molecular Catalysis A: Chemical 150 1999 133–146
139
w
Ž
. x
hydrogenation than the Mo–Rh compound 2 although the Rh OAc 2 2rPPh3 system promoted
hydrogenation to a similar degree. It is of interest that this variation appears to be generally associated
with hydrogenation of the seven-membered enamine 13, e.g., in reaction 3 all of 13 had been
hydrogenated to give 15 whereas only 29% of 12 had been converted into its saturated analogue 14.
None of the catalysts led to hydrogenation of the alkene in the starting material.
Ž
A reaction using the W–Rh compound 1 at 408C gave significantly less saturated material reaction
.
4 and a concurrent formation of dimeric material.
w
x
A reaction was carried out using the bulky, bidentate phosphite ligand BIPHEPHOS 19,31 which
was highly regioselective giving less than 5% of materials derived from the branched aldehyde 10.
(
)
2.4.2. Reactions of N- 1-phenylethyl pent-4-enylamine 9b
Reactions of the homologue 9b of the benzyl compound 9a were carried out under a variety of
Ž
.
conditions. A reaction using BIPHEPHOS as ligand reaction 13 gave a very similar result to that of
Ž
.
the benzyl compound 9a under identical conditions reaction 2 . Again, the products mainly arose
from a highly regioselective linear hydroformylation, the only difference being that no hydrogenated
Ž
.
material 15b was observed. Some hydrogenation occurred during a reaction at 808C reaction 12 .
Other reactions were carried out at 808C in order to make comparisons with the reactions of the
Ž
.
unsaturated phosphines see Table 1 . All reactions went to completion at this temperature with the
exception of reaction 16 using the W–Rh cyclopentadienylphosphine compound 3. This compound
Ž
.
was one of the less active catalysts in reactions of the benzyl homologue 9a see Table 2 . The
reactions all showed products originating from an initial hydroformylation with about 30:70 branched
Ž
.
to linear regioselectivity reactions 9, 14, 15 and 16 . All showed the presence of significant amounts
of at least one of the unsaturated products 12b, 13b and 17b. However, the ease of hydrogenation of
Ž
.
the intermediate enamines 12 and 13, RsCH CH3 Ph showed very different variations with catalyst
to those displayed in reactions of the benzyl analogue 9a. Thus, e.g. the Mo–Rh catalyst 2 gave only
Ž
.
unsaturated material from reactions of 9a reaction 5 but significant amount of saturated products
Ž
.
14b and 15b were obtained from comparable reactions of 9b reaction 15 . In contrast reactions using
the phosphido-bridged W–Rh catalyst 1 gave significantly less saturated material from a reaction of
Ž
.
9b than for a reaction of 9a cf. reactions 14 and 3 . It is unlikely that these differences can be
explained by the different reaction temperatures, 608C for 9a and 808C for 9b. Again preferential
hydrogenation of the seven-ring intermediate was generally observed.

(
)
140
R.S. Dickson et al.rJournal of Molecular Catalysis A: Chemical 150 1999 133–146
Reactions at 808C but using higher pressures of H2rCO were also carried out. Reactions using
w
Ž
.
x
Ž
.
Rh OAc 2 2rPPh3 under these conditions reactions 7 and 8 surprisingly also gave significant
Ž
.
amounts )30% of unsaturated materials 12b, 13b and 17b with only a small amount of
Ž
.
hydrogenated material 14b being observed reaction 9 . In addition, the 6-ring enamine 12b was
Ž
.Ž
.
Ž
found in reactions using 1500 psi, 10.34 MPa of H2rCO with HRh CO PPh3 3 as catalyst reactions
.
Ž
.
10 and 11 . These reactions were carried out by different workers and in one case reaction 10 gave
some uncyclised linear aldehyde 11b together with the aldehyde 16b resulting from a second
hydroformylation of the enamine 12b. In contrast, reaction 11 under apparently identical conditions
gave the saturated azepane 15b as the major product.
2.5. Reactions of N-benzyl-N-methylpent-4-enamide 19
w
Ž
. x
Reactions of the amide 19 with H2rCO using Rh OAc 2 2rPPh3 and the bimetallic W–Rh
Ž
.
catalysts 1 and 3 gave very similar ratios of the linear 20- and branched 21-aldehydes about 60:40 .
Quantitative conversion was obtained at 608C with the first two catalysts but only approx. 50%
conversion was obtained using the cyclopentadienylphosphine compound 3. Reactions using the
bimetallic catalysts 1 and 3 at 408C were very slow but the aldehydes obtained using 1 were still in
the ratio 60:40 for 20:21. Thus, these results show no evidence for chelation of the amido oxygen
atom in the transition state which would lead to a preference for the branched product 21. This result
thus contrasts with reported reactions of pent-4-enamide itself using a range of rhodium catalysts at
1008C which gave only 5-methyl-2,3-dihydropyridin-2-one in excellent yields whose formation was
rationalised in terms of amide-directed chelation control favouring initial formation of the branched
w
x
aldehyde 24 . The influence of N-substitution on the regiochemistry of the initial hydroformylation
reaction is possibly due to steric hindrance to chelation at the amido-oxygen. However, it has been
previously shown that the regioselectivity of the initial hydroformylation of some mono-N-substituted
w
x
but-3-enamides is not sensitive to the bulk of the substituent on nitrogen 24 .
2.6. Conclusions
The variation in catalyst structure as a result of introducing bimetallic systems with phosphido
bridges, i.e., catalysts 1, 2 and 3 leads to very little change in the regioselectivity of hydroformylation
of styrene, 1-hexene, some phosphinoalkenes and N-substituted pentenylamines and pentenylamides.
The possibility that the heterobimetallic compounds were breaking down to monomeric rhodium
species which then acted as catalysts was examined. Reactions using only a low substrate to catalyst
ratio 4:1 allowed isolation of metallic species and the 31P NMR spectrum of the product from a
reaction of 1-hexene using 3 as catalyst showed retention of the W–Rh bond as W–Rh–P coupling
was still observed. The regiochemistry of reactions of the phosphinoalkenes 4; ns2 or 3 was
Ž
.

(
)
R.S. Dickson et al.rJournal of Molecular Catalysis A: Chemical 150 1999 133–146
141
dominated by chelation by phosphorus but reactions of the alkenylamines 9 showed no evidence for
chelation by nitrogen. The results from the reactions with N-benzylpent-4-enylamine 9a contrast with
those reported previously where high regioselectivity for the product 14a arising from N-chelation
Ž
.Ž
.
Ž
was obtained, e.g., using HRh CO PPh3 3 as catalyst at 1008C with 1:1 H2rCO 1800 psi, 12.40
.
w x
MPa 21 . In general, the reactions of the closely related amino substrates 9a and 9b showed
Ž
.
considerable variation in chemoselectivity see Tables 2 and 3 using the same catalyst although the
reaction temperature was 208C higher for reactions of 9b. In view of the complex mixtures of
products obtained from these reactions it appears that they are of very limited preparative value.
It appears that the regiochemistry of reaction is not influenced by the subtle differences in
electronic distribution between the catalysts. In contrast, reactions of alkenylamines 9 were found to
be very susceptible to steric effects associated with the use of the bulky bisphosphite ligand
w
x
BIPHEPHOS 19,31,32 .
3. Experimental
3.1. General methods
1
Most general methods were as described in a recent paper 33 . H, 13C and 31P NMR spectra were
w
x
Ž
.
recorded in CDCl3 solution. Gas chromatography GLC analysis was carried out using a 3700 Varian
gas chromatograph using either a capillary 12QC5rBP5 column or a packed SE-30 column. Gas
Ž
.
chromatography–mass spectroscopy GCMS was carried out using a Hewlett-Packard 5890 gas
Ž
.
Ž
.
chromatograph with a BP5 column 25 m coupled to a VG-TRIO-1 mass spectrometer 70 eV .
3.2. Preparation of substrates
w
x
The phosphinoalkenes 4; ns1,2 and 3 were prepared as described previously 10,22 . N-Phenyl-
w
x
methylpent-4-enylamine 9a was prepared by LiAlH4 reduction of the corresponding amide 21 .
Ž
.
w x
N-1- Phenylethyl pent-4-enylamine 9b b.p. 100–1058Cr0.1 mm 34 was prepared in a similar
Ž
.
manner by LiAlH4 reduction of N- 1-phenylethyl pent-4-enamide.
3.3. Preparation of catalysts
Ž
.
Ž
.Ž
)
Hydridocarbonyltris triphenylphosphine rhodium HRh CO PPh3 3 was purchased from Aldrich.
2 x2
w
Ž
.
w x
Rhodium acetate dimer, Rh OAc
was prepared as described by Legzdins et al. 35 . The
phosphido bridged rhodium–tungsten compound
1 was prepared as
w
x
described by Shulman et al. 25 .
5
1
(
)
3.3.1. Tricarbonylmolybdenum m-h :h -cyclopentadienyldiphenylphosphine dicarbonylrhodium,
5
1
(
)
(
) [ ]
OC 2 Rh-Mo CO 3 m-h :h -C5 H4 PPh2 2
Ž
.
w
Ž
. Ž
.x
Lithium cyclopentadienyldiphenylphosphine tricarbonylmolybdate, Li Mo CO C5H4PPh2 was
3
Ž
.
Ž
.
w x
prepared by reaction of CH3CN Mo CO with lithium diphenylphosphinocyclopentadienide 36 .
Reaction of this salt with m-dichlorotetracarbonyldirhodium, Rh CO Cl 37 was carried out under
conditions described for the preparation of the corresponding bis- p-tolyl phosphine complex 26 and
gave the title compound 2 as yellow-green crystals 49% m.p., 173–1758C decomp. . IR CH2Cl2
3
3
w
Ž
Ž
.
x w
x
2
2
.
w x
Ž
.
Ž
.
Ž
.

(
)
142
R.S. Dickson et al.rJournal of Molecular Catalysis A: Chemical 150 1999 133–146
1
y
1
Ž
.
Ž
.
Ž
.
n CO : 2063 s, 1991 m, 1958 s, 1880 w, 1861 m cm
.
H NMR 400 MHz : d ppm 4.41
Ž
.
Ž
pseudoquartet, Jf2 Hz, 2 H, H3 and H4 of Cp , 5.50 pseudoquartet, Jf2 Hz, 2 H, H2 and H5 of
Cp , 7.52 m, 6 H, m- and p-H of C6H5 , 7.94 m, 4 H, o-H of C6H5 . P NMR 162 MHz : d
31
.
Ž
.
Ž
.
Ž
.
13
Ž
.
Ž
.
Ž
Ž
. .
Ž
.
Ž
ppm q44.3 d, JRh – P 116 Hz . C NMR 50 MHz, CDCl3 qCr acac : d ppm 58.4 d, JP –C
3
.
Ž
.
Ž
Ž
.
Ž
.
53.5 Hz, C1 of Cp , 92.2 m, Cp , 120.0 d, JP –C 49.2 Hz, ArC , 129.3 d, JP –C 11.1 Hz, m-ArCH ,
Ž
.
.
Ž
Ž
. .
Ž
132.2 d, JP –C 2.4 Hz, p-ArCH , 134.0 d, JP –C 14.1 Hz, o-ArCH , 221.6 s, Mo CO , 233.4 s,
Mo CO . Elemental analysis % , found: C, 44.7; H, 2.5; P, 5.0. C22 H14MoO5PRh requires C, 44.9;
2
Ž
..
Ž .
H, 2.4; P, 5.3.
5
1
(
)
3.3.2. Tricarbonyltungsten m-h :h -cyclopentadienyldiphenylphosphine dicarbonylrhodium,
5
1
(
)
(
) [ ]
OC 2 Rh-W CO 3 m-h :h -C5 H4 PPh2 3
w
x
The W–Rh analogue 3 was prepared in a similar manner to that described above 26,36 .
Ž
.
Ž
.
Ž
.
CH3CN W CO was prepared by reaction of W CO and CH3CN under reflux for 120 h and
3
3
6
w
Ž
.
reaction with lithium diphenylphosphinocyclopentadienide gave a THF solvate of Li W CO -
C5H4PPh2 . Reaction of this salt with Rh CO Cl gave the title compound 3 as olive green
3
Ž
.
x
w
Ž
Ž
.
x
2
Ž
.
Ž
.
.
2 Ž
.
Ž .
crystals 41% m.p. 162-1658C decomp . IR Nujol n CO : 2060 s, 1986 s, 1942 s br , 1864 s,
1
1855 s, 1842 s cmy . H NMR C6D6, 400 MHz : d ppm 3.87 pseudoquartet, J 2 Hz, 2 H, H3 and
1
Ž
.
Ž
.
Ž
.
Ž
Ž
H4 of Cp , 4.64 pseudoquartet, J 2 Hz, 2 H, H2 and H5 of Cp, 6.90–7.01 m, 6 H, m- and p-H of
C6H5 , 7.69 dd, JP – H 12.5, JH – H 8.0 Hz, o-H of C6H5 . C NMR C6D6, 100 MHz : d ppm 59.6
d, JP –C 52.5 Hz, C1 of Cp , 89.7 s, C3 of Cp , 89.7 d, JP –C 5.6 Hz, C2 of Cp , 129.3 d, JP –C
11.2 Hz, m-ArCH , 129.4 d, JP –C 48.3 Hz, ArC , 132.0 d, JP –C 2.5 Hz, p-ArCH , 134.1 d, JP –C
13
.
Ž
.
Ž
Ž
.
Ž
.
Ž
.
Ž
.
.
.
Ž
Ž
.
.
Ž
.
Ž
Ž
Ž
.
.
Ž
14.1 Hz, o-ArCH , 184.3 bs, Rh CO , trans to P , 200.2 dd, JRh –C 64.4 Hz, JP –C 113.7 Hz,
Rh CO , trans to W , 210.3 s, with 183W satellites, JW–C 169 Hz, W CO , 223.1 s with 183W
Ž
.
.
Ž
Ž
. .
Ž
2
31
Ž
..
Ž
.
Ž
.
Ž
satellites, JW–C 165 Hz, W CO . P NMR C6D6, 121.5 MHz : d ppm 41.0 d, JRh – P 122.5 Hz,
with 183W satellites, JW– P 21.5 Hz . Elemental analysis % , found: C, 39.2; H, 2.0. C22 H14WO5Rh
.
Ž .
requires C. 39.1; H, 2.1.
3.4. Hydroformylation reactions
3.4.1. General conditions
Ž
.
Reactions were carried out using H2rCO 1:1 in the initial pressure range 400–1500 psi
Ž
.
2.76–10.34 MPa in a Parr 100 ml stainless steel autoclave with a glass liner. The temperature was
measured using a thermocouple inserted between the autoclave and the heating block. All reactions
were carried out at temperatures in the range 40–808C in magnetically stirred ethyl acetate or benzene
solutions. The substrate:Rh ratio was about 100:1. The reactions were allowed to cool and the solvent
removed under vacuum. When air-sensitive samples were hydroformylated, the samples were
transferred to a Schlenk flask under an atmosphere of nitrogen. The solvent was then evaporated
1
under vacuum. Isomer ratios were determined from the H NMR spectra of the crude reaction
mixture. GLC was also used for determining the isomer ratios.
3.4.2. Reactions of styrene and 1-hexene
5
1
Ž
.
Ž
. Ž
.
Ž
.
In a typical reaction, OC RhMo CO m-h :h -C5H4PPh2 2 16.6 mg, 0.0282 mmol and
2
3
Ž
. Ž
.
.
Ž
.
styrene 0.29 g, 0.28 mmol ratio 1:100 in benzene 15 ml were reacted with hydrogen and carbon
monoxide 400 psi, 2.76 MPa according to the general procedure for 22 h at 508C. Evaporation of the
Ž
1
Ž
.
solvent gave the crude product as a dark oil 0.36 g, 97% . A H NMR spectrum indicated that no

(
)
R.S. Dickson et al.rJournal of Molecular Catalysis A: Chemical 150 1999 133–146
143
starting material was present and that the crude product consisted of a mixture of the two aldehyde
isomers 2-phenylpropanal and 3-phenylpropanal in the ratio 95:5.
w
Ž
.
x
Reactions of 1-hexene were carried out under identical conditions using Rh OAc rPPh3 and the
2
cyclopentadienylphosphine 3. The product ratio was based on integration of the CHO absorptions at d
1
Ž
.
Ž
.
9.76 t, J 1.9 Hz for heptanal and at d 9.61 d, J 2.1 Hz for 2-methylhexanal in the H NMR
spectrum of the total product.
Ž
.
A reaction was carried out using 1-hexene 13.6 mg, 0.16 mmol and the cyclopentadienyphos-
Ž
.
Ž
.
Ž
.
phine 3 27.4 mg, 0.04 mmol in benzene 10 ml with H2rCO, 400 psi, 2.76 MPa at 608C for 20 h.
Ž
.
The solvent and product aldehydes were distilled off under reduced pressure 0.1 mm at 208C and the
Ž
.
Ž .
residue, a yellow-brown solid examined spectroscopically. IR C6D6 nCO 2062 s, 1987 s, 1956 s br ,
1
1881 s and 1861 s cmy1. H and 31P NMR spectra were identical to those reported above for 3.
3.4.3. Reactions of the phosphinoalkenes 4; ns1, 2, and 3
Hydroformylation of the phosphinoalkenes 4 was carried out using the appropriate catalyst at 808C
for 20 h and the resulting products analysed by 1H and 31P NMR spectroscopy. Product identification
w
x
was by comparison with published data 22 .
3.4.4. Reactions of aminoalkenes 9a and 9b
Reactions of N-phenylmethylpent-4-enamine 9a were carried out as described above under a range
of conditions as summarised in Table 2. The product ratios were determined using 1H NMR
spectroscopy by integration of the olefinic protons in compounds 12, 13, and 17, the benzyl CH2 of
12, 14 and 15 and the CH3 doublet of 14. The spectra were assigned on the following basis: A sample
of 5-methyl-1-phenylmethyl-1,2,3,4-tetrahydropyridine 12a was isolated from the products of reaction
1
Ž
.
Ž
.
Ž
.
Ž
.
Ž
.
1 Table 2 as an oil, H NMR 400 MHz : d ppm 1.59 s, 3 H, CH3 , 1.93 m, 4 H, H3, H4 , 2.70
13
Ž
Ž
Ž
.
Ž
.
Ž
.
Ž
.
Ž
.
m, 2 H, H2 , 3.80 s, 2 H, C H2 Ph , 5.71 bs, 1 H, H6 , 7.23 m, 5 H, Ph . C NMR 100 MHz : d
ppm 21.25, CH3 , 22.9, 27.2 C3, C4 , 54.2 C2 , 60.2 CH2 Ph , 106.9 C5 , 127.2, 128.3, 128.45
.
Ž
.
Ž
.
Ž
.
Ž
.
Ž
.
X
q
.
Ž
.
Ž
Ž
.
Ž
.
Ž
.
Ž
.
ArCH , 131.7 C6 , 139.0 C1 . MS ESI : mrz 188.1 MqH . HRMS ESI mrz found:
188.1440. Calc. for C13H18N q 188.1439. The spectral data were in agreement with literature values
.
w
x
38 .
1
Ž
.
.
Ž
.
Ž
1-Phenylmethyl-3-methylpiperidine 14a had H NMR 400 MHz : d ppm 0.83 d, J 6.0 Hz, 3 H,
CH3 , 1.68 m, 5 H, H3, H4, H5 , 2.74 m, 4 H, H2, H6 , 3.43 s, 2 H, C H2 Ph , 7.23 m, 5 H, Ph .
C NMR 100 MHz : d ppm 20.0 CH3 , 25.6 C4 or C5 , 31.3 C3 , 33.3 C4 or C5 , 54.2, 62.15,
.
Ž
.
Ž
Ž
.
Ž
.
.
13
Ž
.
Ž
.
Ž
.
Ž
.
Ž
.
Ž
Ž
q
Ž
.
Ž
.
.
Ž
.
.
Ž
.
63.8 C2, C6, CH2 Ph , 126.9, 128.1, 128.3 ArCH . MS ESI : mrz 189.9 MqH . HRMS ESI
q
Ž
mrz found: 190.1595. Calc. for C13H20N : 190.1595. Data were in agreement with literature
w
x
values 21 .
Ž
.
Ž
A sample of 1- phenylmethyl azepane 15a prepared by benzylation of hexamethylenimine 15;
1
.
Ž
.
Ž
.
Ž
.
Ž
RsH had H NMR 400 MHz : d ppm 1.62 bs, 8 H, H3, H4, H5, H6 , 2.69 t, J 5.3 Hz, 4 H,
13
.
Ž
.
Ž
.
Ž
.
Ž
Ž
.
.
Ž
H2, H7 , 3.60 s, 2 H, C H2 Ph , 7.23 m, 5 H, Ph . C NMR 100 MHz : d ppm 26.6, 27.8 C3,
.
Ž
.
Ž
.
Ž
Ž
.
Ž
.
C4, C5, C6 , 55.2 C2, C7 , 62.3 CH2 Ph , 126.3, 127.7, 128.4 ArCH , 139.7 ArC . MS ESI : mrz
q
q
Ž
.
Ž
.
.
189.9 MqH . HRMS ESI mrz found: 190.1595. Calc. for C13H20N 190.1595. Data were in
w
x
agreement with literature values 21 .
1
Ž
.
1- Phenylmethyl -1,2,3,4,5-tetrahydro-1H-azepine 13a not isolated as a pure compound had H
NMR 400 MHz : d ppm 5.18 dt, J 15.3, 7.2 Hz, 1 H, H6 , 5.93 d, J 15.3 Hz, 1 H, H7 . 13C
Ž
Ž
.
.
Ž
.
.
Ž
.
Ž
.
q
Ž
Ž
.
Ž
.
.
Ž
.
Ž
.
NMR 100 MHz : d ppm 119.45 C6 , 134.45 C7 . GCMS: mrz 187 M , 26% , 186 30 , 172
q
Ž
.
Ž
.
Ž
.
Ž
.
Ž
Ž
.
17 , 96 22 , 91 100 , 68 50 . HRMS ESI mrz found: 188.1440. Calc. for C13H18N :

(
)
144
R.S. Dickson et al.rJournal of Molecular Catalysis A: Chemical 150 1999 133–146
Ž
.
188.1439. Structure proof came from hydrogenation of the product from reaction 2 Table 2 using
2 x2
w
Ž
.
Ž
. Ž . Ž .
3.0 mg , BIPHEPHOS 10.0 mg and H2 400 psi at 808C for 20 h. The product was
Rh OAc
shown to be a mixture of the azepane 15a together with the saturated dimer 18a.
1
Ž
.
6- 1-phenylmethylazepan-2-yl -1-phenylmethyl-2,3,4,5-tetrahydro-1H-azepine 17a had H NMR
400 MHz : d ppm 6.04 bs, 1 H, H7 . C NMR 100 MHz : d ppm 114.0 C6 , 141.8 C7 . MS
13
Ž
Ž
.
Ž
Ž
.
Ž
.
Ž
.
Ž
.
Ž
.
Ž
.
q
.
.
Ž
.
Ž
.
Ž
.
EI : mrz 376 M , 4% , 285 12 , 200 38 , 91 100 .
Hydrogenation of the product from reaction 2 as described above gave the saturated dimer 18a.
Ž
.
Similar reactions of N- 1-phenylethyl pent-4-enamine 9b are summarised in Table 3. The product
1
ratios were again established on the basis of H NMR spectra using the olefinic protons of 12b, 13b
and 17b and the C H Ph quartets. Spectral assignments were made on the following basis: 5-methyl-1-
Ž
.
1-phenylethyl -1,2,3,4-tetrahydropyridine 12b was isolated by chromatography of the product from
1
Ž
.
Ž
.
Ž
.
Ž
reaction 7 Table 3 on neutral alumina. H NMR 300 MHz : d ppm 1.44 d, J 6.9 Hz, 3 H,
C H3CH , 1.59 bs, 3 H, CH3–C5 , 1.76–1.88 m, 4 H, H3, H4 , 2.71 m, 2 H, H2 , 3.99 q, J 6.9
Hz, 1 H, C HCH3 , 5.86 bs, 1 H, H6 , 7.22–7.32 m, 5 H, Ph . C NMR 50 MHz : d ppm 18.0
Ž
Ž
.
Ž
.
Ž
.
Ž
.
.
Ž
13
.
Ž
.
Ž
.
Ž
Ž
.
.
Ž
.
Ž
.
Ž
Ž
.
Ž
.
CH3CH , 21.2 CH3–C5 , 22.8, 27.0 C3, C4 , 44.6 C2 , 61.5 CHCH3 , 126.7, 127.2, 128.1
ArCH , 128.9 C6 , 143.4 ArC . MS ESI : mrz 202.2 MqH . MS EI : mrz 201 M , 42% ,
q
q
.
Ž
.
Ž
.
Ž
Ž
.
.
Ž
.
.
Ž
.
Ž .
Ž
.
Ž
.
.
Ž .
Ž
.
Ž
.
Ž
186 9 , 174 14 , 105 100 , 97 11 , 91 7 , 77 12 , 56 18 . HRMS ESI mrz found: 202.1586.
q
Ž
.
Ž
.
Ž
Calc. for C14H20N : 202.1596. HRMS EI mrz found: 201.152"0.002. Calc. for C14H19N:
201.152.
1
Ž
.
Ž
.
Ž
.
Ž
3-Methyl-1- 1-phenylethyl piperidine 14b had H NMR 200 MHz : d ppm 0.9 d, J 7 Hz, 3 H,
.
Ž
CH3 .
.
1- 1-Phenylethyl azepane 15b. A sample of this compound was isolated by chromatography of the
w
Ž
Ž
. x
product from reaction of 9b with H2rCO, 9:1 in the Rh OAc 2 2rBIPHEPHOS at 808 for 20 h as an
Ž
.
.
oil, b.p. 1308 oven r0.4 mm. An authentic sample 90% was prepared by alkylation of hexameth-
1
Ž
.
Ž
.
Ž
.
yleneimine with 1-bromoethylbenzene. H NMR 300 MHz : d ppm 1.35 d, J 6.8 Hz, 3 H, CH3 ,
Ž
.
Ž
.
Ž
.
1.57 bs, 8 H, H3, H4, H5, H6 , 2.62 bs, 4 H, H2, H7 , 3.76 q, J 6.7 Hz, 1 H, CHCH3 , 7.18–7.44
13
Ž
.
Ž
.
Ž
.
Ž
.
Ž
.
Ž
.
m, 5 H, Ph . C NMR 50 MHz : d ppm 18.2 CH3 , 27.0, 28.85 C3, C4, C5, C6 , 52.0 C2, C7 ,
q
Ž
.
.
Ž
.
Ž
.
Ž
.
Ž
Ž
.
63.15 CHCH3 , 126.4, 127.5, 127.9 ArCH , 144.8 ArC . MS ESI : mrz 204.0 MqH .
q
Ž
Ž
.
Ž
.
Ž
.
Ž
.
Ž
.
Ž
.
.
MS EI : mrz 203 M , 9% . 189 13 , 188 100 , 126 35 , 105 56 , 77 20 HRMS ESI mrz
q
Ž
.
found: 204.1746. Calc. for C14H22 N : 204.1752.
1
Ž
.
Ž
.
.
Ž
.
Ž
1- 1-Phenylethyl -2,3,4,5-tetrahydro-1H-azepine 13b had H NMR 300 MHz : d ppm 5.15 dt,
13
.
Ž
Ž
.
Ž
.
Ž
Ž
.
J 15.3, 6.9 Hz, 1 H, H6 , 5.94 d, J 15.3 Hz, 1 H, H7 C 50 MHz : d ppm 118.6 C6 , 134.2
q
Ž
.
Ž
.
Ž
.
Ž
.
Ž
.
Ž
.
Ž
.
Ž
.
Ž
.
C7 . GCMS mrz: 201 M , 40% , 186 47 , 124 10 , 105 100 , 97 40 , 96 23 , 82 90 , 79 40 ,
q
.
Ž
.
Ž
.
77 42 . HRMS ESI mrz found: 202.1583. Calc. for C14H20N : 202.1596. Hydrogenation, as
described above, of the product from reaction 12, rich in this material, gave the azepane 15b.
6- 1- 1-Phenylethyl azepan-2-yl -1- 1-phenylethyl -2,3,4,5-tetrahydro-1H-azepine 17b had 1H
w
Ž
.
x
Ž
.
13
Ž
.
Ž
.
Ž
.
Ž
.
Ž
.
Ž
.
NMR 300 MHz : d ppm 6.18 bs, 1 H, H7 . C 50 MHz : d ppm 139.9 C7 . Hydrogenation of
the product from reaction 12 followed by chromatography silica, 50% ethyl acetaterlight petroleum
gave a sample of 1,1-di 1-phenylethyl -2,3 -bis azepane 18b. Mixture of isomers — H NMR 400
MHz : d ppm 1.04–1.24 m, 16 H , 1.32–1.36 m, 12 H, C H3CHPh , 1.37–1.70 m, 16 H ,
2.15–2.25 m, 2 H, C H2 N 2.35–2.51 m, 5 H, C H2 N , 2.54–2.73 m, 5 H, C H2 N , 3.69–3.85 m,
4 H, C H Ph , 7.16–7.39 m, 20 H, ArH . C NMR 100 MHz : d ppm 16.5, 18.4, 24.35 CH3CH ,
25.18, 25.23 27.32, 27.36, 29.6, 29.8, 29.85, 30.3, 33.7, 33.9, 35.32, 35.35 CH2 , 39.3, 39.8 CH ,
Ž
.
X
X
1
Ž
.
Ž
.
Ž
.
Ž
.
Ž
Ž
.
Ž
.
Ž
.
Ž
.
.
Ž
.
Ž
.
Ž
13
.
Ž
.
Ž
.
Ž
.
Ž
.
.
Ž
.
Ž
Ž
.
Ž
.
Ž
47.9, 52.6, 52.7, 57.6 CH2 N , 58.4, 63.4, 64.0 CHCH3 , 126.43, 126.44, 126.6, 126.8, 127.65,
q
Ž
.
.
Ž
.
Ž
.
127.7, 127.88, 127.93, 128.4 ArCH , 144.8, 145.2, 146.0 ArC . MS ESI : mrz 405.3 MqH ,
443.3 MqK . HRMS ESI mrz found: 405.3265. Calc. for C28H41N2 q: 405.3270. An identical
q
Ž
.
Ž
.
Ž
.

(
)
R.S. Dickson et al.rJournal of Molecular Catalysis A: Chemical 150 1999 133–146
145
w
Ž
. x
sample was obtained from the reaction of 9b with H2rCO, 9:1 using Rh OAc 2 2rBIPHEPHOS
after removal of the azepane 15b by distillation as described above.
3.4.5. Reactions of the amidoalkene 19
Reactions of N-methyl-N-phenylmethylpent-4-enamide 19 were carried out as described above
1
using the appropriate catalyst at 40 or 608C for 20 h and the products analysed by H and 13C NMR
spectroscopy. The product was a mixture of the two aldehyde isomers N-methyl-N-phenylmethyl-6-
oxohexanamide 20 and N-methyl-N-phenylmethyl-4-methyl-5-oxopentanamide 21 in the ratio 60:40
1
1
Ž
.
Ž
.
Ž
and spectral data are reported on this mixture. H NMR 300 MHz : d ppm 1.06 d, J 7.1 Hz,
.
Ž
.
Ž
.
Ž
CCH3, 21b , 1.13 d, J 7.1 Hz, CCH3, 21a , 1.65–1.83 m, H3 and H4, 20; H3, 21 , 2.07 m, H4,
21 , 2.35–2.46 m, H2 and H5, 20; H2, 21 , 2.89 s and 2.92 s C H2 Ph , 4.52 s and 4.57 s
NCH3 , 7.12–7.34 m, Ph , 9.56 d, J 1.7 Hz, H5, 21b , 9.62 d, J 1.8 Hz, H5, 21a , 9.69 t, J 1.6
Hz, H6, 20b , 9.73 t, J 1.6 Hz, H6, 20a . C NMR 100 MHz : d ppm 13.5 CCH3, 21b , 13.6
CCH3, 21a , 21.8, 24.6, 24.8, 25.7, 25.9, 30.1, 32.7, 33.1 C3, C4 and C5, 20; C3, 21 , 34.0, 34.1,
34.8 NCH3 , 43.7, 43.75 C2, 20; C2, 21 , 45.7, 45.8 C4, 21 , 50.8, 50.84, 53.3 CH2 Ph , 126.3,
.
Ž
.
Ž .
Ž . Ž
.
Ž .
Ž .
Ž
.
Ž
.
Ž
.
Ž
.
Ž
13
1
.
Ž
.
Ž
.
Ž
.
Ž
.
Ž
.
Ž
.
Ž
.
Ž
.
Ž
.
Ž
.
Ž
.
Ž
.
Ž
127.3, 127.6, 128.0, 128.4, 128.6, 129.0 ArCH , 136.8, 137.4, 137.5 ArC , 172.2, 172.6, 172.9 C1,
q
.
Ž
.
Ž
.
Ž
.
20; C1, 21 , 202.4, 204.6, 204.7 C6, 20; C5, 21 . GCMS for 20: mrz 233 M , 7% , 205 10 ,
q
Ž
.
Ž
.
Ž
.
.
Ž
.
.
Ž
.
Ž
.
Ž
Ž
.
Ž
.
.
Ž
190 15 , 176 22 , 163 35 , 162 13 , 120 56 , 106 28 , 91 100 , 65 21 . GCMS for 21: mrz 233 M ,
.
Ž .
Ž
Ž
Ž
.
Ž
.
.
Ž
Ž
.
Ž
.
Ž
.
10% , 204 9 , 176 20 , 163 55 , 162 19 , 120 52 , 106 20 , 91 100 , 65 20 , 55 25 . HRMS ESI
mrz found: 234.1483. Calc. for C14H20NO2 q: 234.1494.
Ž
.
Acknowledgements
We thank the Australian Research Council for support and Johnson Matthey for a loan of rhodium.
References
w x
1
w x
2
G.W. Parshall, S.D. Ittel, Homogeneous Catalysis, Wiley-Interscience, New York, 1992.
Ž
.
B. Cornils, W.A. Herrmann Eds. , Applied Homogeneous Catalysis by Organometallic Complexes, VCH VerlagChemie, Weinheim,
1996.
w x
Ž .
M. Beller, B. Cornils, C.D. Frohning, C.W. Kohlpainter, J. Mol. Catal. A: Chem. 104 1995 17, and references therein.
3
w x
4
Ž .
M.J. Doyle, T.J. Duckworth, L. Manojlovic-Muir, M.J. Mays, P.R. Raithby, F.J. Robertson, J. Chem. Soc. Dalton Trans. 1992 2703.
Ž .
Y. Ishii, K. Miyashita, K. Kamita, M. Hidai, J. Am. Chem. Soc. 119 1997 6448.
w x
5
w x
Ž .
6
M. Hidai, A. Fukuoka, Y. Koyasu, Y. Uchida, J. Mol. Catal. 35 1986 29.
w x
Ž .
Z. He, N. Lugan, D. Neibecker, R. Mathieu, J.-J. Bonnet, J. Organomet. Chem. 426 1992 247.
7
w x
Ž .
8
M. Garland, Organometallics 12 1993 535.
w x
Ž .
T. Baig, J. Molinier, P. Kalck, J. Organomet. Chem. 455 1993 219.
9
w
w
x
Ž
.
10 R.S. Dickson, T. De Simone, E.M. Campi, W.R. Jackson, Inorg. Chim. Acta 220 1994 187.
x
11 K.J. Countinho, R.S. Dickson, G.D. Fallon, W.R. Jackson, T. De Simone, B.W. Skelton, A.H. White, J. Chem. Soc. Dalton Trans.
Ž
.
1997 3193.
w
w
w
w
w
x
Ž
.
12 K. Soulantica, S. Sirol, S. Koinis, G. Pneumatikaleis, P. Kalck, J. Organomet. Chem. 498 1995 C10.
x
Ž
.
13 R.C. Matthews, D.K. Howell, W.-J. Peng, S.G. Train, W.D. Treleaven, G.G. Stanley, Angew. Chem., Int. Ed. Engl. 35 1996 2253.
x
Ž
.
14 W.-J. Peng, S.G. Train, D.K. Howell, F.R. Fronczek, G.G. Stanley, Chem. Commun. 1996 2607.
x
Ž
.
15 M.E. Broussard, B. Juma, S.G. Train, W.-J. Peng, S.A. Laneman, G.G. Stanley, Science 260 1993 1784.
x
Ž
.
16 A. Aaliti, A.M. Masdeu, A. Ruiz, C. Claver, J. Organomet. Chem. 489 1995 101.
1
Ž .
Ž .
The amides 20 and 21 were both observed as a 3:2 mixture of rotamers, labelled a and b where appropriate.

(
)
146
R.S. Dickson et al.rJournal of Molecular Catalysis A: Chemical 150 1999 133–146
w
w
w
w
w
w
w
w
w
w
w
w
w
w
w
w
w
w
w
w
w
w
x
Ž
.
17 A.M. Masdeu, A. Orejon, A. Ruiz, S. Castillon, C. Claver, J. Mol. Catal. 94 1994 149.
x
Ž
.
18 G. Suss-Fink, Angew. Chem., Int. Ed. Engl. 33 1994 67.
x
Ž
.
19 D.J. Bergmann, E.M. Campi, W.R. Jackson, Q.J. McCubbin, A.F. Patti, Tetrahedron 53 1997 17449.
x
Ž
.
20 D. Anastasiou, E.M. Campi, H. Chaouk, G.D. Fallon, W.R. Jackson, Q.J. McCubbin, A.E. Trnacek, Aust. J. Chem. 47 1994 1043.
x
Ž
.
21 Z. Zhang, I. Ojima, J. Organomet. Chem. 454 1993 281.
x
Ž
.
22 W.R. Jackson, P. Perlmutter, G.-H. Suh, E.E. Tasdelen, Aust. J. Chem. 44 1991 951.
x
Ž
.
23 I. Ojima, Z. Zhang, J. Organomet. Chem. 417 1991 253.
x
Ž
.
24 I. Ojima, A. Korda, W.R. Shay, J. Org. Chem. 56 1991 2024.
x
Ž
.
25 P.M. Shulman, E.D. Burkhardt, E.G. Lundquist, R.S. Pilato, G.L. Geoffroy, A.L. Rheingold, Organometallics 6 1987 101.
x
Ž
.
26 C.P. Casey, R.M. Bullock, F. Nief, J. Am. Chem. Soc. 105 1983 7574.
x
Ž
.
27 M.M. Doyle, W.R. Jackson, P. Perlmutter, Tetrahedron Lett. 30 1989 5357.
x
Ž
.
28 S. Gladiali, L. Pinna, C.G. Arena, F. Rotondo, F. Faraone, J. Mol. Catal. 66 1991 183.
x
Ž
.
29 N.J. Leonard, F.P. Hauck, J. Am. Chem. Soc. 79 1957 5279.
x
Ž
.
30 S.J. Martinez, J.A. Joule, Tetrahedron 34 1978 3027.
x
Ž
.
31 G.D. Cuny, S.L. Buchwald, J. Am. Chem. Soc. 115 1993 2066.
x
Ž
.
32 I. Ojima, M. Tzamarioudaki, M. Eguchi, J. Org. Chem. 60 1995 7078.
x
Ž
.
33 E.M. Campi, L.K. Eriksson, S.T. Guy, W.R. Jackson, P. Perlmutter, J. Mol. Catal. A: Chem. 143 1999 243.
x
Ž
.
34 D.C. Craig, G.L. Edwards, C.A. Muldoon, Synlett 1997 1441.
x
Ž
.
35 P. Legzdins, R.W. Mitchell, G.L. Rempel, J.D. Ruddick, G. Wilkinson, J. Chem. Soc. A 1970 3322.
x
Ž
.
36 C.P. Casey, R.M. Bullock, W.C. Fultz, A.L. Rheingold, Organometallics 1 1982 1591.
x
Ž
.
37 J.A. McCleverty, G. Wilkinson, L.G. Lipson, M.L. Maddox, H.D. Kaesz, Inorg. Synth. 8 1966 211.
x
Ž
.
38 C. Stevens, N. De Kimpe, Synlett 5 1991 35.