Branched Selectivity in the Pd-Catalysed Methoxycarbonylation of 1-Alkenes

Article abstract of DOI:10.1002/cctc.201501423

The methoxycarbonylation of alkenes by palladium(II) complexes with P,O-donor ligands [(2-methoxyphenyl)diphenylphosphine (L-2), bis(2-methoxyphenyl) phenyl phosphine (L-3) and tris(2-methoxyphenyl) phosphine (L-4)] has been investigated. The results show that the Pd complexes derived from these ligands provide high regioselectivity for the branched esters from 1-pentene and 1-hexene (>80 %). Various parameters (including temperature, pressure, acid concentration) were optimized to improve the performance of the catalyst system. Higher temperatures afforded higher regioselectivity; but effected rapid catalyst decomposition. Acceptable turnover frequencies, conversions as well as catalyst stability could be obtained at higher L/Pd ratios. The dramatic change in regioselectivity is rationalised on the basis of the hemi-lability of the o-methoxy moiety, which may lead to ligand dissociation from L₂PdX₂ (L=ligand, X=Cl) rather than the expected dissociation of X. In support of our hypothesis, direct evidence for the coordination of the o-methoxy to the Pd centre was demonstrated by the crystal structure. To the best of our knowledge, this work provides the first reported route to valuable branched esters through the methoxycarbonylation of alkenes at suitable rates.

Full text of DOI:10.1002/cctc.201501423

DOI: 10.1002/cctc.201501423
Full Papers
Branched Selectivity in the Pd-Catalysed
Methoxycarbonylation of 1-Alkenes
Charmaine Arderne,[a] llia A. Guzei,[a, b] Cedric W. Holzapfel,[a] and Tyler Bredenkamp*[a]
The methoxycarbonylation of alkenes by palladium(II) com-
plexes with P,O-donor ligands [(2-methoxyphenyl)diphenyl-
phosphine (L-2), bis(2-methoxyphenyl) phenyl phosphine (L-3)
and tris(2-methoxyphenyl) phosphine (L-4)] has been investi-
gated. The results show that the Pd complexes derived from
these ligands provide high regioselectivity for the branched
esters from 1-pentene and 1-hexene (>80%). Various parame-
ters (including temperature, pressure, acid concentration) were
optimized to improve the performance of the catalyst system.
Higher temperatures afforded higher regioselectivity; but ef-
fected rapid catalyst decomposition. Acceptable turnover fre-
quencies, conversions as well as catalyst stability could be ob-
tained at higher L/Pd ratios. The dramatic change in regiose-
lectivity is rationalised on the basis of the hemi-lability of the
o-methoxy moiety, which may lead to ligand dissociation from
L2PdX2 (L=ligand, X=Cl) rather than the expected dissociation
of X. In support of our hypothesis, direct evidence for the coor-
dination of the o-methoxy to the Pd centre was demonstrated
by the crystal structure. To the best of our knowledge, this
work provides the first reported route to valuable branched
esters through the methoxycarbonylation of alkenes at suitable
rates.
Introduction
The palladium-catalysed hydromethoxycarbonylation (methox-
ycarbonylation for short) of 1-alkenes is an active area of re-
search (Scheme 1).[1] The products, in particular the alkyl esters,
are useful reagents for various industrial applications (including
the production of solvents, detergents, perfumes, etc.).[2] In
general, the reaction provides a mixture of both linear and
branched esters (Scheme 1) and most of the work to date in-
volves the influence of mono- and bidentate phosphines on
the activity and regioselectivity of the reaction.[1,3]
Scheme 2. Proposed Pd–H-based mechanism for the hydromethoxycarbony-
Two mechanisms (Pd–H and Pd–OMe-based) have been pro-
posed for the reaction, each with much evidence in its fa-
vour.[1o] However, the majority of reports indicate a preference
for the Pd–H-based mechanism (Scheme 2).[1p,q]
lation reaction.
When styrene-derived substrates are employed in the reac-
tion, regioselectivity toward the branched isomer is desired as
these products are used as precursors for non-steroidal anti-in-
flammatory drugs.[1j,3] However, in the case of the a-olefins, the
formation of the linear isomer is most often desired
(Scheme 1) and as such much effort has been directed toward
the synthesis thereof.[1,4] Whereas several methodologies have
been developed for branched ester formation from styrene,[1j,3]
these do not translate into branched ester formation from 1-al-
kenes.[3d] Additionally, methoxycarbonylation methodologies
for branched ester formation from terminal alkenes, methyl
methacrylate in particular, are almost exclusively limited to
patent literature.[5]
Scheme 1. Pd-catalysed methoxycarbonylation of 1-alkenes.
[a] Dr. C. Arderne, Prof. l. A. Guzei, Prof. C. W. Holzapfel, Dr. T. Bredenkamp
Department of Chemistry
University of Johannesburg
Johannesburg, 2006 (South Africa)
E-mail: tylerb@uj.ac.za
Although regioselectivity toward linearity has typically been
the primary goal, branched esters have recently gained interest
for development of new surfactants aimed at enhanced oil re-
covery (EOR) strategies.[6] Purposely, we envisaged the applica-
tion of the methoxycarbonylation reaction to the synthesis of
these valuable branched esters. Alper and co-workers[7] have
reported regioselective conversions of up to 100% in favour of
[b] Prof. l. A. Guzei
Department of Chemistry
University of Wisconsin
1101 University Avenue, Madison, 53706–1322 (United States)
Supporting Information and ORCID(s) from the author(s) for this article
are available on the WWW under http://dx.doi.org/10.1002/
cctc.201501423.
ChemCatChem 2016, 8, 1084 – 1093
1084
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
the branched ester by using the catalyst system shown in
Scheme 3. Although high regioselectivity could be attained,
the reaction rates were extremely poor, requiring as much as 8
days. Interestingly and unexpectedly, the oxidative product
was not observed under the specified conditions.
Table 1. Methoxycarbonylation of 1-pentene in toluene.[a]
Entry
T [8C]
t [h]
Conversion [%]
TOF[b]
Linear/branched[c]
1
2
3
4
125
125
100
100
4
6
4
6
64
72
42
51
48
36
32
25
45:55
43:57
57:43
55:45
We now report reaction conditions and a ligand selection
that result in the productive synthesis of methyl esters from
selected 1-alkenes with >80% selectivity toward the branched
isomer (2-methoxycarbonylated product).
[a] Reaction conditions: PdCl2/6PPh3/4MsOH, PdCl2 =1.67 mmol, 1-pen-
tene=10 mL, MeOH=6 mL, 5.0 MPa CO, 14 mL toluene. [b] TOF=aver-
age mmol ester/mmol Pd/h. [c] Linear/branched ester ratio determined
by GC-FID analysis.
Employing ethers (including THF, THP, DME and anisole) result-
ed in improved reaction rates as well as branched selectivity.
However, except in the case of dioxane, temperatures above
1008C resulted in catalyst decomposition as evidenced by Pd
black formation, which is often a limiting factor in Pd-catalysed
reactions.[9] Fortuitously, the use of dioxane as the solvent al-
lowed the reactions to be carried out at a 1108C for extended
periods of time (12 h) and even at 1258C for shorter periods
(4 h). Furthermore, under the conditions previously used for
toluene at 1258C, branched selectivity increased to 80%.
As indicated in our earlier studies,[1m] the use of o-substitut-
ed triphenylphosphines affects the regioselectivity in favour of
branched ester formation. In line with those findings, a range
of commercially available o-substituted triphenylphosphines
(substituents including Br, OCH3, CN, CHO, COOH, COOCH3)
were evaluated under conditions similar to those described
above. The majority of the ligands evaluated resulted in re-
duced reactions rates and catalyst stability. However, in the
case of the o-methoxy triphenylphosphines (Figure 1) higher
branched selectivity was achieved compared with PPh3 (L-1) as
demonstrated in Table 2.
Scheme 3. Synthesis of branched esters from 1-alkenes.
Results
In general, methanol is both a reagent and a solvent in the
methoxycarbonylation reaction, making the reaction mixture
extremely polar. However, reports have indicated that under
reduced polarity conditions, higher branched selectivity can be
achieved. In a kinetic study conducted by Rosales et al.,[8] they
showed enhanced branched selectivity in the methoxycarbo-
nylation 1-hexene when toluene was employed as the solvent
(Scheme 4).
Scheme 4. Methoxycarbonylation of 1-hexene.
At elevated temperatures (>1208C) and reaction times as
long as 4 hours (4 h), the authors reported that the branched
ester was formed in equal or slightly higher amounts than the
linear esters, in comparison to the corresponding reactions
with MeOH as solvent (3:1, MeOH/toluene). We applied similar
reaction conditions [but with methane sulfonic acid (MsOH)] to
the methoxycarbonylation of 1-pentene. The results shown in
Table 1 confirm a change towards branched selectivity, particu-
larly at elevated temperatures. However, the reaction rates, as
indicated by the average turnover frequencies (TOFs), are low
compared with the corresponding reactions in methanol.[1l]
The possibility of replacing toluene with another solvent
that would support both branched selectivity and higher reac-
tion rates was therefore investigated. Non-polar solvents (in-
cluding 1,2-dichloroethane, benzene and chlorobenzene) gave
results similar to those obtained with toluene, whereas polar
solvents such as acetonitrile (CH3CN), dimethyl sulfoxide
(DMSO) and dimethylformamide (DMF) gave drastically re-
duced rates without improving the branched ester selectivity.
Figure 1. Ligands evaluated in the methoxycarbonylation of 1-pentene.
The results show that an increase in substitution affords
a higher branched selectivity coupled with reduced reaction
rates (compare Table 2, entries 2, 5 and 10). The effect of in-
creased substitution on the reaction rate correlates well with
the evidence[1j,4b,10] for the methanolysis of the Pd–acyl inter-
mediate as being the rate-determining step (Scheme 5). This
step requires the cis orientation of the phosphine ligands,
which becomes increasingly disfavoured by increased steric
hindrance.
Although the exact role of the acid remains a topic of dis-
cussion, its participation in the formation of the active Pd–H
species is certain.[1] This is supported by the observed increase
in rate with 2 equivalents of acid compared with the corre-
sponding reactions without any added acid (compare Table 2,
entries 3 and 4). Further increases in the [acid] initially led to
ChemCatChem 2016, 8, 1084 – 1093
www.chemcatchem.org
1085
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Table 2. Methoxycarbonylation of 1-pentene in dioxane.[a]
Table 3. The effect of [MeOH] in dioxane on branched ester selectivity.[a]
Entry
Ligand
Pd/acid
TOF[b]
Linear/branched[c]
Entry
Ligand
MeOH [%]
TOF[b]
Linear/branched[c]
1
2
3
4
5
6
7
8
9
10
L-1
L-2
L-3
L-3
L-3
L-3
L-3
L-3[d]
L-3[e]
L-4
1:4
1:4
1:0
1:2
1:4
1:6
1:8
1:4
1:4
1:4
158
133
36
79
92
81
68
135
68
34
33:67
33:67
21:79
19:81
18:82
20:80
17:83
14:86
24:76
13:87
1
2
3
4
5
6
7
8
L-3
L-3
L-3
L-3
100
75
50
25
15
15
10
5
15
15
15
10
10
100
55
25
8
265
225
185
142
125
152
92
55:45
51:49
46:54
25:75
20:80
16:84
18:82
18:82
25:75
20:80
12:88
21:79
18:82
98:2
L-3
L-3[d]
L-3
L-3
49
9
L-3[e]
L-3[f]
L-3[g]
L-3[h]
L-3[i]
L-5[j]
L-5[j]
L-5[j]
L-5[j]
145
117
42
63
48
950
623
231
92
10
11
12
13
14
15
16
17
[a] Conditions: PdCl2/9ligand/xMsOH, PdCl2 =0.18 mmol, T=110 8C, PCO
=
5.0 MPa, 1-pentene=10 mL, dioxane=50 mL, MeOH=6 mL, 2 h, 1 mL
toluene as standard. [b] TOF=average mmol ester/mmol Pd/h. [c] Per-
centage of branched ester determined by GC-FID analysis. [d] PdCl2 =
0.09 mmol. [e] PdCl2 =0.36 mmol.
97:3
96:4
96:4
[a] Conditions: PdCl2/9L-3/4MsOH, PdCl2 =0.18 mmol, T=110 8C, PCO
=
5.0 MPa, 1-pentene=20 mL, MeOH/dioxane total volume 50 mL, 2 h.
[b] TOF=average mmol ester/mmol Pd/h. [c] Linear/branched ester ratio
determined by GC-FID analysis. [d] PdCl2 =0.09 mmol. [e] PdCl2 replaced
with Pd(OAc)2. [f] As in [e] but with 2 equiv. Bu4NCl. [g] As in [e] but with
4 equiv. Bu4NCl. [h] EtOH replaced MeOH. [i] PrOH replaced MeOH. [j] Cat-
alyst system: Pd(OAc)2/ 5L-5/6MsOH, 1 h.
Scheme 5. Methanolysis of the Pd–acyl species.
enhanced reaction rates without a significant regioselective
effect followed by a rate reduction owing to catalyst decompo-
sition (Table 2, compare entries 3–7). This may be due to proto-
nation of the ligand at higher acid loadings.
crease in branched selectivity was observed by changing the
Pd/L-3 ratio to 1:3. However, this resulted in rapid catalyst de-
composition.
Although the polarity of the medium affects the regioselec-
tivity of the reaction, it can be reversed by the addition of
a suitable ligand. Employing the catalyst system based on
ligand L-5 (1,2-bis(di-tert-butylphosphinomethyl)benzene), the
benchmark ligand for the conversion of 1-alkenes into linear
esters,[4b] a rapid decrease in reaction rate with decreasing
[MeOH] in dioxane was observed without meaningful change
in linear selectivity (Table 3, compare entries 14–17). This result
demonstrates the critical role that the ligand plays in the selec-
tivity of the reaction.
Higher branched selectivity could be obtained at elevated
temperatures (1258C, b=98%), but this this resulted in lower
conversions owing to rapid catalyst decomposition as evi-
denced by the precipitation of Pd black. Further studies
toward branched selectivity were limited to ligand L-3, which
gave good reaction rates and high branched selectivity with
the catalyst system of 1Pd/9L-3/4MsOH.
Table 3 demonstrates the effect of [MeOH] in dioxane on the
reactions rates and branched selectivity. As seen from the re-
sults, the rate is first order in regard to methanol concentra-
tion, showing a steady decrease in rate as the amount of
methanol is lowered. Replacement of PdCl2 with Pd(OAc)2 re-
sulted in an increase in reaction rate, but a decrease in
branched selectivity (Table 3, entry 9). This effect could be re-
versed by the addition of 2 equivalents of Bu4NCl. However,
further addition of the chloride resulted in high branched se-
lectivity together with a dramatic reduction in reaction rate
(Table 3, compare entries 10 and 11). The remarkable effect of
the chloride counterion on the regioselectivity of styrene me-
thoxycarbonylation is well documented.[11]
Interestingly, when CO pressures
in the range 3.0–10.0 MPa were ex-
amined, it was found that higher
pressures gave higher reaction
rates
without
affecting
the
branched selectivity (Figure 2).
Branched ester selectivity and
the average TOFs were satisfactory in 15% MeOH in dioxane in
the case of reactions involving ligand L-3 with shorter reaction
times (2 h) and lower conversions. However, after approximate-
ly 2 h, reaction rates had declined significantly (as indicated by
CO uptake). At this point, >80% of the remaining pentene
was compromised of cis- and trans-2-pentenes. It is known
that internal alkenes react at least eight times slower than the
corresponding 1-alkenes.[12] Continuation of the reaction for up
to 12 h did not result in more than 75% conversion into esters
owing to catalyst decomposition. Improved substrate conver-
sion could be obtained by increasing the Pd/L-3 ratio to 1:16
as shown in Table 4.
Replacement of MeOH with EtOH or PrOH resulted in a de-
crease in reaction rate with little effect on branched selectivity
(entries 12 and 13). Lowering the [Pd] with an unchanged Pd/
L-3/MsOH ratio afforded higher branched selectivity as well as
reaction rates (compare Table 2, entries 5, 8 and 9 and Table 3,
entries 5 and 6). This may have significance for potential appli-
cations. The possible role of Pd nanoparticles in this effect are
duly noted and are currently under investigation. A similar in-
ChemCatChem 2016, 8, 1084 – 1093
www.chemcatchem.org
1086
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Employing other Brønsted acids such as H2SO4, triflic acid
(TfOH) and HBF4 resulted in enhanced rates with 2 equivalents
of added acid. However, above this threshold rapid decompo-
sition of the catalyst was observed coupled with decreased
rates. As expected,[1l] similar rate enhancements and catalyst
stability were noted with the replacement of the Brønsted acid
with the Lewis acid Al(OTf)3. This is most likely due to the re-
duction in ligand alkylation when Lewis acids are employed in
the reaction, as shown in our previous studies.[1l]
Figure 2. The effect of CO pressure on the TOF by using the Pd complex de-
rived from ligand L-3 (branched selectivity remained unchanged).
Discussion
The observation that the regioselectivity is almost independent
of the Pd/MsOH ratio indicates that the regioselectivity is de-
termined in the catalytic cycle (Scheme 6)[15] earlier than the
final rate-determining step.
Table 4. Effect of ligand/Pd ratio on the selectivity and productivity of
methoxycarbonylation of 1-alkenes.[a]
Entry Substrate
Pd/L TOF[b] Conversion [%] Linear/branched[c]
1
2
3
4
5
6
7
8
1-pentene
1-pentene
1-pentene
1-hexene
1-hexene
1-hexene
1:9
1:12 102
1:16
1:9
1:12 104
118
<75
<80
92
<75
<80
83
19:81
20:80
20:80
21:79
22:78
23:77
20:80
22:78
96
107
1:16
98
–
–
1-octene[d] 1:16
1-decene[d] 1:16
>99
>99
[a] Reaction conditions: PdCl2 =0.18 mmol, Pd/MsOH=1:4, T=110 8C,
CO =5.0 MPa, 1-alkene=25 mL, MeOH=10 mL, dioxane=50 mL, 12 h.
P
[b] TOF=average mmol ester/mmol Pd/h (first 2 h). [c] Branched esters as
determined by GC-FID analysis. [d] Other ester products observed.
Scheme 6. Proposed mechanism for the Pd-catalysed methoxycarbonylation
reaction.
Similar satisfactory results could be obtained with 1-hexene,
1-octene and 1-decene, but with lower TOFs. Methoxycarbony-
lation rates decrease rapidly with increasing chain length[1b]
and as such the goal of converting long chain alkenes (ꢀC10)
into branched esters will require the development of improved
catalytic systems. However, rather unexpectedly, quantitative
conversions for 1-octene and 1-decene were obtained (Table 4,
entries 7 and 8) after 12 h. The higher conversions obtained for
octene and decene may be attributed to the higher concentra-
tion of substrate in the liquid phase under the reaction condi-
tions. Although not shown here, internal alkenes could also be
methoxycarbonylated. However, other unidentified products
were observed in the reaction mixture. This topic is currently
under investigation.
This step involves the addition of Pd–H to the 1-alkene (or
insertion of the alkene into the PdÀH bond). Under standard
conditions, this step is thought to require the dissociation of
a PdÀX bond in the catalytic species L2PdHX (Scheme 7).
It has been reported[13] that HCl plays an important role in
the methoxycarbonylation reaction under similar conditions to
those employed in this study. Firstly, it can serve as an acidic
promoter and it can further be a source of chloride ions, which
can stabilise the palladium species.[14] This stabilising effect
was demonstrated when MsOH was replaced with HCl (pro-
duced in situ by the addition of acetyl chloride), which allowed
the reactions to be carried out by using only 12 equivalents of
the ligand to afford similar conversions and rates to the corre-
sponding reactions done with 16 equivalents of ligand. Fur-
thermore, the reactions could be carried out at 1258C with in-
creased rates and branched selectivity (>85% with 1-hexene)
with catalyst stability up to 2 h.
Scheme 7. Pd–alkyl intermediate formed after ClÀ dissociation from Pd.
The various species shown in Scheme 7 are in rapid equilibri-
um with each other as well as other alkene isomers. In this
equilibrium situation, linear intermediate A is preferred over
the branched intermediate B because of the steric demands of
the ligands (e.g., L-1 and L-5). This intermediate is involved in
a CO insertion reaction, which ultimately furnishes the linear
ester. However, in a less polar solvent, dissociation of the PdÀX
bond will be more difficult and ligand dissociation may present
an alternative route to Pd–alkene association (Scheme 8).[16]
ChemCatChem 2016, 8, 1084 – 1093
www.chemcatchem.org
1087
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Scheme 8. Formation of Pd–alkyl species by ligand dissociation.
Figure 3. Molecular structure of the compound PdCl2L-3 (compound 1) with
thermal displacement ellipsoids drawn at the 50% probability level. All hy-
drogen atoms and solvent molecules are omitted for clarity. See the Sup-
porting Information for the full molecular structure figure showing solvent
and disorder modelling of the structure.
As a result of significantly decreased steric demand in the al-
ternative intermediates C and D, the branched isomer (D) may
be thermodynamically preferred on the basis of electronic con-
siderations.[17] Such a change in reaction mechanism is unlikely
in the case of the bidentate ligand L-5. The beneficial effect of
the o-methoxy substituents on the ligands may be attributed
to the involvement of a bidentate hemi-labile ligand in the
ligand dissociation step (Scheme 9). It is known that catalysed
ligand dissociation and ligand exchange in L2PdCl2 complexes
involve an associative–dissociative pathway and a five-coordi-
nate transition state.[16,18]
of the product mixture. The dimer likely results from the com-
bination of the expected trans-orientated LPdCl2 (trans influ-
ence)[20] and a cis-orientated monomer stabilised by coordina-
tion of the methoxy group to the Pd centre (Scheme 10).
Scheme 10. Proposed formation of the Pd dimer. Other P-substituents omit-
ted for clarity.
Scheme 9. Proposed ligand dissociation from Pd.
On dissolving in DMSO, this complex rapidly disassociates to
the monomeric (L-3)2PdCl2 and PdCl2, whereas complex
[(PPh3)2PdCl2)]2 remained stable for several hours in DMSO. This
rapid dissociation of the L-3-derived dimer may again involve
participation of the methoxy substituent. Additional evidence
that OMe moiety can function as a hemi-labile ligand was pro-
vided by the single-crystal X-ray structure of (L-4)2Pd(OTf)2
(compound 2·solvent; Figure 4).
In support of the suggestion of intramolecular methoxy
group assistance for the ligand dissociation, it was found that
the complex L2PdCl2 with L=L-2, L-3 or L-4 all decompose to
red products on handling in toluene under reflux, whereas
(PPh3)2PdCl2 was stable under the same conditions. The reflec-
tion IR spectra of the sparingly soluble red products were iden-
tical to the complexes obtained by the reaction of 1 mole
equivalent of each of the ligands L-2, L-3 and L-4 with 1 mole
equivalent PdCl2(COD)2 in dichloromethane.
Incidentally, 1,3-bis(di-o-methoxyphenyl)phosphino propane
has been shown to be an extremely active CO–ethene copoly-
The single-crystal X-ray structure of compound 1·2CH2Cl2,
where 1 is a bis[dichloro-di[(methoxyphenyl)phenylphosphine]-
palladium(II)] that incorporates an L-3 ligand, revealed an un-
expected cis orientation of the ligands (Figure 3). Remarkably,
a search of the Cambridge Structural Database (CSD)[19] for
complexes with similar cores showed that this dimer is one of
a total of only 2 others with the P-Pd···Pd-P torsion angle less
than 908, making compound 1 a relatively rare occurrence
with the cis arrangement of the phosphines around the Pd
centres.
Figure 4. Molecular structure of the compound Pd(L-4)(OTf)2 (compound 2)
with thermal displacement ellipsoids drawn at the 50% probability level. All
hydrogen atoms and solvent molecules are omitted for clarity. See the Sup-
porting Information for the full molecular structure figure showing solvent
and disorder modelling of the structure.
The reflection infrared spectrum of the batch of crystals sub-
mitted for X-ray examination was identical to that of the bulk
sample originally obtained in the reaction. Accordingly, this
crystal structure is believed to represent the major component
ChemCatChem 2016, 8, 1084 – 1093
www.chemcatchem.org
1088
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
(Shimadzu GC 2010). A ZB1 30 m analytical column (ID: 0.25 mm,
film thickness: 0.25 mm) was employed and separated various
compounds on the basis of boiling points. Flame ionisation (FID)
was used as the detector, maintained at 3008C, with the following
gas flows: make up flow=30 mLminuteÀ1, H2 flow=40 mLmi-
merisation catalyst and it has been suggested that the hemi-
labile methoxy groups may enter the coordination sphere of
the Pd, although its exact role is unclear.[1j,21a,b] Similarly, Gusev
et al.[21c] reported exclusive copolymerisation selectivity by
using di(o-methoxyphenyl)phosphine ferrocene (doppf), albeit
with low TOF (340), whereas analogue diphenylphosphino fer-
rocene (dppf) furnished copolymers (TOF=1070), methyl prop-
anoate (TOF=1380) as well as other ester products. In their
work, similar coordination of the methoxy oxygen to the Pd
centre was observed in the crystal structure of the Pd(OTf)2
complex of doppf.
nuteÀ1 and air flow=400 mLminuteÀ1
.
Catalysis experiments
A typical experiment involved the addition of a Pd source, the
specified amount of ligand and acid to the reactor. As indicated
above, specific amounts of substrate, MeOH, or other solvents
were added to the reactor. The reactor was flushed 3 times with
CO to remove any remaining air and pressurised to the desired CO
pressure at room temperature. The reaction mixture was then
heated to the specified temperature and was stirred at 600 rpm
with a mechanical stirrer device. The average TOFs were calculated
based on mmols product formed per unit time per mmol Pd used
as determined from GC conversions.
The formation of a monoligated (one ligand) Pd intermedi-
ate has been proposed in enhanced branched regioselectivi-
ty.[3c] Aguirre et al.[3d] described a catalytic system based on
PdCl2 and a P,N-donor ligand [Ph2PHPy], which effects a consid-
erable change towards branched selectivity in the methoxycar-
bonylation of 1-hexene in toluene. It is possible this P,N-donor
may behave as a hemi-labile bidentate ligand under these re-
action conditions. In this regard, we investigated the role of
ligand hemi-lability (and more especially the P, O-donor) under
the conditions described above.
With the exception of 1-pentene (obtained from Sasol R&D) all
chemicals were purchased from Sigma–Aldrich and used with no
further purification.
Conclusions
Synthesis of the metal complexes
The results of the Pd-catalysed methoxycarbonylation of select-
ed 1-alkenes showed a catalytic system based on o-methoxy-
substituted triphenylphosphines as ligands in conjunction with
an appropriate solvent, such as dioxane, under optimal condi-
tions produce branched esters with high selectivities and
yields. The change in regioselectivity from the usual linear to
the branched products is tentatively explained in terms of
a mechanism leading to insertion of the alkene into the PdÀH
bond of an intermediate. This change in the catalytic cycle is
supported by the o-methoxy-substituted triphenylphosphine
ligand functioning as a hemi-labile bidentate ligand. The re-
sults and rationalisation provide a basis for 1-alkene methoxy-
carbonylation with improved branched selectivity and produc-
tivity.
Dichloro[bis(2-methoxyphenyl)phenylphosphine]palladium(II)
cis dimer methylene chloride solvate (1)
Palladium(II) chloride (265 mg, 1.5 mmol) and lithium chloride
(126 mg, 3 mmol) in 10 mL methanol was stirred and heated under
reflux in an atmosphere of nitrogen for 30 min to furnish a clear
orange–brown solution. The solution was allowed to cool to 508C
with continued stirring while a solution of bis(2-methoxyphenyl)-
phenylphosphine (322 mg, 1 mmol) in 8 mL CH2Cl2 was added over
a period of one minute. The resultant dark-orange solution was
heated under reflux for 30 min; then, it was allowed to cool to
room temperature and diluted with 35 mL CH2Cl2. The resultant so-
lution was extracted with water (10 mL) to remove the excess
PdCl2 and the organic phase was dried (over anhydrous Na2SO4)
and then evaporated to dryness. The resultant orange solid
(409 mg) was recrystallized from CH2Cl2 and toluene and the resul-
tant crystalline material (260 mg) was extracted with CH2Cl2 (30 mL
under reflux). The filtered solution was exposed to the vapours of
ether over a period of 24 h, resulting in the formation of orange/
red prisms (110 mg). Suitable crystals were selected for the single-
crystal X-ray diffraction study. M.p.: 245–2478C (decomposed); No
NMR analysis could be performed (insoluble in CDCl3, CD2Cl2, d-
benzene, d-acetone, d-THF; decomposed in d-DMSO).
Experimental Section
Equipment
NMR spectra were recorded on a Bruker 500 MHz spectrometer
and referenced to residual CHCl3 in CDCl3 (1H: d=7.24 ppm, 13C:
d=77.00) unless otherwise stated.
A Hastelloy Parr “Bench Top Reactor” (volume=300 mL, maximum
temperature=2508C and maximum pressure of 300 bar) pur-
chased from Labotec was used for the high-pressure studies. The
reactor was fitted with a mechanical overhead stirrer and its own
heating mantle. The reactor was equipped with a “non-return”
valve, which prevents the reaction mixture from pushing back into
the pipes. This valve had a back-pressure of 0.3 psi. The reactor
was also fitted with a safety valve, which was set at a maximum
pressure of 120 bar.
Bis(tris-o-methoxyphenylphosphine)palladium triflate (2)
Palladium(II) acetate (445 mg, 2.0 mmol) and tris(2-methoxyphe-
nyl)phosphine (1409 mg, 4.0 mmol) was dissolved in 15 mL CH2Cl2
under argon. The resulting yellow solution was stirred vigorously
while trifluoromethane sulfonic acid (TfOH) was added (600 mg,
4.0 mmol) slowly with a micro syringe. The resultant brown solu-
tion was stirred for 1 h at 208C. The solvent was then removed
under vacuum to leave a dark-brown residue (1235 mg). The crude
product was dissolved in CH2Cl2. Crystallisation of the title com-
pound was carried out by diethyl ether vapour diffusion into the
CH2Cl2 solution. The crystals of the title compound were dark-
GC analysis was used to monitor reaction progress and outcomes.
Accordingly, a small aliquot of the reaction mixture was filtered
through silica gel and analysed by using a gas chromatograph (GC)
ChemCatChem 2016, 8, 1084 – 1093
www.chemcatchem.org
1089
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
yellow prisms. Suitable crystals were selected for the single-crystal
X-ray diffraction study.
frames collected at intervals of 0.48 in a 68 range about w with an
exposure time of 20 seconds per frame. The reflections were suc-
cessfully indexed by an automated indexing routine built in the
APEX2 program suite. The final cell constants were calculated from
a set of 9951 strong reflections from the actual data collection.
The data were collected by using the full sphere data collection
routine to survey the reciprocal space to the extent of a full sphere
to a resolution of 0.73 Š. A total of 36879 data were harvested by
collecting five sets of frames with 0.58 scans in w and f with expo-
sure times of 30 s per frame. These highly redundant datasets
were corrected for Lorentz and polarization effects. The absorption
correction was based on fitting a function to the empirical trans-
mission surface as sampled by multiple equivalent measure-
ments.[22] A successful solution by direct methods provided most
non-hydrogen atoms from the E-map. The remaining non-hydro-
gen atoms were located in an alternating series of least-squares
cycles and difference Fourier maps. All non-hydrogen atoms were
refined with anisotropic displacement coefficients.[23] All hydrogen
atoms were included in the structure factor calculation at idealized
positions and were allowed to ride on the neighbouring atoms
with relative isotropic displacement coefficients.
Yield: 675 mg, 67%; m.p.: 168–1708C (decomposed); 1H NMR
(400 MHz, CDCl3, 258C): dH =7.53–7.49 (m, 6H, aromatic H), 7.26–
7.21 (m, 6H, aromatic H), 6.96–6.90 (m, 12H, aromatic H), 3.72 ppm
(s, 18H, OCH3); 13C NMR (100 MHz, CDCl3, 258C): dC =160.1 (3C, ar-
omatic C), 135.7 (6C, aromatic C), 134.9 (3C, aromatic C), 122.3
(3C, aromatic C), 113 (3C, aromatic C), 58.2 ppm (3C, OCH3);
31P NMR (121 MHz, CDCl3): dP =33.2 ppm (s, 2P); 19F NMR (376 MHz,
CDCl3): dF =À78.1 ppm (s, 6F).
Crystal structure determination
The structures of compounds 1·2CH2Cl2 and 2·solvent were deter-
mined similarly. Herein, we describe the representative procedure
for 2·solvent. Crystal data and details for data collection and refine-
ment are given in Table 5.
An orange needle-like crystal of compound 2·solvent with approxi-
mate dimensions 0.44”0.09”0.05 mm3 was selected under oil
under ambient conditions and attached to the tip of a MiTeGen Mi-
croMountꢀ. The crystal was mounted in a stream of cold nitrogen
at 100(1) K and centred in the X-ray beam by using a video
camera. The crystal evaluation and data collection were performed
on a Bruker Quazar SMART APEXII diffractometer with MoKa (l=
0.71073 Š) radiation and the diffractometer to crystal distance was
4.96 cm. The initial cell constants were obtained from three series
of w scans at different starting angles. Each series consisted of 15
Molecular structure determination of complexes 1·2CH2Cl2
and 2·solvent
Compound 1·2CH2Cl2 (as shown in Figure 3) was found to crystal-
lize as a pseudomerohedral twin. The general definition of a twin-
ned crystal is “Twins are regular aggregates consisting of individual
Table 5. Crystal data and structural refinement for 1 and 2.
1
2
Empirical formula
C40H38Cl4O4P2Pd2·2CH2Cl2
C44.15H44Cl0.30F6O12.85P2PdS2
Formula weight
Temperature [K]
Wavelength [Š]
Crystal system
Space group
1169.10
100(2)
0.71069
monoclinic
P21/c
1137.18
100(2)
0.71073
triclinic
¯
P1
Unit cell dimensions
a [Š]
b [Š]
c [Š]
15.3870(15)
14.1010(13)
22.734(2)
11.839(3)
12.783(4)
17.363(5)
a [8]
90
91.97(3)
b [8]
g [8]
109.6720(16)
90
4644.8(8)
4
1.672
1.344
2336
0.32”0.26”0.25
0.95–28.47
101.399(19)
112.268(18)
2366.3(11)
Volume [Š3]
Z
2
Dcalc [MgmÀ3
]
1.596
0.651
1157
Absorption coefficient [mmÀ1
F(000)
]
Crystal size [mm3]
0.44”0.09”0.05
1.73–26.38
À14/14, À15/15, À21/17
36864
q range for data collection [8]
Index ranges
Reflections collected
À20/19, À18/16, À30/30
69923
Independent reflections (Rint
Completeness to q=28.358
Absorption correction
Max. and min. transmission
Refinement method
)
10805 (0.0321)
100.0%
9620 (0.0382)
99.5%
semi-empirical from equivalents
0.7299 and 0.6730
Full-matrix least squares on F2
11 895/3/532
1.037
numerical with SADABS
0.9682 and 0.7630
Full-matrix least squares on F2
9620/19/658
1.036
Data/restraints/parameters
Goodness-of-fit on F2
Final R indices [I>2s(I)]
R indices (all data)
Largest difference peak and hole [eŠ
0.0322
0.0377
1.631 and À1.354
0.0390
0.0484
0.820 and À0.857
À3
]
ChemCatChem 2016, 8, 1084 – 1093
www.chemcatchem.org
1090
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
crystals of the same species joined together in some definite
mutual orientation.”[24] Pseudomerohedral twins are found to have
metric lattice parameters that suggest higher Laue symmetry than
the symmetry appropriate for the Laue group of the sample and
they therefore have a twin element that mimics the symmetry of
the higher Laue group. Depending on how well the higher metric
symmetry is fulfilled, it may be that the reciprocal lattices overlap
exactly and the twinning is not immediately detectable from the
diffraction pattern. This was the case with compound 1·2CH2Cl2.
7.98(3)%. The relationship between the correct and apparent unit
cells is illustrated in Figure 5.
One of the co-crystallized dichloromethane solvent molecules of
compound 1·2CH2Cl2 with a carbon atom C2S was disordered over
two positions with a minor component contribution of 32.7(3)%.
The final refined molecular structure of compound 1·2CH2Cl2 is
shown in Figure 3. The asymmetric unit contains one dichloro[bis-
(2-methoxyphenyl)phenylphosphine]palladium(II)-cis-dimer mole-
cule and two dichloromethane solvent molecules. The PdÀCl
bonds are typical and fall near the average value of 2.375(5) Š for
PdÀCl bonds and the PdÀP bonds are also typical of similar com-
plexes, which have an average value of 2.234(9) Š for PdÀP bonds.
The absolute value of the P-Pd···Pd-P torsion angle in compound
1·2CH2Cl2 is 9.70(4)8. A full geometrical analysis of bond lengths,
bond angles, dihedral angles and ring geometries was carried out
by using Mogul v1.7 (part of the CSD (v5.36)[19]) and only the un-
usual geometrical parameters are reported here. Search criteria
used in ConQuest were restricted to structures with: (a) an R-factor
ꢁ 5%; (b) the heaviest element set to Pd; (c) structures containing
organic compounds only and powders were excluded. The search
returned an unusual dihedral angle of 29.5(4)8 for the o-methoxy-
group on centroid Cg(3) (the 6-membered phenyl ring), which is in
relatively close proximity (2.929(2) Š) to Pd1. A total of 1237 hits of
similar fragments in the CSD were found to have a deviation in the
dihedral angle of this particular fragment of the molecule—corre-
sponding to a total of 826 structures. Of these 826 structures, 180
of them are complexes of PdII. Of these 180 structures, 30 struc-
tures were identified in which the Pd atom was bonded to either
a P or an O atom or both. Of these 30 structures, 6 structures re-
ferred to Pd dimers and of the remaining 6 structures, only 2 struc-
tures are similar to compound 1. The first compound is bis((m2-
chloro)-chloro-(O-(2,6-dimethoxy-4-methylphenyl)diphenylphosphi-
nite)palladium(II)) dichloromethane solvate (Refcode: MIPQIM[28]),
which shows a similar deviation from planarity of the dihedral
angle of À7.9(6)8 for the o-methoxy fragment that is 3.159(3) Š
from Pd1. The second compound is bis((m2-chloro)-chloro-(O-(2,6-
dimethoxy-4-methylphenyl)di-tert-butylphosphinite)palladium(II))
(Refcode: MIPQOS[28]), which also shows a similar deviation from
planarity of the dihedral angle of 19.7(4)8 of the o-methoxy frag-
ment that is 2.958(2) Š from Pd1. Compound 1 then shows a similar
trend in deviation of the dihedral angle of the o-methoxy group
closest to Pd1, making it only the third compound to show similar
geometrical properties to the related materials in the database.
From the start of the structural study of compound 1·2CH2Cl2, we
suspected possible twinning issues. The initial reflection indexing
of the first three crystals selected for the single-crystal X-ray diffrac-
tion experiment produced C-centred orthorhombic unit cells.[25]
The best diffracting specimen of the three was selected for the ex-
periment.
The correct unit cell proved to be monoclinic primitive. The rela-
tionship between the primitive monoclinic unit cell, its twin com-
ponent and the apparent incorrect C-centred orthorhombic cell of
compound 1·2CH2Cl2 is depicted in Figure 5.
Figure 5. The true monoclinic P cell is shown in red (projected along the
b axis). Its pseudomerohedrally twinned unit cell is obtained by a 1808 rota-
tion about vector [102] to generate the green cell on the right. The red
monoclinic primitive lattice can be transformed into the blue C-centred or-
thorhombic lattice with matrix
.
The structure of compound 1·2CH2Cl2 was successfully solved and
all non-hydrogen atoms could be located, however, the refinement
stalled and showed typical signs of twinning: the R-factor stabi-
lized at over 20% and many Fobs values were systematically higher
than the corresponding Fcalc values. We also noted very slight split-
ting in a few of the spots of the diffraction pattern, also hinting
that the crystals of compound 1·2CH2Cl2 were twinned. This was,
however, not observed consistently throughout the diffraction pat-
tern.
For compound 2·solvent (Figure 4), the systematic absences in the
diffraction data were consistent for the space groups P1 and P1.
The E-statistics strongly suggested the centrosymmetric space
group P1, which yielded chemically reasonable and computational-
ly stable results of refinement.[29–31] A successful solution by direct
methods provided most non-hydrogen atoms from the E-map. The
remaining non-hydrogen atoms were located in an alternating
series of least-squares cycles and difference Fourier maps. All non-
hydrogen atoms were refined with anisotropic displacement coeffi-
cients unless specified otherwise. All hydrogen atoms were includ-
ed in the structure factor calculation at idealized positions and
were allowed to ride on the neighbouring atoms with relative iso-
tropic displacement coefficients. The Pd atom coordinates directly
to O1 and O4 (Figure 4), strongly supporting the argument that
the hemi-lability of the o-methoxytriphenylphosphine ligands plays
a role in their function as methoxycarbonylation catalysts.
The program PLATON[26] was then used to detect possible twin-
ning, and indeed the twin law:
0
1
ꢀ
1
0
1
0
1
B
@
C
A
ꢀ
1
0
0
0
corresponding to a 1808 rotation about the [102] vector was iden-
tified. The suggested twin law was incorporated into the instruc-
tion file with a TWIN/BASF combination for the program SHELXL.[23]
Thereafter, the refinement quickly converged to an R-factor of
3.21%. The minor twin component contribution was found to be
The triflate S2 in compound 2·solvent is disordered over two posi-
tions with the major component being present 82.6(3)% of the
ChemCatChem 2016, 8, 1084 – 1093
www.chemcatchem.org
1091
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
time. The minor component of this disorder was refined isotropi-
cally with interatomic distance restraints and thermal displacement
parameter constraints. There were also partially occupied disor-
dered solvent molecules present in the asymmetric unit. A signifi-
cant amount of time was invested into identification and refine-
ment of these diffusely diffracting species. The best model corre-
sponds to the presence of 85.4(4)% of a H2O molecule and
14.6(4)% of a dichloromethane molecule residing in the solvent ac-
cessible void. Atom O13 of the water molecule and atom Cl1 of
the CH2Cl2 molecule almost share the same site. To confirm this re-
finement model, an alternative way of estimating the amount of
electron density corresponding to the diffusely diffracting species
was undertaken. This time the routine SQUEEZE of the PLATON[26]
program was used to account for the solvent molecules. SQUEEZE
computed 12 electrons in the asymmetric unit for the diffusely dif-
fracting species. Our model, (H2O)0.854(CH2Cl2)0.146, contains 14.67
electrons, a satisfactory approximation. The water hydrogen atoms
could not be located and were positioned arbitrarily. The solvent
molecules were refined with anisotropic displacement parameter
constraints. The final least-squares refinement of 658 parameters
against 9620 data resulted in residuals R (based on F2 for Iꢀ2s)
and wR (based on F2 for all data) of 0.0390 and 0.0899, respectively.
The final difference Fourier map was featureless.
and 2.242(4) Š, computed for closely related compounds. A full
geometric analysis of bond lengths, bond angles, dihedral angles
and ring geometries was also carried out on compound 2 by using
Mogul v1.7 and only the unusual geometric parameters are report-
ed here. Search criteria used in Mogul were the same criteria as
that used for compound 1. The search returned four unusual re-
sults worth mentioning. The first two relate to the bond lengths of
P1ÀC8 and P2ÀC29, which measure 1.795(3) Š and 1.793(3) Š, re-
spectively. These distances are significantly shorter than the aver-
age distance of 1.826(2) Š for a PÀC bond. The shortening of the
bond is most probably a result of some eÀ delocalization from the
Pd metal centre through to the P atoms of the complex. A total of
238 hits of similar fragments in the CSD[19] were found, correspond-
ing to 115 structures that showed similarities in terms of shorter
bond lengths. Of these 115 structures, only one structure was
deemed similar to compound 2—this being (acetato)-(tris(2-me-
thoxyphenyl)phosphine)-(2,6-dimethylphenyl)-palladium (Refcode:
MUBFIA[27]), in which there is a similar shortening of the PÀC bond
and the Pd is bonded directly to the O atom of a methoxy group
as in compound 2. The third unusual result relates to the C7-O1-C2
bond angle spanning 115.78(3)8. This angle is significantly smaller
than the average value of 117.5(5)8 for angles with similar atom
types in similar fragments searched in the CSD. We propose that
this is likely due to the C7 shift towards C2, which results in angle
contraction so that the lone pairs on O1 can coordinate to the Pd
metal centre. In this case, only 15 hits of similar fragments in the
CSD were found, corresponding to nine structures showing similar-
ities in terms of a smaller bond angle. Again, only one structure
was deemed similar to compound 2—this also being MUBFIA[27] in
which there is a similar reduction in the C-O-C bond angle value of
the methoxy group bound to the Pd atom. The fourth unusual
result relates to the dihedral angle of À48.62(13)8 for the C1-P1-
C8-C9 fragment of the molecule—referring to the o-methoxy
group on centroid Cg(4) (the six-membered phenyl ring made up
of C8-C9-C10-C11-C12-C13) that is bonded to P1, which in turn is
bonded to C1 on centroid Cg(3) (the six-membered phenyl ring
made up of C1-C2-C3-C4-C5-C6). This unusual dihedral angle is
probably a result of the steric hindrance of the ligand system.
Here, a total of 84 hits of similar fragments in the CSD were found
to have a deviation in the dihedral angle of this particular frag-
ment of the molecule—in 12 structures. Of these 12 structures, the
same MUBFIA structure was similar enough to compound 2. Thus,
compound 2 exhibits similar features in these specific geometric
parameter deviations, making it only the second compound with
such structural properties.
A molecular structure of the Pd cation is shown in Figure 6, clearly
indicating the coordination of the Pd atom to the oxygen atoms
O1 and O4 of the ligand.
Figure 6. Molecular structure of the cation of compound 2, clearly showing
coordination of the Pd atom to O1 and O4. Thermal displacement ellipsoids
drawn at the 50% probability level. All hydrogen atoms are omitted for clari-
ty.
CCDC 1419005 and 1423690 contain the supplementary crystallo-
graphic data for this paper. These data can be obtained free of
charge from The Cambridge Crystallographic Data Centre.
The slightly distorted square-planar coordination environment of
the Pd metal centre is formed by two P and two O atoms. Howev-
er, the two five-membered metallacycles (centroids Cg(1) and
Cg(2) comprising Pd1-P1-C1-C2-O1 and Pd1-P2-C22-C23-O4, re-
spectively) that form as a result of the ligand coordination to Pd1
display ring puckering. The Cremer–Pople[32] ring puckering analy-
sis shows that Cg(1) is distorted into the envelope form on Pd1
with a puckering amplitude (Q) of 0.460(2) Š and a phase angle of
7.3(4)8. Cg(2) is also distorted into the envelope form on Pd1 with
a puckering amplitude Q=0.3827(19) Š and a phase angle of
7.6(5)8. We attribute this ring distortion to the steric factors of the
ligand system on the Pd metal centre.
Acknowledgements
We thank Sasol, the South African National Research Foundation
and the University of Johannesburg for financial assistance with
this project.
Keywords: branched
methoxycarbonylation · palladium
esters
·
hemi-labile
·
[1] a) C. W. Holzapfel, T. Bredenkamp, ChemCatChem 2015, 7, 2598–2606;
b) G. Kiss, Chem. Rev. 2001, 101, 3435–3456 and references therein;
c) G. Walther, A. Martin, A. Kçckritz, J. Am. Oil Chem. Soc. 2013, 90, 141–
145; d) M. R. L. Furst, R. Le Goff, D. Quinzler, S. Mecking, C. H. Botting,
The PdÀO and PdÀO bond lengths are typical and are in good
agreement with the corresponding average values of 2.179(4) Š
ChemCatChem 2016, 8, 1084 – 1093
www.chemcatchem.org
1092
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
D. J. Cole-Hamilton, Green Chem. 2012, 14, 472–477; e) G. Walther, L. R.
Chem. Soc. 2003, 125, 5523–5539; b) S. M. A. Donald, S. A. Macgregor,
V. Settels, D. J. Cole-Hamilton, G. R. Eastman, Chem. Commun. 2007,
562–564.
´
Knçpke, J. Reekah, M. P. Che˛cinski, H. Jiao, U. Bentrup, A. Brüchner, A.
Martin, A. Kçckritz, J. Catal. 2013, 297, 44–55; f) J. T. Christl, P. R. Roelse,
F. Stempfle, P. Wucher, I. Gçttker-Schnetmann, G. Müller, S. Mecking,
Chem. Eur. J. 2013, 19, 17131–17140; g) P. Roesle, C. J. Dürr, H. M.
Mçller, L. Cavallo, L. Caporaso, S. Mecking, J. Am. Chem. Soc. 2012, 134,
17696–17703; h) G. Walther, J. Deutsch, A. Martin, F. E. Baumann, D.
Fridag, R. Franke, A. Kçckritz, ChemSusChem 2011, 4, 1052–1054; i) S.
Chikkali, F. Stempfle, S. Mecking, Macromol. Rapid Commun. 2012, 33,
1126–1129; j) L. Kóllµr, Modern Carbonylation Methods, Wiley-VCH, Wein-
heim, 2008, and references therein; k) D. B. G. Williams, T. M. Breden-
kamp, ChemCatChem 2012, 4, 206–208; l) D. B. G. Williams, M. L. Shaw,
M. J. Green, C. W. Holzapfel, Angew. Chem. Int. Ed. 2008, 47, 560–563;
Angew. Chem. 2008, 120, 570–573; J. Yang, Y. Yuan, Catal. Lett. 2009,
131, 643–648; m) C. van Blerk, C. W. Holzapfel, Acta. Crystallogr. Sect. E
2009, 65, m1341; n) B. Makume, T. Bredenkamp, D. B. G. Williams, Chem-
CatChem 2014, 6, 2801–2804; o) G. Cavinato, L. Toniolo, A. Vavasori, J.
Mol. Catal. A 2004, 219, 233–240; p) R. P. Tooze, K. Whiston, A. P.
Malyan, M. J. Taylor, N. W. Wilson, J. Chem. Soc. Dalton Trans. 2000,
3441–3444; q) G. R. Eastham, R. P. Tooze, M. Kilner, D. F. Foster, D. J. Co-
le Hamilton, J. Chem. Soc. Dalton Trans. 2002, 1613–1617.
[11] a) I. del Rio, N. Ruiz, C. Claver, Inorg. Chem. Commun. 2000, 3, 166–168;
b) M. C. Bonnet, A. L. Monteiro, I. Tkatchenko, J. Mol. Catal. A 1999, 143,
131–136.
[12] E. Drent, W. P. Mul, P. H. M. Budzelaar, Comments Inorg. Chem. 2002, 23,
127–147.
[13] F. Bertoux, Y. Castanet, E. Civade, E. Monflier, A. Morteux, Catal. Lett.
1998, 54, 199–205.
[14] J. Y. Yoon, E. J. Jong, K. H. Lee, J. S. Lee, J. Mol. Catal. A 1997, 118, 181–
187.
[15] a) G. R. Eastman, B. T. Heaton, J. A. Iggo, R. P. Tooze, R. Whyman, S. Zac-
chini, Chem. Commun. 2000, 609–610; b) J. Liu, B. T. Heaton, J. A. Iggo,
R. Whyman, Angew. Chem. Int. Ed. 2004, 43, 90–94; Angew. Chem. 2004,
116, 92–96.
[16] D. A. Redfield, H. H. Nelson, J. Am. Chem. Soc. 1974, 96, 6219–6220.
[17] J. N. Harvey, Organometallics 2001, 20, 4887–4895.
[18] a) D. A. Redfield, H. H. Nelson, Inorg. Chem. 1973, 12, 15–19; b) L. M.
Knight, J. H. Nelson, Inorg. Chem. 1977, 16, 1317–1321; c) A. L. Casado,
P. Espinet, Organometallics 1998, 17, 954–959; d) D. Minniti, Inorg.
Chem. 1994, 33, 2631–2634.
[2] S. Boyde, Green Chem. 2002, 4, 293–307.
[3] a) I. del Rio, C. Claver, L. A. van der Veen, P. W. N. M. van Leeuwen, J. Mol.
Catal. A 2000, 161, 39–48; b) B. K. MuÇoz, A. G. E. Santos, C. Godard, E.
Zangrando, C. Bo, A. Ruiz, C. Claver, Eur. J. Inorg. Chem. 2008, 4625–
4637; c) J. J. R. Frew, K. Damian, H. Van Rensburg, A. Z. Slawin, R. P.
Tooze, M. L. Clarke, Chem. Eur. J. 2009, 15, 10504–10513; d) P. Aguirre,
C. A. Lagos, S. A. Moya, C. Zuniga, C. Vera-Oyrace, E. Soia, G. Peris, J. C.
Bayçn, Dalton Trans. 2007, 5419–5426; e) Y. G. Noskov, E. S. Petrov,
Russ. Chem. Bull. Int. Ed. 2001, 50, 1839–1843; f) T. M. Konrad, J. T. Dur-
rani, C. J. Cobley, M. L. Clarke, Chem. Commun. 2013, 49, 3306–3308;
g) T. M. Konrad, J. A. Fuentes, A. M. Slawin, D. M. L. Clarke, Angew. Chem.
Int. Ed. 2010, 49, 9197–9200; Angew. Chem. 2010, 122, 9383–9386;
h) P. H. Gehrtz, V. Hirshbeck, I. Fleischer, Chem. Commun. 2015, 51,
12574–12577; i) Z. He, Z. Hou, Y. Luo, Y. Dilixiati, W. Eli, Catal. Sci. Tech-
nol. 2014, 4, 1092–1103.
[19] C. R. Groom, F. H. Allen, Angew. Chem. Int. Ed. 2014, 53, 662–671;
Angew. Chem. 2014, 126, 675–684.
[20] A. J. Po, K. Shaw, M. J. Wendt, Inorg. Chim. Acta 1967, 1, 371–377.
[21] a) G. Cavinato, L. Toniolo, Molecules 2014, 19, 15116–15161; b) C. Bian-
chini, A. Meli, W. Oberhauser, C. Claver, E. J. G. Suarez, Eur. J. Inorg.
Chem. 2007, 2702–2710; c) A. M. Kalsin, V. N. Vologdin, T. A. Peganova,
P. V. Petrovskii, K. A. Lyssenko, F. M. Dolgushin, O. V. Gusev, J. Organo-
met. Chem. 2006, 691, 921–927.
[22] Bruker APEX2, SADABS, and SAINT. Bruker AXS Inc., Madison, Wisconsin,
USA, 2009.
[23] G. M. Sheldrick, Acta Crystallogr. Sect. A 2008, 64, 112–122.
[24] Fundamentals in Crystallography, 2nd ed. (Ed.: C. Giacovazzo), Oxford Uni-
versity Press, Oxford, 2002.
[25] I. A. Guzei, R. Herbst-Irmer, A. Munyaneza, J. Darkwa, Acta Crystallogr.
Sect. B 2012, 68, 150–157.
[26] A. L. Spek, Acta Crystallogr. Sect. D 2009, 65, 148–155.
[27] M. Wakioka, Y. Nakamura, M. Montgomery, F. Ozawa, Organometallics
2015, 34, 198–205.
[28] Y. H. Cheng, C. M. Weng, F. E. Hong, Tetrahedron 2007, 63, 12277–
12285.
[29] Bruker APEX2, COSMO, AXScale, and SAINT. Bruker AXS Inc., Madison,
Wisconsin, USA, 2010.
[4] a) T. O. Vieira, M. J. Green, H. Alper, Org. Lett. 2006, 8, 6143–6145; b) C.
Jimenez Rodriguez, D. F. Foster, G. R. Eastman, D. J. Cole-Hamilton,
Chem. Commun. 2004, 1720–1721.
[5] a) E. N. Squire, G. M. Pa, F. J. Waller, W. Del, (DSM N. V. E. I Du Pont De
Nemours and company) US Patent 4292437, 1981; b) W. P. Shum, J. F.
White, (Atlantic Richfield Company) US Patent 4612390, 1986.
[6] a) V. Santanna, F. D. S. Curizelo, T. N. C. Dantas, A. A. Nato, H. S. Albu-
querque, A. I. C. Carnica, J. Pet. Sci. Eng. 2009, 66, 117–120; b) A. Bera,
A. Wandal, B. B. Guha, J. Chem. Eng. Data 2014, 59, 89–96; c) J. R.
Knope, (Union Carbide Corporation, Dow), European Patent 310194A1,
1989.
[30] OLEX2: a complete structure solution, refinement and analysis program,
O. V. Dolomanov, L. J. Bhouris, R. J. Gildea, J. A. K. Howard, H. Pusch-
mann, J. Appl. Crystallogr. 2009, 42, 339–341.
[7] H. Alper, J. B. Woell, B. Despeyroux, D. J. H. Smith, J. Chem. Soc. Chem.
Commun. 1983, 1270–1271.
[31] I. A. Guzei, Internal laboratory computer programs “Inserter”, “FCF_
filter”, “Modicifer” (2006–2008).
[8] M. Rosales, I. Pacheco, J. Medina, J. Fernandµz, A. Gonzµlez, R. Izuierdo,
L. G. Melean, P. J. Baricelli, Catal. Lett. 2014, 144, 1717–1727.
[32] D. Cremer, J. A. Pople, J. Am. Chem. Soc. 1975, 97, 1354–1358.
[9] T. Iwasawa, M. Tokunga, Y. Obora, Y. Tsuji, J. Am. Chem. Soc. 2004, 126,
6554–6555.
Received: December 29, 2015
[10] a) P. W. N. M. van Leeuwen, M. A. Zuideveld, B. H. G. Swennenhuis, Z.
Freixa, P. C. J. Kramer, K. Goubitz, J. Fraanje, M. Lutz, A. L. Spek, J. Am.
Published online on February 17, 2016
ChemCatChem 2016, 8, 1084 – 1093
www.chemcatchem.org
1093
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim