Mechanistic study of the L2Pd-catalyzed reduction of nitrobenzene with CO in methanol: Comparative study between diphosphane and 1,10-phenanthroline complexes

Article abstract of DOI:10.1021/om200882p

The catalytic activity of Pd^II compounds supported by 1,10-phenanthroline (phen) or the bidentate diarylphosphane ligand L4X has been studied in the reaction of nitrobenzene with CO in methanol. Both systems are ～70% selective for azoxybenzene and azobenzene but also produce carbonylation products (methyl phenyl carbamate (MPC) and N,N'-diphenylurea (DPU)) and hydrogenation products (aniline and DPU). The Pd^II(L4X) system also produces methanol oxidation products (dimethyl carbonate, dimethyl oxalate, water). Upon the addition of a catalytic amount of acid, the coupling reaction is suppressed in favor of either the carbonylation reaction (for Pd^II(phen)) or of both the carbonylation and hydrogenation reaction (for Pd^II(L4X)). The palladacycle L₂PdC(O)N(Ph)OC(O) (C7) and palladium-imido species L₂Pd=NPh (C3) were considered as possible carbonylation product-releasing species, where L₂ is phen or the diphosphane ligands L4X and L3X. A ligand exchange experiment of phen-C7 with L4X and L3X, ESI-MS analysis of L3X-C7 and phen-C7, and a DFT study of nitrobenzene deoxygenation intermediates to L₂Pd=NPh all suggest that C7 is not the major product-releasing intermediate; all data suggest that the barrier for C7 decarbonylation (-CO) is lower than that for decarboxylation (-CO₂). C7 is thus thought to be part of an NPh reservoir consisting of palladacycles that are mutually accessible through carbonylation/decarbonylation. Under acidic conditions the decarboxylation barrier is lowered; for phen-C7 apparently to the point where decarboxylative alcoholysis is favored relative to decarbonylation, but for L4X-C7 the decarbonylation barrier still seems lowest due to the destabilizing effect that this bulkier ligand has on such palladacycles. It is thus concluded that the L₂Pd=NPh complex C3 is the prime NPh product-releasing intermediate and only under acidic conditions and in an alcoholic environment may C7-for phenanthroline-become the predominant carbamate product releasing intermediate.

Full text of DOI:10.1021/om200882p

Article
pubs.acs.org/Organometallics
Mechanistic Study of the L2Pd-Catalyzed Reduction of Nitrobenzene
with CO in Methanol: Comparative Study between Diphosphane and
1,10-Phenanthroline Complexes
Tiddo J. Mooibroek, Elisabeth Bouwman,* and Eite Drent
Leiden Institute of Chemistry, Leiden University, P.O. Box 9502, 2300 RA Leiden, The Netherlands
S
* Supporting Information
ABSTRACT: The catalytic activity of PdII compounds supported by
1,10-phenanthroline (phen) or the bidentate diarylphosphane ligand
L4X has been studied in the reaction of nitrobenzene with CO in
methanol. Both systems are ∼70% selective for azoxybenzene and
azobenzene but also produce carbonylation products (methyl phenyl
carbamate (MPC) and N,N’-diphenylurea (DPU)) and hydrogenation
products (aniline and DPU). The PdII(L4X) system also produces
methanol oxidation products (dimethyl carbonate, dimethyl oxalate,
water). Upon the addition of a catalytic amount of acid, the coupling reaction is suppressed in favor of either the carbonylation
reaction (for PdII(phen)) or of both the carbonylation and hydrogenation reaction (for PdII(L4X)). The palladacycle
L2PdC(O)N(Ph)OC(O) (C7) and palladium−imido species L2PdNPh (C3) were considered as possible carbonylation
product-releasing species, where L2 is phen or the diphosphane ligands L4X and L3X. A ligand exchange experiment of phen-C7
with L4X and L3X, ESI-MS analysis of L3X-C7 and phen-C7, and a DFT study of nitrobenzene deoxygenation intermediates to
L2PdNPh all suggest that C7 is not the major product-releasing intermediate; all data suggest that the barrier for C7
decarbonylation (−CO) is lower than that for decarboxylation (−CO2). C7 is thus thought to be part of an “NPh reservoir”
consisting of palladacycles that are mutually accessible through carbonylation/decarbonylation. Under acidic conditions the
decarboxylation barrier is lowered; for phen-C7 apparently to the point where decarboxylative alcoholysis is favored relative to
decarbonylation, but for L4X-C7 the decarbonylation barrier still seems lowest due to the destabilizing effect that this bulkier
ligand has on such palladacycles. It is thus concluded that the L2PdNPh complex C3 is the prime “NPh” product-releasing
intermediate and only under acidic conditions and in an alcoholic environment may C7for phenanthrolinebecome the
predominant carbamate product releasing intermediate.
Scheme 1. Alternative Route To Make Aromatic Isocyanates
by Carbonylation Followed by Pyrolysis
1. INTRODUCTION
Studies in homogeneous catalysis are often motivated by
financial and/or ecological incentives. One process in which
both motivators are in play is the synthesis of aromatic
isocyanates1,2 such as the polymer precursors MDI and TDI,3,4
which are produced on a megaton scale per year. These use
respectively nitrobenzene and dinitrotoluene as the ultimate
feedstocks, while the highly toxic5,6 and corrosive2 phosgene is
employed as the carbonylating reagent.
Phosgene can in principle be replaced by the less toxic
carbon monoxide as the carbonylating reagent, whereby CO
also acts as the reductant/deoxygenating agent for nitro-
benzene; this process is known to proceed in the presence of a
transition-metal catalyst.7 When this reductive carbonylation of
nitrobenzene with carbon monoxide is conducted in a protic
medium such as an alcohol, a carbamate is obtained which can
be pyrolyzed to the isocyanate with the recovery of the alcohol
(Scheme 1).
(in the presence of an acid cocatalyst), which resulted in the
focus of the scientific community on this Pd/phen/H+/
CH3OH catalytic system.7−16 Although a very active catalytic
system has been reported (with ∼105 turnover numbers),13
a
clear-cut and generally agreed upon mechanism for this
reaction involving common intermediates has yet to emerge.
We have recently reported on our mechanistic studies of this
carbonylation reaction in methanol, using bidentate diaryl-
phosphane ligands in combination with a source of
palladium.17,18 During these studies, it became clear that for
this catalytic system there is a catalytic connection between
nitrobenzene reduction chemistry (forming methyl phenyl
carbamate (MPC), N,N′-diphenylurea (DPU), azoxybenzene,
It was disclosed already in the early 1980s that reasonably
active (∼500 turnover numbers) palladium-based catalysts for
this carbonylation reaction are obtained when the Pd center is
stabilized by bidentate N or P ligands.8 Use of the ligand 1,10-
phenanthroline (phen) yielded the most active catalytic systems
Received: September 21, 2011
Published: May 31, 2012
© 2012 American Chemical Society
4142
dx.doi.org/10.1021/om200882p | Organometallics 2012, 31, 4142−4156

Organometallics
Article
aniline, and water) and methanol oxidation chemistry (forming
dimethyl carbonate (DMC), dimethyl oxalate (DMO), methyl
formate (MF), aniline, water, and carbon monoxide).17 This
experimental observation has far-reaching implications for the
underlying mechanisms. Thus, we proposed a stepwise
deoxygenation of nitrobenzene, wherein only CO, only
methanol, or CO and methanol act as the reductant for
nitrobenzene to yield a diphosphane ligand supported
palladium−imido complex (P2PdNPh, left-hand side of
Scheme 2).17 In a methanolic environment, this imido complex
1300) on Pdcomes from the Ragaini group.16 They proposed
that carbamates and ureas are formed as the final carbonylation
products from initially formed isocyanates involving a direct
carbonylation of the corresponding amine. It is speculated that
the direct carbonylation of the amine to the corresponding
isocyanate involves a five-coordinated (phen)Pd(CO)(C(O)-
OCH3)2 catalyst species together with a proposed Pd−
carbenoid type intermediate. The main reasons for this
mechanistic proposal lie in the observed cocatalytic effect of
amine and the observation that carbonylation of a d5-labeled
nitrobenzene in equimolar amounts of aniline produces a
significant amount of unlabeled urea.
In contrast, a more common rationalization of carbamate
formation for the Pd/phen/CH3OH(/H+) system, prior to the
proposal of Ragaini et al., involves a palladacyclic intermediate
as the central product-releasing intermediate (top left in
Scheme 3). This palladacycle can be synthesized and isolated by
Scheme 2. Working Hypothesis of the Interrelated Catalytic
Cycles Operating in the P2Pd-Catalyzed Reaction of
Nitrobenzene with CO in Methanol, Rationalizing the
Genesis of All the Products Observed17
Scheme 3. Palladacycle (Top Left) Proposed To Be the Key
Intermediate MPC (Top Right) Product-Releasing Species
in the Pd/phen/CH3OH(/H+) Catalytic System via a
Pathway Adapted from a Proposal by Osborn et al. (Top)11
a
and van Leeuwen et al. (Bottom)9
can readily form the anilido−methoxido palladium species
P2Pd(NHPh)(OCH3) and dimethoxido species Pd2Pd-
(OCH3)2 with the coproduction of aniline. Under a CO
atmosphere, these anilido−methoxido type palladium com-
plexes naturally will convert into a kinetic mixture of palladium
anilido−carbomethoxylates (e.g., P2Pd(C(O)NHPh)OCH3,
P2Pd(C(O)OCH3)OCH3, and P2Pd(C(O)OCH3)2) that re-
spectively produce MPC, DMC, and DMO upon reductive
elimination.17,18 It is thus our current thinking that the genesis
of nitrobenzene carbonylation products such as MPC follows a
pathway very similar to and competing with the formation of
DMC and DMO: i.e., via a net hydromethoxycarbonylation of
the palladium−imido intermediate. The P2PdNPh inter-
mediate can also rationalize the genesis of products such as
azoxybenzene by a [2 + 2′] cycloaddition type reaction with
nitrobenzene; this places azoxybenzene formation within our
overall mechanistic proposal as well (see right-hand side of
Scheme 2). We have thus constructed a hypothesis centered
around the palladium−imido intermediate that allows ration-
alization of the catalytic formation of all products observed in
nitrobenzene carbonylation experiments using diphosphane-
supported palladium catalysts. We were also able to directly
apply this hypothesis to understand alterations in product
selectivity patterns that were induced by the structure of the
ligand or the choice of additives and reaction conditions
employed.17,18 Although the intermediacy of palladium−imido
complexes has been postulated in earlier reports, mechanistic
details and experimental evidence have not been put forth to
justify these speculations.7,9,10,19
a
N2 = 1,10-phenanthroline.
exposing a Pd/phen mixture in an alcoholic nitrobenzene
medium to mild carbonylation conditions.11,20−22 This pallada-
cycle is thermally very stable and only decomposes to yield 0.2
equiv of isocyanate when heated to 170 °C under 20 bar of CO
in an aprotic medium.11 Addition of 1.6 equiv of a weak acid
(2,4,6-trimethylbenzoic acid) leads to a 3-fold increase in
isocyanate formation, while in the presence of acid and ethanol
as solvent, carbamate/urea formation already proceeds
smoothly at 90 °C with more than 90% isolated yield.11
Osborn et al. proposed (Scheme 3, top) that protonation of the
palladacycle at the carbonyl O may lower the barrier of CO2
extrusion by nucleophilic attack of the alcohol on the amide
carbonyl to produce the carbamate directly.11 Van Leeuwen et
al.9 postulated (Scheme 3, bottom) that protonation occurs on
the ring N atom, followed by CO2 extrusion and [PdC(O)-
NHPh]+ formation. In a subsequent transesterification event
with methanol, aniline is thought to be formed together with
the carbomethoxy cation [PdC(O)OCH3]+; nucleophilic attack
of aniline on this intermediate then leads to MPC formation.
The question naturally arises as to the relationship between
the diphosphane- and phen-based catalytic systems and to the
relationship between proposed palladium−imido and pallada-
cyclic intermediates.
Our mechanistic proposal deviates significantly from those
postulated by others, which are mainly based on phen-
supported palladium catalyst systems. The most recent
hypothesis for the mechanism of the Pd/phen-catalyzed
carbonylation of nitroarenes in methanolic mediausing
large molar excesses of phen (up to 300) and acid (up to
The aim of the present study is therefore to gain crucial
additional catalytic and organometallic data to understand the
differences and similarities between the two systems. In
particular, the relative importance of the proposed palla-
dium−imido and palladacycle intermediates was considered.
4143
dx.doi.org/10.1021/om200882p | Organometallics 2012, 31, 4142−4156

Organometallics
Article
Because we found that the product distribution of the aryl-
containing nitrobenzene reduction products is remarkably
similar (especially for azoxybenzene) when employing the N-
donor ligand phen (Figure 1, left) or the P-donor ligand L4X
is always the major product and Azo the minor product (<2%
selectivity).
2.2. Catalytic Experiments. 2.2.1. Reactivity without
Additives or Cocatalysts. Using GLC analysis of the reaction
mixtures all products could be quantified with calibration lines
made from authentic samples. Selected data are given in Table
1; full analytical details of all reactions are given as Supporting
Information (Table S1). The mass balance of phenyl-
containing products is excellent, showing that all important
products have indeed been identified and quantified (column
∑ø in Table 1).
Figure 1. N-donor ligand phen and P-donor ligands L4X and L3X
used in this study.
When the PdII(phen) system was employed at a reaction
temperature of 110 °C (entry 1), only 17% conversion was
reached, whereas at a reaction temperature of 130 °C (entry 2),
56% nitrobenzene was converted. This conversion at 130 °C is
similar to that of the PdII(L4X) catalytic system at 110 °C
(52%, entry 3). Both catalytic systems are about equally
selective for the coupling product Azoxy (∼70%); PdII(phen) is
more selective toward carbonylation (20%), while PdII(L4X) is
more selective toward hydrogenation (20%). Notably, using
PdII(L4X), large quantities of oxidation products of methanol
are coproduced, whereas these products are hardly formed
when using PdII(phen).
(Figure 1, middle), these ligands have mainly been used for the
catalytic experiments in this study; the ligand L3X (Figure 1,
right) was mainly used for insight supporting (organometallic)
experiments.
2. RESULTS
2.1. General Considerations. In most catalytic reactions,
the catalyst precursor was formed in situ from 0.05 mmol of
Pd(OAc)2 and a certain amount of ligand; testing the catalytic
activity of preformed complex gave identical results. By
thorough drying of the reagents it was ensured that the
reaction conditions were strictly anhydrous (<100 ppm of
H2O).17
The products that were observed in the catalytic reactions are
shown in Scheme 4. The aryl-containing reduction products of
nitrobenzene can be grouped into the carbonylation products
(NCO) methyl phenyl carbamate (MPC) and N,N′-dipheny-
lurea (DPU), the coupling products (NN) azobenzene
(Azo) and azoxybenzene (Azoxy), and the hydrogenation
products (NH) aniline (PhNH2) and also DPU (derived from
aniline).23
Azo and Azoxy are frequently reported side products, as are
aniline and DPU.7 Azo and Azoxy are only derived from
nitrobenzene, but for the formation of aniline and DPU H
atoms are required. We have shown that these H atoms
originate from methanol oxidation, resulting in the oxidative
carbonylation products dimethyl carbonate (DMC) and
dimethyl oxalate (DMO) and in the oxidative dehydrogenation
products methyl formate (MF) and carbon monoxide.17
At reaction temperatures (110−130 °C) DPU in methanol is
involved in a reversible transesterification reaction yielding
MPC and aniline;24 therefore, DPU and MPC should be
considered together as carbonylation products and aniline and
DPU together as hydrogenation products. The coupling
products Azo and Azoxy were almost always detected; Azoxy
2.2.2. Effects of the Initial Nitrobenzene Concentration.
The similar selectivity for reductive coupling products (yielding
predominantly Azoxy) with PdII(phen) and PdII(L4X) catalyst
systems is quite remarkable, given the different structural
characteristics of these catalyst systems. Intrigued by an
observed higher order formation kinetics in nitrobenzene, in
particular with Pd catalyst supported by larger backbone (C4)
diphosphane ligands,17 we investigated the effect of the initial
concentration of the nitrobenzene on product yields. It appears
that the relative ratio of carbonylation (NCO) and hydro-
genation (NH) products over coupling (NN) products
decreases significantly with increasing nitrobenzene concen-
trations (Supporting Information, Figure S1 and Table S1).
These data suggest that also for PdII(phen) catalysts the
formation of Azo(xy) coupling products from nitrobenzene
involves a step competing with carbonylation and hydro-
genation, with higher order kinetics in nitrobenzene.
2.2.3. Effects of Acid Addition. It has been reported that
(over)stoichiometric amounts of acid are beneficial for the
MPC selectivity of the PdII(phen) system.9,12,14 We therefore
decided to investigate the effects of the addition of acid for
PdII(phen) and PdII(L4X) catalyst systems under our reaction
conditions. Thus, some experiments were conducted wherein
(sub)stoichiometric amounts (on Pd) of p-toluenesulfonic acid
Scheme 4. Overview of the Different Products That Are Formed in the Palladium-Catalyzed Reaction of Nitrobenzene with CO
in Methanol
4144
dx.doi.org/10.1021/om200882p | Organometallics 2012, 31, 4142−4156

Organometallics
Article
a
Table 1. Reactions of Nitrobenzene with CO in Methanol, Using PdII(phen)(OAc)2 or PdII(L4X)(OAc)2 as Catalyst Precursor
selectivity (%)
amt (mmol)
H2O
b
c
d
e
entry
ligand
temp (°C)
conversn (%)
∑
NCO
NN
NH
MF
DMC/O
ø
1
2
3
phen
phen
L4X
110
130
110
17
56
52
24.7
24.3
24.4
23
22
8
71
71
73
6
7
0.0
0.2
0.2
0.0
0.3
0.0
0.5
8.2
19
10.1
a
Reaction mixtures were heated for 4 h at the indicated temperature in 25.0 mL of dry and degassed methanol under 50 bar of CO. The catalyst was
generated in situ from 0.05 mmol of Pd(OAc)2. Mole ratios are as follows: Pd(OAc)2:ligand:nitrobenzene = 1:1.5:488. All reactions were performed
b
c
in quadruplicate, and the relative standard deviation of all analytes was <5% in all cases. Conversion = (∑aryl converted)/24.4 × 100%. Selectivity
d
toward carbonylation products = (MPC + DPU)/∑aryl converted × 100%. Selectivity toward coupling products = (2 × Azo + 2 × Azoxy)/∑aryl converted
e
× 100%. Selectivity toward hydrogenation products = (PhNH2 + DPU)/∑aryl converted × 100%.
●
○
Figure 2. Plot of the conversion ( ) and selectivity toward coupling products ( , Azo(xy)), hydrogenation products ( , DPU + PhNH2), and
carbonylation products ( , MPC + DPU) as a function of the amount of p-toluenesulfonic acid added (relative to Pd), when using the ligands L4X
at 110 °C (a) and phen at 130 °C (b). The lines are added as a guide for the eye.
(HOTs) were added, using our standard experimental
conditions (110 °C for PdII(L4X) and 130 °C for PdII(phen)).
As shown in Figure 2a for the PdII(L4X) system, the
conversion increases with the addition of more HOTs, whereas
for the PdII(phen) catalytic system (Figure 2b), the conversion
is more or less constant. Interestingly, in both systems the
coupling reaction is significantly suppressed already when
adding substoichiometric amounts of HOTs. The change in
selectivity, however, is drastically different for the two catalysts:
for PdII(L4X) both the hydrogenation and carbonylation
reactions become more important, whereas for PdII(phen)
the carbonylation reaction becomes highly dominant. This
observed difference in selectivity change upon addition of acid
indicates that dissimilar reaction mechanisms are operative for
the two systems.
2.2.4. Nitrosobenzene as Intermediate toward Azo or
Azoxy? It has been suggested that, in the PdII(phen) catalytic
system, nitrobenzene is first reduced to “free” nitrosobenzene,
which then would react further to form mainly Azoxy and some
Azo.10,15 We have investigated whether elevated amounts of
Azo and Azoxy are detected upon addition of 2.5 mmol of
nitrosobenzene during a catalytic run. However, when either
PdII(L4X) or PdII(phen) is employed, both the conversion and
selectivity are nearly identical with those in a “normal” catalytic
experiment (see Figure S2 in the Supporting Information); no
significant increase in the selectivity for Azo and Azoxy is
observed. Nitrosobenzene is not detected after the catalytic run,
while a significant amount of nitrobenzene is still present. This
confirms that nitrosobenzene is more prone to reductive
carbonylation reactions25 than nitrobenzene, which is con-
sistent with the observation of only traces of nitrosobenzene in
ordinary catalytic experiments.
Azoxy could perhaps also be intermediates to products such as
MPC, DPU, and/or aniline in our catalytic system. The
intermediacy of Azo and Azoxy has been suggested by Mestroni
et al. for the Pd/phen/H+ system at 155 °C26 but was disputed
by Ragaini et al.14 When 2.5 mmol of Azo or Azoxy is added to
the PdII(phen) catalytic system under our standard reaction
conditions, the same amount (plus the amount formed in a
normal run) is detected after the catalytic run, even when 1
equiv of HOTs and 2.5 mmol of Azoxy are used (Table S2,
Supporting Information). This means that Azo and Azoxy are
stable reaction products and are not intermediates toward
MPC, DPU, or aniline employing our standard reaction
conditions of 130 °C and 50 bar of CO.
2.2.6. Is Aniline Directly Carbonylated to Isocyanate for
MPC? Ragaini et al.16 proposed that nitrobenzene, at least with
their catalytic precursor system, [Pd(phen)2](BF4)2 with an
excess of the ligand phen (300 equiv), diphenylphosphinic acid
(Ph2P(O)OH, 1300 equiv), and drying agent (2,2′-dimethox-
ypropane, 4200 equiv), is first fully reduced to aniline. Aniline is
then carbonylated to phenyl isocyanate (via a five-coordinate
(phen)Pd(C(O)OCH3)2CO complex), which in turn forms
MPC in the prevailing methanolic environment.16 To verify
such a scenario with our catalytic system, one could think of
adding labeled aniline (e.g., toluidine) during a catalytic run to
see whether its corresponding carbamate is produced. However,
such an experiment cannot yield meaningful results, as
transesterification and transamidation reactions occur readily
under typical reaction temperatures (i.e., 170 °C employed by
Ragaini et al.). Already at 80 °C, DPU and MPC in methanol
are involved in a reversible transesterification reaction.24 Also, a
labeled aromatic amine and DPU will be involved in a
reversible transamidation reaction; we found that trans-
amidation between 3-methylaniline and DPU occurs to a
significant extent already at 70−80 °C.27
2.2.5. Azo or Azoxy as Intermediate toward MPC, DPU, or
Aniline? We investigated if the coupling products Azo and
4145
dx.doi.org/10.1021/om200882p | Organometallics 2012, 31, 4142−4156

Organometallics
Article
It is well-documented that when aniline is used instead of
methanol as the nucleophilic reactant−solvent to yield DPU,
significantly milder reaction conditions (90−120 °C) can
generally be applied for both phen and diphosphane-based
systems.28−35 With our experimental setup and using
preformed (phen)Pd(OAc)2 as catalyst precursor at 110 °C,
in aniline 76% of nitrobenzene was converted into DPU (53%)
and Azo/Azoxy (47%) (in comparison to the 17% conversion
in methanol at 110 °C, entry 1 in Table 1), but this system
appears to be inactive below 80 °C. On the other hand, nearly
all the Pd/diphosphane catalysts tested in this solvent reach
(nearly) full conversion typically within 1−4 h at 110 °C.36 The
catalyst precursor (L3X)Pd(OAc)2 proved to be particularly
active, reaching full conversion after 4 h at 60 °C. Thus, the
experiment was repeated, using 3-methylnitrobenzene as the
substrate and a large excess of unsubstituted aniline as the
nucleophilic reagent and solvent.37 Full conversion of 3-
methylnitrobenzene was reached; GLC-FID analysis surpris-
ingly revealed the presence of 3-methylaniline (24.4 mmol) as
the only liquid-phase reaction product, while 1H and 13C NMR
analysis of the solid revealed that symmetric, unsubstituted DPU
was produced with 90% selectivity, with the remaining 10%
being diphenyl oxalamide (DPO, verified by ESI-MS). The
formation of large quantities of DPU (and 3-methylaniline) is
in qualitative agreement with reported results by Ragaini et al.16
of an experiment with equimolar quantities of d5-deuterated
nitrobenzene in aniline.
Scheme 5. Envisaged Ligand Exchange for phen-C7 with
L4X or L3X (P2) To Form P2-C7
that ligand exchange indeed took place but that the resulting
product is highly unstable.
In an attempt to slow down the supposed decomposition
reaction, the ligand exchange experiment was repeated, but now
with ligand L3X. Five equivalents of the ligand L3X (dissolved
in deuterated nitrobenzene) was added to 1 equiv of the
complex phen-C7. The yellow suspension was measured with
31P{1H} NMR, showing initially the resonance of ligand L3X
around −25.1 ppm and traces of a new complex around 4−6
ppm (Figure 3a), which was also observed in the reaction
2.3. Considering Two Possible Palladium Intermedi-
ates. 2.3.1. Attempted Synthesis of a Palladacyclic Complex
and DFT Calculations. The five-membered-ring palladacycle
with phen as the supporting ligandoften proposed to be a
key intermediate in the title reaction (i.e., phen-C7 in Scheme
3)7,11,22,26is readily obtained using a literature procedure.11,22
When the same procedure is applied for L4X as the ligand,
however, the anticipated L4X-C7 could not be obtained, nor
could an asymmetric complex attributed to this palladacycle be
observed in 31P NMR spectra of the resulting reaction mixture.
One explanation for the apparent lack of L4X-C7 formation is
that the steric constraints in the plane of coordination, opposite
to the ligand, of such species lead to its destabilization. The
steric constraints are, of course, influenced by the ligand bite
angle, which is much smaller for the ligand phen (∼78°) than
for bidentate diarylphosphane ligands with a butylene backbone
(96°). Indeed, when the above experiment was conducted with
the smaller bite angle ligand L3X (92°), an asymmetric complex
was clearly observed (two doublets in the 31P NMR spectrum
around 3.8 and 5.8 ppm, J = 30 Hz), but this species
decomposed during workup. DFT calculations of the
palladacyclic complexes supported by L4X and L3X (see
section S1 of the Supporting Information for details) indeed
suggest that the used phosphane ligands significantly (ca. 15
kcal mol−1) destabilize the palladacyclic compound relative to
the phen analogue.
Figure 3. 31P{1H} NMR spectra for a solution containing complex
phen-C7 and 5 equiv of L3X under an argon atmosphere in
nitrobenzene: (a) before heating; (b) after heating to 60 °C for
about 1 min and then cooling to room temperature; (c) after standing
for an additional 2 h at room temperature. # denotes the monoxide of
L3X, and * denotes the monophosphazene of L3X.
mixture of our attempt to synthesize L3X-C7 directly via
nitrobenzene carbonylation. After gentle heating to about 60
°C for about 1 min, a clear solution was obtained which was
cooled to room temperature (∼25 °C) and measured with
31P{1H} NMR. As can be seen in Figure 3b, an unsymmetric
diphosphane complex was formed, characterized by two
doublets centered around 3.8 and 5.8 ppm (J = 30 Hz). In
the 1H NMR spectrum of this solution a resonance is observed
around 9.1 ppm, characteristic for uncoordinated phen,
suggesting successful ligand exchange. Assuming that L3X
replaces phen as ligand while keeping the palladacycle intact,
the two doublets can thus reasonably be assigned to the
unsymmetric complex L3X-C7.
The two doublets disappear in about 2 h at room
temperature, while a new broad resonance around 0.5 ppm
appears. This resonance belongs to the neutral bis-ligand
complex [Pd0(L3X)2], as verified by in situ synthesis of
[Pd0(L3X)2] from [Pd2(dba)3] (dba = dibenzylideneacetone)
and 10 equiv of L3X in nitrobenzene (Figure S5, Supporting
Information). Resonances due to some mono-oxidized ligand
(PO; 25.8 and −24.5 ppm) and some dioxidized ligand
(OPPO′; 27.0 ppm) are present as well (Figure 3); two
additional resonances are observed around 14.3 and −26.2 ppm
(marked with * in Figure 3c). These resonances grow equally
2.3.2. Ligand Exchange Experiment of phen-C7 with L4X
and L3X. As our attempts to isolate P2−palladacycle complexes
were unsuccessful, we subsequently considered monitoring
phenanthroline ligand exchange for L4X or L3X of the phen−
palladacycle complex phen-C7 (Scheme 5).11,22
In the case of L4X, the anticipated asymmetric palladacyclic
species L4X-C7 could not be detected with 31P NMR
spectroscopy, but the final product of the reaction, namely
[Pd0(L4X)2] as well as “free” phen, clearly evolved (see section
S2 of the Supporting Information for details). This suggests
4146
dx.doi.org/10.1021/om200882p | Organometallics 2012, 31, 4142−4156

Organometallics
Article
Figure 4. Plots of the 31P{1H} NMR integrals (relative to the ligand and the appearing and disappearing compounds) for the resonances of
0
●
○
[Pd (L3X)2] around 0.5 ppm ( ) and palladacycle L3X-C7 around 5 ppm ( ) as a function of time: (a) under an argon atmosphere; (b) under an
atmosphere of CO.
fast, indicating that they belong to the same species. These
signals cannot arise from a divalent palladium complex;
[PdII(L3X)(OAc)2] is characterized by a sharp singlet around
16 ppm, and [PdII(L3X)2](OTs)2 exhibits a broad resonance
around 5 ppm, as do related complexes.38,39 This leaves as the
only likely option the formation of a phosphazane (PNPh),
which must be formed during the decomposition of the initially
formed unsymmetric complex L3X-C7. The presence of such a
phosphazene was also observed with electron spray ionization
mass spectroscopy (vide infra).
2.3.3. Ligand-Exchange Experiment of phen-C7 with L3X
under a CO Atmosphere. The ligand-exchange experiment was
repeated under an atmosphere of carbon monoxide. The kinetic
data of this experiment are shown in Figure 4b, while the
Figure 5. 31P{1H} NMR spectra for a solution containing phen-C7
and 1 equiv of L3X in nitrobenzene: (a) before heating; (b) after
heating to 60 °C for about 1 min and cooling to room temperature;
(c) after standing for an additional 180 min at room temperature. #
denotes the monoxide of L3X, and * denotes the monophosphazene
of L3X.
kinetic data for the reaction under argon are shown in Figure
4a. The rate of appearance of the resonance assigned to
[Pd0(L3X)2] (0.5 ppm, 0.0021 and 0.0004 min−1) is about
twice the rate of the disappearing signals at 3.8 and 5.8 ppm
(−0.0011 and −0.0002 min−1), which is consistent with the
assignment of the disappearing resonances to a compound
containing one diphosphane ligand (two P atoms, probably
constant (J = 42 Hz). These doublets are indicative of the
formation of another unsymmetric monochelate palladium
complex, which may well be one of the decarbonylated four-
membered-ring complexes L3X-C2 and L3X-C6 shown in
L3X-C7), and the appearing signal to [Pd0(L3X)2] (four P
atoms). The decomposition of the presumed L3X-C7 is
approximately 5 times slower for the reaction under a carbon
monoxide atmosphere (−0.0002 min−1) in comparison to the
Scheme 6 (see section 2.3.5). The large difference in chemical
reaction in argon (−0.0011 min−1). Under a CO atmosphere
shift (31 ppm) between the two phosphorus nuclei is more
less [Pd0(L3X)2] is formed during the initial ligand exchange
consistent with the complex L3X-C2, as the Pd−C(O)O
reaction (the intercept is 0.046 (CO) versus 0.074 (Ar)).
The apparent stabilization by CO suggests that the initial
moiety will cause severe deshielding (23.9 ppm), while the Pd−
N(Ph) moiety will cause significant shielding (−7.1 ppm).
step in the decomposition of L3X-C7 proceeds via a reversible
decarbonylation/carbonylation process. Apparently, extrusion
of CO2 from L3X-C7 has a higher barrier than decarbonylation,
as can be also concluded from the fact that only traces of
isocyanate are formed (vide infra).
After the solution was allowed to stand at room temperature
for an additional 180 min (Figure 5c), the resonances assigned
to complex L3X-C7 largely disappeared, while the presumed
four-membered-ring complex L3X-C2 is still significantly
present (∼20% based on Pd;40 see also Figure S6 in the
Supporting Information). [Pd0(L3X)2] has become the major
species (33% based on Pd). Among the several unknown
species that have evolved (resonances between 14 and 18 ppm;
25% based on Pd) the presence of the palladacyclic complex
L3X-C6 can neither be proven nor be ruled out. Some ligand
monoxide is formed (PO; 25.8 and −24.5 ppm), and some
uncoordinated ligand (−25.1 ppm) is still present, suggesting
that the ligand exchange reaction is not quantitative after 180
min; this is corroborated by 1H NMR, showing that the
complex phen-C7 is also still present in approximately the same
amount as uncoordinated L3X.40
2.3.4. Ligand Exchange Experiment with 1 equiv of L3X. In
an attempt to prevent or slow down the decomposition of L3X-
C7 to [Pd0(L3X)2], which possibly could be accelerated by the
presence of an excess of L3X, and in order to detect possible
intermediate species in the decomposition of L3X-C7, the
ligand exchange experiment was repeated with only 1 equiv of
L3X. The resulting spectra are shown in Figure 5.
The 31P{1H} NMR spectrum (Figure 5b) of the reaction
mixture after heating to 60 °C and cooling shows the
resonances of the anticipated unsymmetric species around 5.0
ppm, the signal of [Pd0(L3X)2] around 0.5 ppm, and two
additional weak doublet signals around 23.9 and −7.1 ppm
which evolved at the same rate and have an identical coupling
2.3.5. ESI-MS Analysis of Ligand Exchange Experiment
with 1 equiv of L3X. The ligand exchange experiment was
4147
dx.doi.org/10.1021/om200882p | Organometallics 2012, 31, 4142−4156

Organometallics
Article
Scheme 6. Proposed Reaction Sequence of the Ligand Exchange between phen-C7 and the Diphosphane Ligand L3X, Followed
by Decomposition of the Palladacycle L3X-C7
repeated, and after gentle heating the clear solution was
analyzed with electron spray ionization mass spectroscopy
(ESI-MS). In the resulting mass spectrum (Figure 6), the
highest observed mass (m/z 986.2) originates from [Pd-
(L3X)2]+ (exact mass 986.3).
imido complex (L3X)PdNPh (L3X-C3). The presence of
(L3X)Pd(ONPh) and (L3X)PdNPh complexes as fragmen-
tation products of L3X-C7 may be reflecting its thermal
decomposition pathway. The monophosphazene [L3X
NPh]+ (m/z 531.8) is clearly present as well (see also
31P{1H} NMR above), while small quantities of aniline and
nitrosobenzene are observed with GLC-MS.
These MS data are thus in agreement with the NMR data
and suggest that the initially formed five-membered P2−
palladacycle L3X-C7 may (at least partially) decompose to a
L3X-C3 imido species, via L3X-C2, as is shown in Scheme 6.
The imido complex (L3X-C3) may, at least partially,
decompose to form the observed phosphazene while liberating
zerovalent palladium, which is trapped by uncoordinated L3X
as [Pd0(L3X)2].
2.4.6. Attempted Identification and Quantification of
PhN-Containing Products. The integrals of L3XNPh and
[Pd0(L3X)2] by the end of the ligand exchange experiment
(Figure 3c) indicate that merely ∼0.4 μmol of the initial
amount of PhN present in the complex phen-C7 (∼10 μmol)
ends up in the L3X-phosphazene.41 Likely products that may
contain the remaining ∼9.6 μmol of the PhN fragment could be
aniline, phenyl isocyanate, nitrosobenzene, and/or azo(xy)-
benzene. Because the 1H NMR resonances of all possible
reaction products containing the PhN moiety are obscured by
the resonances of the abundantly present nitrobenzene, L3X,
and [Pd0(L3X)2], the solution was analyzed with GLC-MS.
However, using this technique we were only able to positively
identify about 10% of the NPh fragment originally present in
complex phen-C7 as aniline, nitrosobenzene, phenyl isocyanate,
and L3XNPh.
Figure 6. ESI mass spectrum of the reaction mixture taken directly
after the ligand exchange of phen-C7 with 1 equiv of L3X. M = L3X-
C7.
The lower mass products observed are absent in an ESI-MS
of pure [Pd0(L3X)2] in nitrobenzene; these lower mass peaks
must originate from other complexes formed in the exchange
reaction. Although the exact mass (709.1) of the presumed
palladacycle L3X-C7 is not observed, various peaks and their
isotope distributions are consistent with solvent adducts of
L3X-C7 (see Figure S7 in the Supporting Information).
A small peak is present with a mass corresponding to the
decarbonylated palladacycle L3X-C7 (i.e., L3X-C2 or L3X-C6;
[M − CO + H+]+ m/z 683.0), whereas the mass of
decarboxylated L3X-C7 ([M − CO2]+; exact mass 665.1) is
not observed. The small peak at m/z 654.0 corresponds to [M
− 2CO + H]+: i.e., nitrosobenzene bound to [(L3X)Pd0], L3X-
C5. Interestingly, the small feature at m/z 637.1 can be
assigned to [M − CO − CO2]+, which corresponds to the
To obtain more unambiguous information about the fate of
the organic PhN moiety, the NMR experiment was repeated in
non-aromatic solvents such as CH3NO2 and CD2Cl2 (see
section 3.1 in the Supporting Information). Although these
experiments strongly suggest the formation of various PhN-
Figure 7. (a) ESI mass spectrum of a saturated solution of palladacycle phen-C7 (=M) in nitrobenzene and (b) simulation of the most prominent
peaks.
4148
dx.doi.org/10.1021/om200882p | Organometallics 2012, 31, 4142−4156

Organometallics
Article
containing products, neither a positive identification nor
quantification could be achieved, mostly due to the interference
of the aromatic resonances of L3X and Pd0(L3X)2. Experiments
using a ligand with pentafluorophenyl groups were also
inconclusive but again showed the formation of various PhN-
containing products (see section 3.2 in the Supporting
Information).
2.3.6. ESI-MS Analysis of phen-C7. The phen-C7 complex
was analyzed with ESI-MS in order to see whether it
decomposes in a manner similar to that for L3X-C7. Due to
the poor solubility of phen-C7 in various common solvents, a
saturated solution of this complex in nitrobenzene was
introduced directly into the ESI source (instead of using an
injector with an eluent). However, when using the same
ionization conditions that were successful for measuring the
fragmentation of L3X-C7 (i.e., 3 kV and 300 °C), no clear mass
spectrum could be obtained. Only when the ionization
temperature was elevated to 450 °C could a proper mass
spectrum be recorded. The mass spectrum of this experiment is
shown in Figure 7a; a simulation of the most prominent peaks
is given in Figure 7b.
However, they did not connect these observations with a
suggestion that phen-C7 preferentially decomposes via
decarbonylation rather than decarboxylation and that an
imido species (phen-C3) might be formed.
2.4. DFT Study of Nitrobenzene Deoxygenation with
CO. The presented data suggest that palladacyclic species such
as phen-C7 and L3X-C7 shown in Scheme 5 may decompose
to the corresponding palladium−imido compounds C3. To
relate these data to the actual catalytic process, the possible
deoxygenation pathways of nitrobenzene to form the
palladayclic and/or palladium−imido complex were studied
using DFT calculations. Because of computational limitations,
attempts to estimate the reaction barrier from one complex to
another were not undertaken. Instead, our approach was to
estimate the Gibbs free energies for the relevant complexes
(Scheme 7), via the methodology explained and validated in
Scheme 7. Possible Intermediates in Nitrobenzene
Deoxygenation with CO as Reductant and Their Relative
Energies As Computed with DFT for L2 = phen (at 130 °C)
a
and L2 = L4X, L3X (Both at 110 °C)
The [M]+ peak of palladacycle phen-C7 (exact mass 449.0) is
not observed. The highest significant mass around m/z 571.8 is
consistent with a nitrobenzene solvent adduct of the phen−
palladacycle phen-C7 (the exact mass is 572.0; the solid state
structure of phen-C7 also contains a nitrobenzene molecule in
its lattice).22 The mass and isotope distributions of the two
features found around m/z 466.7 and 484.9 are in perfect
agreement with [M + H2O]+ and [M + 2 H2O]+, respectively
(see Figure S15 in the Supporting Information). The H2O in
these adducts probably derives from trace amounts of water in
the solvent nitrobenzene.
The monodecarbonylated fragment ([M − CO]+, exact mass
421.0) is not observed, but the mass of the didecarbonylated
fragment [M − 2CO]+ (i.e., the nitrosobenzene complex phen-
C5) is present around m/z 392.7 (exact mass 393.0). The most
abundant mass observed is that of [(phen)Pd]+ around m/z
285.9 (exact mass 286.0); a phenyl adduct of this species seems
also to be present (m/z 362.9; [(phen)Pd + Ph − H]+, exact
mass 363.0). The feature around m/z 407.9 cannot be ascribed
to the anticipated palladium-bound phenyl isocyanate (i.e., [M
− CO2]+), as the exact mass of this species (405.0) differs by 3
units. As the experiment was performed in a nitrobenzene
solvent matrix, the mass around 407.9 may be assigned to a
nitrophenyl adduct of [(phen)Pd]+: i.e., [(phen)Pd + PhNO2
− H]+ (exact mass 408.0). Interestingly, the mass of the very
small peak around m/z 377.9 is in agreement with a
(phen)PdIINPh complex (i.e., [M − CO2 − CO]+, exact
mass 378.0).
The fragmentation pattern surprisingly suggests that phen-
C7 may decompose predominantly via an initial decarbon-
ylation (−CO) instead of a direct decarboxylation (−CO2) to
palladium-bound phenyl isocyanate. After the first decarbon-
ylation, a second decarbonylation may follow, yielding a
palladium-bound nitrosobenzene compound, or CO2 can be
extruded to form the imido intermediate (phen)PdIINPh.
The latter compound may decompose further to [(phen)Pd]+
that will give rise to the observed adducts thereof. Osborn et al.
analyzed the fragmentation of phen-C7 using the fast atom
bombardment ionization technique and also observed the
monodecarbonylated fragment at m/z 422.11 Remarkably, a
fragment at m/z 378 was also reported, but this was assigned to
phen-C7 with loss of CO + NCO instead of loss of CO + CO2.
a
Values are given in kcal mol−1 relative to complex C0.
sections S4 and S5 of the Supporting Information. The
enthalpy differences were estimated using DFT calculations,
and the changes in entropy were estimated at the relevant
reaction temperature from literature data (standard molar
entropy (S°) in J mol−1 K−1). Hence, the data shown in Scheme
7 are the estimated Gibbs free energies at 110 °C for complexes
containing L4X and L3X and at 130 °C for complexes
containing phen as the supporting ligand.
The starting compound for the nitrobenzene reduction
process is the zerovalent complex L2Pd(CO)2 (C0), which is
therefore defined as 0.0 kcal mol−1. The oxidative coupling of
CO with nitrobenzene at C0 can proceed via either C1 or C4.
No preference can be given, as C1 and C4 are similar in energy
for phen, L4X, and L3X. The irreversible CO2 extrusions C1 →
C2 and C4 → C5 are both about equally exothermic (ca. −50
kcal.mol−1), thereby completing the first deoxygenation step of
nitrobenzene.
Complexes C2 and C5 differ in energy only ∼8 kcal mol−1;
they may thus very well interconvert into each other via a
reversible carbonylation/decarbonylation process, possibly at
equilibrium under carbonylation reaction conditions. Likewise,
the reactions C2 ⇆ C7, C5 ⇆ C6, and C6 ⇆ C7 are all likely
to be reversible carbonylation/decarbonylation reactions,
4149
dx.doi.org/10.1021/om200882p | Organometallics 2012, 31, 4142−4156

Organometallics
Article
Scheme 8. Schematic Overview of the Possible Intermediates for the Formation of the Palladacyclic Intermediate C7 (Top) and
the Pd−Imido Intermediate C3 (Center and Bottom) in the Deoxygenation of Nitrobenzene
meaning that C2, C5, C6, and C7 are probably all
interconvertible and possibly at equilibrium. Similar reversible
interconversions are ubiquitous in carbonylation reactions. For
instance, carbonylation−decarbonylation equillibria are known
to exist during the carbonylation of olefins; the carbonylation
step of Pd−alkyl and Pd−O-alkyl intermediates generally is
almost thermoneutral or only slightly exothermic and is
generally fully reversible with a low activation barrier.42 Note
that the palladacycle C7 is the most stable complex of these
(presumably) CO equilibrated species for all three ligands. The
oxidative coupling of nitrosobenzene with CO (i.e., C5 ⇆ C2
and C5 ⇆ C6) is exothermic by ca. −8 kcal mol−1, whereas the
oxidative coupling of nitrobenzene with CO is endothermic
(i.e., C0 ⇆ C1 and C0 ⇆ C4) by ca. +20 kcal mol−1; this is in
line with our finding (vide supra) and literature data7,25 that
nitrosobenzene is more easily reduced than nitrobenzene.
The second deoxygenation of nitrobenzene can only occur
by the irreversible CO2 extrusion from C2 to the imido
complex C3, which is thermodynamically also the most stable
complex (ca. −62 kcal mol−1). The palladacycles that are
formally PdII species and are sterically crowded in the equatorial
positions of Pd (i.e., C1, C2, C4, and C7) are all destabilized by
ca. 10 kcal mol−1 by the bulkier L3X and L4X ligands relative to
phen. The formally zerovalent palladium−nitrosobenzene
complex C5 is more stable when L4X or L3X is the supporting
ligand, due to the increased acidity of palladium, as induced by
the increased π-back-bonding from Pd to P.
followed by an irreversible decarboxylation step (i.e., C1 → C2
or C4 → C5). As argued above, it is likely that carbonylation−
decarbonylation reversible steps connect species C2, C5, C6,
and C7, which all likely act merely as temporary reservoirs of
the PhN fragment; none of these is expected to be a product-
releasing species. Palladacycle C7 should also not be considered
as such, as the ligand exchange and ESI-MS experiments with
L3X suggest that the barrier for decarboxylation (CO2 loss) is
higher than the barrier for decarbonylation (CO loss) of C7.
The methanol-assisted deoxygenation routes are thought to
start with oxidative addition of methanol on L2Pd0(CO)2 to C8
(possibly assisted by coordinated CO)43,44 and most likely
proceed via palladium−hydride chemistry (C8 ⇆ C9 ⇆ C10
→ C11). Eventually the same palladium−imido intermediate
C3 will be formed, as is discussed in more detail elsewhere.17,18
The nitrosobenzene generated in compound C11 may re-enter
the catalytic cycle by binding to C0, forming species C5. The
diphosphane−Pd−imido complex C3 is thought to be the sole
PhN-containing, product-releasing species (i.e., ultimately
producing MPC, DPU, PhNH2, and Azo(xy)).17
3.1.2. Deoxygenation with phen-Based Catalysts. When
the more basic phen is used as the supporting ligand, the CO-
only deoxygenation pathway contributes more than 95%, which
means that the oxidation of methanol is almost completely
suppressed (see section S6 in the Supporting Information for
details). An increased basicity (electron density) of the Pd
center is generally thought to be favorable for oxidative
processes at Pd and thus would be expected to promote both
oxidative coupling of CO with nitrobenzene and oxidative
addition of methanol. Apparently, the oxidative coupling of CO
with nitrobenzene at Pd is significantly more sensitive to Pd
basicity than oxidative addition of methanol, presumably
because of increased electron back-donation of the filled dπ
orbitals of Pd into the antibonding π* orbitals of CO in
L2Pd(CO)2. While increased back-donation from Pd to CO
apparently makes CO more susceptible for attack by the nitro
group, it makes the CO less susceptible for attack by a
methoxide anion possibly necessary for facilitating the oxidative
addition of methanol.
These data thus suggest that it is likely that complexes C2,
C5, C6, and C7 exist in mutual carbonylation/decarbonylation
equillibria, as is indeed also experimentally suggested for L3X-
C7 ⇆ L3X-C2/L3X-C6 (section 2.3.3).
3. DISCUSSION
3.1. Molecular Mechanism of Nitrobenzene Deoxyge-
nation. 3.1.1. Deoxygenation with Diphosphane-Based
Catalysts. We have shown that the fraction of nitrobenzene
that is deoxygenated with only CO or with methanol depends
on the characteristics of the catalyst employed, as determined
by the properties of the supporting ligand; for the catalytic
system based on L4X the CO-only deoxygenation route
contributes merely 9%.17 Our current understanding of the
molecular basis underlying the deoxygenation of nitrobenzene
in methanol for catalysts with phosphanes as the supporting
ligand (L2) is summarized in Scheme 8.17 The CO-only
deoxygenation route starts with an oxidative coupling of
nitrobenzene and CO at L2Pd0(CO)2 (C0) to form C1 or C4,
On the basis of on the ESI-MS fragmentation pattern of
phen-C7, the ligand exchange experiments of phen-C7 with
L4X and L3X, the dependence of the Azo(xy) selectivity on the
nitrobenzene concentration, and our DFT results, we propose
that the molecular mechanism of the CO-only deoxygenation
shown in Scheme 8 is also valid for the Pd/phen catalytic
system. We furthermore propose that phen-C7commonly
4150
dx.doi.org/10.1021/om200882p | Organometallics 2012, 31, 4142−4156

Organometallics
Article
Scheme 9. Schematic Overview of the Possible Intermediates of L2PdIINPh (C3) Reduction to L2Pd0 and the Various Aryl-
Containing Nitrobenzene Reduction Products
thought to be the direct carbonylation product releasing
speciesis not the (main) product-releasing species under
normal (acid-free) catalytic carbonylation conditions. Instead,
phen-C7 merely is part of a network of reversible carbon-
ylation−decarbonylation steps, connecting species (C2, C5,
and C6) that together act as temporary PhN reservoir. This
explains why phen-C7 can be synthesized and isolated under
mild carbonylation conditions (60 °C, ethanol);11 phen-C7 is
the thermodynamically most stable of the CO-connected
species and is therefore thought to be an important resting
state of the catalyst. At elevated temperatures, phen-C7 may be
decarboxylated to phenyl isocyanate (and thus MPC and
DPU), but judging from the product distribution of the
reaction under acid-free conditions, this is not the dominant
reaction pathway; merely 20% MPC but 70% Azoxy is
produced. This must mean that the reaction barrier for phen-
C7 decarboxylation (to PhNCO) is higher than that of its
decarbonylation (to phen-C2 and/or phen-C6). Irreversible
decarboxylation of phen-C2 to phen-C3 then removes the
catalyst from the CO-connected PhN reservoir.
The different temperatures required when phen (130 °C) or
diphosphanes such as L4X (110 °C) are used as the supporting
ligands in catalytic nitrobenzene carbonylation may be ascribed
to the destabilizing effect that the bulkier phosphane ligands
have on palladacycles such as C7 (as indicated by DFT). Both
the decarbonylation barriers (e.g., C7 ⇆ C2) and the barrier
for decarboxylation of C2 → C3 will be lowered, thus allowing
the PhN fragment to escape the temporary reservoir more
easily and thus also allowing the application of milder reaction
temperatures.
kinetics in nitrobenzene. It has been proposed that Azoxy stems
from a reaction between free nitrosobenzene and free nitrene
(PhN)9,10 or from a condensation reaction of nitrobenzene
with aniline.15 We have shown, however, that adding extra
nitrosobenzene during a catalytic run does not result in the
formation of more Azoxy. Nitrosobenzene is more easily
reduced than nitrobenzene, which is suggested by our DFT
studies and finds support in the literature as well.7,10,25,45,46 The
condensation of aniline and nitrosobenzene to azobenzene has
been reported to occur under reaction conditions much more
forcing than those employed by us (hot glacial acetic acid;47
very basic conditions48). Furthermore, it is known that Azoxy
cannot be formed by a condensation reaction of aniline and
nitrobenzene during catalysis, as a carbonylation experiment
with deuterated nitrobenzene and undeuterated aniline gave
only fully deuterated Azoxy when working with the Pd/phen
catalytic system.32
The observed similarity in selectivity for Azoxy when using
both phen and L4X as the supporting ligands can be
understood by considering the electronic and steric effects
that these ligands will have on the proposed imido
intermediate. For phen, the increased basicity is thought to
render the imido complex more polar, thus increasing its
association with nitrobenzene and hence the portion of PhN
that is involved in the disproportionation reaction. L4X, on the
other hand, is far less basic, but here the steric bulk in the plane
of coordination is thought to expedite the reaction with
associated nitrobenzene. Indeed, when the less sterically
demanding ligand L3X is used, hardly any Azoxy is formed,
as detailed elsewhere.17
3.2.2. Formation of Carbonylation Products (MPC, DPU,
DMC, DMO). As an alternative to the disproportionation
reaction, the imido intermediate C3 can be protonated by
methanol to produce PhNH2 and/or MPC (Scheme 9, right).
Direct carbonylation of the highly polarized imido intermediate
with CO to form (palladium-bound) isocyanate is not likely to
occur in a protic environment and also conflicts with our
experimental data using a labeled substrate, 3-methylnitroben-
zene in aniline (see below). Protonation of C3 by methanol
gives C12, which may undergo carbonylation to C13a via an
associative displacement of CH3O− by the smaller and charge-
neutral CO. This is then followed by nucleophilic attack of
nearby CH3O− on coordinated CO and reductive elimination
to produce MPC. Likewise, C13b may be generated, followed
by MPC genesis. DPU may be produced via a similar process in
which aniline replaces methanol as the nucleophilic reagent. As
an alternative to MPC or DPU formation, C12 may be
3.2. Molecular Mechanism of Reactions of the PdII
NPh Intermediate. 3.2.1. Azo- and Azoxybenzene For-
mation. Once the imido intermediate C3 has been formed, it
can undergo a disproportionation reaction with nitrobenzene
and CO to produce azoxybenzene and CO2 as the stable end
products (Scheme 9, left).17
Because formally one N−O bond in nitrobenzene is fully
polarized, nitrobenzene is thought to associate more strongly to
the (also polarized and sterically quite accessible) PdN bond
in C3 (more strongly than the relatively weakly polarized and
charge-neutral methanol). As a result, the disproportionation
reaction dominates, resulting in the high selectivity toward
Azoxy (70%) for both the phen and L4X catalyst systems. The
observed dependence of the Azoxy selectivity on the nitro-
benzene concentration confirms that the formation of Azo(xy)
coupling products from nitrobenzene involves a step competing
with carbonylation and hydrogenation, but with higher order
4151
dx.doi.org/10.1021/om200882p | Organometallics 2012, 31, 4142−4156

Organometallics
Article
protonated to C14 with liberation of aniline. C14 will then be
carbonylated once or twice via associative displacement of
methoxide by CO and subsequent nucleophilic attack on
coordinated CO by methoxide to give C15 and C16, from
which reductive elimination will produce DMC and DMO,
respectively. The strong similarity of the chemistry for
formation of MPC with that of DMC/DMO is an attractive
point for the proposal of a L2PdNPh intermediate in
nitrobenzene carbonylation chemistry. All carbonylation
intermediates, C12−C16, are considered as being involved in
reversible reactions and are possibly at mutual equilibrium.
MPC-, DPU-, DMC-, and DMO-releasing reactions likely
involve irreversible and possibly rate-determining reductive
elimination reactions, thus reproducing a zerovalent palladium
complex. As detailed elsewhere with P2Pd catalysts, an increase
of CO pressure generally leads to an increase in absolute
productivity of all carbonylation products (DMC, DMO, MPC,
DPU).17 This suggests indeed that all intermediate carbon-
ylation complexes in Scheme 9 under the reaction conditions
are interconvertible via CO (de)insertions. The relative
productivity of MPC/DPU vs DMC/DMO depends on the
ligand properties and thus on the exact position of the
equilibria with the particular ligand. With phen as supporting
ligand, only small quantities of DMC/DMO are formed,
implying only a small quantity of H atom production and
therefore of aniline; this could be either a consequence of the
equilibrium position of phen-C15 and phen-C16 in the
reaction mixture or a consequence of a low rate of reductive
elimination of DMC/DMO from these respective complexes
(which are smaller than C13).
As it appears from our data, carbonylation of C12 is favored
when phen is the supporting ligand (20% MPC), whereas
protonation of C12 is favored in case of L4X (20% PhNH2)
with accompanying formation of DMC/DMC (see Table S1 in
the Supporting Information). This effect can be understood by
the difference in basicity of both ligands; the more basic phen
will destabilize the Pd−OCH3 bond in C12, thus facilitating the
associative displacement by CO leading to MPC (via C13).
Our mechanistic proposal for the formation of MPC is
somewhat related to that recently proposed by Ragaini and co-
workers.15,16 Their hypothesis apparently involves first full
reduction of nitrobenzene to aniline while forming a five-
coordinated (phen)Pd(dicarbomethoxy)carbonyl intermediate
species ((phen)Pd(C(O)OCH3)2CO), via an unspecified
mechanism.49 It is noteworthy that the (phen)Pd(C(O)-
OCH3)2 complex is equal to C16 in our mechanism, but as a
direct precursor to DMO and not to MPC. The CO moiety in
the five-coordinate (phen)Pd(C(O)OCH3)2(CO) species is
proposed to be involved in nucleophilic attack by aniline to
ultimately yield phenyl isocyanate via proton transfer from the
aniline moiety via a (hypothetical) five-coordinate palladium−
carbenoid complex. The carbamate then evolves upon reaction
of the isocyanate with solvent methanol.15,16 This proposal
leans heavily on the observed cocatalytic role of added aniline
(as initially reported by the van Leeuwen group)10 and the
observation of the formation of significant amounts of
unlabeled DPU in an experiment with d5-deuterated nitro-
benzene in aniline.16
proposed carbonylation mechanism, as depicted in Scheme 9.
The increased nucleophilicity of aniline relative to methanol
favors urea formation over carbamate formation, by competing
directly with the C3 → C12 → C13 (Scheme 9) reaction,
which thus leads to a higher overall carbonylation rate, as is
indeed observed in aniline. At temperatures above 120 °C DPU
transesterifies rapidly with methanol to give MPC and aniline.24
As the experiments of Ragaini et al. as well as those of van
Leeuwen et al. were carried out above 130 °C (respectively 170
and 135 °C), transesterification of initially formed DPU to
MPC will surely occur. The transesterification of DPU to MPC
will become rate determining at high aniline concentration,
such that the rate of carbonylation (=rate of nitrobenzene
conversion) becomes at a constant level very similar to that in
pure aniline; this also explains the flattening profile of the
aniline-promoting role at high aniline concentrations as
reported by van Leeuwen et al.10 and Ragaini et al.16 We
have observed that when P2Pd systems are applied as catalysts
in reactions with aniline as nucleophile only, carbonylation rates
(giving DPU) even at temperatures below 70 °C can be 100-
fold higher relative to that of carbamate formation.36
Of crucial importance, also in this context, are our
experiments conducted with (L3X)Pd(OAc)2 in aniline as
(excess) solvent using 3-methylnitrobenzene as the substrate
and working below the transamidation temperature (i.e., at 60
°C). Surprisingly, the asymmetric N-(3-methylphenyl)-N′-
phenylurea (MPPU) and symmetric N,N′-bis(3-methylphenyl)-
urea (DMPU) are not observed, whereas a stoichiometric
amount of 3-methylaniline is formed. As transamidation of
MPPU and/or DMPU with solvent aniline cannot occur at the
temperature used, this must mean that the N-aryl groups
derived from 3-methylnitrobenzene at some stage have been
exchanged by those derived from aniline via a mechanism other
than a transamidation. The 3-methylnitrobenzene must interact
with the catalyst in order to be deoxygenated. As indicated
above, the CO-only deoxygenation route will proceed via
species L3X-C7* (Scheme 10).
As proposed above, L3X-C7* cannot be the direct isocyanate
product releasing intermediate, as the unsymmetric MPPU
should then be formed predominantly (which is the very reason
that Ragaini et al. propose a direct aniline carbonylation to
Scheme 10. Proposed Mechanism for the Formation of DPU
in the Pd/Diphosphane-Catalyzed Reduction of 3-
a
Methylnitrobenzene (Ph*NO2) in Aniline
Our mechanistic proposal, however, is distinctly different in
that the precursor for aniline formation (the palladium−imido
intermediate C3) itself is hydromethoxycarbonylated to
ultimately form MPC (see below). The observed cocatalytic
role of aniline can simply be rationalized on the basis of our
a
P2 = L3X.
4152
dx.doi.org/10.1021/om200882p | Organometallics 2012, 31, 4142−4156

Organometallics
Article
Scheme 11. Proposed Mechanism for MPC Genesis under Acidic Conditions for the Pd/phen, Starting with Protonation and
a
Decarboxylation of phen-C7
a
N2 = phen.
isocyanate).16 Decomposition (through initial decarbonylation)
of L3X-C7* will give L3X-C3*. L3X-C3* will, in analogy to our
hypothesis in a methanolic medium (Scheme 9), not be
carbonylated to isocyanate (as no MPPU was observed) but
rapidly react with solvent aniline to produce the dianilido−
palladium complex L3X-C17*. At this point there will be a
competition between CO insertion (and formation of MPPU)
and anion exchange with the solvent aniline to liberate the
observed 3-methylaniline (L3X-C17* + (excess) PhNH2 ⇆
L3X-C17 + Ph*NH2). Clearly this anion exchange reaction
must be faster than the ultimate formation of the urea, as
otherwise the unsymmetrical MPPU would be formed. From
L3X-C17 DPU is formed by a CO insertion (L3X-C17 + CO
⇆ L3X-C18) followed by a reductive elimination (L3X-C18 →
L3X-C0) to return palladium in the catalytic cycle, which
explains why only DPU (and some DPO) was observed. The
appeal of this proposed reaction sequence is that it is analogous
to the evolution of the methanol carbonylation products DMC
(and DMO) and the nitrobenzene carbonylation product MPC
when methanol is the nucleophilic solvent (i.e., C3 → C12 ⇆
C14 ⇆ C15 → C0 and C3 → C12 ⇆ C13 → C0 in Scheme
9).
presence of an added acid (catalyst) it smoothly decomposes to
form ethyl phenyl carbamate with high yield, when heated in
ethanol at just 90 °C.11 It is worth mentioning that
decarboxylative aminolysis, with aniline replacing methanol as
shown in Scheme 11, also rationalizes the observed formation
(at low conversion) of predominantly symmetrical undeuterated
ureas when equimolar amounts of deuterated nitrobenzene and
undeuterated aniline are reacted under acidic conditions with
the Pd/phen catalytic system,16 similar to our observation for
the Pd/L3X system under acid-free conditions in aniline with 3-
methylnitrobenzene as the substrate (vide supra). Whereas the
direct pyrolysis of phen-C7 to isocyanate would result in the
unsymmetrical urea, its decarboxylative aminolysis delivers the
C13b anilide analogue that can be involved in rapid C6D5NH−
anion exchange with undeuterated aniline, while at the same
time being involved in fast decarbonylation−carbonylation
equilibration reactions. This is in complete analogy to the
equillibria C12 ⇆ C13, C12 ⇆ C14 ⇆ C15 (Scheme 9, in
methanol, acid free) and L3X-C17* ⇆ L3X-C17 ⇆ L3X-C18
(Scheme 10, in aniline, acid free), producing respectively MPC,
DMC, and DPU.
Our proposed mechanism shown in Schemes 9 and 10 is also
consistent with the formation of fully deuterated azo(xy)-
benzene (C6D5NN(O)C6D5) side products of reductive
carbonylation of deuterated nitrobenzene in aniline.32 Irrever-
sible Azo(xy) benzene formation (C3* + Ph*NO2 + CO →
P2Pd0 + Ph*NN(O)Ph* + CO2) and irreversible addition of
C3* + PhNH2 → C17* directly compete for C3*, prior to the
anilido group exchange and equilibrating carbonylation steps,
ultimately delivering DPU (Scheme10).
Our mechanistic understanding of nitrobenzene reductive
carbonylation in a CO/CH3OH environment with diphos-
phane-stabilized palladium complexes thus translates directly to
the mechanism of this reaction for the phen-supported
catalysts, as well as to the mechanism of this reaction in
aniline as the nucleophilic reactant/solvent.
3.3. Effect of Added Acid. With our current understanding
of the mechanism detailed above, it can be rationalized why the
addition of (sub)stoichiometric amounts (on Pd) of acid has
such a dramatically different effect on the selectivity of the
PdII(phen) (producing more MPC) and the PdII(L4X)
(producing more PhNH2) catalytic systems. Under acid-free
conditions, the palladacycle C7 (with either phen or
diphosphane ligands such as L4X) is not a (main) product-
releasing species due to the high barrier for decarboxylation
relative to that of decarbonylation. When an acid is added,
however, the decarboxylation barrier is significantly lowered, we
believe by protonation at the amide carbonyl O of the
palladacycle, as suggested by Osborn (top of Scheme 3);11 in
the case of phen-C7 this apparently occurs to the point where
CO2 extrusion is favored over decarbonylation. As a result, in
acidic media phen-C7 becomes the main product-releasing
species, selectively producing MPC. As is shown in Scheme 11,
we believe the mechanism of this decarboxylative methanolysis
proceeds via intermediate C13b (see also Scheme 9).
For the PdII(L4X) system the barrier for L4X-C7
decarboxylation is most probably also lowered by the addition
of acid. However, the barrier for L4X-C7 decarbonylation
apparently is still relatively lower, which is most likely due to
the fact that L4X-C7 is relatively less stable than phen-C7, as
suggested by DFT. As a consequence, L4X-C3 will still be
formed, remaining the main product-releasing species. The
lower azoxybenzene production upon the addition of acid in
favor of the carbonylation and hydrogenation reactions can be
rationalized by an acid-assisted protonation of C3 to C12,
thereby competing with the disproportionation reaction. Why
in the Pd/L4X/H+ system the hydrogenation is promoted
somewhat more than the carbonylation reaction can easily be
understood; protonation of the [PhNH−] anion in C12 to form
aniline will be facilitated by the acid, at the expense of the
associative replacement of the [CH3O−] ligand in C12 by CO.
It is important to note that, in line with this mechanism, upon
addition of an acid to the PdII(phen) system, the amount of
DMC + DMO produced decreases (from 0.5 to 0.2 mmol),
whereas for the catalytic system based on PdII(L4X) the
amount of DMC + DMO increases (from 8.2 to ∼13 mmol, see
Table S1 in the Supporting Information for details).
The imido-intermediate phen-C3 will thus, under acidic
conditions, only scarcely be formed, explaining the strongly
decreased azoxybenzene selectivity (from 71 to ∼15%). This is
in agreement with the findings of Osborn and co-workers, who
reported that phen-C7 is thermally very stable but in the
4153
dx.doi.org/10.1021/om200882p | Organometallics 2012, 31, 4142−4156

Organometallics
Article
1H and 13C NMR spectra were recorded on a Bruker DPX300 (300
MHz) or a Bruker DMX400 (400 MHz) instrument. A Finnigan Aqua
mass spectrometer (MS) with electrospray ionization (ESI) was used
to record mass spectra. Samples were directly introduced into the ESI
source, which was heated at 300 or 450 °C. The voltage of the capillary
and the voltage for the aquamax were set at 3 kV and 50 V,
respectively. High-pressure experiments were conducted in stainless
steel autoclaves (100 mL) equipped with two inlet/outlet valves, a
burst disk, a pressure sensor, and a thermocouple. The autoclaves were
heated by an HEL polyBLOCK electrical heating system. Temper-
atures and pressures were measured with probes connected to a
computer interface, making it possible to record these parameters
throughout the course of the reaction.
5.2. Catalytic/High-Pressure Reactions. In a typical catalytic
experiment, 0.05 mmol of Pd(OAc)2 and 0.075 mmol of the ligand
(and, if relevant, another additive) were weighed and transferred into
an autoclave, together with a magnetic stirring rod. The autoclave was
tightly closed and subsequently filled with argon using a Schlenk
system that was connected to the one of the valves of the autoclave.
Through the other valve was added 2.50 mL (24.4 mmol) of dried and
degassed nitrobenzene, under a continuous flow of argon. In a similar
fashion, 25.0 mL of dried and degassed methanol was then added. This
reaction mixture was stirred at 500 rpm for about 15 min to ensure
that complex formation was complete.38 The autoclave was then
inserted into the heating block and put under 50 bar of carbon
monoxide gas. The reaction mixture was heated to 110 or 130 °C
(within 30 min) with stirring at 500 rpm. After it stood for 4 h at a
certain temperature, the autoclave was cooled to room temperature in
about 1 h. The autoclave was then slowly vented to atmospheric
pressure, and the reaction mixture was analyzed as described
elsewhere.17 To ensure reproducibility, some standard catalytic
reactions were performed in quadruplicate, and the relative standard
deviation was always less than 5% for each analyte.
4. CONCLUSIONS
The palladium−imido species L2PdNPh (C3) and the
palladacycle L2PdC(O)N(Ph)OC(O) (C7) were considered
as possible carbonylation product releasing species for both
phen- and diphosphane-based catalytic systems. The results of
catalytic experiments, supported by ligand exchange experi-
ments of phen-C7 with L4X and L3X, spectroscopic (ESI-MS
and NMR) evidence, and a DFT study, suggest that the
palladacyclic compound C7 in the absence of acid is not a
major product-releasing intermediate. Thus, in the absence of
acid, our proposed mechanism for Pd/diphosphane catalysts
(Schemes 2 and 8)17,18 applies directly to the Pd/phen catalytic
system. Indeed, C7 is thought to be part of a family of CO-
equilibrated palladacycles that together act as temporary PhN
reservoirs (i.e., C7 ⇆ C2 ⇆ C5 ⇆ C6 ⇆ C7). It is our
understanding that this PhN fragment is only released via the
imido complex L2PdNPh (C3), which is thus thought to be
the central product-releasing species, producing mainly the
coupling product Azoxy with phen- and L4X-based systems.
On the other hand, in the presence of acid the palladacyclic
complex phen-C7 becomes the major product-releasing species,
resulting in the selective formation of the nitrobenzene
carbonylation product MPC. Our mechanistic conclusion for
the carbonylation of nitrobenzene in methanol, catalyzed by the
palladium−phenanthroline system, is illustrated in Scheme 12.
Scheme 12. Summarizing Scheme for the Connecting
Catalytic Processes in the Palladium-Catalyzed Reductive
Carbonylation Reactions of Nitrobenzene when 1,10-
Phenanthroline (N2) is the Supporting Ligand
5.3. Adding Nitrosobenzene during a Catalytic Run. The
experiment was first started as a normal high-pressure experiment. A
stainless steel hollow pipe (10 mL), sealed with two valves on each
side, was then put under an argon atmosphere, and nitrosobenzene
(2.5 mmol in 5 mL methanol) was transferred into the hollow pipe
using standard Schlenk techniques. The bottom valve of the hollow
pipe was then mounted onto one of the valves of the autoclave,
connected to the CO supply, and pressurized to about 5 bar above the
pressure inside the autoclave. After about 1 h reaction time,
nitrosobenzene was added to the reaction mixture by opening the
two valves between the autoclave and the hollow pipe. The reaction
was then allowed to run for the remaining reaction time and treated as
any other catalytic experiment.
5.4. NMR Experiments. A 4.50 mg amount (10 μmol) of phen-
C711,22 was weighed into an NMR tube and put under argon. In
another tube, the appropriate amount of phosphane ligand was
dissolved in 1 mL of nitrobenzene(-d5) under an argon atmosphere.
Of this solution, 0.5 mL was added to the phen-C7 complex using a 1
mL syringe, which was dry and flushed with argon. The yellow
suspension thus obtained (20 mM phen-C7) was thoroughly mixed
using a vortex mixer and measured. After the first measurements, the
reaction mixture was carefully heated to about 60 °C, resulting in a
clear yellow-orange solution. This solution was monitored with 1H and
31P{1H} NMR spectroscopy, over a period of 14 h. For the proton
measurements, the number of free inductive decays (FIDs) was 16,
and for the phosphorus NMR spectra, the number of FIDs was 40.
The same procedure was applied for the experiment under a CO
atmosphere, but nitrobenzene was first saturated with CO gas, the
NMR tube was put under a CO atmosphere, and the 1 mL syringe was
flushed with CO.
5.5. DFT Studies. Calculations were done with the SPARTAN ’04
package (Wavefunction, Inc.; www.wavefun.com), using density
functional theory (DFT)53,54 with the Becke and Perdew (BP)
functional.55,56 Geometry optimizations were carried out using Pople’s
6-31G*(d,p) basis set for H, C, O, and P atoms57 and the LANL2DZ
effective core potential for palladium.58−60 All of the geometrical
parameters were fully optimized. No constraints to bonds, angles, or
Osborn et al. already proposed a palladacycle such as phen-C7
as the intermediate for direct isocyanate formation.11 However,
here we propose a crucial modification involving acid-catalyzed
decarboxylative alcoholysis of phen-C7 to produce intermedi-
ates capable of anilide anion exchange to fully rationalize our
catalytic and organometallic observations. Our mechanistic
proposal is also in full agreement with observations reported by
the groups of Ragaini et al. and van Leeuwen et al.
5. EXPERIMENTAL SECTION
5.1. General Remarks. All solids were purchased from Acros
Organics and used as received. Methanol, nitrobenzene, and aniline
were all of analytical reagent purity and were distilled under an argon
atmosphere over the appropriate drying agent.50 After the distillation,
these liquids were saturated with argon. We ensured that no water was
present using an analytical reaction with trimethyl orthoformate
according to a literature procedure.51 Carbon monoxide (>99% pure)
was purchased from Lindegas Benelux BV and used as received.52 The
compound phen-C7 was synthesized according to a literature
procedure;11,22 the analytical data are given in section S7 of the
Supporting Information.
4154
dx.doi.org/10.1021/om200882p | Organometallics 2012, 31, 4142−4156

Organometallics
Article
dihedral angles were applied in the calculations, and all atoms were
free to be optimized. Cartesian coordinates and raw energies (in
hartrees) are collected in an .xls file in the Supporting Information,
along with .mol2 files of all optimized structures C0−C7.
(19) Akazome, M.; Kondo, T.; Watanabe, Y. J. Org. Chem. 1994, 59,
3375.
(20) Paul, F.; Fischer, J.; Ochsenbein, P.; Osborn, J. A. Organo-
metallics 1998, 17, 2199.
(21) Paul, F.; Fischer, J.; Ochsenbein, P.; Osborn, J. A. C. R. Chim.
2002, 5, 267.
ASSOCIATED CONTENT
■
(22) Santi, A. S. o; Milani, B.; Mestroni, G.; Zangrando, E.;
Randaccio, L. J. Organomet. Chem. 1997, 546, 89.
S
* Supporting Information
An Excel file giving Cartesian coordinates and raw energies,
.mol2 files giving all optimized structures C0-C7, and tables,
figures, and text giving additional details of the experiments and
calculations. This material is available free of charge via the
Internet at http://pubs.acs.org.
(23) Note that DPU can be formed not only by the reaction of
nitrobenzene, CO, and (in situ generated) aniline but also by the
transesterification of MPC with aniline. In both cases, aniline is formed
first and DPU is thus derived from aniline.
(24) When 6 mmol DPU was heated for 4 h in methanol under 50
bar of CO, significant amounts of MPC were found: 3.2 mmol (110
°C), 1.5 mmol (100 °C), 1.0 mmol (90 °C), 0.4 mmol (80 °C).
(25) Lund, H. In Organic Electrochemistry. An Introduction and a
Guide, 3rd ed.; Lund, H., Baizer, M. M., Eds.; Marcel Dekker: New
York, 1991; p 411.
AUTHOR INFORMATION
■
Notes
The authors declare no competing financial interest.
(26) o Santi, A. S.; Milani, B.; Zangrando, E.; Mestroni, G. Eur. J.
Inorg. Chem. 2000, 2351.
ACKNOWLEDGMENTS
■
(27) Already at about 70−80 °C a significant (3−7%) conversion was
observed when 6 mmol of DPU was heated in 25 mL of 3-
methylaniline under 50 bar of CO for 2 h.
(28) Kim, H. S.; Kim, Y. J.; Lee, H.; Park, K. Y.; Lee, C.; Chin, C. S.
Angew. Chem. 2002, 41, 4300.
(29) Lee, S. M.; Cho, N. S.; Kim, K. D.; Oh, J. S.; Lee, C. W.; Lee, J.
S. J. Mol. Catal. 1992, 73, 43.
(30) Lee, C. W.; Lee, J. S.; Lee, S. M.; Kim, K. D.; Cho, N. S.; Oh, J.
S. J. Mol. Catal. 1993, 81, 17.
(31) Lee, C. W.; Lee, S. M.; Oh, J. S.; Lee, J. S. Catal. Lett. 1993, 19,
217.
This research has been financially supported by the Council for
Chemical Sciences of The Netherlands Organisation for
Scientific Research (CW-NWO). This work has been
performed under the auspices of the joint NIOK Research
Graduate School of Leiden University and six other Dutch
Universities. Dr. Andreas Ehlers (Vrije Universiteit Amster-
dam) is kindly acknowledged for his assistance with the DFT
calculations. We thank the reviewers for critical reading of the
paper, which has resulted in significant improvements in the
clarity of our message.
(32) Gasperini, M.; Ragaini, F.; Remondini, C.; Caselli, A.; Cenini, S.
J. Organomet. Chem. 2005, 690, 4517.
(33) Wang, X. F.; Li, P.; Yuan, X. H.; Lu, S. W. J. Mol. Catal. A:
Chem. 2006, 253, 261.
(34) Tafesh, A. M.; Weiguny, J. Chem. Rev. 1996, 96, 2035.
(35) Dieck, H. A.; Laine, R. M.; Heck, R. F. J. Org. Chem. 1975, 40,
2819.
REFERENCES
■
(1) Richter, R. H.; Priester, R. D. In Kirk-Othmer Encyclopedia of
Chemical Technology; Kroschmitz, J. L., Howe-Grand, M., Eds.; Wiley:
New York, 1995; Vol. 14.
(2) Ulrich, H. Ullmann’s Encyclopedia of Industrial Chemistry; VCH:
New York, 1989; Vol. A14.
(36) Mooibroek, T. J. Ph.D. Thesis, Leiden University, Leiden, The
(3) Ryan, A. J.; Stanford, J. L. In Comprehensive Polymer Science;
Allen, G., Bevington, J. C., Eds.; Pergamon: New York, 1989; Vol. 5.
(4) Weisermel, K.; Arpe, H. J., Industrielle Organische Chemie, VCH
Verslagsgesellschaft GmbH, Weinheim, Germany, 1988.
(5) Registry of Toxic Effects of Chemical Substances (RTECS, online
database); United States Department of Health and Human Services
(National Toxicology Information, National Library of Medicine),
Bethesda, MD, 1993.
(6) Weast, R. C., Ed. Handbook of Chemistry and Physics, 58th ed.;
CRC Press: Cleveland, OH, 1977−1978; p D-110.
(7) Paul, F. Coord. Chem. Rev. 2000, 203, 269.
(8) Drent, E.; van Leeuwen, P. W. N. M., EU 0086281A1, 1982.
(9) Wehman, P.; Borst, L.; Kamer, P. C. J.; van Leeuwen, P. W. N. M.
J. Mol. Catal. A: Chem. 1996, 112, 23.
Netherlands, 2011; Chapter 7.
(37) These reactions were performed using the same procedure as
our standard carbonylation experiments (see the Experimental
Section), but using 2.5 mL (22.9 mmol) of dry and degassed 3-
methylnitrobenzene as the substrate, 25.0 mL (274.3 mmol) of dry
and degassed aniline as solvent, and 0.05 mmol of (L3X)Pd(OAc)2 as
the catalyst precursor.
(38) Mooibroek, T. J.; Bouwman, E.; Lutz, M.; Spek, A. L.; Drent, E.
Eur. J. Inorg. Chem. 2010, 298.
(39) Mooibroek, T. J.; Lutz, M.; Spek, A. L.; Bouwman, E. Dalton
Trans. 2010, 39, 11027.
(40) After 180 min, the solution contained 17% of phen−palladacycle
and 83% phen (1H NMR), meaning that only 83% of the initially
added palladium can end up as the L3X complex. The percentage of a
specific L3X-Pd complex (relative to palladium) can thus be calculated
by: ∑∫ L3X complex/∑∫ all L3X complexes × 83%. Thus, the species detected
with 31P{1H} NMR that were assigned to Pd-containing species are as
follows (δ (∫ ; assignment; percentage based on Pd)): 1.0 (1.00;
Pd0(L3X)2; 33); 5.0 (0.06; L3X−palladacycle; 4); 13−20 (0.33;
unknown Pd complexes; 22); 24/−7 (0.32; decarbonylated L3X−
palladacycle; 21); 35 (0.05; unknown Pd complex; 3); ∑∫ = 1.76.
The solution also contained L3X (−25 ppm, ∫ = 0.33) and L3XO
(26/−24 ppm, ∫ = 0.30), which is (0.33 + 0.30)/1.76 × 100% = 26%
of the initially added L3X. Hence, 74% of L3X is thus bound to
palladium; this is indeed roughly 1 equiv with respect to the 83% Pd
(10) Wehman, P.; Borst, L.; Kamer, P. C. J.; van Leeuwen, P. W. N.
M. Chem. Ber.-Recl. 1997, 130, 13.
(11) Leconte, P.; Metz, F.; Mortreux, A.; Osborn, J. A.; Paul, F.; Petit,
F.; Pillot, A. J. Chem. Soc., Chem. Commun. 1990, 1616.
(12) Drent, E.; van Leeuwen, P. W. N. M. EU 86281, 1981.
(13) Ragaini, F.; Cognolato, C.; Gasperini, M.; Cenini, S. Angew.
Chem. 2003, 42, 2886.
(14) Ragaini, F.; Gasperini, M.; Cenini, S. Adv. Synth. Catal. 2004,
346, 63.
(15) Ragaini, F. Dalton Trans. 2009, 6251.
(16) Ragaini, F.; Gasperini, M.; Cenini, S.; Arnera, L.; Caselli, A.;
Macchi, P.; Casati, N. Chem. Eur. J. 2009, 15, 8064.
(17) Mooibroek, T. J.; Schoon, L.; Bouwman, E.; Drent, E. Chem.
Eur. J. 2011, 17, 13318.
(18) Mooibroek, T. J.; Bouwman, E.; Drent, E. Eur. J. Inorg. Chem.
2012, 1403.
1
complex, computed on the basis of H NMR.
(41) ∫ 0.5 ppm = 100 (Pd0(L3X)2 = 4P); ∫ 14.3+(−26.2) ppm = 2.2 (L3X
NPh, 2P). Hence, 2.2/(100/2 + 2.2) × 100% = 4.2% and 4.2% of 10
μmol (the amount of phen−palladacycle present) is ∼0.4 μmol.
4155
dx.doi.org/10.1021/om200882p | Organometallics 2012, 31, 4142−4156

Organometallics
Article
(42) Haras, A.; Michalak, A.; Rieger, B.; Ziegler, T. J. Am. Chem. Soc.
2005, 127, 8765.
(43) Aberg, J. B.; Nyhlen, J.; Martin-Matute, B.; Privalov, T.;
Backvall, J. E. J. Am. Chem. Soc. 2009, 131, 9500.
(44) We thank one of the reviewers for suggesting a catalytic role of
CO in the net oxidative addition of methanol on palladium. For
example, a quite conceivable scenario is that coordinated CO at the
Pd0(CO)2 center will be initially attacked by CH3O−H+ to generate a
[Pd(CO)C(O)OCH3]−H+ ion pair. This increases both the acidity of
the proton and the nucleophilicity of the Pd center, thus facilitating the
protonation of the Pd center to provide H−Pd(CO)CO(O)CH3,
which upon decarbonylation results in C8, i.e. a CO-assisted net
oxidative addition of methanol to a Pd0 center. Interestingly, some
methyl formate, with more at the higher CO pressure, is always
formed, which at least partially may originate from reductive
elimination of HCOOCH3 from the HPdCO(O)CH3 intermediate.
(45) Ragaini, F.; Cenini, S.; Fumagalli, A.; Crotti, C. J. Organomet.
Chem. 1992, 428, 401.
(46) Zuman, P.; Fijalek, Z.; Dumanovic, D.; Suznjevic, D.
Electroanalysis 1992, 4, 783.
(47) Ueno, K.; Akiyoshi, S. J. Am. Chem. Soc. 1954, 76, 3670.
(48) Dalmagro, J.; Yunes, R. A.; Simionatto, E. L. J. Phys. Org. Chem.
1994, 7, 399.
(49) Note that, in the formation of this complex from CO/methanol,
two hydrogen atoms must have been produced; in our view, these can
only be transferred to nitrobenzene, giving aniline via a nitrene (Pd
NPh) moiety, as we propose.
(50) Armarego, W. L. F.; Chai, C. L. L. Purification of Laboratory
Chemicals, 5th ed.; Elsevier: Amsterdam, 2003.
(51) Chen, J.; Fritz, J. S. Anal. Chem. 1991, 63, 2016.
(52) customer.linde.com/FIRSTspiritWeb/linde/LGNL/media/
datasheets/NL-PIB-0024.pdf, April 2012.
(53) Koch, W.; Holthausen, M. C. A Chemist’s Guide to Density
Functional Theory, 2nd ed.; Wiley-VCH: Weinheim, Germany, 2000.
(54) Eichkorn, K.; Treutler, O.; Ohm, H.; Haser, M.; Ahlrichs, R.
Chem. Phys. Lett. 1995, 240, 283.
(55) Becke, A. D. Phys. Rev. A 1988, 38, 3098.
(56) Perdew, J. P. Phys. Rev. B 1986, 33, 8822.
(57) Hehre, W. J.; Radom, L.; Schleyer, P. v. R.; Pople, J. A. Ab Initio
Molecular Orbital Theory; Wiley: New York, 1986.
(58) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 270.
(59) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 299.
(60) Wadt, W. R.; Hay, P. J. J. Chem. Phys. 1985, 82, 284.
4156
dx.doi.org/10.1021/om200882p | Organometallics 2012, 31, 4142−4156