Palladium-catalyzed formylation of aryl bromides: Elucidation of the catalytic cycle of an industrially applied coupling reaction

Article abstract of DOI:10.1021/ja804997z

The first comprehensive study of the catalytic cycle of the palladium-catalyzed formylation of aryl bromides with synthesis gas (CO/H ₂, 1:1) is presented. The formylation in the presence of efficient (Pd/PR₂ⁿBu, R = 1-Ad, ^tBu) and nonefficient (Pd/P^tBu₃) catalysts was investigated. The main organometallic complexes involved in the catalytic cycle were synthesized and characterized, and their solution chemistry was studied in detail. Comparison of stoichiometric and catalytic reactions using P(1-Ad)₂ⁿBu, the most efficient ligand known for the formylation of aryl halides, led to two pivotal results: (1) The corresponding carbonylpalladium(O) complex [Pd _n(CO)_mL_n] and the respective hydrobromide complex [Pd(Br)(H)L₂] are resting states of the active catalyst, and they are not directly involved in the catalytic cycle. These complexes maintain the concentration of most active [PdL] species at a low level throughout the reaction, making oxidative addition the rate-determining step, and provide high catalyst longevity. (2) The product-forming step proceeds via base-mediated hydrogenolysis of the corresponding acyl complex, e.g., [Pd(Br)(p-CF ₃C₆H₄CO){P(1-Ad)₂ ⁿBu}]₂ (8), under mild conditions (25-50°C, 5 bar). Stoichiometric studies using the less efficient Pd/P^tBu₃ catalyst resulted in the isolation and characterization of the first stable three-coordinated neutral acylpalladium complex, [Pd(Br)(p-CF₃C ₆H₄CO)(P^tBu₃)] (10). Hydrogenolysis of 10 needed significantly more drastic conditions compared to that of dimeric 8. In the presence of amine base, complex 10 gave a catalytically inactive diamino acyl complex, which explains the low activity of the Pd/P ^tBu₃ catalyst formylation of aryl bromides.

Full text of DOI:10.1021/ja804997z

Published on Web 10/29/2008
Palladium-Catalyzed Formylation of Aryl Bromides:
Elucidation of the Catalytic Cycle of an Industrially Applied
Coupling Reaction
Alexey G. Sergeev, Anke Spannenberg, and Matthias Beller*
Leibniz-Institut fu¨r Katalyse e.V. an der UniVersita¨t Rostock, Albert-Einstein-Strasse 29a,
18059 Rostock, Germany
Received July 14, 2008; E-mail: matthias.beller@catalysis.de
Abstract: The first comprehensive study of the catalytic cycle of the palladium-catalyzed formylation of
aryl bromides with synthesis gas (CO/H2, 1:1) is presented. The formylation in the presence of efficient
(Pd/PR2nBu, R ) 1-Ad, tBu) and nonefficient (Pd/PtBu3) catalysts was investigated. The main organometallic
complexes involved in the catalytic cycle were synthesized and characterized, and their solution chemistry
was studied in detail. Comparison of stoichiometric and catalytic reactions using P(1-Ad)2nBu, the most
efficient ligand known for the formylation of aryl halides, led to two pivotal results: (1) The corresponding
carbonylpalladium(0) complex [Pdn(CO)mLn] and the respective hydrobromide complex [Pd(Br)(H)L2] are
resting states of the active catalyst, and they are not directly involved in the catalytic cycle. These complexes
maintain the concentration of most active [PdL] species at a low level throughout the reaction, making
oxidative addition the rate-determining step, and provide high catalyst longevity. (2) The product-forming
step proceeds via base-mediated hydrogenolysis of the corresponding acyl complex, e.g., [Pd(Br)(p-
CF3C6H4CO){P(1-Ad)2nBu}]2 (8), under mild conditions (25-50 °C, 5 bar). Stoichiometric studies using the
less efficient Pd/PtBu3 catalyst resulted in the isolation and characterization of the first stable three-
coordinated neutral acylpalladium complex, [Pd(Br)(p-CF3C6H4CO)(PtBu3)] (10). Hydrogenolysis of 10
needed significantly more drastic conditions compared to that of dimeric 8. In the presence of amine base,
complex 10 gave a catalytically inactive diamino acyl complex, which explains the low activity of the Pd/
PtBu3 catalyst formylation of aryl bromides.
n
Introduction
Pd(OAc)2 and P(1-Ad)2 Bu (cataCXium A, Evonik Degussa
GmbH) enables a more efficient (0.1-0.25 mol % of palladium)
The palladium-catalyzed formylation of aryl halides consti-
tutes an attractive method for the preparation of aromatic
aldehydes. In general, synthetic protocols for catalytic formyl-
ation of aryl halides use combinations of CO2,3 or acetic formic
anhydride1 as carbonylation agents and silyl2 or tin3 hydrides
as well as formate salts4 as hydrogen donors. Unfortunately,
side reactions such as reductive dehalogenation, limited func-
tional group tolerance, high catalyst loadings, and environmen-
tally nonbenign reagents restricted this methodology. During
our search for more reliable protocols to overcome these
limitations, we turned our attention to synthesis gas (CO/H2
mixture), the simplest and cheapest formylation agent widely
used in numerous industrial applications. Before our work, only
a few reports on the use of synthesis gas for formylation of
aryl halides were published. In all of them, [Pd(PPh3)2Cl2] was
employed as catalyst, and the reactions proceeded at high
pressure (30-100 bar) and temperature (125-150 °C).5,6
Recently, we demonstrated that a catalytic system comprising
formylation of aryl bromides under milder conditions (5 bar
pressure of synthesis gas).7 The high selectivity and efficiency
of the catalytic system allowed it to be used in the first industrial-
scale formylation of aryl halides (>1000 kg scale).
n
Notably, P(1-Ad)2 Bu is also known to be an efficient ligand
for other palladium-catalyzed carbonylations, e.g., alkoxy-
carbonylation of aryl bromides8 and carbonylative Suzuki
coupling,9 as well as Buchwald-Hartwig aminations10 and
Mizoroki-Heck,11 Suzuki,12 Sonogashira,13 and ketone aryl-
ation reactions.14 Comparing the catalytic performance of P(1-
n
Ad)2 Bu with that of symmetrically substituted bulky phosphines
(6) Palladium-catalyzed formylation of aryl halides activated with coor-
dination with tricarbonylchromium: Mutin, R.; Lucas, C.; Thivolle-
Cazat, J.; Dufaud, V.; Dany, F.; Basset, J. M. J. Chem. Soc., Chem.
Commun. 1988, 896.
(7) (a) Klaus, S.; Neumann, H.; Zapf, A.; Stru¨bing, D.; Hu¨bner, S.;
Almena, J.; Riermeier, T.; Groꢀ, P.; Sarich, M.; Krahnert, W.-R.;
Rossen, K.; Beller, M. Angew. Chem., Int. Ed. 2006, 45, 154. (b)
Brennfu¨hrer, A.; Neumann, H.; Klaus, S.; Riermeier, T.; Almena, J.;
Beller, M. Tetrahedron 2007, 65, 6252.
(1) Cacchi, S.; Fabrizi, G.; Goggiamani, A. J. Comb. Chem. 2004, 6, 692.
(2) Pri-Bar, I.; Buchman, O. J. Org. Chem. 1984, 49, 4009.
(3) (a) Baillargeon, V. P.; Stille, J. K. J. Am. Chem. Soc. 1974, 96, 7761.
(b) Misumi, Y.; Ishii, Y.; Hidai, M. Organometallics 1995, 14, 1770.
(4) Ben-David, Y.; Portnoy, M.; Milstein, D. J. Chem. Soc., Chem.
Commun. 1989, 1816.
(8) Neumann, H.; Brennfu¨hrer, A.; Groꢀ, P.; Riermeier, T.; Almena, J.;
Beller, M. AdV. Synth. Catal. 2006, 348, 1255.
(9) Neumann, H.; Brennfu¨hrer, A.; Beller, M. Chem. Eur. J. 2008, 14,
3645.
(10) (a) Ehrentraut, A.; Zapf, A.; Beller, M. J. Mol. Catal. 2002, 182-
183, 515. (b) Tewari, A.; Hein, M.; Zapf, A.; Beller, M. Tetrahedron
2005, 61, 9705.
(5) First use of synthesis gas in palladium-catalyzed formylation: Schoenberg,
A.; Heck, R. F. J. Am. Chem. Soc. 1974, 96, 7761.
(11) Ehrentraut, A.; Zapf, A.; Beller, M. Synlett 2000, 1589.
9
10.1021/ja804997z CCC: $40.75
2008 American Chemical Society
J. AM. CHEM. SOC. 2008, 130, 15549–15563 15549

A R T I C L E S
Sergeev et al.
Scheme 1. Proposed Catalytic Cycle for the Pd-Catalyzed
Formylation of Aryl Bromides with Synthesis Gas
formylation of aryl halides is established, which contributes to
a better understanding of palladium-catalyzed carbonylations
in general.
Results and Discussion
As a model for our stoichiometric studies, we chose the
catalytic formylation of 4-bromobenzotrifluoride in the presence
n
n
of three bulky phosphine ligands: P(1-Ad)2 Bu, PtBu2 Bu, and
PtBu3. The reactions were carried out using a combination of
0.25 mol % of palladium acetate and 0.75 mol % of the ligand
in the presence of TMEDA as a base in toluene at 100 °C and
5 bar pressure of synthesis gas (a 1:1 CO/H2 mixture) (eq 1),
which is comparable to the industrially applied conditions.
such as PCy3 and PtBu3, significant differences are observed.7
Hence, the specific substitution pattern including two bulky alkyl
groups and the n-butyl tail seems to be essential for the high
efficiency of the catalytic system. Replacement of the n-butyl
tail with other substituents such as methyl or benzyl led to a
significant drop in activity. In order to understand these issues,
we studied the synthesis and organometallic chemistry of
palladium complexes bearing unsymmetrically substituted bulky
alkylphosphines. More specifically, we focused our attention
on the organometallic elementary steps of the catalytic formyl-
n
Use of Pd/P(1-Ad)2 Bu, the commercially applied formylation
catalyst, resulted in 4-trifluorobenzaldehyde in 86% yield.7 A
catalyst based on PtBu2 Bu of similar structure also gave the
n
corresponding aldehyde in high yield (92%). In contrast,
application of the PtBu3-ligated catalyst led to low conversion
and only 4% yield of the product! Apparently, the nature of
n
ation of aryl halides in the presence of P(1-Ad)2 Bu-derived
palladium catalysts, which surpass the efficiency of all known
catalytic systems. To the best of our knowledge, no systematic
stoichiometric studies on palladium-catalyzed formylation of aryl
halides have been reported previously. To date, only two
examples concerning the product-forming step, namely, hydro-
genolysis of acylpalladium complexes of the type [Pd{(O)CR}-
(X)(PPh3)2], where X ) Hal15 or OC(O)R′,16 have been
published. The same is partly true for carbonylations of aryl
halides in general: despite the importance and wide use of
palladium complexes with bulky alkylphosphine ligands, there
is a lack of mechanistic details about these complexes involved
in the catalytic cycle.17
n
the bulky substituents R in PR2 Bu does not appreciably change
the catalytic activity, whereas replacement of the n-butyl group
led to severe deterioration of the catalyst performance. In order
to understand this different behavior, we studied all elementary
steps of the catalytic cycle using this set of ligands.
As the basis for our work, the generally accepted mechanism
shown in Scheme 1 was used. It includes oxidative addition of
the aryl bromide to the active palladium(0) species, migratory
insertion of CO into the Ar-Pd bond, and hydrogenolysis of
the resulting acyl complex to give the desired aldehyde. Finally,
the formed palladium hydrobromide complex is regenerated to
Pd(0) after reaction with base.
1. Synthesis and Structure of Palladium(0) Complexes. Ini-
tially, palladium(0) complexes [PdL2] were synthesized by reaction
In the present paper we describe the synthesis, characteriza-
tion, and crystal structures as well as the solution chemistry of
most of the complexes involved in the palladium-catalyzed
formylation of aryl halides with synthesis gas. All key steps of
the catalytic cycle were investigated one by one, and the results
of these studies were combined with the data on the catalytic
reaction. As a result, a catalytic cycle for palladium-catalyzed
n
n
of [AllylPdCp] with an excess of P(1-Ad)2 Bu or PtBu2 Bu in
heptane solution (eq 2). Complex 1 directly precipitated from the
reaction mixture and was isolated by simple filtration followed by
crystallization from a toluene/methanol mixture to give white
n
crystals in 77% yield. The palladium(0) complex of PtBu2 Bu (2)
was isolated in 82% yield via evaporation of the reaction mixture
and subsequent crystallization from a mixture of ether/methanol.
(12) (a) Zapf, A.; Ehrentraut, A.; Beller, M. Angew. Chem. Int. Ed 2000,
39, 4153. (b) Tewari, A.; Hein, M.; Zapf, A.; Beller, M. Synthesis
2004, 935.
(13) (a) Ko¨llhofer, A.; Pullmann, T.; Plenio, H. Angew. Chem., Int. Ed.
2003, 42, 1056. (b) Hillerich, J.; Plenio, H. Chem. Commun. 2003,
3024. (c) Remmele, H.; Ko¨llhofer, A.; Plenio, H. Organometallics
2003, 22, 4098. (d) Ko¨llhofer, A.; Plenio, H. Chem. Eur. J. 2003, 9,
1416.
The structure of the obtained complexes was confirmed by
X-ray analysis (Figures 1 and 2). Selected bond lengths and
angles are given in Table 1.
(14) Ehrentraut, A.; Zapf, A.; Beller, M. AdV. Synth. Catal. 2002, 344,
209.
Complex 2 shows a linear P1-Pd1-P1′ angle (180°),
whereas 1 exhibits a slightly bent geometry with a P1-Pd1-P2
angle of 172°. Interestingly, the conformations of the different
alkyl substituents along the P-Pd-P axis in both complexes
vary markedly. Complex 2 adopts a staggered conformation
(15) Heaton, B. T.; He´bert, S. P. A.; Iggo, J. A.; Metz, F.; Whyman, R.
J. Chem. Soc., Dalton Trans. 1993, 3081.
(16) Nagayama, K.; Kawastaka, F.; Sakamoto, M.; Shimizu, I.; Yamamoto,
A. Bull. Chem. Soc. Jpn. 1999, 72, 573.
(17) Martinelli, J. R.; Clark, T. P.; Watson, D. A.; Munday, R. H.;
Buchwald, S. L. Angew. Chem., Int. Ed. 2007, 46, 8460–8463.
9
15550 J. AM. CHEM. SOC. VOL. 130, NO. 46, 2008

Pd-Catalyzed Formylation of Aryl Bromides
A R T I C L E S
Figure 1. Molecular structure of 1.26 Hydrogen atoms are omitted for
clarity. The thermal ellipsoids correspond to 30% probability.
Figure 3. Molecular structure of 3·4C7H8.26 Hydrogen atoms are omitted
for clarity. The thermal ellipsoids correspond to 30% probability.
26
Table 2. Selected Bond Lengths (Å) and Angles (°) for 3·4C7H8
Pd1-Br1
2.5618(10)
2.5326(9)
2.012(4)
2.302(1)
83.03(4)
98.70(5)
C1-Pd1-P1
C1-Pd1-Br1
Pd1-Br1-Pd2
C8-P1-Pd1-C1
Ra
91.8(1)
169.5(1)
81.55(4)
-9.8(2)
120.16(7)
79.2(1)
Pd1-Br2
Pd1-C1
Pd1-P1
Br1-Pd1-Br2
P1-Pd1-Br1
ꢀb
a Folding angle between planes defined by P1,C1,Pd1,Br1,Br2 and
P2,C32,Pd2,Br1,Br2 respectively. b Angle between planes defined by
P1,C1,Pd1,Br1,Br2 and C1-C6, respectively.
the large distance between substituents bound to different
phosphorus atoms. Furthermore, complex 1 bearing the bulky
1-adamantyl groups adopts an eclipsed arrangement, whereas
2 having the less bulky tert-butyl substituents prefers to take
up a staggered conformation.
Figure 2. Molecular structure of 2. Hydrogen atoms are omitted for clarity.
The asymmetric unit contains half of the molecule; the remaining part is
generated by the following symmetry operator: -x + 2, -y, -z + 1. The
thermal ellipsoids correspond to 30% probability.
2. Synthesis, Structure, and Solution Behavior of Aryl-
palladium Bromide Complexes. Next, oxidative addition of aryl
bromides to the palladium(0) complexes 1 and 2 led to
arylpalladium bromide dimers 3 and 4 in 55-80% isolated
yields. Both complexes precipitated from the reaction mixture
or after addition of heptane (eq 3).
Table 1. Selected Bond Lengths (Å) and Angles (°) for 1 and 2
complex 126
complex 2
Pd1-P1
2.286(2)
2.286(2)
172.18(5)
15.4(3)
5.1(1)
17.1(3)
Pd1-P1
2.2850(4)
2.2849(4)
179.999(1)
180.0(1)
180.0(1)
180.0(1)
Pd1-P2
Pd1-P1′
P2-Pd1-P1
P1-Pd1-P1′
C1-P1-P1′-C1′
C5-P1-P1′-C5′
C9-P1-P1′-C9′
Ra
ꢀb
γc
a Angle between planes defined by C1,P1,P2 and P1,P2,C29
respectively. b Angle between planes defined by C5,P1,P2 and P1,P2,C39
respectively. c Angle between planes defined by C15,P1,P2 and P1,P2,C25
respectively.
It is noteworthy that 31P NMR spectra of the reaction mixtures
showed that the dimeric complexes are present in solution only
in small concentrations, whereas the corresponding monomeric
bisphosphine arylpalladium complexes are the major species.
Details on the reversible reaction of dimeric complexes 3 and
4 with phosphine ligands, as well as the structure of the
corresponding bisphosphine complexes, are discussed below.
In contrast to reactions with 1 and 2 (eq 3), the oxidative addition
to [Pd(PtBu3)2] gave monomeric T-shaped species as shown by
Hartwig.23,24 Unlike the synthesis of complexes 3 and 4, the
reaction toward [Pd(Br)(p-CF3C6H4)(PtBu3)] (5) (eq 4) was
which is typical for the majority of bisphosphine palladium(0)
complexes,18,19 including [Pd(PtBu3)2].20 In contrast, complex
1 takes an eclipsed arrangement of substituents about the
P1-Pd1-P2 direction, which is known for [Pd(PCy3)2]21 and
[Pd(PPhtBu2)2].22 As for geometries of 1 and of 2, the effect of
steric factors on the conformation should be ruled out due to
(18) (a) Reid, S. M.; Boyle, R. C.; Mague, J. T.; Fink, M. J. J. Am. Chem.
Soc. 2003, 125, 7816. (b) Weng, Z.; Teo, S.; Koh, L. L.; Hor, T. S. A.
Organometallics 2004, 23, 4342. (c) Grotjahn, D. B.; Gong, Y.;
Zakharov, L.; Golen, J. A.; Rheingold, A. L. J. Am. Chem. Soc. 2006,
128, 438.
(19) Paul, F.; Patt, J.; Hartwig, J. F. Organometallics 1995, 14, 3030.
(20) Tanaka, M. Acta Crystallogr. 1992, C48, 739.
(22) (a) Matsumoto, M.; Yoshioka, H.; Nakatsu, K.; Yoshida, T.; Otsuka,
S. J. Am. Chem. Soc. 1974, 96, 3322. (b) Otsuka, S.; Yoshida, T.;
Matsumoto, M.; Nakatsu, K. J. Am. Chem. Soc. 1976, 98, 5850.
(21) Immirzi, A.; Musco, A. Chem. Commun. 1974, 400.
9
J. AM. CHEM. SOC. VOL. 130, NO. 46, 2008 15551

A R T I C L E S
Sergeev et al.
in either of the two palladium square planar units the n-butyl
group of the coordinated phosphine and the aromatic group
bound to palladium are in eclipsed positions, so that the n-butyl
group lies roughly parallel to the plane of the aromatic ring
and the γ-methylene group is placed nearly above the center of
the ring. Such a conformation might be explained on the basis
of steric effects.28 In solution, complexes 3 and 4 exhibit
fluxional behavior due to cis-trans isomerization and restricted
rotation about the Pd-P bonds. At room temperature the
31P{1H} NMR spectra of the complexes in THF-d8 displayed
two broad singlets at 47.4 and 44.4 ppm for 3 and 51.6 and
50.4 ppm for 4 in a ratio of 74:26 and 60:40, respectively. The
signals of major and minor forms are assigned to trans and cis
isomers of dimers, respectively.29 Domination of the trans form
in THF is explained by steric reasons. A larger trans-to-cis ratio
for 3 (74:26) as compared to that of 4 (60:40) is a consequence
n
of the increased steric bulk of P(1-Ad)2 Bu compared to
n
PtBu2 Bu and the nonplanar structure of 3 that might additionally
increase the steric strain in the cis form. In contrast to the
31P{1H} NMR spectra of 3 and 4, displaying two distinctive
signals for cis and trans forms, in the proton spectra both forms
are in fast exchange, as indicated by the presence of only one
pair of broadened aromatic doublets at 7.59 and 7.11 ppm for
3 and 7.59 and 7.13 ppm for 4. On cooling of the solutions to
-40 (complex 3) and -60 °C (complex 4), these broadened
signals resolved into two pairs of doublets corresponding to the
cis and trans forms. A closer look at the aliphatic part of the
proton spectra revealed that both isomers of 3 and 4 show
pronounced upfield shifts of the n-butyl protons. This effect is
explained by the aromatic ring current effect30 if we suppose
that both cis and trans forms of the complexes possess the same
eclipsed orientations of n-butyl groups and aromatic rings that
were found in the solid state. The shielding of the methylene
Figure 4. Molecular structure of 4. Hydrogen atoms are omitted for clarity.
The asymmetric unit contains half of the molecule; the remaining part is
generated by the following symmetry operator: -x + 1, -y + 1, -z + 1.
The thermal ellipsoids correspond to 30% probability.
Table 3. Selected Bond Lengths (Å) and Angles (°) for 4
Pd1-C1
1.984(3)
2.2891(8)
2.5176(4)
2.5812(4)
85.86(9)
Br1-Pd1-Br1′
C1-Pd1-P1
C8-P1-Pd1-C1
Pd1-Br-Pd1′
Ra
82.85(1)
90.04(9)
1.1(2)
97.15(1)
180.0(1)
72.76(9)
Pd1-P1
Pd1-Br1
Pd1-Br′
C1-Pd1-Br1
P1-Pd1-Br1′
101.26(2)
ꢀb
a Folding angle between planes defined by P1,C1,Pd1,Br1,Br1′ and
P1′,C1′,Pd1′,Br1,Br1′ respectively. b Angle between planes defined by
P1,C1,Pd1,Br1,Br1′ and C1-C6 respectively.
1
protons is also seen by comparison of the H NMR spectra of
arylpalladium complexes 3 and 4 with those of the analogous
bisphosphine palladium(0) complexes 1 and 2 (Table 4).
Chemical shifts of the adamantyl and tert-butyl protons are
changed only slightly on going from [PdL2] to [Pd(Ar)(Br)L]2,
whereas signals of the n-butyl protons are shifted by up to 1.3
ppm toward high field. Moreover, the γ-methylene protons of
the n-butyl group show signals at even higher field than the
corresponding methyl protons! This is a result of the closer
location of γ-methylene protons to the centers of the aromatic
rings, where the shielding has a peak value.
accompanied with the formation of several side products
presumably due to decomposition of 5.25 Therefore, the reaction
time should be minimized to obtain acceptable yields of 5.
Complexes 3 and 4, assumed to be involved in the catalytic
cycle, exist as dimers both in the solid state and in solution, as
shown by X-ray and NMR data.26 Molecular structures of 3
and 4 are depicted in Figures 3 and 4, and selected bond lengths
and angles are given in Tables 2 and 3, respectively.
Both structures show a trans geometry of the complexes with
two palladium square planar units linked by two bromine atoms.
The units are located to each other at a folding angle of 120.2°
in complex 3 and 180° in complex 4.27 The complexes exhibit
a peculiar arrangement of the substituents about P-Pd bonds:
(27) Most dimeric arylpalladium complexes possess nearly coplanar
orientations of the two palladium square planes. For selected structures,
see: (a) Dufaud, V.; Thivolle-Cazat, J.; Basset, J.-M.; Mathieu, R.;
Jaud, J.; Waissermann, J. Organometallics 1991, 10, 4005. (b) Paul,
F.; Patt, J.; Hartwig, J. F. J. Am. Chem. Soc. 1994, 116, 5969. (c) Lin,
S.-T.; Cheo, H.-S.; Liu, L.-S.; Wang, J.-C. Organometallics 1997, 16,
1803. (d) Marshall, W. J.; Young, R. J., Jr.; Grushin, V. V.
Organometallics 2001, 20, 523. (e) Bartolome´, C.; Espinet, P.; Mart´ın-
Alvarez, J. M.; Villafan˜e, F. Eur. J. Inorg. Chem. 2003, 3127. (f) Teo,
S.; Weng, Z.; Hor, T. S. A. Organometallics 2006, 25, 1199.
(28) For comparison of steric parameters of substituents in organometallic
compounds: White, D. P.; Anthony, J. C.; Oyefeso, A. O. J. Org.
Chem. 1999, 64, 7707, and references therein.
(23) Stambuli, J. P.; Bu¨hl, M.; Hartwig, J. F. J. Am. Chem. Soc. 2002,
124, 9346.
(29) On cooling of the solutions of 3 and 4 to -60 and -80 °C, these
signals sharpen and two additional marginal singlets appear. New
signals probably correspond to various minor conformers of cis
and trans form.
(24) Stambuli, J. P.; Incarvito, C. D.; Bu¨hl, M.; Hartwig, J. F. J. Am. Chem.
Soc. 2004, 126, 1184.
(25) Barrios-Landeros, F.; Carrow, B. P.; Hartwig, J. F. J. Am. Chem. Soc.
2008, 130, 5842.
(30) (a) March, J. AdVanced Organic Chemistry. Reactions, Mechanisms,
and Structure, 6th ed.; John Wiley & Sons: New York, 2007; pp 54-
92. (b) Chen, J.; Cammers-Goodwin, A. Eur. J. Org. Chem. 2003,
3861. (c) Pascal, R. A., Jr.; Winans, C. G.; Engen, D. V. J. Am. Chem.
Soc. 1989, 111, 3007. (d) Boekelheide, V. Pure Appl. Chem. 1975,
44, 751.
(26) Recently we reported elucidation of dimeric structure of P(1-Ad)2nBu-
ligated arylpalladium bromide complexes in solution as well as the
existence of cis-trans equilibrium between dimers: Sergeev, A. G.;
Zapf, A.; Spannenberg, A.; Beller, M. Organometallics 2008, 27, 297.
9
15552 J. AM. CHEM. SOC. VOL. 130, NO. 46, 2008

Pd-Catalyzed Formylation of Aryl Bromides
A R T I C L E S
Table 4. 1H NMR Chemical Shifts of the n-Butyl Protons in
Arylpalladium Complexes 3 and 4 and Corresponding Palladium(0)
Complexes 1 and 2
complex
δH(CH3), ppm
δH(γ-CH2), ppm
δH(ꢀ-CH2), ppm
δH(R-CH2), ppm
3a
1a
4b
2b
0.62
1.09
0.55
0.91
0.43
1.73
0.42
1.61
0.98
1.98
0.99
1.77
1.23
1.46
1.42
1.44
a 1H NMR spectrum recorded in toluene-d8 at -36 °C. b 1H NMR
spectrum recorded in THF-d8 at -60 °C.
The 31P{1H} NMR spectra of the reaction mixtures of 1 and
2 with 4-bromobenzotrifluoride (eq 3) showed formation of
monomeric bisphosphine aryl complexes as the main species
instead of dimers 3 and 4 that were finally isolated as products
via precipitation. This implies that the dimer complexes revers-
ibly react with phosphines to give the corresponding bisphos-
phine species. To prove this assumption, dimer 3 was reacted
Figure 5. Molecular structure of 7. Hydrogen atoms are omitted for clarity.
The thermal ellipsoids correspond to 30% probability.
Table 5. Selected Bond Lengths (Å) and Angles (°) for 7
Pd1-C1
1.992(4)
2.383(1)
2.384(1)
2.374(1)
88.5(1)
Cl1-Pd1-P1
Cl1-Pd1-P2
P1-Pd1-P2
C8A-P1-Pd1-C1
C20-P2-Pd1-C1
Ra
91.48(4)
90.68(5)
170.77(5)
11.5(3)
-5.7(2)
86.8(1)
Pd1-P1
Pd1-P2
Pd1-Cl1
C1-Pd1-P1
C1-Pd1-P2
n
with 2 equiv of P(1-Ad)2 Bu in toluene-d8 at room temperature
(eq 5).
90.9(1)
a Angle between planes defined by P1,Cl1,P2,C1,Pd1 and C1-C6
respectively.
solution, 6 and 7 exist in trans form and exhibit an equilibrium
between two conformers in a ratio of 4.6:1 (-36 °C in toluene-
d8) for 6 and 6:1 (-60 °C in THF-d8) for 7, respectively. This
1
is illustrated by variable-temperature 31P{1H} and H NMR
Indeed, the reaction resulted in an equilibrium mixture
consisting of starting complex 3, free phosphine ligand, and
complex 6 in ca. 1:1:3.9 ratio as evidenced by 31P{1H} NMR
spectra. Due to reversibility of the reaction, we failed to isolate
complex 6; however, it was characterized in the mixture by 1H,
31P{1H}, and 13C{1H} NMR and HR-ESI mass spectra.
Unlike the bisphosphine aryl bromide complexes of P(1-
spectra of 7 in THF-d8 (Figure 6). At room temperature, the
31P{1H} NMR spectrum of 7 contains a broad singlet at 33.2
ppm. Upon cooling of the solution to -60 °C, the peak
transforms into a singlet at 29.8 ppm (major form) and two
doublets at 39.3 and 31.6 ppm (minor form, 2JPP ) 378 Hz) in
a ratio of 6:1.
n
n
Ad)2 Bu and PtBu2 Bu, the corresponding chloride complex of
The equivalence of the phosphorus atoms in the major form
is consistent with the solid-state structure of 7 with a sym-
metrical arrangement of butyl groups above and below the aryl
ring. Such a conformation of the major form is again evidenced
by the shielding of the aliphatic protons due to the aromatic
ring current effect. At low temperature the signals of the
γ-methylene groups of 7 appear at higher field with respect to
neighboring methyl group (Figure 6). Additionally, the methyl
groups of the tert-butyl substituents of the coordinated ligands
exhibit a triplet at 1.52 ppm (3JPH ) 6.2 Hz) due to the virtual
P-P coupling, which is a distinctive feature of trans-diphos-
phine complexes.31 The minor form of 7 contains two non-
equivalent phosphines, as evidenced by the presence of two
n
PtBu2 Bu (7) was successfully prepared and characterized by
NMR and X-ray analysis. 7 was obtained via oxidative addition
of 4-chlorobenzotrifluoride to 2 in toluene at 70 °C to give the
product in 69% yield (eq 6).
2
doublets at 39.3 and 31.6 ppm with JPP ) 378 Hz. This
magnitude of P-P coupling constant is indicative of complexes
with mutually trans-located phosphine ligands.32 A closer
1
Single crystals of 7 suitable for X-ray analysis were obtained
via solvent diffusion of methanol into a solution of the complex
in dichloromethane at -28 °C. The molecular structure of 7 is
shown in Figure 5, and selected bond lengths and angles are
listed in Table 5.
inspection of the aliphatic region in the H NMR spectrum at
-60 °C revealed that the minor form shows a triplet of the
methyl group of the n-butyl substituent at 0.90 ppm, which is
close to the chemical shift of the unshielded methyl group in
palladium(0) complex 2 bearing the same phosphine ligand
Similarly to arylpalladium complexes 3 and 4, bisphosphine
complex 7 possesses an eclipsed arrangement of the aryl
substituent and the n-butyl groups along the Pd-P bond. In
(31) Crabtree, R. H. The Organometallic Chemistry of the Transition Metals,
4th ed.; John Wiley & Sons, Inc.: New York, 2005; pp 276-279.
9
J. AM. CHEM. SOC. VOL. 130, NO. 46, 2008 15553

A R T I C L E S
Sergeev et al.
1
Figure 6. Variable-temperature 31P{1H} and H NMR spectra of 7 in THF-d8.
(Table 4). In contrast, the corresponding triplet of the methyl
group of the major form appears at 0.63 ppm. Unshielding of
the n-butyl protons as well as nonequivalence of phosphorus
atoms in the minor form agrees with structure 7b having two
pairs of tert-butyl substituents in trans arrangement (eq 6a).
Therefore, the fluxional behavior of bisphosphine complexes 6
and 7 is attributed to the equilibrium between two rotamers about
the Pd-P bonds.33
tion; selected bond lengths and angles are given in Tables 6
and 7, respectively.
Each complex contains two palladium square planes bridged
by the bromine atoms. The folding angle between these planes
is 147.3° for 8 and 110.9° for 9. In both complexes the acyl
groups are approximately orthogonal to the respective palladium
square plane and located syn to each other (Figure 7). In
accordance with arylpalladium complexes 3, 4, and 7 the acyl
complexes 8 and 9 contain an eclipsed arrangement of the
n-butyl groups and the acyl substituents about the Pd-P bonds.
Consequently, the methyl groups of the n-butyl substituents take
a position above the plane of the aromatic ring.
As shown in Figure 9, the acyl complex of PtBu3 (10)
possesses a distorted T-shaped geometry; see Table 8 for
selected bond lengths and angles. Here, one of the hydrogen
atoms (H20A) of the phosphine ligand is located close to
the metal center (2.47 Å), which implies a weak agostic
interaction.34 The reason for the preference for this particular
configuration can hardly be explained by steric effects, since
tri-tert-butylphosphine occupies a position cis to the acyl
ligand, rather than to the less bulky bromine atom.28
However, the placement of the acyl substituent opposite to
the free coordination site may be a result of the large trans
influence. Clearly, in complex 10, aroyl has the largest trans
influence among all three ligands.35 A similar configuration
is known for the three-coordinate arylpalladium halide
complexes of bulky phosphines23,24 and carbene ligands.36
To our surprise, the coordinative unsaturated acyl complex
10 is stable in solution (THF-d8) and does not decompose to a
considerable extent within 2 days, as shown by 1H and 31P{1H}
NMR spectra. This is in contrast to the common belief that three-
coordinate palladium acyl complexes readily decompose via
decarbonylation reactions (eq 9).37
3. Synthesis, Structure, and Solution Behavior of Acylpalla-
dium Bromide Complexes. The dimeric arylpalladium com-
plexes 3 and 4 smoothly reacted with CO leading to dimeric
acyl complexes 8 and 9 in 92 and 93% isolated yields,
respectively (eq 7). Applying the same conditions, carbony-
lation of 5 bearing PtBu3 yielded the monomeric acyl complex
10 (eq 8). To the best of our knowledge, this is the first
example of a three-coordinated acylpalladium complex
described to date.
Since the acylpalladium complexes 8-10 are crucial species
in the product-forming step of the formylation of aryl bromides,
their solution behavior is especially important. Hence, we
investigated the structures of 8-10 in detail using 1D and 2D
The solid-state structures of all acyl complexes were deter-
mined by X-ray crystallography, and the solution behavior of
the complexes was investigated by means of 1H, 31P{1H}, and
13C{1H} NMR spectroscopy. Single crystals of 8 and 9 were
grown via slow evaporation of ether and ether/dichloromethane
solutions, respectively, in a vacuum at +4 °C. As shown in
Figures 7 and 8, both acyl complexes exist in trans configura-
2
(32) JPP for trans-diphosphine palladium complexes usually exceeds 200
Hz, whereas for cis-ones 2JPP is less than 80 Hz: (a) Berger, S.; Braun,
S.; Kalinowsky, H.-O. NMR-Spectroskopie Von Nichtmetallen. 31P
NMR-Spectroscopie, Band 3; Thieme; Stuttgart/New York, 1993; pp
159-160. (b) Verkade, J. G. Coord. Chem. ReV. 1972, 9, 1.
9
15554 J. AM. CHEM. SOC. VOL. 130, NO. 46, 2008

Pd-Catalyzed Formylation of Aryl Bromides
A R T I C L E S
Figure 7. Molecular structure of 8·0.3Et2O. Hydrogen atoms are omitted for clarity. The asymmetric unit contains half of the molecule; the remaining part
is generated by the following symmetry operator: -x + 5/4, -y + 1/4, -z + 5/4.
Table 6. Selected Bond Lengths (Å) and Angles (°) for 8·0.3Et2O
Pd1-C32
1.967(5)
2.306(1)
2.5254(5)
2.6048(6)
168.9(2)
84.75(2)
C32-Pd1-Br1
P1-Pd1-Br1′
Pd1-Br1-Pd1′
C8-P1-Pd1-C32
Ra
85.0(1)
100.22(3)
89.39(2)
3.3(2)
147.3(1)
77.7.(1)
Pd1-P1
Pd1-Br1
Pd1-Br1′
C32-Pd1-Br1′
Br1-Pd1-Br1′
ꢀb
Indeed, in the 31P{1H} NMR spectrum of the mixture, two
new broad singlets (52.3 and 47.8 ppm) of 11 appeared in
addition to signals of the starting complexes 8 (52.9 ppm) and
9 (47.8 ppm). More specifically, 8 and 9 show a fluxional
behavior which is consistent with three dynamic processes:
cis-trans isomerization and hindered rotation about both Pd-P
and Pd-C bonds. Since the trends observed in the temperature-
dependent NMR spectra of 8 an 9 are very similar, here we
discuss only the NMR behavior of 8 as an example. At room
temperature the 31P{1H} NMR spectrum of 8 in THF-d8 displays
a broad singlet at 47.3 ppm. On cooling of the solution to -60
°C, this signal transforms into four singlets with a relative
intensity of 85.5:9.6:3:2.8. The major singlet at 47.0 ppm is
assigned to the trans form of the complex. Other minor signals
are attributed to various conformers of trans and cis isomers.
Analysis of the variable-temperature 1H NMR spectra of 8
shows that, in contrast to the corresponding aryl complex 3,
the acyl complex 8 exhibits hindered rotation about the Pd-C
bond, which results in nonequivalence of four aromatic protons
at -100 °C (Figure 10).
a Folding angle between planes defined by P1,C32,Pd1,Br1,Br1′ and
P1′,C32′,Pd1′,Br1,Br1′ respectively. b An angle between planes defined
by P1,C32,Pd1,Br1,Br1′ and C32,C1-C6 respectively.
More specifically, at room temperature, rotation of the
1
p-trifluoromethylbenzoyl group is fast on the H NMR time
Figure 8. Molecular structure of 9. Hydrogen atoms are omitted for clarity.
The asymmetric unit contains half of the molecule; the remaining part is
generated by the following symmetry operator: -x - 1/2, y, -z + 3/2.
The thermal ellipsoids correspond to 30% probability.
scale, and 8 displays two doublets at 7.76 and 8.46 ppm,
corresponding to protons in positions ortho to the CF3 (H3, H5)
and C(O) groups (H2, H6), respectively (Figure 11). However,
when the temperature is decreased to -20 °C, the doublet of
aromatic protons adjacent to the carbonyl group (H2, H6) turns
into a broad singlet, whereas the doublet of protons H3, H5
broadens only slightly. Clearly, protons H2 and H6 are closer
to the sterically congested palladium atom compared to protons
H3 and H5. Therefore, slowing down the rotation of the acyl
group has a more pronounced influence on the chemical shifts
of H2 and H6. At -60°, two broad singlets at 7.87 and 9.41
ppm are observed, which at -100 °C transform into four
broadened doublets at 7.59, 7.76, 8.11, and 9.43 ppm corre-
sponding to the major trans form of 8. Additionally, a small
broad doublet at 9.65 ppm corresponds to one of the minor
isomers, which is in ca. 1:10 ratio to the major form. Considering
the proton spectra of the trans form, one can see that the doublet
Table 7. Selected Bond Lengths (Å) and Angles (°) for 9
Pd1-C20
1.979(3)
2.3016(9)
2.5274(5)
2.6329(4)
90.7(1)
C20-Pd1-Br1
P1-Pd1-Br1′
Pd1-Br1-Pd1′
C20-Pd1-P1-C8
Ra
85.9(1)
100.85(3)
78.70(1)
-10.9(2)
110.91(2)
69.59(8)
Pd1-P1
Pd1-Br1
Pd1-Br1′
C20-Pd1-P1
Br1-Pd1-Br1′
82.87(2)
ꢀb
a Folding angle between planes defined by P1,C20,Pd1,Br1,Br1′ and
P1′,C20′,Pd1′,Br1,Br1′ respectively. b Angle between planes defined by
P1,C20,Pd1,Br1,Br1′ and C20,C1-C6 respectively.
NMR experiments. The dimeric structures of 8 and 9 in solution
were confirmed by the formation of the mixed dimer 11 upon
mixing 8 and 9 in THF-d8 (eq 10).
9
J. AM. CHEM. SOC. VOL. 130, NO. 46, 2008 15555

A R T I C L E S
Sergeev et al.
Figure 9. Molecular structure of 10. Hydrogen atoms (except those attached
to C20) are omitted for clarity. The thermal ellipsoids correspond to 30%
probability.
Figure 10. Variable-temperature 1H NMR spectrum of 8 in THF-d8. The
assignment of the aromatic protons at -100 °C is confirmed by COSY and
HMQC.
Table 8. Selected Bond Lengths (Å) and Angles (°) for 10
Pd1-C8
1.946(4)
2.3069(10)
2.4484(5)
O1-C8-C1-C6
C8-Pd1-Br1
P1-Pd1-Br1
Ra
-16.9(6)
93.0(1)
165.19(3)
74.8(1)
Pd1-P1
Pd1-Br1
C8-Pd1-P1
101.1(1)
a Angle between planes defined by P1,Pd1,Br1,C8 and C1-C6,C8
respectively.
at 9.43 ppm exhibits a significant low-field shift comparing to
the other three doublets of the trifluoromethylbenzoyl group.
This low-field doublet is assigned to proton H2 located above
the coordination plane in close proximity to the metal center
(Figure 11). The pronounced deshielding of this proton is due
to the paramagnetic anisotropy of the palladium atom.38
Similar to the aryl complexes, the acyl complexes 8 and 9
show a comparable arrangement of the n-butyl moiety of the
phosphine ligand, lying over the aromatic ring of the acyl group,
as evidenced by the shielding of n-butyl protons. The methyl
and γ-methylene protons of the n-butyl groups are located close
to the centers of the aromatic rings and display a pronounced
downfield chemical shift as compared to the corresponding
protons of the phosphine ligand in palladium(0) complexes 1
and 2. For example, at -100 °C the γ-methylene and methyl
protons of the n-butyl group of 8 are shifted by 0.86 and 0.83
ppm toward high field with respect to the corresponding signals
of 1. Notably, the monomeric acyl complex 10 does not exhibit
a dynamic behavior in the 1H and 31P{1H} NMR spectra at room
temperature.
Figure 11. Orientation of the trifluoromethylbenzoyl moiety in 8. Number-
ing of aromatic protons corresponds to the crystal structure of 8.
excess of ligand is used. Interestingly, according to the NMR
spectra, the three-coordinate complex 10 did not react with
excess of the tri-tert-butylphosphine.39 Addition of 2 equiv of
the phosphine did not lead to the formation of any new signals
in the 1H and 31P{1H} NMR or broadening of existing ones.40
In contrast, 8 and 9 react reversibly with the phosphine ligands
to give an equilibrium mixture of the corresponding trans-
bisphosphine acyl complexes 12 and 13 (eq 11).
Next, we studied the reactivity of 8-10 in the presence of
n
Thus, addition of 2 equiv of P(1-Ad)2 Bu to complex 8 led
additional phosphines because, under catalytic conditions, an
to an equilibrium mixture of the starting dimer 8 and the
bisphosphine complex 12 in a 1:2.2 ratio. When the temperature
decreases, the equilibrium shifts toward 12, implying that the
(33) Analogous rotamers were reported for other square-planar trans
complexes with similar phosphine ligands, trans-[Pd(Hal)2(PtBu2H)2],
trans-[Pd(Hal)(CO)(PtBu2H)2], and trans-[M(Cl)(CO)(PtBu2R)2] (M
) Ir, Rh): (a) Mann, B. E.; Masters, C.; Shaw, B. L.; Stainbank, R. E.
J. Chem. Soc. D 1971, 1103. (b) Bright, A.; Mann, B. E.; Masters,
C.; Shaw, B. L.; Slade, R. M.; Stainbank, R. E. J. Chem. Soc. A 1971,
1826.
(36) Computational calculations of all possible isomers of [Pd(ItBu)(Ph)(Cl)]
showed that the most stable configuration is that with phenyl trans to
the free position: Green, J. C.; Herbert, B. J.; Lonsdale, R. J.
Organomet. Chem. 2005, 690, 6054.
(34) Sum of van der Waals radii of palladium and hydrogen is 2.8 Å: Bondi,
A. J. Phys. Chem. 1964, 68, 441. Some reviews on agnostic
interaction: (a) Brookhart, M.; Green, M. L. H. J. Organomet. Chem.
1983, 250, 395. (b) Brookhart, M.; Green, M. L. H.; Wong, L.-L.
Prog. Inorg. Chem. 1988, 36, 1.
(37) (a) To´th, I.; Elsevier, C. J. J. Am. Chem. Soc. 1993, 115, 10388. (b)
Shultz, C. S.; Ledford, J.; DeSimone, J. M.; Brookhart, M. J. Am.
Chem. Soc. 2000, 122, 6351. (c) Ledford, J.; Shultz, C. S.; Gates,
D. P.; White, P. S.; DeSimone, J. M.; Brookhart, M. Organometallics
2001, 20, 5266. (d) Bianchini, C.; Meli, A.; Mu¨ller, G.; Oberhauser,
W.; Passaglia, E. Organometallics 2002, 21, 4965.
(35) Appleton, T. G.; Clark, H. C.; Manzer, L. E. Coord. Chem. ReV. 1973,
10, 335.
9
15556 J. AM. CHEM. SOC. VOL. 130, NO. 46, 2008

Pd-Catalyzed Formylation of Aryl Bromides
A R T I C L E S
Figure 12. Variable-temperature 1H and 31P{1H} NMR spectra of 12 formed by mixing 8 with P(1-Ad)2nBu (5 equiv) in toluene-d8. The assignment of the
aromatic protons was confirmed by COSY.
reaction is exothermic. Thus, by lowering the temperature from
+9 to -51 °C, the ratio of 8 to 12 is increased from 1:8.7 to
1:100.
9.97, 7.66, 7.40, 7.05 ppm and 9.99, 7.69, 7.36, 6.98 ppm,
corresponding to a pair of conformers in ca. 2.2:1 ratio. In the
31P{1H} NMR spectrum, major and minor conformers display
two doublets at 29.6 and 26.3 ppm (2JPP ) 256 Hz) and one
broadened singlet at 23.0 ppm, respectively (see Figure 12).
Whereas the major form displays two doublets in the temper-
ature range starting from the coalescence point to -81 °C, the
singlet of the minor form at 23 ppm transforms into two broad
doublets at 24.4 and 21.4 ppm (2JPP ) 260 Hz) when the
temperature decreases below -51 °C. Major and minor
conformers have 2JPP constants exceeding 200 Hz, which allows
for an unambiguous assignment of the trans configuration to
both species. Finally, we suppose that the dynamic behavior of
12 is explained by an equilibrium between the two rotamers
depicted in Figure 13. The major form is assigned to rotamer
12a, whereas the minor species corresponds to two degenerated
rotamers 12b.
Both complexes 12 and 13 exist in solution as two conformers
due to the hindered rotation about the Pd-P bonds. This is
illustrated by variable-temperature NMR spectra of 12 in
toluene-d8 (Figure 12). At room temperature, both conformers
are in fast equilibrium on the 1H NMR time scale. However, in
contrast to 8, both conformers of 12 exhibit a restricted rotation
about the Pd-C bond already at room temperature. This is
1
evidenced in the H NMR spectrum by the presence of four
broadened doublets at 9.98, 7.73, 7.47, and 7.23 ppm, corre-
sponding to the four nonequivalent aromatic protons of the
p-trifluoromethylbenzoyl group (Figure 12).
In 12 the aroyl group is orthogonal to the palladium square
plane. As a result, the appearance of a low-field doublet at 9.98
ppm is indicative of the aromatic proton H2 directed toward
the palladium center. Upon cooling of the solution to -51 °C,
the set of four doublets splits into two groups of doublets at
(38) (a) Miller, R. G.; Stauffer, R. D.; Fahey, D. R.; Parnell, D. R. J. Am.
Chem. Soc. 1970, 92, 1511. (b) Deeming, A. J.; Rothwell, I. P.;
Hursthouse, M. B.; New, L. J. Chem. Soc., Dalton Trans. 1978, 1490.
(c) Widenhoefer, R. A.; Zhong, H. A.; Buchwald, S. L. Organome-
tallics 1996, 15, 2745. (d) Albert, J.; Cadena, J. M.; Granell, J. R.;
Solans, X.; Font-Barida, M. Tetrahedron Asymmetry 2000, 11, 1943.
(e) Barloy, L.; Gauvin, R. M.; Osborn, J. A.; Sizun, C.; Graff, R.;
Kyritsakas, N. Eur. J. Inorg. Chem. 2001, 1699. (f) Riera, X.; Moreno,
V.; Feisinger, E.; Lippert, B. Inorg, Chim. Acta 2002, 339, 253.
(39) Similarly, Hartwig et al. did not observe any reaction of the three
coordinate aryl complex [Pd(Br)(Ph){P(1-Ad)tBu2}] with added P(1-
Ad)tBu2, see ref 23.
Figure 13. Newman projections of two rotamers of complexes 12 and 13
along the P-Pd-P direction.
In addition, when the temperature decreases to -66 °C, one
of the two doublets of 12a at 26.3 ppm broadens to a much
greater extent that the second one. This is attributed to a hindered
rotation of one of the sterically encumbered adamantyl groups
about the P-C bond. As shown in the Newman projection of
12a (Figure 13), one adamantyl group bound to the rear
phosphorus atom is “clutched” between the oxygen atom of the
acyl substituent and another adamantyl group. Complex 13
shows a similar dynamic behavior, except that the coalescence
temperature in 31P{1H} NMR for the equilibrium between 13a
and 13b is lower (0 °C) and the two degenerated rotamers 13b
(40) In contrast, complex 10 reacts with 1 equiv of P(1-Ad)2nBu at room
temperature in toluene-d8 yielding the acyl complex 8 and free PtBu3
in quantitative yield. This result illustrates (1) the possibility of
phosphine exchange processes in 10, which could proceed via four
coordinate bisphosphine intermediates, and (2) stronger binding of P(1-
Ad)2nBu to palladium(II) as compared to PtBu3.
(41) A similar decelerating effect of phosphine additives was observed in
the hydrogenolysis of cobalt and ruthenium acyl complexes that
proceeds via formation of coordinatively unsaturated species. See for
example: Joshi, A. M.; James, B. R. Organometallics 1990, 9, 199,
and references therein.
9
J. AM. CHEM. SOC. VOL. 130, NO. 46, 2008 15557

A R T I C L E S
Sergeev et al.
n
n
are in fast exchange, even at -80 °C, which is in agreement
with the lower steric bulk of tert-butyl groups as compared to
adamantyl ones.
Ad)2 Bu and PtBu2 Bu on the one hand and PtBu3 on the other
hand differed drastically. Intrigued by this pronounced discrep-
ancy, we studied the hydrogenation of the different acylpalla-
dium complexes, which is the product-forming step of the
formylation of aryl bromides. As a starting point we chose the
hydrogenation of 8 at 5 bar of synthesis gas in toluene-d8
(eq 13). Here, we examined the influence of phosphine ligands,
base, and temperature on the reaction course (Table 10).
4. Synthesis and Structure of Palladium Hydride Complexes.
Palladium hydride complexes are assumed to result from the
hydrogenation of the acyl complexes.15 Hence, we synthesized
novel palladium hydrobromide complexes having di-1-adaman-
tylalkylphosphine (14) and di-tert-butylalkylphosphine (15)
ligands. They were smoothly prepared by addition of pyridine
hydrobromide in acetonitrile to a toluene solution of the
corresponding [PdL2] complexes at room temperature (eq 12).
1
The reaction mixtures were analyzed by H and 31P{1H}
NMR; conversions and yields were measured using 1,4-
dimethoxybenzene as an internal standard for H NMR. At 50
1
Single crystals of 14 suitable for X-ray analysis were obtained
by diffusion of ether into a solution of 14 in dichloromethane
at room temperature. The structure of 14 is depicted in Figure
14. Selected bond lengths and angles are given in Table 9.
Complex 14 exhibits a gauche conformation of the sub-
stituents. The P1-Pd-P2 angle of 166.9° is appreciably
distorted from the linear one. Interestingly, the angle is bent
in a direction that brings two gauche-adamantyl groups close
to each other. This geometry may be explained by minimiza-
tion of steric repulsion between the bulky adamantyl sub-
stituents and the bromine atom. As a result, the hydrogen
atom is effectively shrouded by two adamantyl groups.
5. Hydrogenolysis of Acylpalladium Complexes with Synthesis
Gas. Notably, all performed stoichiometric transformations of
°C, hydrogenolysis of 8 did not proceed without additives, and
the starting complex precipitated from the reaction mixture
(Table 10, entry 4). Interestingly, addition of 4 equiv of TMEDA
dramatically accelerated the desired reaction, yielding 4-tri-
fluoromethylbenzaldehyde in 76% yield at full conversion (Table
10, entry 6). A similar result was obtained in the presence of a
100-fold excess of TMEDA (Table 10, entry 7). Under the same
n
conditions, the PtBu2 Bu ligated complex 9 gave the aldehyde
in 74% yield (Table 10, entry 10). In the presence of base,
hydrogenation of 8 can be effected even at room temperature
(Table 10, entry 2). An increase in the concentration of the base
improves both the yield and the conversion (Table 10, entries
1-3), while an increase in phosphine concentration hampers
the reaction and decreases the aldehyde yield to 54% (Table
10, entries 6, 8-9). The decelerating effect of phosphine is
consistent with the formation of coordinatively unsaturated
species prior to hydrogenolysis of the Pd-Cacyl bond.41 In this
respect the activation effect of the amine seems rather unusual,
since TMEDA is also prone to bidentate coordination, thus
blocking free coordination sites. In order to clarify if the
activation effect of TMEDA is due to its reaction with the acyl
n
n
the palladium complexes bearing either P(1-Ad)2 Bu, PtBu2 Bu,
or PtBu3 up to the corresponding acyl complexes proceeded with
high yields whatever the ligand used. However, product yields
and efficiency of the catalytic reactions utilizing ligands P(1-
1
complex, we compared the H and 31P{1H} NMR spectra of
the dimeric acyl complex 8 in toluene-d8 in the presence and
in the absence of a 4-fold molar excess of TMEDA. However,
the 1H and 31P{1H} NMR spectra did not reveal any appreciable
differences.42
To rationalize the accelerating effect of the base, the possible
pathways for hydrogenolysis of acyl complexes are shown in
Figure 15. In agreement with known mechanisms,43,44 we
suppose that, prior to hydrogenolysis, the dimeric acyl com-
plexes 8 and 9 dissociate to give coordinatively unsaturated
Figure 14. Molecular structure of 14. The asymmetric unit contains two
molecules with similar bond lengths and angles; therefore, only one molecule
is depicted. All hydrogen atoms except H1 are omitted for clarity. The
thermal ellipsoids correspond to 30% probability.
(42) In contrast, NMR spectra of the mixture in THF-d8 at room temperature
display a rather complicated picture including equilibrium between
starting complex 8 and at least four new species. 31P{1H} NMR
spectrum recorded at-40°C showed the presence of four major groups
of multiplets: two doublets at 26.5 and 29.9 ppm (2JPP ) 256 Hz) and
at 27.1 and 30.8 ppm (2JPP ) 265 Hz) and two broad singlets at 23.8
and 22.7 ppm in ca. 3:1.3:1.3:1 ratio.
(43) Stanley, G. G. In ComprehensiVe Organometallic Chemistry III;
Mingos, D. M., Crabtree, R. H., Canty, A., Eds.; Elsevier: Oxford,
2007; Vol. 1, pp 119-140.
(44) Jessop, P. G.; Morris, R. H. Coord. Chem. ReV. 1992, 121, 155.
(45) In contrast to many other transition metals, palladium does not form
η2-dihydrogen complexes sufficiently stable to be detected at room
temperature, but their existence was supported by theoretical
calculations: Kubas, G. J. Organomet. Chem. 2001, 635, 37, and
references therein.
Table 9. Selected Bond Lengths (Å) and Angles (°) for 14
Pd1-P1
2.3096(7)
2.3082(7)
2.5448(5)
C15-P1-Pd1-Br1 -90.6(1)
C39-P2-Pd1-Br1 -90.1(1)
Pd1-P2
Pd1-Br1
C5-P1-Pd1-Br1
144.0(1)
P1-Pd1-P2
P1-Pd1-Br1
P2-Pd1-Br1
C1-P1-Pd1-Br1
C25-P2-Pd1-Br1
166.89(3)
C29-P2-Pd1-Br1 142.7(1)
96.12(2)
96.99(2)
23.7(1)
23.2(1)
C15-P1-P2-C39
C1-P1-P2-C29
C1-P1-P2-C25
C5-P1-P2-C29
174.9(1)
169.2(2)
49.4(2)
-70.2(1)
9
15558 J. AM. CHEM. SOC. VOL. 130, NO. 46, 2008

Pd-Catalyzed Formylation of Aryl Bromides
A R T I C L E S
Table 10. Hydrogenolysis of Acylpalladium Complexes 8-10a
entry
complex
T (°C)
TMEDA (equiv)
ligand (equiv)
conversion (%)
yield ArCHO (%)
main organometallic products
1
2
3
4
5
6
7
8
8
8
8
8
8
8
8
8
8
24
24
24
50
50
50
50
50
50
50
50
50
50
100
100
100
100
100
100
100
100
100
100
100
100
100
0
4
20
0
0
4
100
4
4
100
0
2
0
0
4
20
100
0
4
4
0
0
0
0
4
0
0
4
12
0
0
0
2
0
0
0
0
4
4
8
0
0
4
4
2
0
0
79
97
0
11
100
100
100
83
100
31
0
34
56
0
-
nd
17, 14 (13.2:1)b
-
nd
17
0
76
77
73
54
74
17
0
11
25
43
62
73
61
67
70
28
67
68
74
57
52
17
17, 14 (4:1)b
17, 14 (1:1.3)b
nd
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
9
10
10
10
8
8
8
8
8
8
8
9
9
9
9
10
10
nd
100
28
16 (precipitate)
nd
100
100
100
100
100
100
100
100
100
100
100
100
100
14+ Pd black
Pd black + P(1-Ad)2nBu
Pd black + P(1-Ad)2nBu
Pd black + P(1-Ad)2nBu
14 + Pd black
14 + Pd black
14 + Pd black
nd
0
4
0
4
0
2
Pd black + PtBu2nBu
15+ Pd black
15+ Pd black
Pd black + PtBu3
Pd black + PtBu3
a Reaction conditions: 5 bar of CO/H2 (1:1), toluene-d8, 100 °C, 16 h. Conversions and yields were measured by 1H NMR spectra using
1,4-dimethoxybenzene as internal standard. b The ratio of the signals in the 31P{1H} NMR spectra.
Figure 15. Possible pathways for the hydrogenolysis of acylpalladium bromide complexes 8 and 9.
monomers that reversibly react with hydrogen, resulting in η2-
dihydrogen complexes.45 These complexes can give rise to the
aldehyde via four different pathways. Pathway A includes direct
σ-metathesis of H-H and Pd-C bonds, furnishing the aldehyde
in one step. Pathway B consists of formation of an intermediate
dihydride Pd(IV) species,46 followed by reductive elimination
of aldehyde. Moreover, the η2-dihydrogen complex might
undergo a base-assisted heterolytic cleavage of the coordinated
dihydrogen to give the aldehyde directly (pathway C) or via
formation of an anionic intermediate (pathway D).
η2-coordinated dihydrogen has increased acidity. For example,
η2-dihydrogen ruthenium complexes are deprotonated even with
mild bases, resulting in the corresponding monohydride com-
plexes.47 This process is believed to be involved in base-
mediated ruthenium-catalyzed hydrogenation.48 In addition,
intramolecular heterolytic cleavage of η2-dihydrogen was
proposed as a plausible pathway for the hydrogenolysis of acyl
carboxylato palladium complexes. However, this pathway is
difficult to distinguish from A and B.49 So far, to the best of
Clearly, pathway A is not consistent with the participation
of the base. Pathway B includes formation of highly unstable
dihydrido spices of palladium(IV) that should rapidly undergo
reductive elimination; therefore, it is unlikely that base can
mediate this process. In contrast, pathways C and D, including
heterolytic cleavage of the coordinated dihydrogen, agree with
the accelerating role of the base. It is well known that
(47) (a) Tye, J. W.; Hall, M. B. In ActiVation of Small Molecules.
Organometallic and Bioinorganic PerspectiVes; Tolman, W. B., Ed.;
Wiley-VCH: Weinheim, 2006; pp 121-150. (b) Henderson, R. A.
Transition Met. Chem. 1988, 13, 474. (c) Esturelas, M. A.; Oro, L. A.
Chem. ReV. 1998, 98, 577. (d) Kubas, G. J. Metal Dihydrogen and
σ-bond Complexes; Kluwer Academic: New York, 2001; pp 268-
295.
(48) See for example: (a) Kitamura, M,; Noyori, R. In Ruthenium in Organic
Synthesis; Murahashi, S.-I., Ed.; Wiley-VCH: Weinheim, 2004; pp
3-52. (b) Stanley, G. G. In ComprehensiVe Organometallic Chemistry
III; Mingos, D. M., Crabtree, R. H., Parkin, G., Eds.; Elsevier: Oxford,
2007; Vol. 1, pp 119-163.
(46) It is known that η2-dihydrogen complexes are often in equilibrium
with their dihydrido tautomers. For more information, see ref 47.
9
J. AM. CHEM. SOC. VOL. 130, NO. 46, 2008 15559

A R T I C L E S
Sergeev et al.
our knowledge, no base-assisted intermolecular heterolytic
cleavage of dihydrogen is reported for palladium acyl complexes
nor for the corresponding complexes of any other transition
metal. In view of this, the base-mediated hydrogenation of
complexes 8 and 9 seems rather remarkable.
a TMEDA/Pd ratio of 100:1 (close to the real catalytic reaction)
and a combination of TMEDA and phosphine ligand (Table
10, entries 17, 20, and 24).
To our surprise, at 100 °C the less reactive PtBu3-ligated acyl
complex 10 also became prone to hydrogenolysis and gave
4-trifluoromethylbenzaldehyde in moderate yields (52-57%;
Table 10, entries 25-26). Moreover, in the presence of 2 equiv
of phosphine or TMEDA the product yields are similar (57 and
52%), suggesting that formation of the catalytically inactive
complex 16 is reversible at 100 °C (eq 14). The discrepancy
between stoichiometric hydrogenation (52-57%) and catalytic
formylation (4%) is explained by the larger ratio of TMEDA
to palladium (300:1) used in the catalytic reaction. Appar-
ently, the reversible formation of the active complex 10 is
efficiently blocked under catalytic conditions. Indeed, close in-
spection of a catalytic formylation in the presence of 1 mol %
of [Pd(PtBu3)2] after 1.5 h revealed formation of a precipitate
of 16 and PtBu3 as the only phosphorus-containing compound
in solution.
In contrast to 8 and 9, the PtBu3-ligated acyl complex 10 is
completely inactive under similar conditions (toluene, TMEDA,
5 bar of synthesis gas, 50 °C; Table 10, entry 12). Under these
conditions, 10 formed the diamino complex 16 via displacement
of PtBu3 (eq 14). In a separate experiment we showed that
addition of 2 equiv of TMEDA to a toluene solution of 10 led
to a precipitate of pale yellow crystals of 16 in 98% yield. A
similar substitution of PtBu3 in arylpalladium complexes [Pd(Ar)-
(Br)(PtBu3)] was observed by Hartwig and co-workers.50
6. The Fate of Palladium Species after Hydrogenolysis:
Regeneration of Palladium(0) Species and Catalyst Stability.
Hydrogenolysis of the acylpalladium complex 8 in the presence
of TMEDA gave the carbonyl complex 17 as the sole palladium-
containing product, displaying a singlet at 46.6 ppm in the
31P{1H} NMR spectrum of the reaction mixture (eq 15).
The obtained data for the stoichiometric hydrogenolysis
experiments of acyl complexes 8 and 10 are in good agreement
with results of the catalytic formylation (eq 1). Indeed, poor
activity of the Pd/PtBu3 system is explained by the formation
n
of the insoluble diamino complex 16. In contrast, P(1-Ad)2 Bu,
The same carbonyl species formed quantitatively when
palladium(0) complex 1 was reacted with CO in toluene at room
temperature (eq 16). This carbonylation proceeded rapidly with
liberation of 1 equiv of phosphine ligand. Heating of the same
reaction mixture in the absence of CO atmosphere gave the
starting [PdL2] complex, indicating the reversibility of this
reaction (eq 16).
PtBu2 Bu ligated complexes 8 and 9 are stable toward the amine.
n
Hence, they operate under catalytic reaction conditions, where
300-fold excess of the base is used. Furthermore, an increase
in base concentration promotes the hydrogenolysis. The smooth
hydrogenolysis of complexes 8 and 9 explains the rather low
pressure of synthesis gas (5 bar) needed for the catalytic
n
formylation in the presence of Pd/P(1-Ad)2 Bu. For comparison,
the previously reported catalyst based on [Pd(PPh3)Cl2] enabled
formylation of bromobenzene with synthesis gas at 93 bar at
150 °C,5 which is in agreement with the high pressure of
hydrogen (60 bar at 60 °C) needed for hydrogenolysis of trans-
[Pd(Br){C(O)Ph}(PPh3)2].15
To compare the reactivity of the acyl complexes under
reaction conditions, we further studied hydrogenation of 8-10
at 100 °C (Table 10, entries 14-26). At 100 °C acyl complexes
8 and 9 undergo hydrogenolysis in the absence of any additives
to give the aldehyde in low yields of 25-28% (Table 10, entries
14 and 21).51 However, addition of TMEDA led to a significant
increase in the aldehyde yield (Table 10, entries 14-17, 21-22).
In general, at 100 °C the product yields are somewhat lower
compared to those at 50 °C (Table 10, entries 6-8 and 15, 17,
19), probably due to concurrent decomposition of the starting
complex. Indeed, addition of 4 equiv of phosphine ligand, which
stabilizes the catalyst system, considerably improved the yields
(Table 10, entries 14 and 18-20, 21, 23). Finally, the best yields
for hydrogenolysis of complexes 8 and 9 are obtained by using
Complex 17 was isolated from the reaction mixture by
precipitation with heptane in 73% yield. Although we failed to
obtain single crystals suitable for X-ray analysis, 17 was
characterized by IR and 1H, 31P{1H}, and 13C{1H} NMR
spectra. 17 contains one type of coordinated phosphine ligand,
as evidenced by a singlet at 46.6 ppm in the 31P{1H} NMR
spectrum. The presence of a bridged CO group is indicated by
a strong band at 1806 cm-1 in the IR-ATR spectrum52 and a
broad singlet at δ 241 ppm in the 13C{1H} NMR.53 These data
clearly suggest a multinuclear cluster structure.54 Carbonyl
complex 17 is thermally unstable and completely decomposes
to give palladium black and free ligand when heated under
catalytic reaction conditions. Since palladium(0) complex 1
(52) Most carbonyl complexes having bridge carbonyls exhibit a ν(CO)
band at 1800-1900 cm-1: Nakamoto, K. Infrared and Raman Spectra
of Inorganic and Coordination Compounds; Wiley-Interscience: New
York, 1986; pp 291-301.
(49) Yamamoto, A.; Kayaki, Y.; Nagayama, K.; Shimizu, I. Synlett 2000,
925.
(50) Shen, Q.; Shekhar, S.; Stambuli, J. P.; Hartwig, J. F. Angew. Chem.,
Int. Ed. 2004, 44, 1371.
(53) The complex exhibits dynamic behavior in the 13C{1H} NMR
spectrum. Detailed investigation of solution structure of this complex
was not performed.
(51) This shows that at 100 °C hydrogenolysis of complexes 8 and 9 can
partially proceed via base-free pathways A and B.
9
15560 J. AM. CHEM. SOC. VOL. 130, NO. 46, 2008

Pd-Catalyzed Formylation of Aryl Bromides
A R T I C L E S
rapidly transforms into 17 in CO atmosphere, the stability of 1
under the reaction conditions is the same as that of 17. Therefore,
it is not surprising that hydrogenation of complexes 8 and 9
carried out in the presence of TMEDA at 100 °C yielded
palladium black and free phosphine. The same result was
obtained for complex 10 when the reaction was carried out in
the presence of 2 equiv of TMEDA or PtBu3 (eq 17).
the other hand, the reaction of [Pd(PtBu3)2] with TMEDA·HBr
in toluene-d8 at 70 °C did not proceed to any extent (eq 20).
These results imply that the reactivity of these hydrido
complexes toward the base is governed by thermodynamics,
but not kinetics. Therefore, the reason for the pronounced
difference in the reactivity between 14, 15, and 18 may be
revealed by comparison of the ground states of the [PdL2] and
[Pd(Br)(H)L2] complexes. It is evident that PtBu3 is more
sterically demanding than PtBu2 Bu, and probably P(1-Ad)2 Bu.
Since the hydridobromide complexes are more sterically en-
cumbered as compared to the corresponding bisphosphine
palladium(0) complexes, the increased steric bulk of PtBu3
facilitates reductive elimination from 18. In addition, the
divergent reactivity may be caused by the different electron-
donating properties of the phosphines. Clearly, electron-rich
phosphines stabilize a hydrido palladium(II) complex and
increase the basicity of the respective palladium(0) complex,
thus making dehydrobromination less favorable. To evaluate
the σ-donor ability of the phosphines, we compared phosphorus-
selenium (1JPSe) coupling constants of the corresponding phos-
n
n
A completely different reaction pattern was observed when
the hydrogenation of acyl complexes 8 and 9 was effected in
the presence of additional phosphine and TMEDA (catalytic
conditions). Here, the corresponding hydride complexes 14 and
15 are formed as main products. This suggests that hydride
species 14 and 15, unlike the respective carbonyl clusters, are
stable at 100 °C for 16 h under synthesis gas. Moreover, it is
rather surprising that these complexes were observed even when
an excess of base was used (eq 18)!
1
phine selenides. The magnitude of JPSe is known to increase
with decrease in basicity of the phosphine.56 We prepared
n
n
selenides of PtBu2 Bu and P(1-Ad)2 Bu by stirring of the
phosphines with 2 equiv of Se in CDCl3 at 60 °C for 40 min.
The resulting phosphine selenides formed quantitatively, and
1
the coupling constants JPSe were measured without isolation.
Interestingly, 1JPSe values obtained for PtBu2 Bu (690 Hz) and
n
P(1-Ad)2 Bu (680 Hz) are lower than that of PtBu3 (712 Hz),57
n
indicating that the first two phosphines are more basic. This
may imply that these ligands form more stable hydrido
complexes 14 and 15 with respect to the PtBu3-ligated com-
1
plexes. However, care must be taken in correlating JPSe
1
constants with electronic properties of phosphines since JPSe
also may depend on steric factors.58 Thus, the poor reactivity
Intrigued by these results, we examined the reactivity of
n
n
of PtBu2 Bu- and P(1-Ad)2 Bu-ligated hydrobromide complexes
toward the amine bases as compared to 18 may be explained
by both steric and electronic factors.
n
hydrido palladium complexes [Pd(H)(Br)L2], L) P(1-Ad)2 Bu
n
(14), PtBu2 Bu (15), and PtBu3 (18), toward TMEDA in detail.
When 14 or 15 was mixed with a 2-fold excess of TMEDA in
toluene-d8 at room temperature, the 1H and 31P{1H} NMR
spectra showed no appreciable reaction. Moreover, neither
heating of 14 with 2 equiv of TMEDA at 100 °C nor use of a
100-fold excess (!) of base at room temperature showed any
formation of palladium(0) complex 1 in the 31P{1H} NMR
spectra (eq 19).
Palladium hydride complexes are assumed to be crucial
intermediates in many palladium-catalyzed coupling reactions
of aryl halides.59 In general, the main role of base in these
reactions consists of trapping the hydrogen halide from [Pd(H)-
(54) Most phosphine-containing palladium carbonyl clusters possesses from
3 to 69 palladium atoms: (a) Cavell, K. J. In ComprehensiVe
Organometallic Chemistry III; Mingos, D. M., Crabtree, R. H., Canty,
A., Eds.; Elsevier: Oxford, 2007; Vol. 8, pp 206-210. (b) Kudo, K.;
Hidai, M.; Uchida, Y. J. Organomet. Chem. 1971, 33, 393. (c) Yoshida,
T.; Otsuka, S. J. Am. Chem. Soc. 1977, 99, 2134. (d) Mednikov, E. V.;
Eremenko, N. K.; Mikhalkov, V. A.; Gubin, S. P.; Slovokhotov, Y. L.;
Struchkov, Y. T. J. Chem. Soc., Chem. Commun. 1981, 989. (e)
Goddard, R.; Jolly, P. W.; Kru¨ger, C.; Schick, K.-P.; Wilke, G.
Organometallics 1982, 1, 1709. (f) Manojloviæ-Muir, L.; Muir, K. W.;
Lloyd, B. R.; Puddephatt, P. J. J. Chem. Soc., Chem. Commun. 1985,
536. (g) Klein, H.-F.; Mager, M. Organometallics 1992, 11, 3915.
(h) Tran, N. T.; Kawano, M.; Dahl, L. F. J. Chem. Soc., Dalton Trans.
2001, 2731.
In contrast to 14 and 15, the reaction of the PtBu3-ligated
hydrido palladium complex 18 with TMEDA proceeded smoothly
within 10 min at room temperature to give the respective
palladium(0) complex quantitatively (eq 20).
(55) At room temperature reaction does not occur due to very low solubility
of TMEDA ·HBr in toluene.
(56) (a) Allen, D. W.; Taylor, B. F. J. Chem. Soc., Dalton Trans. 1982,
51. (b) Socol, S. M.; Verkade, J. G. Inorg. Chem. 1984, 23, 3487. (c)
Sua´rez, A.; Me´ndez-Rojas, M. A.; Pizzano, A. Organometallics 2002,
21, 4611.
(57) Du Mont, W.-W.; Kroth, H.-J. J. Organomet. Chem. 1976, 113, C35.
(58) Pinnell, R. P.; Megerle, C. A.; Manatt, S. L.; Kroon, P. A. Inorg.
Chem. 1973, 95, 977.
The same result was obtained when the reactions depicted in
esq 19 and 20 were effected from opposite directions. For
example, heating of 1 with TMEDA·HBr in toluene-d8 at 70
°C55 gave the hydrido complex 14 in 93% yield (eq 19). On
(59) Reviews on palladium hydride complexes: (a) Grushin, V. V. Chem.
ReV. 1996, 96, 2011. (b) Hii, K. K. In Handbook of Organopalladium
Chemistry for Organic Synthesis; Negishi, E. I., Ed.; Wiley-Inter-
science: New York, 2002; Vol. 1, pp 81-90.
9
J. AM. CHEM. SOC. VOL. 130, NO. 46, 2008 15561

A R T I C L E S
Sergeev et al.
Table 11. Test of Isolated Complexes in Catalytic Reactions
(Hal)L2] species to regenerate the active palladium(0) catalyst.
The importance of such dehydrohalogenations was clearly
illustrated by Fu and co-workers, who showed that the low
catalytic activity of Pd/PCy3 compared to Pd/PtBu3 in the
Mizoroki-Heck vinylation of aryl chlorides is attributed to the
formation of a stable hydrido palladium complex, [Pd(H)-
(Cl)(PCy3)2], that is reluctant to react with the base used in
catalytic reaction.60 In contrast, we observed here that the
palladium-catalyzed formylation of aryl bromides displays a
completely different trend: the most efficient catalytic systems
generate the most stable palladium hydride complexes 14 and
15. To ensure that these hydride complexes are relevant to
catalysis, we further studied their generation and ability to
catalyze the formylation. As shown above, hydrogenolysis of
the acyl complex 8 in the presence of the base at 50 °C gave
carbonyl cluster 17 as the only organometallic product (eq 17,
Table 10, entry 6). In the presence of phosphine the reaction
slowed down and a mixture of 14 and 17 was formed. The ratio
14 to 17 increased with increasing concentration of the
phosphine ligand (Table 10, entries 6-8). At 100 °C the
hydrobromide complex 14 was obtained as the major palladium-
containing species, providing the phosphine ligand was added
(eq 18). These results suggest that hydride complex 14 is not
necessarily the direct product of hydrogenolysis, but a result of
a subsequent transformation of carbonyl complex 17. Indeed,
17 was completely transformed into 14 under conditions similar
to those used for the catalytic reaction (toluene-d8 solution, 1
equiv of phosphine, excess of TMEDA ·HBr under CO/H2
atmosphere at 100 °C for 16 h) (eq 21).
Pd/L conversion of
ArBr, %a
yield of
ArCHO, %a
selectivity,
%
expt
complex
Pd(OAc)2/L
[PdL2] (1)
[Pd(Br)(Ar)L]2(3)
[Pd(Br)(Ar)L]2(3)/2L
[Pd(Br){C(O)Ar}L]2(8)
ratio
1
2
3
4
5
6
7
8
1:3
1:2
1:1
1:2
1:1
100
100
94
100
93
86
87
80
85
82
84
87
87
86
87
85
85
88
84
87
87
[Pd(Br){C(O)Ar}L]2(8)/2L 1:2
100
100
100
[Pd(Br)(H)L2](14)
[Pdn(CO)mLn](17)/nL
1:2
1:2
b
a Conversion and yield were determined by GC using hexadecane as
internal standard. b The catalyst of this composition was obtained by
bubbling CO via solution of [PdL2] in toluene.
4-bromobenzotrifluoride in the presence of 2 equiv of TMEDA
in toluene-d8 to give the oxidative addition product 6. However,
this reaction proceeded under more drastic conditions (>1.5 h
at 100 °C) as compared to the oxidative addition of the
respective palladium(0) complex 1 (1.5 h at 70 °C). The
oxidative addition product 7 is not observed when 14 is heated
with 4-bromobenzotrifluoride in the absence of base.63 It appears
that regeneration of palladium(0) from the palladium hydride
species in the presence of the base and subsequent oxidative
addition are the rate-determining steps. These transformations
require at least 100 °C to proceed with reasonable rate, while
all steps of the catalytic cycle from oxidative addition to
hydrogenolysis occur at 70 °C or below. At the same time, the
catalytic reaction was shown to proceed at 100 °C but not at
70 °C.
The data obtained for the stoichiometric studies on the
n
formylation of aryl halides with Pd/P(1-Ad)2 Bu catalyst are
The 31P{1H} NMR spectrum of the filtered reaction mixture
revealed disappearance of 17 and appearance of a singlet at 57.4
ppm belonging to the hydride species 14. Additionally, the 1H
NMR displayed the characteristic high-field triplet at -13.2 ppm
(JPH ) 7.9 Hz) corresponding to the hydride proton of 14.
Notably, this transformation is also reversible.61 It is likely that
interconversion between 14 and 17 occurs via the bisphosphine
palladium complex 1. In fact, 17 reversibly reacts with free
ligand to give 1 (eq 16), which in turn reacts with TMEDA ·HBr
to give 14 (eq 19). Moreover, a mixture of [Pdn(CO)mLn]/L
catalyzed the test reaction, providing the same yield (87%) and
selectivity (87%) as obtained with [PdL2] (Table 11, entry 8).
Apparently, the carbonyl cluster [Pdn(CO)mLn] serves as a
[PdL2] source for the catalytic reaction.62 The same is true for
summarized in Figure 16. An important point of the catalytic
cycle is that hydride complex 14 and cluster compound 17 are
considered as sources for catalytically active [PdL]/[PdL2]
species. Both complexes 14 and 17, being in equilibrium,
maintain the concentration of [PdL]/[PdL2] at such a low level
that oxidative addition becomes the rate-determining step of the
reaction. Indeed, when the formylation of 4-bromobenzotrif-
luoride catalyzed by 1 mol % of 1 is stopped after 1.5 h at 61%
conversion, the 31P{1H} NMR spectrum displayed three main
n
resonances corresponding to 14, 17, and P(1-Ad)2 Bu in a
1.1:1:2.1 ratio.64 In other words, complexes 14 and 17 are
“reservoirs” of highly active [PdL] species, releasing them in
low concentration throughout the reaction, thus making the Pd/
n
P(1-Ad)2 Bu catalytic system highly efficient. The same is also
n
the bromohydride complex [Pd(Br)(H){P(1-Ad)2 Bu}2] (14).
n
true for the formylation catalyzed by Pd/PtBu2 Bu. The impor-
Although being “inert” toward the amine as shown by NMR,
14 is nevertheless able to generate highly reactive palladium(0)
species in the presence of TMEDA. Hence, complex 14
catalyzed the formylation of 4-bromobenzotrifluoride to give
the desired aldehyde in 87% yield, which is comparable (86%)
to the original catalytic system comprising Pd(OAc)2/P(1-
(62) Generally, in catalytic cycles of carbonylation reactions carbonyl
palladium(0) complexes are postulated to be monomeric and consid-
ered as direct intermediates involved in oxidation addition. See for
example: (a) Ozawa, F.; Kawasaki, N.; Okamoto, H.; Yamamoto, T.;
Yamamoto, A. Organometallics 1987, 6, 1640. (b) Carpentier, J.-F.;
Castanet, Y.; Mortreux, A.; Petit, F. J. Organomet. Chem. 1994, 482,
31.
n
Ad)2 Bu (Table 10, entry 7). Compound 14 also reacted with
(63) In contrast, recently Hartwig reported that oxidative addition of
bromobenzene to hydride complex [Pd(Br)(H)(PtBu3)2] proceeds even
faster than to [Pd(PtBu3)2], see ref 25.
(60) Hills, I. D.; Fu, G. C. J. Am. Chem. Soc. 2004, 126, 13178.
(61) According to 31P{1H} NMR spectrum, the reaction of 14 with 4 equiv
of TMEDA in toluene-d8 at 5 bar pressure of CO/H2 at room
temperature after 16 h yielded a mixture of carbonyl compound 17,
free phosphine ligand, and starting complex 14 as indicated by three
singlets in at 46.6, 24.4 and 57.4 ppm in a 1:1:4.65 ratio.
(64) When the reaction was carried out with Pd(OAc)2/P(1-Ad)2nBu
additional resonances were observed. These could be attributed to some
complexes containing coordinated acetoxy ligand. See for example
ref 65e.
9
15562 J. AM. CHEM. SOC. VOL. 130, NO. 46, 2008

Pd-Catalyzed Formylation of Aryl Bromides
A R T I C L E S
aryl halides with synthesis gas. This methodology has been
recently applied by industry, and we have elucidated the
mechanism of this synthetically useful reaction. More specif-
ically, most relevant catalytic intermediates have been
unambiguously characterized, and the synthesis and solution
n
n
chemistry of complexes bearing P(1-Ad)2 Bu, PtBu2 Bu and
PtBu3 ligands are described in detail. Using Pd/P(1-Ad)2 Bu,
n
the most efficient catalyst known to date for formylation of
aryl bromides, we compared results of stoichiometric studies
with those of the catalytic reaction. The data obtained imply
that [Pdn(CO)mLn] (17) and [Pd(Br)(H)L2] (14) are resting
states of the active catalyst and probably may not lie within
the catalytic cycle. During the reaction course, complexes
14 and 17 keep the concentration of the catalytically active
species at a low level, which has two consequences: (1)
oxidative addition of the aryl bromide becomes the rate-
determining step of the reaction and (2) the high efficiency
n
and longevity of the Pd/P(1-Ad)2 Bu catalyst system. Nota-
bly, the product-forming step, hydrogenolysis of the acyl
n
complex [Pd(Br)(p-CF3C6H4CO){P(1-Ad)2 Bu}]2, is signifi-
cantly accelerated by a base and occurs under mild conditions
(25 °C, 5bar of CO/H2), thus making the catalytic reaction
n
in the presence of Pd/P(1-Ad)2 Bu feasible at lower pressures
Figure 16. Proposed catalytic cycle for formylation of aryl bromides with
Pd/P(1-Ad)2nBu.
compared to all other previously reported catalysts. These
conclusions are also valid for the catalyst comprising a similar
unsymmetrical alkylphosphine ligand, PtBu2 Bu. Addition-
n
tance of maintaining a low concentration of palladium(0) species
for catalyst longevity can be rationalized as follows: the
efficiency of the Pd(0) catalyst is dictated by competition of
oxidative addition and decomposition via agglomeration of Pd(0)
species, leading eventually to the formation of palladium black.
The rate of this latter process is second order in Pd or higher,
whereas oxidative addition is usually first order in pal-
ladium(0).65 Therefore, catalyst systems maintaining concentra-
tion of active palladium(0) species on a low level during the
respective reaction often possess prolonged existence.66
In contrast, the low activity of the Pd/PtBu3 system in the
formylation of aryl bromides is attributed to formation of the
poorly soluble complex [Pd(Br){C(O)Ar}(TMEDA)] (16). The
low longevity of the Pd/PtBu3 catalyst is a consequence of
the limited stability of the corresponding palladium(0) com-
plexes in presence of base and carbon monoxide.
ally, stoichiometric studies of the poorly active Pd/PtBu3
catalytic system led to the synthesis of the first stable three-
coordinate neutral palladium acyl complex. This complex is
n
less prone to hydrogenolysis as compared to P(1-Ad)2 Bu-
and PtBu2 Bu-ligated acyl complexes and, in the presence
n
of amine base, gives a catalytically inactive diamine acyl
complex, which results in poor performance of the Pd/PtBu3
catalyst in the catalytic formylation. Finally, this study led
to elucidation of the special role of unsymmetrical trialkyl-
phosphine ligands, which have been largely neglected in
coupling chemistry.
Acknowledgment. This research was supported by the State
of Mecklenburg-Vorpommern, the BMBF, the DFG (Leibniz Prize),
and the Fonds der Chemischen Industrie (FCI). We thank Dipl.
Ing. A. Koch and Ms. S. Giertz (LIKAT) for their excellent
analytical and technical support.
Conclusion
In summary, we report the first comprehensive study on
the catalytic cycle of a palladium-catalyzed formylation of
Supporting Information Available: All experimental proce-
dures and spectroscopic data of new compounds. This material
is available free of charge via the Internet at http://pubs.acs.org.
(65) See for example: (a) Fitton, P.; Rick, E. A. J. Organomet. Chem. 1971,
28, 287. (b) Fauvarque, F.-F.; Pflu¨ger, F.; Troupel, M. J. Organomet.
Chem. 1981, 208, 419. (c) Amatore, C.; Pflu¨ger, F. Organometallics
1990, 9, 2276. (d) Hartwig, J. F.; Paul, F. J. Am. Chem. Soc. 1995,
117, 5373. (e) Amatore, C.; Carre´, E.; Jutand, A.; M′Barki, M. A.
Organometallics 1995, 14, 1818. (f) Alami, M.; Amatore, C.;
Bensalem, S.; Choukchou-Brahim, A.; Jutand, A. Eur. J. Inorg. Chem.
2001, 2675.
JA804997Z
(66) (a) Beletskaya, I. P.; Cheprakov, A. V. J. Organomet. Chem. 2004,
689, 4055. (b) Dupont, J.; Consorti, C. S.; Spencer, J. Chem. ReV.
2005, 105, 2527.
9
J. AM. CHEM. SOC. VOL. 130, NO. 46, 2008 15563