Transition metal and hydrogen bond donor hybrids: Catalysts for the activation of alkylidene malonates

Article abstract of DOI:10.1002/chem.201201206

Palladium ureaka! Urea palladacycles are introduced to activate alkylidene malonates for nucleophilic attack. The strategic incorporation of palladium on a urea scaffold give rise to urea catalysts with enhanced reactivity when compared to conventional ur

Full text of DOI:10.1002/chem.201201206

DOI: 10.1002/chem.201201206
Transition Metal and Hydrogen Bond Donor Hybrids:
Catalysts for the Activation of Alkylidene Malonates
David M. Nickerson and Anita E. Mattson*[a]
Advances within synthetic organic chemistry have largely
led to the development of transition metal catalysis and or-
ganic catalysis into two distinct catalyst systems.[1,2] The
combination of these two catalyst systems is a relatively
recent, new direction enabling access to useful bond-forming
processes that are inaccessible with either catalyst system
alone.[3,4] Driven by the advantageous prospect of combined
catalyst systems, we anticipated the design of hybrid cata-
lysts: single molecules containing both transition metal and
organocatalytic sites. More specifically, we reasoned that the
strategic merger of transition metals and hydrogen bond
donor (HBD) catalysts would present opportunities for the
development of single catalyst systems benefitting from en-
hanced activity and unprecedented reactivity. This account
describes the successful and strategic incorporation of palla-
dium into a urea scaffold to generate urea palladacycle cata-
lysts (1) for the activation of alkylidene malonates.
Figure 1. Urea palladacycles as hybrid transition metal and hydrogen
bond donor catalysts.
5) the promising nature of future, bifunctional catalysis in-
volving the enhanced hydrogen bonding site and the transi-
tion metal center. To explore the promise of urea palladacy-
cles as a new family of hybrid catalysts, we set out to probe
their ability to catalyze the conjugate addition of indole to
alkylidene malonates.
The inspiration for the development of urea palladacycle
catalysis grew from the separate and exceptional successes
in the catalytic development of the two key components:
palladium and the hydrogen bonding functionality of the
urea. Individually, the fields of palladium catalysis and urea
catalysis have allowed significant advances in synthetic
chemistry, and we reasoned a hybrid palladium–urea catalyst
might offer an opportunity to coalesce the best of both
fields. Our urea palladacycle design (1) was inspired by bor-
onate ureas (Figure 1), a recently disclosed family of tunable
urea catalysts that exhibit enhanced activity when compared
with conventional urea catalysts.[5–7] This improvement in ac-
tivity is proposed to arise from the increased polarization of
the urea functionality as a result of internal Lewis acid coor-
dination. We speculated that similarly placed transition
metals on the urea scaffold would give rise to hybrid cata-
lysts with interesting reactive properties. We were particular-
ly excited to explore several specific features of our new
hybrid catalysts including: 1) enhanced hydrogen bond
donor catalyst activity; 2) tunable ligands and their effect on
catalyst reactivity; 3) ease of preparation; 4) stability, and
The activation of alkylidene and arylidene malonates for
nucleophilic attack is a useful synthesis technique typically
achieved with traditional Lewis acid catalysts like Cu-
AHCTUNGTRENNUNG
(OTf)2.[8] Conventional ureas and thioureas, although easily
able to activate nitroalkenes[9] (2) and a,b-unsaturated
imides[10] (3) through proposed hydrogen bonding interac-
tions, have rarely been reported as catalysts for alkylidene
malonate (4) activation (Figure 2).[11] These few reports rely
on bifunctional urea or thiourea catalysts that may operate
to activate both the nucleophile and electrophile; no simple
ureas have been reported in the activation of alkylidene
malonates. Prompted by this lack of reactivity, we consid-
ered this process an ideal testing ground to study urea palla-
[a] D. M. Nickerson, Dr. A. E. Mattson
Department of Chemistry, The Ohio State University
100 W. 18th Ave., Columbus, OH 43210 (USA)
Fax : (+1)614-292-1685
E-mail: mattson@chemistry.ohio-state.edu
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201201206.
Figure 2. Urea palladacycle hybrid catalysts for the activation of alkyli-
dene malonates.
8310
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 8310 – 8314

COMMUNICATION
dacycle catalysis. If our hypothesis that urea palladacycles
would have enhanced activity compared to conventional
ureas held true, then we would be able to effect a transfor-
mation that is difficult to achieve with traditional urea and
thiourea catalysts, such as the activation of alkylidene malo-
nates (5). With this in mind, investigations were initiated by
using urea palladacycles (1) as enhanced HBDs to catalyze
the addition of indoles (7) to alkylidene malonates (6).
Our early studies focused on the addition of indole (7a)
to dimethyl 2-(cyclohexylmethylene)malonate (6a) in the
presence of 15 mol% urea palladacycle (1a, Table 1). The
palladacycle (1a) was rapidly synthesized in gram quantities
by using procedures adapted from the literature.[12,13] Initial
results confirmed the formation of the desired bonds and
after optimization of the reaction conditions, a 69% yield of
desired product 8a was observed after 24 h at 508C in tolu-
ene (entry 1). The proposed mode by which 1a activates 6a
is through hydrogen bonding to the 1,3-dicarbonyl function-
ality, perhaps through a species similar to that depicted as 9.
Solvent effects were substantial in this process. After ach-
ieving only modest yields with chloroform and dioxane (47
and 72%, respectively, entries 2 and 3), methanol was found
to provide nearly quantitative yields of 8a (99%, entry 4).
At this time we propose that the dramatic solvent effects
are linked to catalyst solubility: the catalyst is sparingly
soluble in aprotic solvents, such as chloroform, 1,4-dioxane
and toluene, whereas the urea palladacycle is easily dis-
solved in methanol. With an excellent solvent identified,
a control experiment to determine the background reaction
rate in optimized conditions showed no observable reactivity
of the alkylidene malonate with indole nucleophile in the
absence of 1a (entry 9). Convinced that our palladacycle
urea was indeed catalyzing the reaction, we proceeded to in-
vestigate the limits of the reaction with respect to catalyst
loading. Reducing the loading of 1a by half to 7.5 mol% af-
forded a 91% yield of 8a after 24 h (entry 5). Even just
2.5 mol% of 1a gave rise to a 70% yield of 8a after 72 h
(entry 8). Given adequate time for completion, reactions at
room temperature also proceeded in high yield with higher
catalyst loadings serving to decrease reaction times. For ex-
ample, at 20 mol%, a reaction at 238C gave a 98% isolated
yield after 96 h (entry 11).
A survey of several catalysts revealed that ligands play an
important role in the reactivity of urea palladacycle catalysts
Table 2. Optimization of reaction conditions for the addition of indole to
6a catalyzed by urea palladacycles (1).[a]
Entry
Catalyst
Yield
[%][b]
Table 1. Optimization of reaction conditions for the addition of indole to
alkylidene malonates catalyzed by urea palladacycle 1a.[a]
1
1a
1b
1c
1d
91
2
3
4
6[c]
NR
24
Entry 1a
[mol%]
Solvent
T
t
Yield
ACHTUNGTRENNUNG
[8C] [h] [%][b]
1
2
3
4
5
6
7
8
15
15
15
15
7.5
5
5
2.5
0
toluene
CHCl3
1,4-dioxane 50
50
50
24 69
24 47
24 72
24 99
24 91
24 82
48 92
72 70
solvent screen
MeOH
MeOH
MeOH
MeOH
MeOH
MeOH
MeOH
MeOH
50
50
50
50
50
50
23
23
catalyst loading study
RT example
5
1e
15[c]
9
10
11
24
0
20
20
48 63
96 98
[a] Reactions were performed by using 1.5 equiv indole at a concentration
of 2m; see the Supporting Information for detailed experimental proce-
dures. [b] Isolated yield unless otherwise noted. [c] Yield obtained by
1H NMR spectroscopy analysis.
[a] Reactions were performed by using 1.5 equiv indole at a concentration
of 2m; see the Supporting Information for detailed experimental proce-
dures. [b] Isolated yield.
Chem. Eur. J. 2012, 18, 8310 – 8314
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8311

A. E. Mattson and D. M. Nickerson
in the addition of indole to 6a (Table 2). A series of urea
palladacycles containing various ligands including 2,2,-bipyr-
idine (1b), 1,2-bis(phenylthio)ethane (1c), and 1,2-bis(di-
phenylphosphino)ethane (DPPE, 1d) were prepared and
compared as catalysts (entries 2–4). Ureas 1b and 1d gave
rise to low yields of product 8a, 6 and 24% yields, respec-
tively, whereas 1c resulted in no reaction after 24 h in meth-
anol at 508C. A chiral urea palladacycle derived from cis-
aminoindanol was also examined and found to afford low
yields of 8a after 24 h at 508C; this suggests that the bistri-
fluoromethylphenyl substituent on the urea working along
with the internal palladium coordination is critical for opti-
mal catalyst activity (15%, entry 5).[13] Initial attempts to
promote enantioselective reactions by using chiral 1e were
not successful.
comparing palladacycle 1a with traditional HBD urea cata-
lysts (10 and 11) revealed that no reaction occurred with
20 mol% of 10 or 11 in otherwise identical reaction condi-
tions. Clearly, palladium was playing a major role in catalyst
reactivity.
In order to support our proposed HBD mode of action
(9), we synthesized methylated control ureas 1 f and 1g
(Scheme 1). We were surprised to see that 1 f, with only
a single available hydrogen-bonding site, still promoted the
reaction, albeit with reduced reactivity. This result initially
suggested the possibility that palladium could be playing
a role as a Lewis acid in its own right; however, testing 1g,
a homologous control urea with no hydrogen bond capabili-
ties resulted in no observable reaction. These results suggest
À
that N H urea components are necessary for catalysis and
The comparison of crystal structures of 1a and 1d helped
us gain a better understanding of the properties of the
newly designed urea palladacycle catalysts (Figure 3).[14] Of
particular interest is the length of the urea carbonyl as it can
be correlated to the urea polarization due to internal Lewis
acid coordination. The urea carbonyl length is slightly
longer in TMEDA catalyst 1a (1.258 ꢁ) versus that of
DPPE catalyst 1d (1.255 ꢁ). Interestingly, these lengths fall
between those of a conventional urea (about 1.23 ꢁ)[15] and
a boronate urea (about 1.28 ꢁ).[6] A more significant differ-
that the reactive intermediate, though benefitting from
a dual hydrogen bond interaction, is also viable with a single
hydrogen bond event.[16] Alternative explanations for the
variations in observed catalytic activity of 1a, 1 f and 1g,
such as conformational constraints present in dimethylated
1g but not 1a and 1 f, have not been explicitly ruled out at
this time and are the source of ongoing studies in our labo-
ratory. All of our attempts to catalyze the reaction with vari-
ous PdII sources or triflic acid resulted in no observed addi-
tion products.[14]
À
ence between 1a and 1d exists when comparing the Pd O
With new urea palladacycle 1a operating catalytically
under the optimized reaction conditions, we set out to deter-
mine how tolerant the process would be with respect to dif-
ferent alkylidene and arylidene malonates (Table 3). A sub-
strate scope of various conjugated diesters derived from di-
methyl malonate were prepared and tested. Substrates de-
rived from aliphatic aldehydes (i.e., cyclohexanecarboxalde-
hyde, hexanal and hydrocinnamaldehyde) provided
excellent yields of desired product (>90%) after 24 h with
10 mol% of 1a (entries 1–3). Diesters derived from aromat-
bond lengths (2.023 vs. 2.108 ꢁ, respectively). Chiral urea
palladacycle 1e crystallized as two unique palladacycle com-
plexes within the asymmetric unit. Its carbonyl bond lengths
À
were close to that of 1a (1.259 and 1.257 ꢁ) and its Pd O
bond lengths were in between those of 1a and 1d (2.029
and 2.037 ꢁ).[14]
Studies to probe the role of the palladium Lewis acid and
its effect on urea activity in the conjugate addition reaction
were initiated next (Scheme 1). An experiment directly
Figure 3. ORTEP representations of 1a, 1d and 1e. Ellipsoids are displayed at 50% probability. For 1e, only one complex from the asymmetric unit is
shown.[13]
8312
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 8310 – 8314

Transition Metal and Hydrogen Bond Donor Hybrids
COMMUNICATION
Scheme 1. Control experiments support the enhanced HBD activity of catalyst 1a.
ic aldehydes were more sluggish in the reactions. A benzal-
dehyde-derived diester gave rise to 46% of 8d after 24 h at
a 20 mol% loading, whereas the 2-furfural derivative pro-
vided a 42% yield of product 8e at 20 mol% loading (en-
tries 4 and 5). Alkylidene malonates containing alcohols
protected with p-methoxybenzyl (PMB) or tert-butyldime-
thylsilyl (TBS) were successfully incorporated into
the process giving rise to the desired adducts 8 f
and 8g in excellent yields (entry 6). The addition of
indole to lactones 6h and 6i occurred in excellent
yield affording product 8i and 8j as >20:1 mixture
of diastereomers favoring a trans relationship be-
tween the two stereocenters (entries 7 and 8).
Table 3. Survey of reaction scope with respect to electrophile 6.[a]
Reaction scope screening was continued with
a survey of nucleophilic heterocycles by assessing
a variety of substituted indoles for reactivity in the
Entry
Substrate
Product
1a
[mol%]
Yield
G
[%][b]
new urea catalyzed process (Table 4). Excellent
yields of product were obtained with most substitut-
ed indoles. N-Methylindole operated well as a nucle-
1
2
3
4
5
6a
6b
6c
6d
6e
8a
8b
8c
8d
8e
10
10
10
20
20
98
ophile, requiring just 5 mol% of 1a to yield 82% of
8j after 24 h in methanol (entry 2). Electron-rich 5-
methoxyindole gave rise to 90% of 8k after 24 h in
methanol (entry 3). As expected, less electron-rich
species, such as 5-chloroindole and methyl indole-4-
carboxylate, required either longer reaction times
or higher catalyst loadings to perform similarly to
more electron-rich species (entries 4 and 5).
94
98
46
42
In summary, urea palladacycles have been dis-
closed as a hybrid class of organometallic hydrogen
bond donors benefitting from enhanced activity.
The improved activity of urea palladacycles over
conventional urea and thiourea hydrogen bond
donor catalysts has enabled the activation of alkyli-
dene malonates for nucleophilic attack by indoles.
The activation of alkylidene malonates has not pre-
viously been achieved with conventional urea cata-
lysts. We are excited about the promise of metallic–
hydrogen bond donor hybrid catalysts and we look
forward to reporting results from our ongoing ef-
forts dedicated toward exploring their potential as
bifunctional catalysts.
6 f, R2 =PMB
6g, R2 =TBS
8 f
8g
10
10
94
96
6
7
6h
6i
8h
8i
10
20
94[c]
8
87[c]
[a] Reactions were performed by using 1.5 equiv indole at a concentration of 2m; see
the Supporting Information for detailed experimental procedures. [b] Isolated yield
unless otherwise noted. [c] 20:1 diastereoselectivity.
Chem. Eur. J. 2012, 18, 8310 – 8314
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8313

A. E. Mattson and D. M. Nickerson
Table 4. Survey of reaction scope with respect to indoles 7.[a]
[6] For ground-breaking examples of internal Lewis acid assisted ureas
in the context of molecular recognition see: a) M. P. Hughes, M. Y.
Shang, B. D. Smith, J. Org. Chem. 1996, 61, 4510; b) M. P. Hughes,
B. D. Smith, J. Org. Chem. 1997, 62, 4492.
[7] For additional recent, interesting examples of alternate strategies to
access hydrogen bond donors with enhanced activity please see:
a) M. T. Robak, M. Trincado, J. A. Ellman, J. Am. Chem. Soc. 2007,
129, 15110; b) M. Ganesh, D. Seidel, J. Am. Chem. Soc. 2008, 130,
16464; c) C. R. Jones, G. D. Pantos, A. J. Morrison, M. D. Smith,
Angew. Chem. 2009, 121, 7527; Angew. Chem. Int. Ed. 2009, 48,
7391.
[8] For examples of Lewis acid catalyzed indole additions to alkylidene
and/or arylidene malonates see: a) W. Zhuang, T. Hansen, K. A. Jor-
gensen, Chem. Commun. 2001, 347; b) J. Zhou, Y. Tang, J. Am.
Chem. Soc. 2002, 124, 9030; c) K. A. Jorgensen, Synthesis 2003, 7,
1117; d) J. Zhou, Y. Tang, Chem. Commun. 2004, 432; e) J. Zhou,
M. C. Ye, Z. Huang, Y. Tang, J. Org. Chem. 2004, 69, 1309; f) S. Ya-
mazaki, Y. Iwata, J. Org. Chem. 2006, 71, 739; g) Y. Sun, N. Li, Z.
Zheng, L. Liu, Y. Yu, Z. Qin, B. Fun, Adv. Synth. Catal. 2009, 351,
3113; h) Y. Liu, D. Shang, X. Zhou, X. Liu, X. Feng, Chem. Eur. J.
2009, 15, 2055; i) Y. Liu, X. Zhou, D. Shang, X. Liu, X. Feng, Tetra-
hedron 2010, 66, 1447; j) L. Wen, Q. Shen, X. Wan, L. Lu, J. Org.
Chem. 2011, 76, 2282.
Entry
7
Product 1a [mol%] Yield [%][b]
1
2
3
4
5
indole
8a
8j
8k
8l
10
5
10
15
20
98
82
90
N-methylindole
5-methoxyindole
5-chloroindole
87
methyl indole-4-carboxylate 8m
58[c]
[a] Reactions were performed by using 1.5 equiv indole 7 at a concentra-
tion of 2m; see the Supporting Information for detailed experimental
procedures. [b] Isolated yield. [c] 48 h reaction time.
Acknowledgements
Support for this work was provided by the OSU Department of Chemis-
try. We are grateful to have insightful discussions with Professor James
Stambuli (OSU). We acknowledge the Ohio BioProducts Innovations
Center (OBIC) for helping support our mass spectrometry facility. Dr.
Judith Gallucci (OSU) is thanked for solving our crystal structures.
[9] For examples of nitroalkene activation with ureas and thioureas,
see: a) T. Okino, Y. Hoashi, T. Furukawa, X. N. Xu, Y. Takemoto, J.
Am. Chem. Soc. 2005, 127, 119; b) R. P. Herrera, V. Sgarzani, L.
Bernardi, A. Ricci, Angew. Chem. 2005, 117, 6734; Angew. Chem.
Int. Ed. 2005, 44, 6576; c) K. L. Kimmel, J. D. Weaver, J. A. Ellman,
Chem. Sci. 2012, 3, 121.
[10] For examples of a,b-unsaturated imide activation with ureas and thi-
oureas, see: a) T. Inokuma, Y. Hoashi, Y. Takemoto, J. Am. Chem.
Soc. 2006, 128, 9413; b) Y. Hoashi, T. Okino, Y. Takemoto, Angew.
Chem. 2005, 117, 4100; Angew. Chem. Int. Ed. 2005, 44, 4032;
c) P. R. Schreiner, A. Wittkopp, Org. Lett. 2002, 4, 217.
Keywords: alkylidene malonate · catalysis · hydrogen bond
donor · organocatalysis · palladacycles
[11] For examples of urea derivatives involved in addition reactions to
alkylidene malonates, see: a) C.-L. Cao, Z.-L. Sun, J.-L. Zhou, Y.
Tong, J. Org. Chem. 2007, 72, 4073; b) M. Chiarucci, M. Lombardo,
C. Trombini, A. Quintavalla, Adv. Synth. Catal. 2012, 354, 364.
[12] J. Vicente, J. Abad, J. Lopez-Serrano, P. G. Jones, C. Najera, L. Bo-
tella-Segura, Organometallics 2005, 24, 5044.
[13] For early work demonstrating that diaryl ureas with trifluoromethyl
substituents have enhanced activity, see: a) ref. [10c]; b) A. Witt-
kopp, P. R. Schreiner, Chem. Eur. J. 2003, 9, 407.
[1] E. Negishi, A. de Meijere, Handbook of Organopalladium Chemis-
try for Organic Synthesis, Wiley, New York, 2002.
[2] A. Berkessel, H. Grçger, Asymmetric Organocatalysis, Wiley-VCH,
Weinheim, 2009. For recent reviews, see: a) Y. Takemoto, Chem.
Pharm. Bull. 2010, 58, 593; b) S. J. Connon, S. A. Kavanagh, A. Pic-
cinini, Org. Biomol. Chem. 2008, 6, 1339; c) Z. G. Zhang, P. R.
Schreiner, Chem. Soc. Rev. 2009, 38, 1187; d) Y. Takemoto, Org.
Biomol. Chem. 2005, 3, 4299.
[3] For recent reviews see: a) C. Zhong, X. Shi, Eur. J. Org. Chem.
2010, 2999; b) H. Zhang, Z. Shao, Chem. Soc. Rev. 2009, 38, 2745.
[4] Pioneering reports demonstrating the promise of combined transi-
tion metal and organic catalyst systems include: a) I. Ibrahem, A.
Cordova, Angew. Chem. 2006, 118, 1986; Angew. Chem. Int. Ed.
2006, 45, 1952; b) G. L. Hamilton, E. J. Kang, M. Mba, F. D. Toste,
Science 2007, 317, 496; c) S. Mukherjee, B. List, J. Am. Chem. Soc.
2007, 129, 11336.
[14] See the Supporting Information.
[15] M. C. Etter, Z. Urbanczyk-Lipkowska, M. Ziaebrahimi, T. W. Pan-
unto, J. Am. Chem. Soc. 1990, 112, 8415.
[16] For proposed single hydrogen bond donor activation of alkylidene
malonate for nucleophilic attack, see: a) ref. [11a]; b) J. Liu, Z.
Yang, X. Liu, Z. Wang, Y. Liu, S. Bai, L. Lin, X. Feng, Org. Biomol.
Chem. 2009, 7, 4120; c) D. R. Magar, C. Chang, T.-F. Ting, K. Chen,
Eur. J. Org. Chem. 2010, 2062.
[5] Difluoroboronate ureas as enhanced urea catalysts have been previ-
ously reported, see: a) S. S. So, T. J. Auvil, V. J. Garza, A. E. Matt-
son, Org. Lett. 2012, 14, 444; b) S. S. So, J. A. Burkett, A. E. Matt-
son, Org. Lett. 2011, 13, 716.
Received: April 9, 2012
Published online: June 12, 2012
8314
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 8310 – 8314