Noncovalently functionalized dendrimers as recyclable catalysts

Article abstract of DOI:10.1021/ja005774u

The efficient reversible functionalization of the periphery of urea adamantyl poly(propylene imine) dendrimers with catalytic sites using noncovalent interactions is described. Phosphine ligands equipped with urea acetic groups, a binding motive complemen

Full text of DOI:10.1021/ja005774u

J. Am. Chem. Soc. 2001, 123, 8453-8458
8453
Noncovalently Functionalized Dendrimers as Recyclable Catalysts
Debby de Groot,§ Bas F. M. de Waal,‡ Joost N. H. Reek,*,§ Albertus P. H. J. Schenning,‡
Paul C. J. Kamer,§ E. W. Meijer,*,‡ and Piet W. N. M. van Leeuwen§
Contribution from the Institute of Molecular Chemistry, UniVersity of Amsterdam, Nieuwe Achtergracht
166, 1018 WV Amsterdam, The Netherlands, and Laboratory of Macromolecular and Organic Chemistry,
EindhoVen UniVersity of Technology, Post Office Box 513, 5600 MD EindhoVen, The Netherlands
ReceiVed NoVember 9, 2000. ReVised Manuscript ReceiVed May 30, 2001
Abstract: The efficient reversible functionalization of the periphery of urea adamantyl poly(propylene imine)
dendrimers with catalytic sites using noncovalent interactions is described. Phosphine ligands equipped with
urea acetic groups, a binding motive complementary to that of the dendrimer host, have been prepared and
assembled to the dendrimer support. The resulting supramolecular complex has been used as a multidentate
ligand system in the palladium-catalyzed allylic amination reaction in a batch process and in a continuous-
flow membrane reactor. We found that the activity and selectivity of the dendrimeric complex is similar to
that of the monomer complex, which indicates that the catalytic centers act as independent sites. The size of
the supramolecular system is sufficiently large and the binding of the guests is strong enabling a good separation
of the catalyst components from the reaction mixture using nanofiltration techniques.
Introduction
Scheme 1. The Concept of Supramolecular Anchoring of
Catalysts to a Soluble Support
The immobilization of transition metal catalysts by anchoring
them to organic or inorganic solid supports combines the good
catalytic performance of a tailor-made system with a facile
separation of the catalyst from the product phase.1 Disadvantages
of these heterogenized systems include a decreased activity due
to mass transfer limitations, lowered selectivity, and metal
leaching. Metal leaching can be suppressed by using properly
chosen ligands that coordinate strongly to the metal center. The
use of soluble supports such as small (hyperbranched) poly-
mers,2 dendrimers,3-5 and hybride materials6 result in homo-
geneous, immobilized catalysts that do not suffer from mass
transfer limitations. In most systems reported so far the catalyst
was covalently linked to the support or via ionic interactions.
An interesting alternative approach would be the noncovalent
anchoring of the catalyst to the soluble support using well-
defined binding sites. The reversible nature of this type of
anchoring allows controlled de- and re-functionalization of the
support, which enables the easy reuse of the support and
simplifies the variation of catalyst loading even during catalysis.
Furthermore, using a noncovalent approach, multicomponent
assemblies can be envisaged that are interesting for tandem
reactions and combinatorial techniques. To achieve the su-
pramolecular anchoring of catalysts, a soluble support must
contain well-defined binding sites to which the tailor-made
transition metal catalysts with a complementary binding motive
can be assembled (Scheme 1). Another requirement is that the
dimension of the support allows the use of nanofiltration
techniques. Dendrimers are excellent supports for developing
this type of material because of their well-defined size and
structure.
Recently, a fifth generation poly(propylene imine) dendrimer
functionalized with urea adamantyl units at the periphery (1)
has been reported.7 This dendrimer provided directional binding
(2) (a) Bergbreiter, D. E. Catal. Today 1998, 42, 389. (b) Bergbreiter,
D. E.; Osburn, P. L.; Liu, Y.-S. J. Am. Chem. Soc. 1999, 121, 9531. (c)
Wentworth, P., Jr.; Janda, K. D. Chem. Commun. 1999, 1917. (d)
Buchmeiser, M. R.; Wurst, K. J. Am. Chem. Soc. 1999, 121, 11101. (e)
Schlenk, C.; Kleij, A. W.; Frey, H.; van Koten, G. Angew. Chem., Int. Ed.
2000, 39, 3445.
(3) General reviews on dendrimers: (a) Newkome, G. R.; Moorefield,
C. N.; Vo¨gtle, F. Dendritic Molecules; Verlag-Chemie, Weinheim, Germany,
1996. (b) Zeng, F.; Zimmerman, S. C. Chem. ReV. 1997, 97, 1681-1712.
(c) Bosman, A. W.; Janssen, H. M.; Meijer, E. W. Chem. ReV. 1999, 99,
1665-1688. (d) Majoral, J.-P.; Caminade, A.-M. Chem. ReV. 1999, 99,
845-880. (e) Hearshaw, M. A.; Moss J. R. Chem. Commun. 1999, 1-8.
(f) Cuadrado, I.; Mora´n, M.; Casado, C. M.; Alonso, B.; Losada, J. Coord.
Chem. ReV. 1999, 193-195, 395-445.
(4) Recent review on dendrimers in catalysis: Oosterom, G. E.; Reek,
J. N. H.; Kamer, P. C. J.; Van Leeuwen, P. W. N. M. Angew. Chem., Int.
Ed. 2001, 40, 1829.
(5) (a) Keijsper, J. J.; Van Leeuwen, P. W. N. M.; Van der Made, A. W.
European Patent 0,456,317, 1991, and U.S. Patent 5,243,079, 1993, to Shell
Int. Research; Chem. Abstr. 1992, 116, 129, 870. (b) Knapen, J. W. J.; Van
der Made, A. W.; De Wilde, J. C.; Van Leeuwen, P. W. N. M.; Wijkens,
P.; Grove, D. M.; Van Koten, G. Nature 1994, 372, 659-662. (c) Bhyrappa,
P.; Young, J. K.; Moore, J. S.; Suslick, K. S. J. Am. Chem. Soc. 1996, 118,
5708-5711. (d) Seebach, D.; Marti, R. E.; Hintermann, T. HelV. Chim.
§ University of Amsterdam.
‡ Eindhoven University of Technology.
(1) Reviews on polymer immobilized catalysts: (a) Hartley, F. R.; Vezey,
P. N. AdV. Organomet. Chem. 1977, 15, 189. (b) Hartley, F. R. Supported
Metal Complexes. A New Generation of Catalysts; Reidel: Dordrecht, The
Netherlands, 1985. (c) Iwasawa, Y. Tailored Metal Catalysts; Ugo, R.,
James, B. R., Eds.; Reidel: Dordrecht, The Netherlands, 1986. (d) Keim,
W.; Driessen-Ho¨lscher, B. Handbook of Heterogeneous Catalysis; Ertl, G.,
Kno¨zinger, H., Weitkamp, J., Eds.; Wiley-VCH: Weinheim, Germany1997;
Vol 1, p 231.
10.1021/ja005774u CCC: $20.00 © 2001 American Chemical Society
Published on Web 08/07/2001

8454 J. Am. Chem. Soc., Vol. 123, No. 35, 2001
de Groot et al.
Figure 1. Phosphine ligands assembled to the periphery of a urea adamantyl functionalized poly(propylene imine) dendrimer (1).
Scheme 2. Synthesis of a Phosphine-Containing Guest Molecule (3) in Two Steps and the Structure of the Schematic Structure
of the Urea Adamantly Functionalized Dendrimeric Host (1)
sites for the strong but reversible binding of 32 guest molecules
functionalized with the complementary binding motive. The
binding is based on a combination of ionic interactions and the
formation of multiple hydrogen bonds. Here we report for the
first time the noncovalent anchoring of phosphine ligands and
their transition metal complexes, which can be used as catalysts,
to the periphery of a soluble dendrimeric support (Figure 1).
The resulting recyclable homogeneous supramolecular complex
is used in the palladium-catalyzed allylic amination reaction in
a batch process and in a continuous-flow membrane reactor.
the introduction of the binding motive in a ligand system must
be straightforward to have a general strategy that can also be
used for more complicated ligand systems. A reaction of
p-(diphenylphosphino)benzylamine with commercially available
ethyl isocyanatoacetate yielded ester derivative 2 (Scheme 2).
After hydrolysis of the ester and recrystallization from chloro-
form pure 3 was isolated as a white powder. Phosphine guest 3
is thus prepared in two simple reaction steps.
The binding of the guest molecule into the periphery of 1
was studied by NMR spectroscopy. In contrast to acid ligand
3, ester ligand 2 showed no affinity for the binding motive of
the dendrimer. Addition of 32 equiv of acid(3) to 1 (in CDCl3)
resulted in a shift of the urea protons of the dendrimer from
6.20 and 5.43 ppm to 6.34 and 5.64 ppm, respectively. This
shift is similar to that previously reported for other guest
molecules.7 Addition of an extra portion of dendrimer (1 equiv)
resulted in a random distribution of the guest molecules over
the dendrimers and average signals for the urea protons of the
dendrimer were observed in the 1H NMR spectrum at 6.30 and
5.57 ppm, respectively. This clearly shows that the exchange
between the bound and unbound species is rapid on the NMR
Results and Discussion
Synthesis and Complex Formation. It was previously found
that both the urea and the carboxylic acid functionality must
be present in the guest molecules to achieve the required strong
binding to the dendrimeric host 1.7,8 We also anticipated that
Acta 1996, 79, 1710-1740. (e) Reetz, M. T.; Lohmer, G.; Schwickardi, R.
Angew. Chem., Int. Ed. Engl. 1997, 36, 1526-1529. (f) Bourque, S. C.;
Maltais, F.; Xiao, W.-J.; Tardif, O.; Alper, H.; Arya, P.; Manzer, L. E. J.
Am. Chem. Soc. 1999, 121, 3035-3038. (g) Kleij, A. W.; Gossage, R. A.;
Jastrzebski, J. T. B. H.; Boersma, J.; Van Koten, G. Angew. Chem., Int.
Ed. 2000, 39, 176-178. (h) Breinbauer, R.; Jacobsen, E. N. Angew. Chem.,
Int. Ed. 2000, 39, 3604-3607.
(6) Lindner, E.; Schneller, T.; Auer, F.; Mayer, H. A. Angew. Chem.,
Int. Ed. 1999, 38, 2154.
(7) Baars, M. W. P. L.; Karlsson, A. J.; Sorokin, V.; De Waal, B. F. M.;
Meijer, E. W. Angew. Chem., Int. Ed. 2000, 39, 4262.
1
time scale. In a H-NOE NMR experiment of the dendrimer
with 32 equiv of acid(3) the aromatic protons of the guest were
selectively irradiated and a NOE effect was observed for the
adamantyl protons. This shows that the diphenylphosphine
groups of the guest and the adamantyl endgroups of the
dendrimer are in close proximity. A 2D-NOESY NMR spectrum
(8) Boas, U.; Karlsson, A. J.; De Waal, B. F. M.; Meijer, E. W. J. Org.
Chem. 2001, 66, 2136.

NoncoValently Functionalized Dendrimers as Recyclable Catalysts
J. Am. Chem. Soc., Vol. 123, No. 35, 2001 8455
Figure 2. 2D-NOESY NMR spectrum of guest ligands acid(3) assembled to the periphery of urea adamantyl poly(propylene imine) dendrimers
(1).
also shows the interaction of the aromatic protons of the guest
molecule with the protons of the adamantyl groups (Figure 2).
These experiments suggest that, as a result of the multiple
hydrogen bond interactions, the guest molecule is positioned
in the periphery of the dendrimeric host in a well-defined way,
as is shown schematically in Figure 1.
the dendrimer have broadened significantly and are shifted with
respect to the free dendrimeric host.
A third route toward Pd-functionalized dendrimers involves
the mixing of all the reactants simultaneously (Scheme 3c). In
this way we have prepared [{acid(3)}Pd(crotyl)Cl]32-dendrimer
complex, which gives a characteristic broad singlet in the 31P-
{1H} NMR spectrum at 24.4 ppm and [{acid(3)}2Pd(crotyl)-
Cl]16-dendrimer complex (broad singlet at 24.9 ppm). These
complexes were used for catalysis (vide supra).
Size exclusion chromatography (SEC) was used to check the
stability of the acid(3)-Pd-dendrimer complex. NMR showed
that the [{acid(3)}2Pd(allyl)Cl]16-dendrimer complex remained
intact after SEC, in contrast to the {ester(2)}2Pd-dendrimer
mixture. This indicates that the binding constant of the acid-
(3)-PdCl(allyl)-complex to the periphery of the dendrimer as
well as that of the Pd to the ligand is very high. This is in line
with the high binding constants (104 M-1) that were found
previously for similar supramolecular systems.8
After the guest ligands were bound into the periphery of the
dendrimer they were complexed to palladium by addition of
Pd(COD)MeCl or [(crotyl)PdCl]2 (Scheme 3a). A second
approach toward the formation of metal functionalized systems
involves the synthesis of the metal complex of acid(3) prior to
the noncovalent anchoring of the complex to the periphery of
the dendrimer, leading to the same product (Scheme 3b). Upon
the addition of 0.5 equiv of (COD)PdMeCl with respect to the
phosphine ligand, trans complexes are formed. In the 31P-{1H}
NMR spectrum a broad singlet at 30.4 ppm is observed, which
is comparable to the 30.7 ppm found for the model compound
trans-(PPh3)2PdMeCl. The PdMe signal in the 1H NMR
spectrum is at -0.04 ppm, compared to -0.03 ppm for trans-
(PPh3)2PdMeCl. Importantly, the 1H NMR spectrum shows that
the guest molecules and the dendrimeric host are still assembled
after metal complexation; the signals for the urea protons of
Catalysis
The dendrimeric host containing 32 phosphine ligands
assembled to the periphery of 1 was used as a multidentate

8456 J. Am. Chem. Soc., Vol. 123, No. 35, 2001
de Groot et al.
Scheme 3. Schematic Presentation of the Preparation of Transition Metal Complexes Using Ligands Noncovalently Anchored to
the Periphery of a Dendrimer
Table 1. Results of the Pd-Catalyzed Allylic Amination of Crotyl
Acetate and Piperidine, Comparing the Supramolecular Dendrimeric
Catalyst with the Monomera
Scheme 4. Allylic Amination of Crotyl Acetate and
Piperidine
conversionb trans cis branched
catalyst
P/Pd
(%)
(%) (%)
(%)
{ester(2)}2Pd(crotyl)Cla
{acid(3)}2Pd(crotyl)Cl-
dendrimera
2
2
89
91
38
37
14
15
48
48
{ester(2)}Pd(crotyl)Clc
{acid(3)}Pd(crotyl)Cl-
dendrimerc
1
1
80
72
33
33
6
6
61
61
ligand in the Pd-catalyzed allylic amination9 with crotyl acetate
and piperidine as the substrate molecules (Scheme 4). The
reaction has been performed under various conditions with
[{acid(3)}2Pd(crotyl)Cl]16-dendrimer (P/Pd ) 2) and [{acid-
(3)}Pd(crotyl)Cl]32-dendrimer (P/Pd ) 1) as a catalyst. The
reaction was started by mixing the substrates and the catalyst
solution and appeared to be fast; using the [{acid(3)}2Pd(crotyl)-
Cl]16-dendrimer (P/Pd ) 2) and a substrate/palladium ratio of
100 the conversion was 91% after 5 min. Approximately the
same rate was observed for the Pd complex of ester(2) in the
absence of the dendrimer (Table 1). This indicates that every
active site on the dendrimer acts as an independent catalyst.10
These experiments show clearly that the supramolecular anchor-
ing of the catalysts does not decrease the activity, which
generally is observed for catalysts immobilized on an insoluble
support. Moreover, the product selectivity generated by the
{acid(3)}2-Pd-dendrimer complex is the same as that induced
by the {ester(2)}2-Pd complex. The noncovalently function-
alized dendrimeric catalyst used in these batchwise reactions
was recycled by using SEC. The conversion measured after 1
h, however, was lower in the second run, suggesting that the
catalyst partly decomposed during the recycling procedure.
a Room temperature; solvent, CH2Cl2; volume, 5 mL; [crotyl acetate]
) 0.2 M, [piperidine] ) 0.4 M, [Pd] ) 0.002 M. b Conversion after 5
min. c Room temperature; solvent, CH2Cl2; volume, 5 mL; [crotyl
acetate] ) 0.12 M, [piperidine] ) 0.24 M, [Pd] ) 0.0040 M.
Upon using the [{acid(3)}Pd(crotyl)Cl]32-dendrimer complex
(P/Pd ) 1) as a catalyst, a similarly fast reaction was observed
(Table 1).11 Again the results for the {acid(3)}Pd(crotyl)Cl-
dendrimer complex are similar to those for the {ester(2)}Pd-
(crotyl)Cl complex. The selectivity is slightly different from
that obtained with a P/Pd ratio of 2.12
Optimal advantage of dendrimers as soluble supports for
catalysts can be obtained when they are applied in a continuous-
flow membrane reactor.10,13,14 For this purpose we used a Koch
MPF-60 NF membrane (molecular weight cutoff ) 400 dalton)
that was placed in a homemade reactor. The retention measured
for the ester complex (2)Pd(crotyl)Cl (MW ) 617.4) in the
(11) Since the reaction conditions are different, these results cannot be
compared to the results of Table 1.
(12) The selectivity of this reaction strongly depends on the ligands used,
see for example: (a) Rieck, H.; Helmchen, G. Angew. Chem., Int. Ed. Engl.
1995, 34, 2687. (b) Pre´toˆt, R.; Lloyd-Jones, G. C.; Pfaltz, A. Pure Appl.
Chem. 1998, 70, 1035. (c) van Haaren, R. J.; Druijven, C. J. M.; van
Strijdonck, G. P. F.; Oevering, H.; Reek, J. N. H.; Kamer, P. C. J.; van
Leeuwen, P. W. N. M. J. Chem. Soc., Dalton Trans. 2000, 1549.
(13) Hovestad, N. J.; Eggeling, E. B.; Heidbu¨chel, H. J.; Jastrzebski, J.
T. B. H.; Kragl, U.; Keim, W.; Vogt, D.; Van Koten, G. Angew. Chem.,
Int. Ed. 1999, 38, 1655.
(9) Johannsen, M.; Jørgensen, K. A. Chem. ReV. 1998, 98, 1689.
(10) This has previously been observed for covalently functionalized
dendrimers: De Groot, D.; Eggeling, E. B.; De Wilde, J. C.; Kooijman,
H.; Van Haaren, R. J.; Van der Made, A. W.; Spek, A. L.; Vogt, D.; Reek,
J. N. H.; Kamer, P. C. J.; Van Leeuwen, P. W. N. M. Chem. Commun.
1999, 1623.
(14) Brinkmann, N.; Giebel, D.; Lohmer, G.; Reetz, M. T.; Kragl, U. J.
Catal. 1999, 183, 163.

NoncoValently Functionalized Dendrimers as Recyclable Catalysts
J. Am. Chem. Soc., Vol. 123, No. 35, 2001 8457
acid(3) clearly demonstrate that the noncovalently functionalized
dendrimers are suitable as soluble and recyclable supports for
catalysts. We are currently studying the potential of the concept
in other reactions.
Conclusions
In conclusion, we have developed a new homogeneous
catalyst that is anchored to a soluble dendrimer support using
supramolecular interactions. The catalytic system is non-
covalently attached to the periphery of a urea adamantyl poly-
(propylene imine) dendrimer by ionic interactions in combina-
tion with multiple hydrogen bonds, which positions the guest
ligand at the periphery of the dendrimer in a well-defined way.
The supramolecular dendrimeric system shows the same activity
and selectivity in the Pd-catalyzed allylic amination as its
unbound monomeric analogue, which indicates that every active
site on the dendrimer acts as an independent catalyst and is
easily accessible to the substrate. Moreover, the catalyst is
strongly bound such that the system can be operated in a
continuous setup, which results in efficient separation of the
catalyst from the reaction mixture. Employing the concept of
noncovalent anchoring simplifies the route toward sophisticated
dendrimeric catalysts since ligand modification with the binding
motive is straightforward. One of the limitations in dendrimeric
catalysis is the troublesome synthesis of functionalized den-
drimers, since quantitative coupling of ligands to the periphery
is not always possible. With this strategy these problems are
circumvented. Moreover, this approach opens the way toward
the use of multipurpose supports, not only for dendrimers but
also other supports, which can be functionalized and refunc-
tionalized with multiple complex catalytic systems containing
a relatively simple binding motive.
Figure 3. Application of noncovalently functionalized dendrimers in
a continuous-flow membrane reactor using acid(3) as a ligand (a) and
using ester(2) as a ligand (b).
presence of the host dendrimer is 97%, which is too low for
practical application since one-third of the ligand will be washed
out of the reactor after 13 reactor volumes of solvent have been
pumped through. In contrast, the supramolecular acid-den-
drimer complex [(3)Pd(crotyl)Cl]32-dendrimer has a retention
as high as 99.4%. Interestingly, upon using a lower palladium
loading (P/Pd ) 2) the retention further increased to 99.9%,
suggesting that the palladium diphosphine complexes are bound
more strongly due to cooperative effects. These results indicate
that this type of supramolecular complexe will be efficiently
separated from smaller compounds such as the reactants and
products in a continuous-flow membrane reactor. The applica-
tion in catalysis was studied by using the allylic amination
reaction.
A solution of crotyl acetate and piperidine in dichloromethane
(including n-decane as an internal standard) was pumped through
the reactor. The Pd complex of acid(3) assembled to the
dendrimer was used as a catalyst and compared to the Pd
complex of ester(2) in the presence of the dendrimer.15
In Figure 3 the conversion is plotted as a function of the
substrate flow.16 In both experiments the conversion has
increased to its maximum (ca. 80%) after approximately 1 h
(which is equivalent to 1-2 reactor volumes of substrate
solution pumped through the reactor). Upon using acid(3) the
conversion remains fairly constant during the first 10 h of the
experiment (Figure 3a).17 The slight decrease observed is
presumably a result of slow deactivation of the catalyst, which
has also been observed using covalently functionalized den-
drimers.10,14 Upon using ester(2) the activity of the catalytic
system drops more rapidly, since ester(2), which is not bound
to the dendrimer, is washed out of the reactor (retention ) 97%).
In this experiment slow deactivation of the catalyst occurs as
well. Although these deactivation processes slightly obscure the
results, the differences observed between ligands ester(2) and
Experimental Section
General Data. All reactions were carried out under an atmosphere
of purified nitrogen with standard Schlenk techniques. Solvents were
distilled under N2 from sodium/benzophenone (THF, hexane) or calcium
hydride (dichloromethane) prior to use. Water and CDCl3 were degassed
and stored under nitrogen. Chemicals were purchased from Aldrich
Chemical Co. and Acros Chimica and were used without further
purification. The piperidine was filtered over neutral alumina prior to
use. For the size exclusion chromatography Bio-Beads S-X1 Beads
(Gel Permeation Gel 200-400 mesh, Bio-Rad Laboratories, Hercules,
USA) were used. p-(Diphenylphosphino)benzylamine,18 (COD)Pd-
MeCl,19 [(allyl)PdCl]2,20 and [(crotyl)PdCl]220 were prepared according
to literature procedures. 1H- and 31P-{1H} NMR spectra were recorded
on a Varian Mercury 300. 13C-{1H} NMR spectra were recorded on
a Varian Inova 500. The chemical shifts are given in ppm relative to
1
TMS for H and 13C NMR and relative to H3PO4 for 31P NMR. Fast
Atom bombardment (FAB) mass spectrometry was carried out using a
JEOL SX/SX 102A (Tokyo, Japan) four-sector tandem mass spec-
trometer (B1E1B2E2 geometry), coupled to a JEOL MS/MP9021D/UPD
data system. Gas chromatography was performed on an Interscience
HR GC Mega 2 apparatus (split/splitless injector, J&W Scientific, DB1
30 m column, film thickness 3.0 mm, carrier gas: 70 kPa He, F.I.D.
detector).
Synthesis of the phosphine ligands. (a) Ester(2). To a solution of
0.4417 g of (diphenylphosphino)benzylamine (1.516 mmol) in 5 mL
of dichloromethane was added 0.17 mL of ethyl isocyanatoacetate
(1.515 mmol). After the mixture was stirred for 2 h at room temperature
the solvent was evaporated. The product was recrystallized from
dichloromethane/hexane and obtained as a white solid in 89% yield
(15) Room temperature; reactor volume, 5 mL; solvent, CH2Cl2; [crotyl
acetate] ) 0.1 M, [piperidine] ) 0.2 M, [Pd] ) 0.004 M; flow rate, 9 mL/
h.
(16) The selectivity observed during these continuous experiments is
similar to that of the batch processes.
(17) The activity of the catalyst observed in the continuous process cannot
be compared directly to that observed in the batch reaction because of the
large differences in the reactor setup.
(18) Hingst, M.; Tepper, M.; Stelzer, O. Eur. J. Inorg. Chem. 1998, 73.
(19) Ru¨lke, R. E.; Han, I. M.; Elsevier: C. J.; Vrieze, K.; Van Leeuwen,
P. W. N. M.; Roobeek, C. F.; Zoutberg, M. C.; Wang, Y. F.; Stam, C. H.
Inorg. Chim. Acta 1991, 169, 5.
(20) Dent, W. T.; Long, R.; Wilkinson, A. J. J. Chem. Soc. 1964, 1585.

8458 J. Am. Chem. Soc., Vol. 123, No. 35, 2001
de Groot et al.
(0.5690 g, 1.353 mmol). Mp: 100-101 °C. 1H NMR (300 MHz,
CDCl3): δ 7.4-7.2 (m, 14H, ArH), 5.0 (s (br.), 2H, NH), 4.36 (d, 2H,
(300 MHz, CDCl3): δ 7.8-7.1 (m, 8H, ArH), 6.56 (s (br), 1H, NH),
6.24 (s (br), 1H, NH), 5.58 (s, 4H, COD), 4.18 (s (br), 2H, ArCH2N),
3.68 (s (br), 2H, NCH2CO), 2.36 (s, 8H, COD). 31P-{1H} NMR (121.5
MHz, CDCl3): δ 32.3 (s), 30.4 (s).
3
3
ArCH2N, JHH ) 5.7 Hz), 4.15 (q, 2H, OCH2CH3, JHH ) 7.2 Hz),
3.97 (d, 2H, NCH2CO, 3JHH ) 5.1 Hz), 1.23 (t, 3H, OCH2CH3, 3JHH
)
7.2 Hz). 31P-{1H} NMR (121.5 MHz, CDCl3): δ -5.59 (s). 13C-
(d) [{Acid(3)}2-PdMeCl]32-Dendrimer. Method A. A solution
of 3.352 mg of (COD)PdMeCl (12.6 µmol) in 1 mL of CDCl3 was
added to a solution of {acid(3)}32-dendrimer (25.2 µmol of 3 + 0.802
µmol of dendrimer (1)) in 1 mL of CDCl3. After overnight stirring the
{1H} NMR (125.8 MHz, CDCl3): δ 171.4 (s, COOEt), 157.8 (s,
1
NHCONH), 140.0 (s, ipso-PhCH2N), 137.4 (d, ipso-PhP, JCP ) 10.6
1
2
Hz), 136.4 (d, ipso-PhP, JCP ) 10.6 Hz), 134.3 (d, m-PhCH2N, JCP
) 19.4 Hz), 133.9 (d, o-PhP, 2JCP ) 19.4 Hz), 128.9 (s, p-PhP), 128.7
1
mixture was analyzed by H- and 31P-{1H} NMR.
3
3
(d, m-PhP, JCP ) 6.7 Hz), 127.7 (d, o-PhCH2N, JCP ) 7.2 Hz), 61.6
(s, OCH2CH3), 44.5 (s, ArCH2N), 42.5 (s, NCH2CO), 14.4 (s, CH3).
IR (KBr) υ (cm-1) 3338 (NH), 3052 (CH, arom.), 2982 (CH, alif.),
1748 (CO), 1630 (CO), 1576 (CO). FAB-MS m/z 437.2 ([M + O +
H]+), HRMS (FAB+) m/z calcd for C24H26O4N2P [M + O + H]+
437.1630. Found: 437.1619. Anal. Calcd for C24H25N2O3.25P (partly
oxidized): C 67.92; H 5.94; N 6.60. Found: C 67.85; H 5.82; N 6.45.
(b) Acid(3). A solution of 0.0233 g of NaOH (0.583 mmol) in 4
mL of water was added to a solution of 0.2197 g of 2 (0.5225 mmol)
in 5 mL of THF. After overnight stirring the THF was evaporated and
the reaction mixture was neutralized by addition of 1 mL of 0.58 M
aqueous HCl. The solvent was decanted and the crude product was
washed with water. After recrystallization from chloroform a white
powder was obtained in 40% yield (0.0824 g, 0.210 mmol). Mp: 129-
Method B. A solution of 0.0146 g of dendrimer (0.789 µmol) in
0.5 mL of CDCl3 was added to a solution of [{acid(3)}2-PdMeCl]
(12.7 µmol of Pd, 25 µmol of 3) in 1 mL of CDCl3. After stirring for
1
15 min the mixture was analyzed by H- and 31P-{1H} NMR.
1H NMR (500 MHz, CDCl3): δ 7.8-7.0 (m, 448H, ArH of guest
ligand), 6.34 (br, 64H, CH2NHCONH of dendrimer), 5.69 (br, 64H,
CH2NHCONH of dendrimer), 5.59 (s, 128H, COD), 4.4 (br, 64H,
ArCH2N of guest), 3.7 (br, 64H, NCH2CO of guest), 3.08 (s (br), 128H,
CH2NHCONH of dendrimer), 2.2-2.8 (br, 372H, NCH2CH2CH2NHCO
{128H} + NCH2CH2CH2N {240H} + NCH2CH2CH2CH2N {4H}),
2.37 (s, 256H, COD), 1.96 (s (br), 192H, adamantyl), 1.90 (s (br), 384H,
adamantyl), 1.59 (m, 636H, adamantyl {384H} + NCH2CH2CH2NHCO
{128H} + NCH2CH2CH2N {120H} + NCH2CH2CH2CH2N {4H}),
-0.04 (br, 192H, PdMe). 31P-{1H} NMR (121.5 MHz, CDCl3): δ
30.5 (br).
1
130 °C. H NMR (300 MHz, DMSO-d6): δ 7.5-7.2 (m, 14H, ArH),
6.88 (t, 1H, NH, 3JHH ) 6 Hz), 6.20 (t, 1H, NH, 3JHH ) 6 Hz), 4.22 (d,
2H, ArCH2N, 3JHH ) 5.7 Hz), 3.69 (d, 2H, NCH2CO, 3JHH ) 5.4 Hz).
31P-{1H} NMR (121.5 MHz, DMSO-d6): δ -2.26 (s). 13C-{1H} NMR
(125.8 MHz, DMSO-d6): δ 172.5 (s, COOH), 158.0 (s, N-CO-N),
142.0 (s, ipso-PhCH2N), 136.8 (d, ipso-PhP, 1JCP ) 9.7 Hz), 134.4 (d,
Retention Measurements. A solution of 0.0231 g of fifth generation
adamantyl-urea functionalized poly(propylene imine) dendrimer (1)
(1.25 µmol), 15.692 mg of 3 (40.0 µmol) or 16.812 mg of 2 (40.0
µmol), and 7.879 mg of [(crotyl)PdCl]2 (20.0 µmol) in 2 mL of
dichloromethane was stirred overnight. The membrane was cut to the
correct size for the reactor and stored in acetone during one night before
storing it in methanol (for at least one night). When it was adjusted in
the membrane reactor, the membrane was flushed overnight with
dichloromethane. The ligand-Pd-dendrimer mixture was transferred
into the reactor and the solvent was pumped through for 25 h at a flow
rate of 6 mL/h to flush the reactor 30 times. The solvent was evaporated
from both the contents of the reactor and the solution that had been
pumped through the membrane. The residue was weighed and analyzed
by NMR.
1
2
ipso-PhP, JCP ) 10.2 Hz), 133.4 (d, m-PhCH2N, JCP ) 21.1 Hz),
2
133.1 (d, o-PhP, JCP ) 19.4 Hz), 128.9 (s, p-PhP), 128.7 (d, m-PhP,
3JCP ) 6.8 Hz), 127.4 (d, o-PhCH2N, 3JCP ) 5.9 Hz), 42.6 (s, ArCH2N),
41.6 (s, NCH2CO). IR (KBr) υ (cm-1) 3412 (OH), 3376 (NH), 3052
(CH), 2924 (CH), 1724 (CO), 1624 (CO), 1575 (CO). FAB-MS m/z
409.1 ([M + O + H]+) HRMS (FAB+) m/z calcd for C22H22O4N2P [M
+ O + H]+ 409.1317. Found: 409.1240. Anal. Calcd for C22H21N2O3.6
P
(partly oxidized): C 65.73; H 5.27; N 6.97. Found: C 65.63; H 5.52;
N 7.04.
Binding of the Guest into the Periphery of the Dendrimer. (a)
{Acid(3)}32-Dendrimer. A solution of 0.0198 g of 3 (50.5 µmol) in 1
mLof CDCl3 was added to a solution of 0.0297 g of fifth generation
adamantyl-urea functionalized poly(propylene imine) dendrimer (1)
(1.60 µmol) in 1 mL of CDCl3. The resulting mixture was analyzed by
Catalysis. The allylic amination experiments were performed under
N2 atmosphere at room temperature.
Batch Process. The catalyst solution of the model compound was
prepared by overnight stirring of a mixture of the fifth generation of
adamantyl-urea functionalized poly(propylene imine) dendrimer, the
ligand, and [(crotyl)PdCl]2 in 2 mL of dichloromethane. To the catalyst
solution were added 2.0 mL of CH2Cl2 and 1.0 mL of substrate solution.
The reaction was monitored in time by quenching samples in DBA/
Et2O solution (DBA ) dibenzylideneacetone). Conversions and product
distribution were determined by using GC analysis.
1
1H- and 31P-{1H} NMR. H NMR (500 MHz, CDCl3): δ 7.6-7.2
(m, 448H, ArH of guest ligand), 6.5 (broad shoulder, NH of guest),
6.34 (s (br), 64H, CH2NHCONH of dendrimer), 5.64 (s (br), 64H, CH2-
NHCONH of dendrimer), 4.32 (s (br), 64H, ArCH2N of guest), 3.73 (s
(br), 64H, NCH2CO of guest), 3.07 (s (br), 128H, CH2NHCONH of
dendrimer), 2.2-2.8 (br, 372H, NCH2CH2CH2NHCO {128H} + NCH2-
CH2CH2N {240H} + NCH2CH2CH2CH2N {4H}), 1.93 (s (br), 192H,
adamantyl), 1.86 (s (br), 384H, adamantyl), 1.56 (m, 636H, adamantyl
{384H} + NCH2CH2CH2NHCO {128H} + NCH2CH2CH2N {120H}
+ NCH2CH2CH2CH2N {4H}). 31P-{1H} NMR (121.5 MHz, CDCl3):
δ -5.73 (s).
Continuous Process. The catalyst solutions were prepared similar
to those for the batch reactions. For the model reaction 0.0117 g of
dendrimer (0.632 µmol), 8.408 mg of 2 (20.0 µmol), and 3.915 mg of
[(crotyl)PdCl]2 (9.94 µmol, 19.9 µmol of Pd) were stirred overnight in
2 mL of dichloromethane. For the dendrimeric catalyst solution 0.0116
g of dendrimer (0.627 µmol), 7.836 mg of 3 (20.0 µmol), and 3.905
mg of [(crotyl)PdCl]2 (9.91 µmol, 19.8 µmol of Pd) were stirred
overnight in 2 mL of dichloromethane. The membrane was cut to the
correct size for the reactor, stored in acetone for one night, and then
stored in methanol (for at least one night). After it was transferred into
the membrane reactor, the membrane was first flushed overnight with
CH2Cl2 and then with substrate solution (approximately two reactor
volumes). The substrate solution was prepared by mixing 1.5 mL of
crotyl acetate, 2.47 mL of piperidine, and 2.44 mL of n-decane (as
internal standard) in CH2Cl2 (total volume ) 100 mL) and was pumped
through the reactor with a flow rate of 9 mL/h. The reaction was started
by transferring the catalyst solution in the membrane reactor. Samples
of the solution coming out of the reactor were quenched in DBA/Et2O
and analyzed by GC.
(b) {Acid(3)}16-Dendrimer. Half of the {acid(3)}32-dendrimer
solution (25.2 µmol of 3, 0.802 µmol of dendrimer) was added to 0.0147
1
g of dendrimer (1) (0.794 µmol). The mixture was analyzed by H
1
NMR. H NMR (500 MHz, CDCl3): δ 7.5-7.1 (m, 224H, ArH of
guest ligand), 6.30 (s (br), 64H, CH2NHCONH of dendrimer), 5.57 (s
(br), 64H, CH2NHCONH of dendrimer), 4.65 (s (br), 32H, ArCH2N
of guest), 3.74 (s (br), 32H, NCH2CO of guest), 3.08 (s (br), 128H,
CH2NHCONH of dendrimer), 2.2-2.8 (br, 372H, NCH2CH2CH2NHCO
{128H} + NCH2CH2CH2N {240H} + NCH2CH2CH2CH2N {4H}),
1.98 (s (br), 192H, adamantyl), 1.91 (s (br), 384H, adamantyl), 1.60
(m, 636H, adamantyl {384H} + NCH2CH2CH2NHCO {128H} +
NCH2CH2CH2N {120H} + NCH2CH2CH2CH2N {4H}).
(c) [{Acid(3)}2-PdMeCl]. A solution of 3.371 mg of (COD)-
PdMeCl (12.7 µmol) in 0.7 mL of CDCl3 was added to a solution of
0.0098 g of 3 (25 µmol) in 0.3 mL of CDCl3. After being stirred for 3
1
1
h the mixture was analyzed by H- and 31P-{1H} NMR. H NMR
JA005774U