A divergent, solid-phase approach to dendritic ligands on beads. Heterogeneous catalysis for hydroformylation reactions

Full text of DOI:10.1021/jo991621h

J . Org. Chem. 2000, 65, 1881-1885
1881
tested them for homogeneous catalysis.9,10 Due to the
large size of such molecules, they could be separated from
the reaction mixture using various size exclusion tech-
niques. In addition, the dendritic ligands could also
exhibit high reactivities as a result of the cooperative
behavior.11,13
A Diver gen t, Solid -P h a se Ap p r oa ch to
Den d r itic Liga n d s on Bea d s.
Heter ogen eou s Ca ta lysis for
Hyd r ofor m yla tion Rea ction s1a
Prabhat Arya,* N. Venugopal Rao, and
J irada Singkhonrat1b
Resu lts a n d Discu ssion
Chemical Biology Program, Steacie Institute for Molecular
Sciences, National Research Council of Canada,
Recently, we reported the application of polyamino-
amido diphosphonated dendrimers anchored onto silica
gel for hydroformylation reactions.12 In contrast to what
is commonly known for heterogeneous catalysts, these
systems, when complexed to Rh, were excellent catalysts.
It was interesting to note that the hydroformylation of
styrenes and vinyl esters gave a high selectivity for the
branched products. Branched phenyl propionaldehydes
are important because they could lead to several useful
intermediates for the pharmaceutical industry (e.g.,
oxidation to nonsteroidal antiinflammatory agents).
Herein, we report a solid-phase synthetic approach to
obtain dendritic ligands anchored onto beads and their
application for the hydroformylation of several olefins.
There are several advantages with the use of polystyrene-
based beads as a solid support: (i) the ease of solid-phase
synthesis using a building block approach, (ii) charac-
terization of products anchored onto beads after cleavage,
(iii) better swelling properties in most solvents, and (iv)
flexible polymeric backbones (Figure 1).14-17 This could
further be extended to develop a library approach to
catalysis by high-throughput synthesis. Parallel to this
approach in pharmaceutical research, combinatorial
chemistry toward material sciences is relatively a new
field and could reduce the time required to find lead
catalysts.18-20 As observed with the silica gel supported
systems,12 dendritic phosphine ligands anchored onto
beads are excellent catalysts (i.e., highly reactive after
several cycles, highly regioselective for the branched
aldehyde of styrene and vinyl ester sytems) for the
hydroformylation reaction of several olefins. To our
knowledge, this is the first study that utilizes a solid-phase
100 Sussex Drive, Ottawa, Ontario, Canada, K1A 0R6
Howard Alper* and S. Christine Bourque
Department of Chemistry, University of Ottawa,
10 Marie Curie Street, Ottawa, Ontario, Canada, K1N 6N5
Leo E. Manzer
DuPont Central Research & Development,
Experimental Station, Wilmington, Delaware 19880-0262
Received October 18, 1999
In tr od u ction
Over the years, the interest in supporting homogeneous
catalysts has grown significantly for several reasons.2-4
Highly complex ligands and metal catalysts have become
very expensive and must be recycled. Separation of high-
boiling products from nonvolatile catalysts has also
limited the commercialization of many excellent homo-
geneous catalysts. We began this work to develop new
approaches for the heterogenization of metal catalysts.
Most of the research in this area utilizes polymer-
supported catalysts. In general, such systems are found
to be more stable, but significantly less reactive. For
industrial catalytic processes, there is a need for develop-
ing systems that function like homogeneous catalysis (i.e.,
high reactivity) and are easy to separate from the
reaction mixture.
With this goal, we decided to explore the scope of
dendritic, multivalent ligands anchored onto beads for
heterogeneous catalysis. Dendrimers are relatively well-
defined macromolecules with emerging applications in
the area of material and biological sciences.5-8 Several
groups have utilized the hyper-branched nature of den-
dritic materials to obtain multivalent ligands and have
(9) Reetz, M. T.; Lohmer, G.; Schwickardi, R. Angew. Chem., Int.
Engl. 1997, 36, 1526-1529.
(10) Petrucci-Samija, M.; Guillemette, V.; Dasgupta, M.; Kakkar,
A. K. J . Am. Chem. Soc. 1999, 121, 1968-1969.
(11) (a) Knapen, J . W. J .; van der Made, A.; de Wilde, J . C.; van
Leeuwen, P. W. N. M.; Wijkens, P.; Grove, D. M.; van Koten, G. Nature
1994, 372, 659-663. (b) Annis, D. A.; J acobsen, E. N. J . Am. Chem.
Soc. 1999, 121, 4147-4154.
(12) 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.
(13) Brussard, M. E.; J uma, B.; Train, S. G.; Peng, W.-J .; Laneman,
S. A.; Stanley, G. G. Science 1993, 260, 1784-1788.
(14) Obrecht, D.; Villalgordo, J . M. Solid-supported combinatorial
and parallel synthesis of small-molecular weight compound libraries;
Pergamon: Oxford, 1998; Vol. 17.
(15) Brown, R. C. D. J . Chem. Soc., Perkin Trans. 1 1998, 3293-
3320.
(16) Hermkens, P. H. H.; Ottenheijm, H. C. J .; Rees, D. C. Tetra-
hedron 1997, 53, 5643-5678.
(17) Hermkens, P. H. H.; Ottenheijm, H. C. J .; Rees, D. Tetrahedron
1996, 52, 4527-4554.
* To whom correspondence should be addressed. (P.A.)Tel: (613)
993 7014. Fax: (613) 952 0068. E-mail: Prabhat.Arya@nrc.ca. (H.A.)
Tel: (613) 562 5189. Fax: (613) 562 5871. E-mail: halper@uottawa.ca.
(1) (a) NRC publication no. 43823. (b) Undergraduate student from
Thailand on exchange program, 1997-98.
(2) (a) Halm, C.; Kurth, M. J . Angew. Chem., Int. Ed. Engl. 1998,
37, 7, 510-512. (b) Kobayashi, S.; Nagayama, S. J . Am. Chem. Soc.
1998, 120, 2985-2986.
(3) Mehnert, C. P.; Weaver, D. W.; Ying, J . Y. J . Am. Chem. Soc.
1998, 120, 12289-12296.
(4) Herrmann, W. A.; Cornils, B. Angew. Chem., Int. Engl. 1997,
36, 1047-1067.
(5) Chang, H.-T.; Frechet, J . M. J . J . Am. Chem. Soc. 1999, 121,
2313-2314.
(6) Frechet, J . M. J . Science 1994, 263, 1710-1714 and references
therein.
(18) Cong, P.; Doolen, R. D.; Fan, Q.; Giaquinta, D. M.; Guan, S.;
McFarland, E. W.; Poojary, D. M.; Self, K.; Turner, H. W.; Weinberg,
W. H. Angew. Chem., Int. Ed. Engl. 1990, 29, 484-488.
(19) Senkan, S. M. Nature 1998, 394, 350-353.
(20) Porte, A. M.; Reibenspies, J .; Burgess, K. J . Am. Chem. Soc.
1998, 120, 9180-9187.
(7) Tomalia, D. A. Aldrichimica Acta 1993, 26, 91-101 and refer-
ences therein.
(8) McElhanon, J . R.; McGarth, D. V. J . Am. Chem. Soc. 1998, 120,
1647-1656.
10.1021/jo991621h CCC: $19.00 © 2000 American Chemical Society
Published on Web 03/01/2000

1882 J . Org. Chem., Vol. 65, No. 6, 2000
Notes
F igu r e 1. G3 Generation, octavalent dendritic ligands on beads.
synthesis of dendritic materials having phosphine ligands
to explore for hydroformylation applications.
compound 1-G1, four with 2-G2, and eight with 3-G3.
Amino groups at each generation were then modified to
obtain mutlivalent phosphine ligands at the surface of
beads.
For dendritic molecules, multistep solution chemistry
presents a synthetic challenge because of the need to
purify products at every stage. With few exceptions, most
of the efforts toward dendritic materials have been made
using solution-phase approaches. No purification is re-
quired in solid-phase synthesis, a major advantage over
the traditional solution chemistry. The reactions are
driven to completion by addition of the excess reagents.
The product is anchored onto the support, and after the
reaction is complete, the excess reagents are removed by
filtration. A pseudopeptide-based building block (Scheme
1, 4b) was utilized for solid-phase synthesis. The building
block, 4b, has two Fmoc-protected amino groups and a
carboxylic acid, and was synthesized on large scale using
solution chemistry. Various generations (1-G1, 2-G2, and
3-G3) of branched, dendritic molecules were then planned
for assembly by solid-phase synthesis. Using this ap-
proach, it is possible to obtain two amino groups with
3,5-Diaminobenzoic acid was coupled with glycine-
benzyl ester using DIC/HOBt at room temperature to
give the benzyl ester derivative in 72% yield. It was then
coupled with Fmoc-phe-OH using similar reaction condi-
tions. Di-Fmoc protected pseudopeptide benzyl ester was
purified by silica gel chromatography (85% yield) and
then subjected to hydrogenation using 10% Pd/C to obtain
di-Fmoc-protected pseudopeptide carboxylic acid, 4b, that
was directly used for solid-phase synthesis. All the
compounds were well characterized using NMR and MS-
electrospray. Fmoc-Rink amide MBHA resin (Novabio-
chem, loading 0.45 mmol/gm) was treated with 20%
piperidine in DMF to remove the Fmoc group. Di-Fmoc-
protected pseudopeptide building block 4b was coupled
with the free amino group on the resin using a DIC/HOBt
coupling method (4.0 equiv of 4b, 4.0 equiv of DIC, 4.0

Notes
J . Org. Chem., Vol. 65, No. 6, 2000 1883
Sch em e 1. Retr osyn th etic a n a lysis: solid -p h a se syn th esis of d en d r itic liga n d s on bea d s
equiv of HOBt, 8.0 equiv of DIPEA in DMF, 12 h) to give
compound 6 (G1) (Scheme 2). Compound 6 was cleaved
from a known amount of resin, isolated and purified by
reverse-phase HPLC. The yield of the coupling was
determined to be 85% calculated from the cleavage of the
known amount of the resin. Second generation based
compound 8 (G2) was synthesized from 6 on the solid
phase. Compound 6 was treated with 20% piperidine in
DMF to remove the Fmoc groups. Free amino groups
were then coupled with the building block 4b as described
earlier (8.0 equiv of 5, 8.0 equiv of DIC, 8.0 equiv of
HOBt, 16.0 equiv of DIPEA in DMF, 16 h) to obtain the
second-generation pseudopeptidic dendrimer, 8 (G2). As
in the case of the first-generation compound 6, the yield
was calculated to be 82% from the isolated product
obtained after the cleavage from the resin followed by
purification by reversed-phase HPLC. A repeat process
gave the third-generation product 10 (G3) respectively.
Phosphonated rhodium complexed, pseudopeptide den-
drimers 7, 9, and 11 were prepared from 6, 8, and 10 as
follows. Gen er al P r ocedu r e: A solution of diphenylphos-
phine (10 mmol) and paraformaldehyde (10 mmol) in
degassed methanol (20 mL) was stirred at 70 °C for 45
min for in situ preparation of diphenylphosphinometha-
nol. The reaction mixture was cooled to room tempera-
ture. A degassed toluene suspension of each generation
dendrimer, 6, 8, and 10 with free amino groups anchored
on solid support (Fmoc group was first removed from 6,
8, and 10 by the treatment with 20% piperidine), was
added to the above mixture (6.0 equiv of diphenylphos-
phinomethanol per free amino group). The mixture was
refluxed for 2 h and left at room temperature for 17 h.
The resin was washed with degassed methanol. The
resulting phosphonated dendrimers, 7, 9, and 11 were
characterized by 31P NMR (chemical shifts: -26 to -28
ppm). No sign of the phosphine oxide was detected from
31P NMR. Each generation of the phosphonated den-
drimer, 7, 9, and 11, was then independently complexed
to rhodium by the reaction with choloro(dicarbonyl)-
rhodium(I) dimer in dichloromethane. Different genera-
tion-based rhodium complexed dendrimers 1 (G1), 2 (G2),
and 3 (G3) on resins were obtained by filtration. Once
again, the formation of a rhodium-phosphine complex
was characterized by 31P NMR (chemical shifts: 24-26
ppm). The 31P NMR confirmed the complete coordination
of rhodium to phosphonated ligands in all the three
generations. Rhodium complex dendrimers anchored onto
beads, 1 (G1), 2 (G2), and 3 (G3) were tested as catalysts
for the hydroformylation reaction of several olefins.
Initial studies were performed with styrene as a sub-
strate (Table 1). In all the cases, the experiments were
carried out on 10.0 mmol scale of the substrate with 25
mg of the catalyst in dichloromethane at 1000 psi total
pressure CO/H2 (1:1). Using catalyst 2 (G2) at 45 °C,
complete conversion occurred to the product (>99%) with
a high selectivity for the branched isomer (branched:
linear, 16:1) was obtained. Second-generation dendrimer
on beads, 2 (G2), was found to be more reactive than the
first generation. The catalytic activity for the second-
generation dendrimer remained the same up to the fifth
cycle; however, a significant drop in the reactivity with
the first generation was observed during the fifth cycle.
In a second set of experiments, the hydroformylation
of several olefinic substrates was studied using compound
2 (G2) as a catalyst. The results are summarized in Table
2. Among all of the olefins tested, product formation was
very high up to the third cycle. Significantly, high
selectivity for the branched product (first cycle, 40:1;
second cycle, 60:1; third cycle, 22:1) was obtained with
vinyl benzoate as a substrate. As with styrene, the

1884 J . Org. Chem., Vol. 65, No. 6, 2000
Sch em e 2. Solid -P h a se Syn th esis of Den d r itic Liga n d s on Bea d sa
Notes
a
(a) (i) 20% piperidine, DMF, (ii) 4.0 equiv of 4b, 4.0 equiv of DIC, 4.0 equiv of HOBt, 8.0 equiv of DIPEA, DMF, 12 h; (b) repeat step
a (i), (ii) 8.0 equiv of 4b, 8.0 equiv of DIC, 8.0 equiv of HOBt, 16.0 equiv of DIPEA, DMF, 16 h; (c) repeat step a (i); (ii) 16.0 equiv of 4b,
16.0 equiv of DIC, 16.0 equiv of HOBt, 24.0 equiv of DIPEA, DMF, 20 h; (d) repeat step a (i); (ii) 10 mmol of HCHO and Ph2PH in
degassed methanol at 70 °C, 45 min (in situ preparation of diphenylphosphinomethanol), mix with the dendrimer with free amino groups
on beads, 2 h (reflux), 24 h (stirred at room temperature); (e) [Rh(CO)2Cl]2, (1.0 equiv per bisphosphinomethylamino group) dichloromethane,
room temperature, 14 h.
Ta ble 1. Hyd r ofor m yla tion of Styr en e (10 m m ol) w ith 25
Ta ble 2. Hyd r ofor m yla tion of Olefin s (10 m m ol) w ith 25
m g of Rh od iu m -Com p lexed Den d r im er s An ch or ed on to
m g of Rh od iu m Com p lexed Den d r im er s An ch or ed on to
Bea d s 1 (G1), 2 (G2), a n d 3 (G3)a
Bea d s 2 (G2) a t 65 °Ca
substrate
cycle time (h) conversion (%) B/L ratiob
vinyl acetate
vinyl acetate
vinyl acetate
vinyl benzoate
vinyl benzoate
vinyl benzoate
p-methoxy styrene 1st
p-methoxy styrene 2nd
t-butyl styrene
1st
10
22
22
10
36
22
10
36
22
>99
>99
>99
>99
>99
>99
>99
>99
>99
15:1
15:1
15:1
40:1
60:1
22:1
10:1
11:1
8:1
2nd
3rd
1st
2nd
3rd
time
(h)
T
(°C)
conversion
(%)
catalyst
cycle
branched/linearb
1 (G1)
1 (G1)
1 (G1)
1 (G1)
1 (G1)
1 (G1)
2 (G2)
2 (G2)
2 (G2)
2 (G2)
2 (G2)
2 (G2)
2 (G2)
2 (G2)
3 (G3)
3 (G3)
3 (G3)
3 (G3)
3 (G3)
3 (G3)
1st
5
18
23
24
24
24
21
5
16
22
22
22
22
22
22
22
22
22
22
22
65
65
65
65
65
65
25
65
45
65
65
65
65
65
65
65
65
65
65
65
42
70
85
51
22
8
35
57
>99
>99
>99
>99
98
88
99
99
99
13:1
12:1
10:1
15:1
9:1
10:1
35:1
14:1
16:1
11:1
12:1
12:1
11:1
12:1
10:1
9:1
3rd
2nd
3rd
4th
5th
6th
1st
a
Reaction conditions: substrate (10.0 mmol), CO(1000 psi), H2
b
(1000 psi), catalyst (25 mg). Ratio of branched/linear aldehydes
was determined by 1H NMR.
reactions were carried out using 25 mg of the catalyst
on 10.0 mmol substrate.
1st
1st
To summarize, rhodium complexes with dendritic,
phosphine ligands anchored onto beads were found to be
excellent catalysts for the hydroformylation of several
olefins. In some cases (i.e., 2, second generation) catalysts
are reactive up to several cycles. It is generally believed
that polymer supported catalysts are relatively less
reactive compared to homogeneous catalysis. Contrary
to this, our studies strongly support the possibility of
achieving high reactivities in heterogeneous systems. The
origin of the high reactivity is not clear at this stage. One
may speculate that well-exposed ligands on the outer-
core may account for the high reactivity as compared to
the irregular-shaped polymeric materials. Cooperativity
2nd
3rd
4th
5th
6th
1st
2nd
3rd
4th
5th
6th
12:1
10:1
12:1
12:1
99
78
47
a
Reaction conditions: styrene (10.0 mmol), CO(1000 psi), H2
b
(1000 psi), catalyst (25 mg). Ratio of branched/linear aldehydes
was determined by 1H NMR.

Notes
J . Org. Chem., Vol. 65, No. 6, 2000 1885
the solvent was evaporated to give the building block 4b in 95%
yield: 1H NMR (DMSO-d6) δ 2.88-2.94 (m, 2H), 3.08-3.17 (m,
2H), 3.76 (s, 2H), 4.10-4.18 (m, 6H), 4.45-4.54 (m, 2H), 7.08-
7.94 (m, 29H), 8.31 (bs, 2H) and 10.60 (bs, 2H); 13C NMR (DMSO-
d6) δ 39.8, 45.0, 47.4, 58.0, 66.6, 114.1, 120.9, 122.3, 128.9, 129.7,
130.2, 138.2, 140.2, 141.5, 141.5, 144.5, 144.6, 156.8 and 171.7;
LRMS (electrospray, positive ion mode, m/z) for C57H49N5O9 948
(MH+).
may be the other factor playing a role for high reactivi-
ties. This will be the subject of further investigations.11
Exp er im en ta l Section
Gen er a l Rem a r k s. All reagents (DMF, DIPEA, CH2Cl2,
MeOH, piperidine, and hexanes) were purchased from Aldrich
as anhydrous grade and used as supplied. The reactions were
carried out under argon or nitrogen atmosphere. Purification of
compounds by flash chromatography was performed using
recycled silica gel (230-400 mesh, 60 Å) supplied by Silicycle
(Que´bec, Canada).
Gen er a l P r oced u r e for Solid -P h a se Syn th esis. Rink
amide MBHA resin (loading 0.4-0.5 mmol/g, 5) was suspended
in DMF for 30-45 min under argon. After filtration, a solution
of 20% piperidine in DMF was added and the mixture strirred
for 40 min (two cycles). This was followed by the addition of the
building block (4b, 4.0 equiv), DIC (4.0 equiv), HOBt (4.0 equiv),
and DIPEA (8.0 equiv), and it was stirred for 12 h. After
filtration, compound 6 (G1) was subjected to the Fmoc group
removal as discussed before. In a separated reaction flask,
formaldehyde (10 mmol) and diphenylphosphine (10 mmol) in
degassed MeOH (20 mL) were stirred at 70 °C for 45 min under
argon (in situ preparation of diphenylphosphinomethanol).
Beads from the solid-phase synthesis were transferred to this
solution, and the mixture was stirred for 24 h at room temper-
ature. The beads were filtered under argon and further stirred
with [Rh(CO)2Cl2] (1.0 equiv per bisphosphinomethylamino
group) in CH2Cl2 for 14 h at room temperature. The synthesis
of the higher generation dendritic compounds anchored on to
beads 8 (G2) and 9 (G3) was obtained from 6 (G1), followed by
the alkylation of the amino groups to obtain phosphonated
ligands, and subsequently, complexation with the Rh as dis-
cussed before.
Gen er a l P r oced u r e for Hyd r ofor m yla tion of Styr en e
a n d Oth er Su bstr a tes. A solution of styrene (10.0 mmol) in
CH2Cl2 (10-20 mL) was treated with a 1:1 mixture of carbon
monoxide and hydrogen in the presence of Rh-based catalytsts
(25 mg) at 1000 psi total pressure. The reaction was studied at
several temperatures with variation in time period, and the ratio
of the products(s) was determined by 1H NMR and GC (see,
Table 1 for the effect of the temperature and time on the product-
(s) yields as well as for the ratio of the branched vs linear
aldehydes as products).
Syn th esis of Bu ild in g Block 4b. (i) A solution of 3,5-
Diaminobenzoic acid (50 mmol), GlyOBn (60 mmol), DCC (60
mmol), HOBt (60 mmol) and DIPEA (60 mmol) in CH2Cl2/DMF
(1:1, 100 mL) was stirred under nitrogen at room temperature
for 14-16 h. The precipitated solid was filtered and washed
thoroughly with CH2Cl2 and the mother liquor collected, washed
with saturated solution of NaHCO3, and evaporated to give the
crude product. 3,5-Diamino-N-benzamide glycine benzyl ester
was obtained (72% yield) after purification by flash chromatog-
raphy (CH2Cl2/MeOH, 20:1): 1H NMR (DMSO-d6) δ 7.35 (m, 5H,
Ar-H), 5.75 (s, 2H, -OCH2Ar), 4.95 (bs, 1H, OdCNHCH2), 3.95
(bs, 4H, ArNH2) and 3.3 (m, 2H, -NHCH2CdO-); LRMS (FAB,
positive ion mode, m/z) C16H17N3O3 299.1 (M+). (ii) To a solution
of Fmoc-Phe-OH (60 mmol) in CH2Cl2 (100 mL) were added DCC
(60 mmol), HOBt (60 mmol) and 3,5-diamino-N-benzamide
glycine benzyl ester (20 mmol). The mixture was stirred under
nitrogen at the room temperature for 30-40 h. The precipitated
solid was filtered and washed thoroughly with CH2Cl2 and the
mother liquor collected, washed with a saturated solution of
NaHCO3, and evaporated to give the crude product. The benzyl
ester derivative of the building block (4a ) was obtained in 76%
isolated yield after purification: 1H NMR (DMSO-d6) δ 2.89 (m,
2H), 3.05-3.10 (m, 2H), 4.00-4.24 (m, 2H), 4.42-4.48 (m, 2H),
5.17 (s, 2H), 7.08-7.88 (m, 34H), 8.31 (bs, 2H) and 10.34 (bs,
2H); 13C NMR (DMSO-d6) δ 38.2, 42.4, 47.4, 60.6, 66.6, 66.7,
114.5, 120.9, 126.1, 126.2, 127.3, 127.9, 128.5, 128.7, 128.9, 129.0,
129.3, 136.1, 136.8, 138.8, 140.0, 141.5, 141.5, 144.6, 144.6, 156.8,
167.8, 170.6 and 171.7; LRMS (electrospray, positive ion mode,
m/z) for C64H55N5O9 1038 (MH+). This was subjected to the
debenzylation conditions to obtain the building block 4b, re-
quired for solid-phase synthesis. (iii) The benzyl derivative of
the building block 4a (10 mmol) was dissolved in DMF (50 mL)
and hydrogenated in the presence of 10% Pd over carbon (10
mol %) for 30-45 min. The catalyst was filtered over Celite, and
Ack n ow led gm en t. We thank DuPont Canada and
NRC/NSERC Research partnership for the financial
support.
J O991621H