First synthesis of a highly basic dendrimer and its catalytic application in organic methodology

Article abstract of DOI:10.1002/adsc.200404089

Sixteen OCH₂CH₂N₃P(i-BuNCH ₂CH₂)₃N substituents (A) containing the highly basic bicyclic azido-phosphine moiety shown, have been incorporated into the dendrimer [CH₂CH₂

Full text of DOI:10.1002/adsc.200404089

COMMUNICATIONS
First Synthesis of a Highly Basic Dendrimer and its Catalytic
Application in Organic Methodology
Arunkanti Sarkar, Palanichamy Ilankumaran, Philip Kisanga, John G. Verkade*
Iowa State University, Department of Chemistry, 1275 Gilman Hall, Ames, Iowa 50011, USA
Fax: (þ1)-515-294-0105, e-mail:jverkade@iastate.edu
Received: March 19, 2004; Accepted: July 29, 2004
Abstract: Sixteen OCH2CH2N3P(i-BuNCH2CH2)3N
substituents (A) containing the highly basic bicyclic
azido-phosphine moiety shown, have been incorpo-
rated into the dendrimer [CH2CH2N[(CH2)3
N[(CH2)3NHC(O)-3,5-bis-(A)-C6H3]2]2]2 in five steps
from commercially available starting materials with
an overall yield of 22.5%. Also presented are exam-
ples of Michael additions, nitroaldol reactions and
aryl isocyanate trimerizations that are efficiently cat-
alyzed by the dendrimer.
bromoethane in the presence of 18-crown-6 and
K2CO3 gave the dibromoester 4 in 87% yield.
Although a few procedures have been reported[12] for
the conversion of a meta-dihydroxybenzene (namely, 1-
benzyloxy-3,5-dihydroxybenzene) to the corresponding
3,5-bis(2-bromoethyloxy) derivative, none of them af-
forded the 3,5-bis(2-bromoethyloxy) derivative 4 as a
clean product. We therefore developed a modified pro-
cedure in which dibromoethane is used as the solvent in
Keywords: catalyst; dendrimers; organic catalysis;
phosphine imine; proazaphosphatrane
Dendrimer synthesis has fascinated chemists for more the presence of finely powdered K2CO3 to give 4, which
than a decade.[1] Phosphorus-containing dendrimers was pure by 1H and 13C NMR spectroscopy. Conversion
make up an active subarea[2] which have materials, bio- of 4 to 5 in 93% yield, followed by saponification and
logical and catalysis applications.[3] To our knowledge, acidification gave diazido acid 6 in 90% yield, which
however, no reports of catalytic reactions promoted by was then condensed with DAB-Am-8 dendrimer to af-
phosphorus-containing dendrimers have appeared in ford 7 in 35% yield. The hexadeca-azido dendrimer 7
which the phosphorus-containing moiety itself functions was found to be pure by NMR spectroscopy, although
as the active site. We disclose here the first such example its polydispersity value (Mw/Mn ¼1.159 versus a poly-
and show that it is an efficient catalyst for useful organic styrene standard) obtained by GPC analysis was not ide-
transformations such as Michael addition, aldol conden- al (1.00), presumably due to aggregation that occurred
sation and isocyanate trimerization. Acyclic phosphine in the carrier solvent (THF).[13] Molecular weights esti-
¼
imines [RN P(NR2)3] constitute a class of strong non- mated by GPC relative to a polystyrene standard de-
ionic bases[4] wherein C6H5N P(NMe2)3, for example, pend on the shape and flexibility of the molecules and
¼
displays a pKa of 20.9 in acetonitrile.[5] Such phosphorus were considerably lower than calculated values, as was
imines have been found to be very useful in amino acid previously experienced with a variety of dendrimers.[14]
synthesis.[6,7] We have found that bicyclic 1a is a powerful
Attempts to convert compound 7 to the dendrimeric
catalyst for acylation,[8] Michael addition[9] and isocya- base 8a were not successful because combination of sol-
nate trimerization.[10] We were thus attracted to the pos- utions of 2a and azide 7 in benzene [Equation (1)] pro-
sibility of grafting such a bicyclic base to a dendrimeric duced an intractable precipitate. Using an excess of 2a
framework. The synthetic approach we report here is or changing the solvent to THF did not solve this prob-
based upon a previous study[11] in which we found that lem.
1a is stable in the azide form[11a] whereas the analogous
Subsequently we found that compound 2a is a suffi-
ꢀ
reaction with phenyl azide gives the iminophosphorane ciently strong base to abstract an N H proton from aro-
1b.[11b]
The strategy we chose for assembling the dendrimeric
framework is shown in Scheme 1. In order to realize a
substantial number of basic sites in the molecule, we de-
cided to employ the diazido acid 6 to effect branching.
Methyl 3,5-dihydroxybenzoate 3 on treatment with di-
Adv. Synth. Catal. 2004, 346, 1093–1096
DOI: 10.1002/adsc.200404089
ꢀ 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1093

Arunkanti Sarkar et al.
COMMUNICATIONS
steric requirements in the four-membered ring transi-
tion state are much more rigorous than those in the ad-
dition transition state.[17] Donor character on the part of
the P(III) substituents stabilizes azidophosphine inter-
mediates and this factor also apparently operates in 8b
to give it thermal stability. The unusual resistance of
8b to thermolysis may also be in part due to the rigidity
of the cage structure, which by virtue ofthe planargeom-
etry around the PN3 cage nitrogens,[18] maintains an iso-
butyl group in close proximity to the phosphorus-azide
moiety.
Scheme 1.
matic amide functions in the dendrimer.[15] To circum-
To our knowledge, 8b is the first example of a den-
vent this difficulty, 2b was employed, hoping that the drimer bearing highly basic phosphine sites. With this
greater bulk provided by the three iso-butyl groups dendrimer in hand, we carried out preliminary studies
would inhibit penetration of this base to amide functions on its efficacyin catalysis. Michaeladdition, a highly use-
within the dendrimer molecule, but would still allow 2b ful tool in organic synthesis, is efficiently catalyzed by
to react with the azide functions on the dendrimer sur- 8b. Hence the reaction of methyl acrylate with ketoester
face. We were successful in this regard, although the 9 (Table 1) and with methyl diethyl malonate (11) gave
product of this reaction is the azidophosphine 8b which the Michael adducts 10 and 12, respectively, in nearly
even upon heating in refluxing benzene did not decom- quantitative yields. A tandem Michael/aldol reaction
pose to the corresponding phosphine imine. The NMR between ketoester 13 and acrolein were also catalyzed
spectra (1H, 13C, 31P), polydispersity value (Mw/Mn ¼ by dendrimer 8b to give bicyclic 14 in good yield. Den-
1.115 versus a polystyrene standard) and the elemental drimer 8b catalyzed the aldol reaction of nitromethane
analysis of dendrimer 8b are all consistent with its with benzaldehyde very efficiently to give 16. We also in-
good purity and proposed formulation.[16] The lack of a vestigated 8b for nucleophilic catalytic properties in the
peak for 2b in the 31P NMR spectrum of the dendrimer trimerization of isocyanates to isocyanurates.[10] In addi-
is consistent with the absence of 2b trapped within the tion to their use as an additive in nylon-6 manufacture,
dendrimer matrix.
the polymer industry actively seeks isocyanurate-form-
The Staudinger reaction is a two-step process involv- ing catalysts for the preparation of heat and fire-resist-
ing the initial electrophilic addition of an alkyl or aryl ant polymers. Here we find that the azido dendrimer
azide to a P(III) center followed by N2 elimination 8b catalyzes the trimerization of phenyl isocyanate, as
from the intermediate phosphazide to give the corre- well as trimerization of p-chlorophenyl isocyanate to
sponding iminophosphine.[17] Steric hindrance at the the corresponding isocyanurates in 90 and 89% yields,
P(III) center does not interfere with the electrophilic ad- respectively (Table 1). The reactions in Table 1 required
dition step, but it does suppress decomposition, because only 0.24 mol % of catalyst 8b, indicating that this den-
1094
ꢀ 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2004, 346, 1093–1096

Synthesis of a Highly Basic Dendrimer
COMMUNICATIONS
chloride (200 mg) in 75 mL of acetone was refluxed for 24 h.
Acetone was evaporated under vacuum, and then the residue
was dissolved in water and extracted twice with 200 mL of
ether. The ether layer was dried over Na2SO4 and the volatiles
were removed under reduced pressure to afford the diazido es-
ter 5; yield: 5.7 g (93%); 1H NMR (CDCl3, 300 MHz): d¼7.23
(d, J¼3.0 Hz, 2H), 6.70 (t, J¼3.0 Hz, 1H), 4.18 (t, J¼6.0 Hz,
4H), 3.91 (s, 3H), 3.61 (t, J¼6.0 Hz, 4H); 13C NMR (CDCl3,
75.5 MHz): d¼166.4, 159.2, 132.2, 108.2, 106.8, 67.2, 52.3,
50.0; LR-MS: m/z¼324 (MþNH4), 307, 281, 255; HR-MS:
calcd. for C12H14N6O4: 306.1077; found: 306.1080.
drimer is a more potent catalyst than 2a or 1a, of which
10 mol % and 0.33 mol % were required for Michael[19]
and isocyanate trimerization reactions,[9] respectively.
The increased potency of 8b is not entirely unexpected
in view of the presence of sixteen catalytic sites per mol-
ecule.
Table 1. Reactions catalyzed by dendrimer 8b.[a]
Diazido Acid (6)
To a stirred solution of diazido ester 5 (5.7 g, 19 mmol) in
MeOH (30 mL) was added NaOH (4.5 g) in water (15 mL). Af-
ter stirring for 6 h, MeOH was evaporated under reduced pres-
sure, and then water (50 mL) was added. The aqueous layer, on
careful acidification with 2 N HCl, precipitated acid 6, which
was extracted with dichloromethane (50ꢁ3 mL). After drying
the extract with Na2SO4, removal of the volatiles under vacuum
afforded diazido acid 6 as a pale yellow solid; yield: 4.9 g (90%);
1H NMR (CDCl3, 300 MHz): d¼7.29 (d, J¼3.0 Hz, 2H), 6.76
(t, J¼3.0 Hz, 1H), 4.20 (t, J¼3.0 Hz, 4H), 3.63 (t, J¼3.0 Hz,
4H); 13C NMR (CDCl3, 75.5 MHz): d¼171.7, 159.3, 131.3,
108.7, 107.9, 67.3, 50.0; LR-MS: m/z¼310 (MþNH4), 267,
241, 198; HR-MS: calcd. for C11H12N6O4: 292.0920; found:
292.0924.
Hexadeca-azido Dendrimer (7)
[a]
To a stirred solution of DCC (988.0 mg, 4.80 mmol) and azido
acid 6 (1.4 g, 4.8 mmol) in dichloromethane (10 mL) was added
DAB-Am-8 Generation 2 polypropyleneimine octamine den-
drimer (309.2 mg, 0.400 mmol, Aldrich). After stirring for
24 h, precipitated dicyclohexylurea was filtered and washed
with dichloromethane (50 mL) and then the solvent was re-
moved from the filtrate under reduced pressure. The residue
was purified by silica gel chromatography by eluting succes-
sively with ethyl acetate, a 20:80 mixture of MeOH/ethyl ace-
tate and then a 5:25:70 mixture of triethylamine/MeOH/ethyl
acetate to give product 7 as a glassy solid; yield: 419 mg (35%);
1H NMR (CDCl3, 300 MHz): d¼7.81 (s, 8N-H), 7.0 (d, J¼
3.0 Hz, 16H), 6.56 (t, J¼3.0 Hz, 8H), 4.07 (t, J¼3.0 Hz,
32H), 3.44 (t, J¼3.0 Hz, 32H) 3.43–3.60 (m, 48H), 2.20–2.60
(m, 36H), 1.24–1.70 (m, 28H); 13C NMR (CDCl3, 75.5 MHz):
d (aromatic carbons only)¼166.9, 159.4, 137.0, 106.1, 104.8;
LR-MS (ESI): m/z¼2965, 1483 (M2þ), 989 (M3þ), 741 (M4þ).
Reactions were carried out in acetonitrile at room tem-
perature using 0.24 mol % of 8b unless stated otherwise.
[b]
No solvent was used (see Ref.[9]).
Experimental Section
Dibromo Ester (4)
A stirred suspension of methyl 3,5-dihydroxybenzoate (10.1 g,
60.0 mmol), finely powdered K2CO3 (20.7 g, 150 mmol) and
18-crown-6 (1 g) in 120 mL of 1,2-dibromoethane was heated
at 808C for 36 h. The reaction mixture was cooled, filtered,
washed with methylene chloride (100 mL) and then the filtrate
was evaporated under vacuum giving a residue that was puri-
fied by column chromatography using silica gel with 20% ethyl
acetate in hexanes as eluent to give pure dibromo ester 4; yield:
1
14.2 g (87%); H NMR (CDCl3, 300 MHz): d¼7.22 (d, J¼
¼
3.0 Hz, 2H), 6.70 (t, J 3.0 Hz, 1H), 4.32 (t, J¼6.0 Hz, 4H),
Azidophosphine Dendrimer (8b)
3.91 (s, 3H), 3.64 (t, J¼6.0 Hz, 4H); 13C NMR (CDCl3,
75.5 MHz): d¼166.4, 159.2, 132.3, 108.4, 107.2, 68.1, 52.3,
28.9; LR-MS: m/z¼382 (Mþ), 274, 195, 109, 107; HR-MS:
calcd. for C12H14Br2O4: 381.9240; found: 381.9244.
To azide 7 (41 mg, 0.013 mmol) dissolved in 5 mL of dry ben-
zene was added 2b (945.0 mg, 0.55 mmol) in 2 mL of dry ben-
zene under argon. The reaction mixture was stirred at room
temperature for 36 h and then benzene was evaporated to
give a paste. Dry pentane (10 mL) was added and stirring
was continued for 1 h to give a slurry. The solvent was removed
with a cannula, an additional 10 mL of dry pentane were added
and stirring was continued for 30 min. This process was repeat-
ed three times after which the solid residue was dried under
Diazido Ester (5)
A stirred solution of dibromo ester 4 (7.64 g, 20.0 mmol), so-
dium azide (7.8 g, 120 mmol) and benzyltriethylammonium
Adv. Synth. Catal. 2004, 346, 1093–1096
asc.wiley-vch.de
ꢀ 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1095

Arunkanti Sarkar et al.
COMMUNICATIONS
1
vacuum to give the product; yield: 91 mg (88%); H NMR
(CDCl3, 300 MHz): d¼9.27 (broad s, 8H), 7.75 (s, 16H), 6.96
(s, 8H), 4.57 (broad s, 32H), 4.30 (broad s, 32H), 3.65–3.80
(m, 16H), 2.30–3.0 (m, 340H), 2.20 (septet, 48H), 2.03 (broad s,
8H), 1.76 (broad s, 4H), 1.06 (d, 288H); IR (nujol): n¼3235,
2923, 1650, 1591, 1541, 1431, 1386, 1332, 1115, 1076, 1025,
[6] M. J. OꢁDonnell, C. Zhow, W. L. Scott, J. Am. Chem. Soc.
1996, 118, 6070.
[7] E. Dominguez, M. J. OꢁDonnell, W. L. Scott, Tetrahe-
dron Lett. 1998, 39, 2167.
[8] P. Ilankumaran, J. G. Verkade, J. Org. Chem. 1999, 64,
9063.
[9] J. Tang, T. Mohan, J. G. Verkade, J. Org. Chem. 1994, 59,
4931.
[10] J. Tang, J. G. Verkade, Angew. Chem. Int. Ed. Engl. 1993,
32, 896.
[11] a) B. F. Fetterly, J. G. Verkade, in preparation; b) J. Tang,
J. Dopke, J. G. Verkade, J. Am. Chem. Soc. 1993, 115,
5015.
861 cmꢀ1
;
31P NMR (C6D6): d¼36.0; elemental analysis: calcd.
C 59.08, H 9.46, N 20.98; found: C 58.03, H 9.39, N 19.91.
The 1H NMR spectrum of 10 compared favorably with that
of an authentic sample. The H NMR spectra of 12,[20] 14,[21]
1
16,[22] 18[23] and 20[23] compared favorably with those reported
in the cited references.
[12] H. F. Chow, Z. Y. Wang, Y. F. Lau, Tetrahedron 1998, 54,
13813.
Acknowledgements
[13] a) M. J. Laufersweiler, J. M. Rohde, J. L. Chaumette, D.
Sarazin, J. R. Parquette, J. Org. Chem. 2001, 66, 6440;
b) I. S. Stupp, R. E. Zubarev, J. Am. Chem. Soc. 2002,
124, 5763.
[14] E. R. Zubarev, S. I. Stupp, J. Am. Chem. Soc. 2002, 124,
5762.
The authors thank the National Science Foundation and also the
Donors of the Petroleum Research Fund administered by the
American Chemical Society for grant support of this work.
They also thank ISU Chemistry Department members Mr.
Erik Hagberg in Prof. Valerie Sheares-Ashbyꢀs group and Mr.
Mike Rodriquez in Prof. Dan Armstrongꢀs group for obtaining
the GPC and the mass spectrometric data.
[15] While a 1H NMR spectrum of N-methylbenzamide
shows a doublet for the CH3 protons due to coupling
with the N-H proton, a 1:1 mixture of N-methylbenza-
mide and 2a shows a singlet for the CH3 protons. A
31P NMR spectrum of this mixture displays a single
peak for 2aHþ.
[16] Although success was attained in obtaining the mo-
lecular weight of 7 by ESI-mass spectrometry, this was
not the case for 8b despite exhaustive attempts with
MALDI and ESI experiments using a variety of condi-
tions and matrices.
[17] a) Y. G. Gololobov, I. N. Zhmurova, L. F. Kasukhin, Tet-
rahedron 1981, 37, 437 and references cited therein;
b) Y. G. Gololobov, L. F. Kasukhin, Tetrahedron, 1992,
48, 1353 and references cited therein; c) C. Widauer,
H. Grutzmacher, I. Shevchenko, V. Gramlich, Eur. J. In-
org. Chem. 1999, #38#1659.
[18] The postulate of virtual planarity around the PN3 nitro-
gens in 8b is supported by the near planarity of the
PN3 nitrogens observed in the structures (determined
by X-ray means) of the isoelectronic compounds
References and Notes
[1] For leading reviews on dendrimers, see: a) G. R. New-
kome, C. N. Morrefield, F. Vögtle, Dendritic Molecules:
Concepts Synthesis, Perspectives; VCH: Weinheim, 1996;
b) M. Fischer, F. Vögtle, Angew. Chem. Int. Ed. 1999, 38,
884; c) A. M. Bosman, H. M. Janssen, E. W. Meijer,
Chem. Rev. 1999, 99, 1665; d) S. Hecht, J. M. J. Frechet,
Angew. Chem. Int. Ed. 2001, 40, 74; e) S. E. Stiriba, H.
Frey, R. Haag, Angew. Chem. Int. Ed. 2002, 41, 1329
and references cited therein.
[2] a) J. P. Majoral, A. M. Caminade, Chem. Rev. 1999, 99,
845; b) F. L. Derf, E. Levillain, G. Trippe, A. Gorgues,
M. Salle, R. M. Sebastian, A. M. Caminade, J. P. Majoral,
Angew. Chem. Int. Ed. 2001, 40, 224; c) A. Balueva, S.
Merino, A. M. Caminade, J. P. Majoral, J. Organometallic
Chem. 2002, 112, 643 and references cited therein; d) J. P.
Majoral, A. M. Caminade, V. Maraval, Chem. Commun.
2002, 2929; e) A. M. Caminade, V. Maraval, R. Laurent,
C. O. Turrin, P. Sutra, J. Leclaire, L. Griffe, P. Marchand,
C. Baudoin-Dehoux, C. Rebout, J. P. Majoral, Compt.
Rend. Chim. 2003, 6, 791, this reference is a review on
phosphorus dendrimers and the other reviews in this
journal issue are devoted to other aspects of dendrimers.
[3] a) I. Cuadrado, M. Moran, C. M. Casado, B. Alomso, J.
Losada, Coord. Chem. Rev. 1999, 193, 395; b) C. Kollner,
B. Pugin, A. Togni, J. Am. Chem. Soc. 1998, 120, 10274;
c) V. Maraval, R. Laurent, A. M. Caminade, J. P. Majo-
ral, Organometallics 2000, 19, 4025.
¼
O P(MeNCH2CH2)3N: J.-S. Tang, T. Mohan, J. G. Ver-
¼
kade, J. Org. Chem., 1994, 59, 4931; and S P(MeNCH2
CH2)3N: S.-K. Xi, C. Lensink, J. G. Verkade, Inorg.
Chem. 1990, 29, 2214.
[19] P. B. Kisanga, P. Ilankumaran, B. M. Fetterly, J. G. Ver-
kade, J. Org. Chem. 2002, 67, 3555.
[20] L. S. Hegedus, W. H. Darlington, J. Am. Chem. Soc.
1980, 102, 4980.
[21] M. Filippini, R. Faure, J. Rodriguez, J. Org. Chem. 1995,
60, 6872.
[22] S. Kiyooka, T. Tsutsui, H. Maeda, Y. Kaneko, K. Isobe,
Tetrahedron Lett. 1995, 36, 6531.
[4] a) R. Schwesinger, Chimia, 1985, 39, 296; b) R. Schwe-
singer, Nachr. Chem. Tech. Lab. 1990, 38, 1214.
[23] See first citation in Ref.[18]
[5] T. Rodina, V. Maemets, I. Koppel, J. Chem. Soc. Perkin
Trans. 1, 2000, 2637.
1096
ꢀ 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2004, 346, 1093–1096