An approach to glycerol dendrimers and pseudo-dendritic polyglycerols [6]

Full text of DOI:10.1021/ja994363e

2954
J. Am. Chem. Soc. 2000, 122, 2954-2955
An Approach to Glycerol Dendrimers and
Pseudo-Dendritic Polyglycerols
conversion of all linear units into dendritic units by postsynthetic
modification.11 An often faced problem remains the realization
12
of exactly analogous structures and sufficient characterization
Rainer Haag,* Alexander Sunder, and Jean-Fran c¸ ois Stumb e´
8a
of the compounds.
Freiburger Materialforschungszentrum und Institut f u¨ r
Makromolekulare Chemie, Albert-Ludwigs-UniVersit a¨ t,
Stefan-Meier-Strasse 21, 79104 Freiburg, Germany
In this contribution we present a new divergent synthesis
leading to dendrimers, with glycerol as building unit. Furthermore,
we describe a two-step approach to the analogous, perfectly
branched structure, designated “pseudo-dendrimer”, by modifica-
tion of well-defined hyperbranched polyglycerols. This strategy
permits a direct comparison between a perfect dendritic molecule
and a pseudo-dendritic polymer derived from precisely the same
building blocks, i.e., a core unit with only dendritic and terminal
units attached.
ReceiVed December 14, 1999
Macromolecules with a branch-on-branch structure, such as
dendrimers and hyperbranched polymers, have attracted increasing
interest over the past few years due to their unique molecular
features and properties.1,2 A promising class of these highly
branched molecules are aliphatic polyether polyols. Due to their
inert building blocks and multiple reactive chain ends they possess
promising potential as supports for catalysts and functional organic
molecules. In addition, their excellent water solubility and
The synthesis of aliphatic polyether polyol dendrimers requires
4
an efficient strategy for the formation of the ether linkages.
Classical Williamson ether synthesis (NaH, alkylhalogenide,
DMF) is often incomplete when polyols are used. This problem,
biocompatibility renders them as valuable compounds for polymer
1
3
therapeutics.3
however, can be overcome by using a phase-transfer catalyst.
For the synthesis of the [G-3] glycerol dendrimer 1 we have
selected a simple iterative two-step process (Scheme 1) based on
allylation of an alcohol and catalytic dihydroxylation of the allylic
double bond. The allylation under phase-transfer conditions
reproducibly leads to quantitative formation of the product even
when polyols are employed.14 Catalytic dihydroxylation of the
double bond with N-methyl morpholine oxide (NMO) as cooxi-
dant15 completes the sequence and leads to the formation of new
glycerol units on every available alcohol functionality (Scheme
So far, only three examples of aliphatic polyether dendrimers
have been reported in the literature. They have been prepared in
4
tedious repetitive multistep syntheses, which are obviously a
limiting factor for many applications. Hyperbranched aliphatic
polyethers, on the other hand, are conveniently prepared in one
step.5,6 Due to the statistics of the polymerization reaction they
are a mixture of isomers, possess a certain polydispersity, and
usually contain varying amounts of linear units in addition to
dendritic and terminal building units. To date, two monomers for
the preparation of hyperbranched aliphatic polyether polyols are
1). Both transformations can be carried out in aqueous phase and
5
6
consume only inexpensive reagents, i.e., allyl chloride and
hydrogen peroxide for the recycling of NMO.15 Starting from
trimethylolpropane (TMP) we obtained a [G-3] glycerol dendrimer
known: 3-hydroxymethyl-3-ethyl oxetane and glycidol. Only
the latter one is commercially available and permits control in
terms of initiator (core) incorporation, molecular weight (M
000-10000) and low polydispersities (M /M < 1.5).
A challenging issue is the comparison of structure-property
n
:
1, with 24 hydroxyl end groups after three cycles (six synthetic
1
w
n
steps) in 75% overall yield. Purification by column chromatog-
raphy was necessary only after the allylation steps. The perally-
lated intermediates tend to cross-link within several days;
therefore, oxygen should be excluded for their storage. The final
product 1 was further purified by dialysis in methanol (benzylated
cellulose membrane, MWCO ) 1000).
profiles of hyperbranched polymers and dendrimers based on the
difference in topology and polydispersity.7,8 To this end, hyper-
branched polymers without linear units are required. So far, two
strategies have been described in the literature: (i) the use of
peculiar monomers, in which the formation of linear units is less
favorable than the formation of dendritic ones9 and (ii) the
,10
The glycerol dendrimer 1 (degree of branching (DB) ) 100%)
possesses only two types of structural units (dendritic and
terminal), while the hyperbranched polyglycerol 2 contains
additional linear groups (either bearing primary or secondary
*
Author for correspondence. E-Mail: haag@fmf.uni-freiburg.de
(
1) For recent reviews, see: (a) Newkome, G. R.; Moorefield, C. N.; V o¨ gtle
F. Dendritic Macromolecules: Concepts, Syntheses, PerspectiVes; VCH:
Weinheim, Germany, 1996. (b) Fischer, M.; V o¨ gtle, F. Angew. Chem., Int.
Ed. 1999, 38, 884. (c) Bosman, A. W.; Janssen, H. M.; Meijer, E. W. Chem.
ReV. 1999, 99, 1665.
6a
hydroxy groups) leading to its lower DB, typically about 60%.
When the above-described allylation and dihydroxylation se-
quence is applied to the hyperbranched polymer all linear and
terminal units are transformed into dendritic units (Scheme 1),
and hence, the polymer 3 consists of terminal and dendritic units
only. We used a hyperbranched polyglycerol 2 with a molecular
(2) For recent reviews, see: (a) Kim, Y. H. J. Polym. Sci. Polym. Chem.
Ed. 1998, 36, 1685. (b) Voit, B. I. Acta Polym. 1995, 46, 87. (c) Fr e´ chet, J.
M. J.; Hawker, C. J. Synthesis and Properties of Dendrimers and Hyper-
branched Polymers. In ComprehensiVe Polymer Chemistry; Agarwal, S. L.,
Russo, S., Eds.; Pergamon Press: Oxford, 1996; 2nd Supplement, p 71.
(3) For example: (a) Ferruti, P.; Knobloch, S.; Ranucci, E.; Duncan, R.;
weight of M
) 1300 (DP
) 15) with a polydispersity M
w n
/M )
n
n
Gianasi, Macromol. Chem. Phys. 1998, 199, 2565. (b) Wroblewski, S.;
Kopeckova, J. Macromol. Chem. Phys. 1998, 199, 2601. (c) Mammen, M.;
Choi, S.-K.; Whitesides, G. M. Angew. Chem., Int. Ed. 1998, 37, 2754.
1.2 and a degree of branching DB ) 58% (thus containing 40%
linear units). The material was prepared from TMP in a one-pot
synthesis, applying the base-catalyzed ring-opening multibranch-
ing polymerization of glycidol on a 500 g scale as described
(4) (a) Padias, A. B.; Hall, H. K.; Tomalia, D. A.; McConnell, J. R. J.
Org. Chem. 1987, 52, 5305. (b) Gitsov, I.; Wu, S.; Ivanova, P. T. Polm. Mater.
Sci. Eng. 1997, 77, 214. (c) Jayaraman, M.; Fr e´ chet, J. M. J. J. Am. Chem.
Soc. 1998, 120, 12996. (d) Grayson, S. M.; Jayaraman, M.; Fr e´ chet, J. M. J.
Chem. Commun. 1999, 1329.
(9) (a) Hobson, L. J.; Feast, W. J. Chem. Commun. 1997, 1877. (b) Hobson,
L. J.; Feast, W. J. Chem. Commun. 1997, 2067. (c) Hobson, L. J.; Feast, W.
J. Polymer 1999, 40, 1279.
(10) Maier, G.; Zech, C.; Voit, B.; Komber, H. Macromol. Chem. Phys.
1998, 199, 2655.
(11) A comparison of dendritic and modified hyperbranched polycarbosi-
lanes was reported in Lach, C.; Frey, H. Macromolecules 1998, 31, 2381.
(12) Comparison of dendrimers with their linear analogues was reported
in Hawker, C. J.; Malmst o¨ m, E.; Frank, C. W.; Kampf, J. P. J. Am. Chem.
Soc. 1997, 119, 9903.
(13) Demlow, E. V.; Demlow, S. S. Phase Transfer Catalysis; Verlag
Chemie: Weinheim, 1980; p 87.
(5) (a) Bednarek, M.; Biedron, T.; Helinski, J.; Kaluzynski, K.; Kubisa,
P.; Penczek, S. Macromol. Rapid Commun. 1999, 20, 369. (b) Magnusson,
H.; Malmstr o¨ m, E.; Hult, A. Macromol. Rapid Commun. 1999, 20, 453.
(6) (a) Sunder, A.; Hanselmann, R.; Frey, H.; M u¨ lhaupt, R. Macromolecules
1
999, 32, 4230. (b) Sunder, A.; M u¨ lhaupt, R.; Haag; R.; Frey, H. Macromol-
ecules 2000, 33, 253-254.
7) Synthesis of dendritic and hyperbranched polycarbosilanes was reported
(
in (a) Lorenz, K.; M u¨ lhaupt, R.; Frey, H.; Rapp, U.; Mayer-Posner, F. J.
Macromolecules 1995, 28, 6657. (b) Lach, C.; M u¨ ller, P.; Frey, H.; M u¨ lhaupt,
R. Macromol. Rapid Commun. 1997, 18, 253.
(8) Synthesis of dendritic and hyperbranched aliphatic polyesters was
reported without comparison in: (a) Malmstr o¨ m, E.; Johansson, M.; Hult, A.
Macromolecules 1995, 28, 1698. (b) Ihre, H.; Hult, A.; S o¨ derlind, E. J. Am.
Chem. Soc. 1996, 118, 6388.
(14) Nouguier, R. M.; Mchich, M. J. Org. Chem. 1985, 50, 3296.
(15) VanRheenen, V.; Kelly, R. C.; Cha, D. Y. Tetrahedron Lett. 1976,
1973.
1
0.1021/ja994363e CCC: $19.00 © 2000 American Chemical Society
Published on Web 03/10/2000

Communications to the Editor
J. Am. Chem. Soc., Vol. 122, No. 12, 2000 2955
Scheme 1. Synthesis of Glycerol Dendrimer 1,
Hyperbranched Polyglycerol 2, and Pseudo-Dendritic
Polyglycerol 3a
Figure 1. 13C NMR spectra of (A) glycerol dendrimer 1 and (B) pseudo-
dendritic polyglycerol 3. Signals of characteristic carbons are assigned
according to Scheme 1. For simplicity, signals of the TMP-core at higher
field (43.6, 23.4, 8.1 ppm) are not shown.
Figure 2. MALDI-TOF mass spectra of (A) [G-3] glycerol dendrimer 1
and (B) pseudo-dendritic polyglycerol 3. The spectrum of dendrimer 1
contains traces of imperfect dendrimer structures with one and two missing
glycerol units (<5% abundance).
Although structural equivalence is proven by NMR, MALDI-
TOF spectra show the difference between the two compounds
(Figure 2). While the [G-3] glycerol dendrimer 1 shows only one
major peak at the expected mass (m/z ) 1690, i.e., TMP + 21
monomer units), the pseudo-dendrimer 3 exhibits a distribution
of molecular weights with the mass of the monomer unit (M )
74) as peak distance.
We have developed a new dendrimer synthesis based on
a
(
4
i) Allyl chloride, NaOH, TBAB, water; (ii) NMO, OsO (cat.), water,
chemically stable polyether linkages. By this divergent approach
a third generation [G-3] glycerol dendrimer with 24 functional
end groups was prepared in an efficient way from inexpensive
starting materials and reagents. We also introduced a new strategy
to generate a polyglycerol pseudo-dendrimer 3, which also consists
of dendritic and terminal units only, resembling the perfect
glycerol dendrimer 1. The advantages of the second approach
are obvious: (i) only three synthetic steps (polymerization,
allylation, dihydroxylation) are required to tailor any molecular
1
3
acetone, t-BuOH. Characteristic carbon signals in the C NMR spectra
are assigned a, b, c, d.
previously.6a Subsequently, all hydroxy groups were converted
into allyl ethers while completion of the reaction was monitored
by the disappearance of the OH band in the IR spectrum. In the
second step all allyl groups were converted into terminal glycerol
units, and the conversion was monitored by the disappearance of
1
the allyl signals in the H NMR spectrum. The conversions for
both reactions were higher than 95%, and the overall isolated
yield was 89% after dialysis. SEC (DMF, poly(propylene oxide)
calibration) revealed a significant increase in molecular weight
weight between M
n
) 1000 and 20 000 with low polydispersities
(M /M < 1.5), (ii) the synthesis can be carried out on a multigram
w
n
scale, (iii) the capacity of 1,2-diols () terminal units) on the
polymer is increased from 4.5 mmol/g for the hyperbranched
polymer 2 to 7.1 mmol/g for the fully dendritic structure 3, and
(iv) all reported advantages of the polyether scaffold are preserved.
(
hyperbranched polymer 2 M
polymer 3 M ) 4760) maintaining the narrow polydispersity of
/M ) 1.2.
We have used 13C NMR spectroscopy for a detailed structural
n
) 2730; modified hyperbranched
n
M
w
n
Acknowledgment. The authors would like to thank Dr. H. Frey for
fruitful discussions, Dr. R. Hanselmann for the MALDI measurements,
and Professor Dr. R. M u¨ lhaupt for his generous support. R. H. is indebted
to the Fonds der Chemischen Industrie for financial support.
characterization of the dendritic polyglycerols. Figure 1 shows
the spectra of the glycerol dendrimer 1 and the transformed
hyperbranched polymer 3, confirming the structural perfection
of 3. The signals of both compounds differ only in their line width,
which arises from the isomerism and the polydispersity of 3. On
the basis of the structural similarity and the fact that exactly one
core and exclusively dendritic and terminal units are present we
suggest the name “pseudo-dendrimer” for this fully branched
polyglycerol 3.
Supporting Information Available: Experimental procedures and
1
13
full characterization ( H, C, IR, MS) for all glycerol dendrimers [G-1]
through [G-3] and the pseudo-dendritic polyglycerol (PDF). This material
is available free of charge via the Internet at http://pubs.acs.org.
JA994363E