Synthesis of an anionically chargeable, high-molar-mass, second-generation dendronized polymer and the observation of branching by scanning force microscopy

Article abstract of DOI:10.1021/ja057964g

An efficient synthesis of a methacrylate-based, second-generation (G2) dendronized macromonomer and its free radical polymerization to the corresponding high-molar-mass G2 dendronized polymer are described. The molar mass is determined by gel permeation c

Full text of DOI:10.1021/ja057964g

Published on Web 03/18/2006
Synthesis of an Anionically Chargeable, High-Molar-Mass,
Second-Generation Dendronized Polymer and the Observation
of Branching by Scanning Force Microscopy
Edis Kase¨mi,§ Wei Zhuang,‡ Ju¨rgen P. Rabe,*,‡ Karl Fischer, Manfred Schmidt,
Martin Colussi,§ Helmut Keul,| Ding Yi,# Helmut Co¨lfen,# and A. Dieter Schlu¨ter*,§
Contribution from the Department of Materials, Institute of Polymers, Swiss Federal Institute of
Technology, ETH-Ho¨nggerberg, HCI J 541, 8093 Zu¨rich, Switzerland, Department of Physics,
Humboldt UniVersity Berlin, Newtonstrasse 15, D-12489 Berlin, Germany, Institute for Physical
Chemistry, UniVersity of Mainz, Jakob-Welder-Weg 11, D-55099 Mainz, Germany, Institute for
Technical and Macromolecular Chemistry, RWTH Aachen, Paulwelsstrasse 8,
D-52056 Aachen, Germany, and Max Planck Institute of Colloids and
Interfaces, Am Mu¨hlenberg, D-14476 Golm-Potsdam, Germany
Received November 23, 2005; E-mail: rabe@physik.hu-berlin.de; dieter.schluter@mat.ethz.ch
Abstract: An efficient synthesis of a methacrylate-based, second-generation (G2) dendronized mac-
romonomer and its free radical polymerization to the corresponding high-molar-mass G2 dendronized
polymer are described. The molar mass is determined by gel permeation chromatography (GPC), light-
scattering, and analytical ultracentrifugation and compared with values estimated from a scanning force
microscopy (SFM) contour lengths analysis of individualized polymer strands on mica. The polymer carries
terminal tert-butyl-protected carboxyl groups, the degree of deprotection of which with trifluoroacetic acid
is quantified by NMR spectroscopy using the highest molar mass sample. SFM imaging of both protected
(noncharged) and unprotected (charged) dendronized polymers on solid substrates reveals mostly linear
chains but also some with main-chain branches. The nature of these branches is investigated and the
degree roughly estimated to which they are formed. Finally, a synthetic model experiment is described
which sheds some light on the aspect of whether chain transfer, a process that could lead to covalent
branching, is of importance in the synthesis of the present dendronized polymers.
pattern solid surfaces,3 (b) move individualized macromolecules
on solid substrates by use of a the scanning force microscope
Introduction
Dendronized polymers are a class of comb polymers in which
regularly branched units (dendrons) with a specific number of
branching layers (generation numbers) are attached to a linear
backbone at every repeat unit.1 The higher the generation
numbers are, the higher is the dendrons’ space demand and,
thus, the interaction between the tightly spaced (typically 2.5
Å) consecutive dendrons. This increased dendron/dendron
interaction leads to an increased shape persistence which is a
key difference between such macromolecules and those with
the same backbone but without dendritic layer. The dendritic
layer, however, not only impacts the shape persistence but also
has the advantage that its thickness can be systematically varied
by synthetic measures. Together with the ability of many
dendronized polymers to be surface engineered,2 these polymers
exhibit interesting application options, which include to (a)
(SFM),4 (c) do first steps toward a bottom-up approach to
ordered functional arrays,5 (d) apply a shielding concept to those
representatives with fluorescent or electrically conducting
backbones,6 (e) let representatives substituted with mesogenic
dendrons self-assemble into ordered bulk material,7 and (f) let
positively charged representatives self-assemble into huge
ordered molecular aggregates.8 Specifically the latter aspect, in
which we see some future potential for the construction of
hierarchically structured molecular objects of unprecedented
dimensions, led the authors to develop dendronized polymers
(2) For example, see: (a) Ouali, N.; Me´ry, S.; Skoulios, A. Macromolecules
2000, 33, 6185-6193. (b) Grayson, S. M.; Fre´chet, J. M. J. Macromolecules
2001, 34, 6542-6544. (c) Shu, L.; Go¨ssl, I.; Rabe, J. P.; Schlu¨ter, A. D.
Macromol. Chem. Phys. 2002, 203, 2540-2550. (d) Suijkerbuijk, B. M. J.
M.; Shu, L.; Klein Gebbink, R. J. M.; Schlu¨ter, A. D.; van Koten, G.
Organometallics 2003, 22, 4175-4177. (e) Zhang, A.; Barner, J.: Go¨ssl,
I.; Rabe, J. P.; Schlu¨ter, A. D. Angew. Chem., Int. Ed. 2004, 43, 5185-
5188.
§ Swiss Federal Institute of Technology.
(3) (a) Stocker, W.; Schu¨rmann, B. L.; Rabe, J. P.; Fo¨rster, S.; Lindner, P.;
Neubert, I.; Schlu¨ter, A. D. AdV. Mater. 1998, 10, 793-797. (b) Percec,
V.; Ahn, C.-H.; Ungar, G.; Yeardley, D. J. P.; Mo¨ller, M.; Sheiko, S. S.
Nature 1998, 39, 161-164.
‡ Humboldt University Berlin.
University of Mainz.
| RWTH Aachen.
# Max Planck Institute of Colloids and Interfaces.
(1) Schlu¨ter, A. D.; Rabe, J. P. Angew. Chem., Int. Ed. 2000, 39, 864-883.
Zhang, A.; Shu, L.; Bo, Z.; Schlu¨ter, A. D. Macromol. Chem. Phys. 2003,
204, 328-339. Frauenrath, H. Prog. Polym. Sci. 2005, 30, 325-384.
Schlu¨ter, A. D. Top. Curr. Chem. 2005, 245, 151-191.
(4) Shu, L.; Schlu¨ter, A. D.; Ecker, C.; Severin, N.; Rabe, J. P. Angew. Chem.,
Int. Ed. 2001, 40, 4666-4669.
(5) Barner, J.; Mallwitz, F.; Shu, L.; Schlu¨ter, A. D.; J. P. Rabe, Angew. Chem.,
Int. Ed. 2003, 42, 1932-1935. Barner, J.; Al-Hellani, R.; Schlu¨ter, A. D.;
Rabe, J. P., submitted.
9
10.1021/ja057964g CCC: $33.50 © 2006 American Chemical Society
J. AM. CHEM. SOC. 2006, 128, 5091-5099
5091

A R T I C L E S
Kase¨mi et al.
Scheme 1 a
with such peripheral functional groups that can be turned into
negatively charged ones. In this way, the already existing
polyelectrolytes, based on dendronized polymers with hundreds
or even thousands of positive peripheral charges per macro-
molecule, would have interesting counterparts to explore their
individual and joint aggregation behavior.
This paper describes the synthesis of a G2 macromonomer
with peripheral tert-butyl-protected carboxylates and a meth-
acrylate polymerizing unit, as well as its free radical polymer-
ization using azobisisobutyronitrile (AIBN) as initiator precursor.
The molar masses of the corresponding G2 dendronized
polymers were determined by three methods: gel permeation
chromatography (GPC) using a calibration based on two-angle
light-scattering and viscometry, static and dynamic multiangle
light-scattering, and analytical ultracentrifugation. The obtained
values were compared with those estimated on the basis of SFM
contour lengths analysis. The SFM images of both protected
and unprotected dendronized polymers on mica and polyorni-
thine-coated mica exhibit primarily linear but also a trace amount
of main-chain branched structures. Because branches had not
been observed since the discovery of dendronized polymers
some 10 years ago, their nature was investigated by a careful
height analysis. This SFM work is complemented by a synthetic
model study aiming at a quantification of an eventual covalent
branching through chain-transfer events. Finally, the deprotec-
tion of the new G2 dendronized polymer will be described
whereby the efficiency of this process was monitored by high-
field NMR spectroscopy. For these experiments, the highest
molar mass sample was used.
a Reagents and conditions: (a) 1a, LAH, THF, 0 °C, 16 h (86%); (b)
1b, 1c, HCl/THF, rt, 6 h (95%); (c) HOBt, EDC, DCM, -20 °C, 16 h
(76%); (d) 4a, Pd(PPh3)4, C7H7NaO2S‚H2O, DCM/MeOH, rt, 4 h (80%).
well-known Pd-catalyzed protocol for allylic esters.11 G2
dendron 5 was then obtained by reacting an excess of 4b with
2a via standard peptide methods. Finally, macromonomer 6 was
obtained by reacting this G2 dendron with freshly distilled
methacrylic acid chloride. All neutral compounds were purified
by column chromatography on silica gel and obtained as
analytically pure, completely colorless materials. Monomer 6
was obtained on the 6 g scale as a foamy material. All structures
were fully confirmed by NMR spectroscopic and mass spec-
trometric data.
Results and Discussion
Synthesis. The synthetic steps are presented in Schemes 1-3.
Macromonomer 6 uses the same main dendritic skeleton which
proved successful in previous work.9 It differs from this,
however, by both the peripheral groups (tert-butyl-protected
carboxylates instead of protected amines) and the polymerizable
unit (acrylate instead of styrene). Its synthesis starts from the
known G1 dendrons 1a and 1c.10 Ester 1a was reduced with
LAH to give alcohol 1b, which was then deprotected at its
peripheral amines to furnish the ammonium chloride 2a.
Similarly, dendron 1c was deprotected to 2b, which was then
reacted with the succinic acid mono tert-butyl ester 3 to yield
the tert-butyl-protected carboxylate G1 unit 4a. Its deprotection
at the focal point to give 4b was carried out according to the
The polymerizations were performed under nitrogen in highly
concentrated DMF solution [2:1 (w/w) 6:DMF; for example,
2.0 g of 6 in 1 mL of solvent] at 65 °C using 0.5-1.0 mol %
of recrystallized AIBN. The initially clear and homogeneous
solutions after 2-3 h showed a viscosity increase, and polym-
erization was continued for a further 8-10 h when the mixture
had completely solidified. After three dissolution/precipitation
cycles, polymer 7a was obtained as a colorless, amorphous solid
in 75-90% yield. Several experiments were performed, the
conditions and results of which are collected in Tables 1 and 2.
The molar mass data were obtained by GPC using a calibration
based on two-angle light-scattering and viscometry, multiple-
angle light-scattering, and analytical ultracentrifugation and were
compared with those derived from SFM contour lengths
analysis. (For a discussion, see the following section.)
(6) For example, see: Bao, Z.; Amundson, K. R.; Lovinger, A. J. Macromol-
ecules 1998, 31, 8647-8649. Sato, T.; Jiang, D.-L.; Aida, T. J. Am. Chem.
Soc. 1999, 121, 10658-10659. Malenfant, P. R. L.; Fre´chet, J. M. J.
Macromolecules 2000, 33, 3634-3640. Jakubiak, R.; Bao, Z.; Rothberg,
L. Synth. Met. 2000, 61, 114-118. Percec, V.; Obata, M.; Rudick, J. G.;
De, M. Gloode, B. B.; Bera, T. K.; Magonov, S. N.; Balagurusamy, V. S.
K.; Heiney, P. A. J. Polym. Sci., Polym. Chem. 2002, 40, 3509-3533.
Pongantsch, A.; Wenzl, F. P.; List, E. J. W.; Leising, G.; Grimsdale, A.
C.; Mu¨llen, K. AdV. Mater. 2002, 14, 1061-1064.
(7) Kwon, Y. K.; Chvalun, S.; Schneider, A.-I.; Blackwell, J.; Percec, V.; Heck,
J. A. Macromolecules 1994, 27, 6129-6132. Kwon, Y. K.; Chvalun, S.;
Blackwell, J.; Percec, V.; Heck, J. A. Macromolecules 1995, 28, 1552-
1558. Rapp, A.; Schnell, I.; Sebastiani, D.; Brown, S. P.; Percec, V.; Spiess,
H. W. J. Am. Chem. Soc. 2003, 125, 13284-13297.
(8) (a) Go¨ssl, I.; Shu, L.; Schlu¨ter, A. D.; Rabe, J. P. J. Am. Chem. Soc. 2002,
124, 6860-6865. (b) Bo¨ttcher, C.; Schade, B.; Ecker, C.; Rabe, J. P.; Shu,
L.; Schlu¨ter, A. D. Chem. Eur. J. 2005, 11, 2923-2928.
The deprotection of dendronized polymer 7a (Table 1, entry
5) to give 7b was achieved with trifluoroacetic acid (TFA).12 If
a large excess of the acid per protecting group was applied at
room temperature to the polymer in dichloromethane, a tert-
butyl signal with a very small intensity compared to the original
one of Boc remained in the 1H NMR spectrum (see Supporting
Information). If the deprotection was first done in neat TFA,
(9) Neubert, I.; Schlu¨ter, A. D. Macromolecules 1998, 31, 9372-9378.
(10) Ingerl, A.; Neubert, I.; Klopsch, R.; Schlu¨ter, A. D. Eur. J. Org. Chem.
1998, 2551-2556. Klopsch, R.; Koch, S.; Schlu¨ter, A. D. Eur. J. Org.
Chem. 1998, 1275-1283. Zistler, A.; Koch, S.; Schlu¨ter, A. D. J. Chem.
Soc., Perkin Trans. 1 1999, 501-508.
(11) Honda, M.; Morita, H.; Nagakura, I. J. Org. Chem. 1997, 62, 8932-8936.
(12) Girbasova, N.; Aseyev, V.; Saratovsky, S.; Moukhina, I.; Tenhu, H.; Bilibin,
A. Macromol. Chem. Phys. 2003, 204, 2258-2264.
9
5092 J. AM. CHEM. SOC. VOL. 128, NO. 15, 2006

Synthesis of a High-Molar-Mass G2 Dendronized Polymer
A R T I C L E S
Scheme 2 a
a Reagents and conditions: (e) HOBt, EDC, DCM, -20 °C, 16 h (60%); (f) MAC, DMAP, THF, rt, 16 h (85%).
Scheme 3 a
Molar Mass Determination by MALLS, AUC, and SFM
Contour Lengths Analysis. (a) MALLS. The static and
dynamic light-scattering results are summarized in Tables 1 and
2. The virial coefficients are positive, indicating that DMF is a
good solvent for the G2 polymers. This is confirmed by the
positive concentration dependence of the diffusion coefficients
(see Supporting Information). Nonetheless, stiff polymers have
the notorious tendency to form aggregates, a possible influence
of which cannot be excluded for the present characterization,
particularly for the light-scattering results. The Zimm plots of
the two higher molar mass samples (see Supporting Information)
do show a slight downward curvature at small q values which
could be caused either by the broad molar mass distribution
(i.e., by some extremely high molar mass chains) or by some
aggregates. In our data analysis, this downward curvature was
ignored; i.e., the molar masses and radii given in Tables 1 and
2 may be smaller than the true ones. It is also conceivable that
some of the larger aggregated or nonaggregated chains are
removed by filtration. However, such a concentration loss is
assumed to be small and leads to smaller measured molar
masses. Therefore, it is concluded that the molar masses
represent a lower limit.
a Reagents and conditions: (g) AlBN, DMF, 65 °C, 12 h (90%); (h)
TFA, rt, 24 h (95%); (i) KOH, MeOH, rt, 3 h.
followed by a second treatment in solution, even a highly
amplified NMR spectrum did not show any indication of a tert-
butyl signal anymore. It can be concluded that the deprotection
is virtually quantitative. Given the molar mass of the sample of
approximately Mw ) 7 × 106, approximately 20 000 esters per
macromolecule were hydrolyzed. The deprotection procedure
was not further optimized. The deprotonation of 7b’s carboxylic
acid functions to give the anionically charged polyelectrolyte
7c was easily done with potassium hydroxide. Polymer 7c was
fully soluble in deionized water with pH ) 8.5.
An inspection of the data given in Table 2 reveals that the
reported radii do not fit the molar masses well; i.e., the data
scatter significantly, which prohibits a more quantitative inter-
pretation, e.g., in terms of chain stiffness.
(b) AUC. The molar mass of polymer 7a (Table 1, entry 5)
was also investigated by analytical ultracentrifugation. Sedi-
mentation velocity experiments yielded the sedimentation coef-
ficient distribution, together with the density distribution
according to the procedure by Ma¨chtle and Mu¨ller,13 which is
essentially based on the Schachmann density variation tech-
nique,14 yielding a polymer density of 1.197 g/mL. It can be
seen that the sedimentation coefficient distribution is broad
(Figure 2), with the typical tailing toward higher molar masses,
The glass transition temperatures of the polymers 7a-c were
determined by differential scanning calorimetry as 52 (7a), 53
(7b), and 125 °C (7c), respectively, on the basis of the second
heating curves (see Supporting Information). Thermogravimetric
analysis of 7a under nitrogen indicated an abrupt and complete
loss of the tert-butyl groups at 214 °C (see Supporting
Information).
(13) Ma¨chtle, W. Makromol. Chem. 1984, 185, 1025-1039. Mu¨ller, H. G.;
Herrmann, F. Prog. Colloid Polym. Sci. 1995, 99, 114-119.
(14) Edelstein, S. J.; Schachman, H. K. J. Biol. Chem. 1967, 242, 306-311.
9
J. AM. CHEM. SOC. VOL. 128, NO. 15, 2006 5093

A R T I C L E S
Kase¨mi et al.
Table 1. Conditions for and Results of the Radical Polymerization of Monomer 6 (Polymerization Time, 16 h)
a
a
b
c
d
d
7
DMF
L)
AIBN
yield
(%)
Mn
106)
Mw
106)
Mw
106)
Mw
106)
Mn
106)
Mw
entry
(g)
(
µ
(mol %)
(
×
(
×
PDIa
(
×
(
×
(
×
(
×
106)
1
2
3
4
5
0.21
0.23
0.44
1.00
2.22
70
80
150
450
0.6
0.8
0.8
1.0
0.7
85
78
78
87
90
1.3
4.5
3.8
1.8
3.1
2.9
2.9
1.3
1.6
2.6
2.9
3.3
3.7
7.8
9.3
4.2
5.8
7.6
1.1
1.8
3.1
5.8
1000
6.5
a GPC results calculated by light-scattering signals only. b MALLS in DMF without LiBr. c Analytical ultracentrifugation in NMP. d SFM contour length
analysis.
Table 2. Summary of the Light-Scattering Results
1
1
)
+ 2A2c
Mw
Rg
Rh
A2
Mw,app Mw
1
entrya
(×
106 g mol-
)
(nm)
(nm)
(×
104 cm3 g-2 mol)
From this evaluation (Figure 2), a weight-average molar mass
of 6.54 × 106 g/mol is calculated with a second osmotic virial
coefficient of 1.69 × 10-5 cm3‚mol/g2. The positive virial
coefficient in N-methylpyrrolidone (NMP) indicates the absence
of self-association and is a measure of nonideality. From it the
excluded volume can be calculated to be 221 mL/g. This value
is by a factor of 4 lower than the one from light-scattering, which
can be attributed to the different applied solvents and also a
slightly different molar mass (Tables 1 and 2).
3
4
5
4.2
5.8
7.6
61
94
97
35
45
56
0.13
0.20
0.57
a The entry numbers refer to Table 1.
(c) SFM Contour Length Analysis. Individualized den-
dronized polymer molecules (Table 1, entries 4 and 5) were
spin-coated from dilute chloroform solutions (2-10 mg/L) onto
freshly cleaved mica and visualized via tapping-mode SFM. The
observed feature heights (apparent heights) were determined to
be 1.6 nm ((10%), which is lower than the expected geometrical
thickness, as typically observed in tapping-mode SFM of single
macromolecules in air.15 From their contour length distributions,
the averages of their molar masses, Mn and Mw, were calculated
assuming a length of the repeat unit of 0.25 nm (see Table 1).
The absolute values of both Mn and Mw for the samples of
entries 4 and 5 of Table 1 are generally lower than those
obtained by GPC, light-scattering, and ultracentrifugation. There
are different possible reasons for this, one being that the
backbone may not be in its all-trans conformation, which
reduces the average length of the repeat unit to a value below
0.25 nm. Moreover, molecules were not counted for the contour
analysis if they did not exhibit two clearly recognizable ends,2c
e.g., if they aggregated intra- or intermolecularly. While the
fraction of objects which could not be included in the statistical
analysis is below approximately 10%, it is the very long
molecules that are particularly prone to aggregate, thereby
lowering the determined average molar masses. For linear brush
polymers consisting of a long main chain with densely grafted
linear side chains, several studies exist which compare the
absolute molar mass as determined by scattering methods and
the length of the polymers as determined by SFM.16-20 The
interpretation in terms of apparent length per main-chain repeat
unit, lm, on the surface gave strongly varying results, i.e., 0.07
Figure 1. Differential sedimentation coefficient distribution, g(s*), vs
sedimentation coefficient, s. The apparent sedimentation coefficients were
corrected to their apparent standard values in water at 20 °C.
Figure 2. Plot of the inverse apparent weight-average molar mass, 1/Mw,app
,
vs the concentration in NMP from sedimentation equilibrium for the
extrapolation of Mw and the second osmotic virial coefficient A2.
(15) Zhuang, W.; Ecker, C.; Metselaar, G. A.; Rowan, A. E.; Nolte, R. J. M.;
Samor´ı, P.; Rabe, J. P. Macromolecules 2005, 38, 473-480 and references
therein.
which can be expected for the applied synthesis conditions,
consistent with the Schulz-Flory distribution. A slight bimodal-
ity is indicated for the part of the distribution with smaller
sedimentation coefficients.
(16) Gerle, M.; Fischer, K.; Mu¨ller, A. H. E.; Schmidt, M.; Sheiko, S. S.;
Prokhorova, S.; Mo¨ller, M. Macromolecules 1999, 32, 2629-2635.
(17) Bo¨rner, H. G.; Beers, K.; Matyjaszewski, K.; Sheiko, S. S.; Mo¨ller, M.
Macromolecules 2001, 34, 4375-4383.
(18) Dziezok P.; Fischer, K.; Schmidt, M.; Sheiko, S.; Mo¨ller, M. Angew. Chem.,
Int. Ed. 1997, 36, 2812-2815.
Sedimentation equilibrium experiments allowed for the de-
termination of Mw and the second osmotic virial coefficient A2
from an extrapolation of the inverse apparent weight-average
molar mass Mw,app to infinite dilution according to
(19) Li, C.; Gunari, N.; Fischer, K.; Janshoff, A.; Schmidt, M. Angew. Chem.
2004, 116, 1121-1124.
(20) Sun, F.; Sheiko, S. S.; Mo¨ller, M.; Beers, K.; Matyjaszewski, K. J. Phys.
Chem. 2004, 108, 9682-9686.
9
5094 J. AM. CHEM. SOC. VOL. 128, NO. 15, 2006

Synthesis of a High-Molar-Mass G2 Dendronized Polymer
A R T I C L E S
Figure 3. Tapping-mode SFM images of individualized protected, uncharged polymers (7a) on mica, some of which exhibit branches: samples of (a) entry
2, (b) entry 4, and (c) entry 5. The height traces along the contours in image a indicate chain crossing at the left and branching at the right junction. The
white blobs on images b and c stem from unidentified material.
nm < lm < 0.23 nm, depending on side-chain length and grafting
density, which are not always precisely known. The correspond-
ing values for the dendronized polymers of entries 4 and 5 of
Table 1 are lm ) 0.13 and lm ) 0.19, respectively. It should be
noted that the reported light-scattering molar masses represent
a lower limit, and accordingly the lm values could become even
smaller.
It seems that brush and dendronized polymers show a similar
contraction behavior upon adsorption to solid substrates. The
origin of this effect is not yet understood. Reports on lm for
cylindrical brush polymers in solution are contradictory.16,21-27
The analysis of combined SLS/SANS data28,29 has now provided
some evidence that, irrespective of the side-chain lengths, the
l
m values for cylindrical brushes are lm ) 0.25 and 0.23 nm in
good and bad solvents, respectively. While many arguments
exist as to why the conformation of the polymers at a surface
in the dry state is different from that in solution, at least one
example is reported where no difference was observed between
cylindrical brush polymers in the dry state and adsorbed at a
surface in the presence of solvent.19
By far most chains were linear; however, occasionally (on
the order of 1% of all cases) main-chain branching was
observed. Figure 3 displays three examples. While at crossing
points of chains the thickness almost doubles, at the branches
the height of the molecules does not vary significantly more
than along the contours (Figure 3a). The accidental formation
(21) Fischer, K.; Gerle, M.; Schmidt, M. Proc. ACS, PMSE Anaheim 1999, 30,
133-134.
(22) Fischer, K.; Schmidt, M. Macromol. Rapid Commun. 2001, 22, 787-791.
(23) Terao, K.; Takeo, Y.; Tazaki, M.; Nakamura, Y. Polym. J. 1999, 31, 193-
196.
(24) Terao, K.; Nakamura, Y.; Norisuye, T. Macromolecules 1999, 32, 711-
716.
(27) Hokajo, T.; Terao, K.; Nakamura, Y.; Norisuye, T. Polym. J. 2001, 33,
481-485.
(25) Terao, K.; Hokajo, T.; Nakamura, Y.; Norisuye, T. Macromolecules 1999,
32, 3690-3694.
(28) Rathgeber, S.; Pakula, T.; Wilk, A.; Matyjaszewski, K.; Beers, K. L. J.
Chem. Phys. 2005, 122, 124904-1-124904-12.
(26) Terao, K.; Hayashi, S.; Nakamura, Y.; Norisuye, T. Polym. Bull. 2000,
44, 309-316.
(29) Zhang, B.; Gro¨hn, F.; Pedersen, J. S.; Schmidt, M., in preparation.
9
J. AM. CHEM. SOC. VOL. 128, NO. 15, 2006 5095

A R T I C L E S
Kase¨mi et al.
Figure 4. Tapping-mode SFM images of individualized unprotected, negatively charged polymers on polyornithine-coated mica before (a) and after (b)
manipulation in contact mode. The three arms of a supposedly branched molecule are colored, and black arrows indicate the path the SFM tip followed
during the manipulation attempts.
of T-like structures due to the adsorption of two different
molecules meeting at one point is extremely unlikely. Provided
there is no particular attractive interaction between a chain end
and the main chain, this cannot explain the occurrence of about
1% of branched configurations in the diluted submonolayers
investigated here.
for the abstraction would be the CH bonds R to the ether oxygen
atoms, the tert-butyl ester groups, or the benzylic groups. These
bonds have lower dissociation energies than normal CH’s.32 To
complement the above SFM findings from the synthetic side,
an experiment was devised to prove or disprove the occurrence
of chain transfer. Methyl methacrylate (MMA) was polymerized
in the presence of a large amount of the third-generation (G3)
dendron 8 (Figure 5) under conditions similar to the ones applied
for macromonomer 6 (for details, see Supporting Information).
This dendron has the same repeat units and protecting groups
as the ones in macromonomer 6 and dendronized polymer 7a
and, thus, was considered an ideal compound to provoke chain
transfer during the MMA polymerization. If it acted as a chain-
transfer agent, formation of a PMMA was to be expected which
carries this very dendron as an end group. In two independent
We also considered whether it was possible to distinguish
between covalently and noncovalently connected branches by
exerting a force on the junction in order to estimate its strength.5
However, while one can drag long dendronized polymers across
a graphite surface using an SFM tip,4 the same is not possible
on mica, since the friction during manipulation on this substrate
is so large that even short molecules break immediately. On
the other hand, it was not possible to immobilize the molecules
investigated here on graphite. We also tried to manipulate the
deprotected, negatively charged polymers 7c by spin-coating
them from dilute aqueous solutions (5-10 mg/L) on a positively
charged substrate, poly-L-ornithine-coated mica, which had been
employed previously for the deposition of DNA.8a Figure 4
displays SFM images before and after manipulation, indicating
that also in this case the interaction with the substrate is too
strong to drag a long chain across the surface without breaking
it immediately.
Experiment Regarding Chain Transfer. Chain transfer is
one of the ways through which branch formation can occur
during polymerization.30 An active (radical) chain end abstracts,
e.g., a hydrogen atom from its own or another independent
chain,31 with formation of an inactive chain end and a new
radical center. If this is active enough to carry on polymerization,
this leads to main-chain branching, and the radical’s position
marks the branch point. Supposed branching through chain
transfer plays a role in the synthesis of dendronized polymer
7a, and given the steric crowding around its backbone, hydrogen
abstraction is considered more likely to take place at the
dendritic shell rather than at the backbone. Preferred positions
experiments A and B, PMMAs with molar masses of Mw
)
65 000 (A) and 450 000 g/mol (B) were obtained. The strongly
1
amplified aromatic regions of the H NMR spectra of these
samples are compared (Figure 6a,b) with that of dendron 8
(Figure 6c) to evaluate potential dendron incorporation. The fact
that there is no indication of dendron-typical signals in the
PMMAs is a strong indication that, within the sensitivity of the
NMR experiment, dendron 8 does not serve as a chain-transfer
agent.33 Spectra b and c of Figure 6 contain signals which cannot
be associated with parent PMMA. To see whether they are
somehow related to dendron 8 present in the polymerization
mixture, MMA was also polymerized under the exact same
conditions but without any dendron present. The NMR spectrum
of the corresponding PMMA is shown in Figure 6d. As can be
seen, most of the signals of unclear origin also appear there34
and, therefore, cannot have anything to do with chain transfer
by the dendron. Since the remaining signals’ intensities by
comparison with the 13C satellites of the main PMMA signals
(32) McMillan, D. F.; Golden, D. M. Annu. ReV. Phys. Chem. 1982, 33, 493-
532.
(30) Odian, G. Principles of Polymerization; Wiley: New York, 1991; pp 255-
(33) This holds true at least as far as no decomposition of the dendron during
chain transfer occurs.
259.
(31) Chain transfer to solvent, monomer, radical initiator precursor, and alike
is not considered here.
(34) The differences in chemical shifts are assumed to be due to the different
solvents used.
9
5096 J. AM. CHEM. SOC. VOL. 128, NO. 15, 2006

Synthesis of a High-Molar-Mass G2 Dendronized Polymer
A R T I C L E S
Figure 5. Chemical structure of the G3 dendron 8 used to provoke chain transfer.
1
Figure 6. Highly amplified H NMR spectra of two independent preparations of PMMA (Mw ) 450 000 and 5000 g/mol, respectively) synthesized in the
presence of G3 dendron 8 as a potential chain-transfer agent (a,b), the G3 dendron 8 (c), and a PMMA obtained under the same conditions but without any
8 present (d). The chloroform signal appears at δ ) 7.25; its 13C satellites can be seen in spectra a and d.
(not shown) are in the sub-percent range, their significance was
considered too low to be further investigated.
here) showed that branches also occur (in the sub-% range) at
low conversions (e.g. 13%),
This model study supports the estimation based on SFM
measurements that branching takes place in the sub-1% range.
An eventual argument that a significant amount of branched
structures may be contained in the approximately 10% of
unidentifiable aggregates observed by SFM can be ruled out. It
should also be noted that the branches do not seem to form
exclusively by a gel effect at high conversion. A recent study
using a G3 macromonomer based on dendron 8 (not described
Conclusion and Outlook
The free radical polymerization of macromonomer 6 gave
the G2 denpol 7a, whose molar mass (Mw) was determined by
several methods to be 8.5 × 106 (GPC), 7.6 × 106 (MALLS),
6.5 × 106 (AUC), and 5.8 × 106 (SFM). These values
correspond to the highest chain lengths ever observed for a G2
9
J. AM. CHEM. SOC. VOL. 128, NO. 15, 2006 5097

A R T I C L E S
Kase¨mi et al.
15° and 90° angle) detectors (DMF + 1 g L-1 LiBr as eluent at 80
°C). Universal calibration was done using PMMA standards in a range
of Mp ) 2 680-3 900 000 (Polymer Labs Ltd., UK).
dendronized polymer,35 which indicates that it may be possible
to also achieve longer chains for the corresponding G3 and G4
macromonomers than previously observed in related systems.35a
If this is true, the key disadvantage of the macromonomer route
to dendronized polymers could be overcome, which is that for
high generations (G4) only relatively short chains have been
so far accessible. Polymer 7a was prepared on the 2 g scale.
Deprotection of its approximately 20 000 tert-butyl ester groups
to give 7c was achieved under acidic conditions, and its degree
was determined by high-field 1H NMR spectroscopy to be
virtually quantitative. Even under extreme amplification, no
residual tert-butyl signal could be seen. The dendritic layer
around the backbone stays intact under the deprotection condi-
tions.
High-resolution thermogravimetric analysis (TGA) was performed
on a Q500 thermogravimetric analyzer (TA Instruments, New Castle,
DE). All measurements were carried out in an air stream under the
same conditions. The mass loss with increasing temperature and its
first derivative (DTG), which represents the change in decomposition
rate, were plotted. Differential scanning calorimetry (DSC) was carried
out under nitrogen at a heating or cooling rate of 10 °C/min on a DSC
7 instrument (Perkin-Elmer, Norwalk, CT). Two heating runs and one
cooling run were consecutively carried out in a cycle, and the peak
maxima were considered as the transition temperatures.
MALLS. Static and dynamic light-scattering measurements were
performed with an ALV-SP86 goniometer, a Uniphase HeNe laser (25
mW output power at 632.8 nm wavelength), an ALV/High QE APD
avalanche diode fiber optic detection system, and an ALV-3000
correlator. The static scattering intensities were analyzed according to
Given the respective limitations of the methods used, the
molar mass values for polymer 7a (Table 1, entry 5) match
reasonably well. SFM has been used to observe branching of
polymer chains, which occurs in the present systems at a
concentration of approximately 1% only. Presently there is no
other method available with this kind of sensitivity and the
potential for positional identification of both the branching site
and lengths of branches. The charged dendronized polymers
will now be used to study their aggregation behavior, aiming
at hierarchically structured three-dimensional matter.
2
2
mean square radius of gyration, Rg
) Rg z, and the second virial
coefficient A2. The correlation functions showed a broad but monomodal
decay and were fitted by a sum of two exponentials, from which the
first cumulant Γ was calculated. The z-average diffusion coefficient
Dz was obtained by extrapolation of Γ/q2 to q ) 0 and to infinite
-1
dilution. The inverse z-average hydrodynamic radius, Rh ) 1/Rh
,
z
was evaluated by formal application of Stokes law. The dilute polymer
solutions in DMF (typically 4-5 concentrations 0.05 e c e 0.5 g/L)
were measured from 30° to 150° in steps of 5° (SLS) or in steps of
10° (DLS). Prior to measurement, the solutions were filtered through
0.2 µm pore size Dimex filters (Millipore LG).
Experimental Section
General. Compounds 1a,9 1c,9 and 2b10 as well as the G3 dendron
836 were synthesized according to literature methods. Other reagents
were purchased from Aldrich, Across, or Fluka. Methacryloyl chloride
(MAC) was freshly distilled before use. Tetrahydrofuran (THF) and
triethylamine (TEA) were refluxed over Na with benzophenone as
indicator. Dichloromethane (DCM) was dried by distilling over CaH2.
All other reagents and solvents were used as received. All reactions
were performed under nitrogen atmosphere. Silica gel 60 M (Macherey-
Nagel, 0.04-0.063 mm/ 230-400 mesh) was used as the stationary
phase for column chromatography. Whenever possible, reactions were
monitored by thin-layer chromatography (TLC) using TLC silica gel
coated aluminum plates 60F254 (Merck). Compounds were detected by
UV light (254 or 366 nm) and/or by treatment with a solution of
The refractive index increment was measured by a home-built
Michelson interferometer as described elsewhere37 and determined to
be dn/dc ) 0.0962 cm3/g in DMF.
Analytical Ultracentrifugation. Sedimentation experiments were
performed on an Optima XL-I instrument (Beckman Coulter, Palo Alto,
CA) at 25 °C in cells with 12 mm 2.5° titanium center pieces
(Nanolytics GmbH, Dallgow, Germany). For all experiments, the
Rayleigh interference optics was applied. For the sedimentation
equilibrium experiments, NMP was used as the solvent and the
centrifugal speed was 3000 rpm. The sedimentation velocity experi-
ments were performed in NMP (30 000 rpm) and DMF, the latter in
the protonated (30 000 rpm as well as the deuterated form (20 000 rpm).
For the determination of the sedimentation coefficient distribution
simultaneously with the density distribution of the polymer, sedimenta-
tion velocity data in DMF and DMF-d7 were combined following an
algorithm published for latexes by Ma¨chtle and Mu¨ller et al.13 For the
polymer density, the value at the maximum of the distribution was
taken to be 1.197 g/mL (partial specific volume Vj ) 0.835 mL/g).
Sedimentation equilibrium experiments were evaluated using the
MSTAR algorithm,38 and the weight-average apparent molar masses
were extrapolated to infinite dilution to yield Mw as well as the second
osmotic virial coefficient and excluded volume.
1
ninhydrin in ethanol followed by heating. If not otherwise noted, H
and 13C NMR spectra were recorded on Bruker AM 300 (1H, 300 MHz;
13C, 75 MHz) and AV 500 (1H, 500 MHz; 13C, 125 MHz) spectrometers
at room temperature using chloroform-d as a solvent. High-resolution
mass spectral (HRMS) and ESI-MS analyses were performed by the
MS service of the Laboratorium fu¨r Organische Chemie at ETH Zu¨rich.
ESI-MS and MALDI-MS were run on an IonSpec Ultra instrument. In
the case of MALDI-MS, 2,5-dihydroxybenzoic acid (DHB), 2-[(2E)-
(4-tert-butylphenyl)-2-methylprop-2-enylidene]malononitrile (DCTB),
or 3-hydroxypyridine-2-carboxylic acid (3-HPA) served as the matrix.
The FAB experiments were carried out with 3-nitrobenzyl alcohol
(MNBA)/CH2Cl2. Elemental analyses were performed by the Mikro-
labor of the Laboratorium fu¨r Organische Chemie, ETH Zu¨rich. The
samples were dried rigorously under vacuum prior to analysis to remove
strongly adhering solvent molecules. GPC measurements were carried
out using a PL-GPC 220 instrument with a 2x PL-Gel Mix-B LS column
set equipped with refractive index, viscosity, and light-scattering (with
Scanning Force Microscopy. SFM images were recorded using a
MultiMode scanning probe microscope (Digital Instruments, Inc., Santa
Barbara, CA) operated in tapping mode. Olympus etched silicon
cantilevers were used with a typical resonance frequency in the range
of 200-400 kHz and a spring constant around 42 N/m. All samples
were investigated at room temperature in air environment. As substrates
we used freshly cleaved mica (PLANO W. Plannet GmbH, Wetzlar,
Germany) or poly-L-ornithine (molar mass 30 000-70 000 g/mol,
Sigma, St. Louis, MO)-coated mica. Contour lengths were determined
taking into account the tip broadening, assuming a tip radius of 15
nm.
(35) For other high-molar-mass cases, see: (a) Zhang, A.; Zhang, B.; Wa¨cht-
ersbach, E.; Schmidt, M.; Schlu¨ter, A. D. Chem. Eur. J. 2003, 9, 6083-
6092. (b) Percec, V.; Ahn, C.-H.; Cho W.-D.; Jamieson, A. M.; Kim J.;
Leman, T.; Schmidt, M.; Gerle, M.; Mo¨ller, M.; Prokhorova, S. A.; Sheiko,
S. S.; Cheng, S. Z. D.; Zhang, A.; Ungar, G.; Yeardley D. J. P. J. Am.
Chem. Soc. 1998, 120, 8619-8631.
(37) Becker, A.; Ko¨hler, W.; Mu¨ller, B. Ber. Bunsenges. 1995, 89, 600-605.
(38) Co¨lfen, H.; Harding, S. E. Eur. Biophys. J. 1997, 25, 333-346.
(36) Kase¨mi, E.; Schlu¨ter, A. D., in preparation.
9
5098 J. AM. CHEM. SOC. VOL. 128, NO. 15, 2006

Synthesis of a High-Molar-Mass G2 Dendronized Polymer
A R T I C L E S
Acknowledgment. This work was supported by the Deutsche
Forschungsgemeinschaft (Sfb 448, TP A1 and A11), which is
gratefully acknowledged.
NMR spectra of 6, 1H NMR spectra of 7a’s deprotection, GPC
elution curve of 7a (Table 1, entry 5), and Zimm plots. This
material is available free of charge via the Internet at
http://pubs.acs.org.
Supporting Information Available: All synthetic procedures
and analytical data, DSC and TGA curves of 7a-c, 1H and 13
C
JA057964G
9
J. AM. CHEM. SOC. VOL. 128, NO. 15, 2006 5099