Structural defects in polyallylcarbosilane dendrimers and their polyol derivatives characterized by NMR and MALDI-TOF mass spectrometry

Article abstract of DOI:10.1021/ma100315w

A series of polyallylcarbosilane dendrimers and carbosilane-based dendritic polyols up to third generation was analyzed by means of matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) mass spectrometry and multinuclear NMR spectroscopy to determine the character, origin, and number of structural defects. Besides occasionally reported α-hydrosilylation, several side reactions accompanying hydrosilylation, such as isomerization of terminal double bonds, were detected during the synthesis of carbosilane skeleton. Despite increased steric hindrance, internal double bonds react in subsequent addition reactions. Depending on the synthetic sequence applied, the retained reactivity of the internal double bonds can lead either to suppression of the defect in the next generation or to creation of more significant defects such as dendrimer dimers. Hydroboration of allyl groups using dicyclohexylborane proceeded with near quantitative conversion; a small amount of hydrolysis accompanying the following oxidation step producing nonreactive alkyl groups at the periphery was detected.

Full text of DOI:10.1021/ma100315w

Macromolecules 2010, 43, 4511–4519 4511
DOI: 10.1021/ma100315w
Structural Defects in Polyallylcarbosilane Dendrimers and Their Polyol
Derivatives Characterized by NMR and MALDI-TOF Mass Spectrometry
ꢁ
Alena Krupkov ꢀa * and Jan Cerm ꢀa k
Institute of Chemical Process Fundamentals, v.v.i., Academy of Sciences of the Czech Republic, Rozvojov aꢀ 135,
165 02 Prague 6, Czech Republic
Zuzana Walterov ꢀa and Ji ꢁr ı
´
Horskꢀy
Institute of Macromolecular Chemistry, v.v.i., Academy of Sciences of the Czech Republic, Heyrovsk eꢀ ho n aꢀ m. 2,
62 06 Prague 6, Czech Republic
1
Received February 8, 2010; Revised Manuscript Received April 8, 2010
ABSTRACT: A series of polyallylcarbosilane dendrimers and carbosilane-based dendritic polyols up to
third generation was analyzed by means of matrix-assisted laser desorption ionization time-of-flight
(
MALDI-TOF) mass spectrometry and multinuclear NMR spectroscopy to determine the character, origin,
and number of structural defects. Besides occasionally reported R-hydrosilylation, several side reactions
accompanying hydrosilylation, such as isomerization of terminal double bonds, were detected during the
synthesis of carbosilane skeleton. Despite increased steric hindrance, internal double bonds react in
subsequent addition reactions. Depending on the synthetic sequence applied, the retained reactivity of the
internal double bonds can lead either to suppression of the defect in the next generation or to creation of more
significant defects such as dendrimer dimers. Hydroboration of allyl groups using dicyclohexylborane
proceeded with near quantitative conversion; a small amount of hydrolysis accompanying the following
oxidation step producing nonreactive alkyl groups at the periphery was detected.
Introduction
processes accompanying the reactions employed is crucial. For
1
polyvinylcarbosilanes, several possible sources of defects were
Since their first synthesis in 1992, carbosilanedendrimers have
2
identified, such as R-hydrosilylation of the vinyl group leading to
gained considerable attention because of their unique properties.
6
branched alkyl spacers in the structure and “deactivation” of the
Nonpolar, highly flexible, and both chemically and thermally
stable skeleton together with variability of the synthetic pathway
and ease of peripheral functionalization make them suitable
for many applications ranging from catalysis to receptors and
biomedicine. However, despite numerous papers on the pre-
paration and functionalization of carbosilane dendrimers, de-
tailed information on the synthesis of the polyvinylcarbosilane
or polyallylcarbosilane skeleton itself is scarce. This applies
even more to the full-scale characterization of products that
could disclose the character and extent of possible structural
defects. These can vary in their impact ranging from small
disturbance of the uniform structure to detrimental effect on
polydispersity due to interference of the defect structure in
subsequent reaction steps.
Carbosilane dendrimers are typically prepared via divergent
approach, starting from a suitable core and growing the molecule
outward by alternation of hydrosilylation of double bonds
with chlorosilanes and nucleophilic substitution of chlorine with
branch by introducing ethyl instead of vinyl moiety during the
reaction with Grignard reagent prepared from vinylbromide
8
3
contaminated with ethylbromide. For polyallylcarbosilanes,
4
no specific defects were described; imperfect structures present
in the published matrix-assisted laser desorption ionization time-
of-flight (MALDI-TOF) mass spectra were ascribed to incom-
7
5
,6
plete hydrosilylation.
7
,8
We have recently constructed a theoretical model of dendrimer
formation to compare MALDI-TOF mass spectra of a small
family of polyallylcarbosilane dendrimers with predicted values
and thus confirm the validity of quantitative interpretation of MS
1
1
data for dendrimers. In this article, we present both qualitative
and quantitative analysis of the MALDI-TOF mass spectra of
two sets of polyallylcarbosilane dendrimers built on tetraallylsi-
lane core and their polyol derivatives together with NMR
characterization of the defect structures and the consideration
of possible consequences these defects can have in the synthesis of
more elaborated structures.
2
alkenylmagnesium halides. Because the divergent synthesis
leaves little space for the purification of products, high selectivity
and quantitative conversion of each iteration step is needed. In
9
Experimental Section
particular, hydrosilylation, which is known to be accompanied
by various types of side reactions and to be highly sensitive to
reagents and conditions applied, requires careful optimization
for each particular combination of silane and olefin type. For
the development of the synthetic protocol, understanding of
General. All reactions were performed under an atmosphere
of dry nitrogen using standard Schlenk techniques. Diethyl
ether was distilled from the solution of sodium benzophenone
ketyl and stored under nitrogen. Borane-dimethylsulfide com-
plex and cyclohexene were purchased from Aldrich and used
without further purification. Functionalized dendrimers 2G-
1
0
6
,12
13
SiH
sly published methods. H, C{ H}, and Si{ H} (INEPT)
8
and 2G-ArOH
8
were prepared according to previou-
1
13
1
29
1
*Corresponding author.
r 2010 American Chemical Society
Published on Web 04/22/2010
pubs.acs.org/Macromolecules

4
512 Macromolecules, Vol. 43, No. 10, 2010
Krupkov ꢀa et al.
1
1
Si ); 0.40 (s, 48H,
spectra were measured at 25 °C in CDCl
00 spectrometer at 299.98, 75.44, and 59.60 MHz, respectively.
Chemical shifts are reported in ppm relative to TMS and
3
on Varian Mercury
2G-C2 l
CH
(m, 8H, Si CH CH CH Si ); 1.45 (m, 16H, Si CH CH -
8
. H NMR: δ -0.05 (s,12H, CH
3
0,1
2
3
3
Si ); 0.57 (m, 32H, CH
2
Si ); 0.89 (m, 16H, CH
2
1
Si ); 1.30
0
1
2
2
2
2
2
1
29
2
13
1
2
0
referenced to internal standard hexamethyldisilane ( H, Si)
or residual solvent peak ( C). Silicon atoms are numbered in
CH
17.7 (C ); 18.1 (C ); 18.6 (C ); 19.0 (C ); 23.5 (CH
2
Si ). C NMR: δ -5.0 (CH
3
Si ); 1.9 (CH
3
Si ); 17.6 (C );
1
3
1
1
0
0
2
29
Si ). Si
2
0
ascending order from the core (Si ) to the periphery, and carbon
0
1
2
NMR: δ 0.53 (Si ); 1.24 (Si ); 31.16 (Si ).
n
atoms are labeled according to generation (C stands for spacer
1
1
2G-Allyl . H NMR: δ -0.07 (s, 12H, CH Si ); -0.02 (s, 48H,
8
2 0,1,2 0,1
3
n
atoms between Si and Si ).
nþ1
3 2 2
CH Si ); 0.57 (m, 48H, CH Si ); 1.33 (m, 24H, Si CH -
1,2
3
4
MALDI-TOF MS. The samples were prepared by the dried
droplet method: tetrahydrofuran (THF) solutions of the den-
drimer (10 mg/mL), matrix (anthracene-1,8,9-triol; 20 mg/mL),
and ionizing agent (silver trifluoroacetate for polyallylcarbosi-
lanes, sodium trifluoroacetate for polyols, 10 mg/mL) were
mixed in the volume ratio 4:20:1, and 0.5 to 1 μL of the mixture
was deposited on the target plate. MALDI TOF MS spectra
were acquired with a Biflex III mass spectrometer (Bruker
Daltonics) in the positive ion reflectron mode using delayed
extraction and accelerating voltage 19 kV. The spectra were
CH CH Si ); 1.51 (ddd, JHH = 8.2 Hz, JHH = 1.2 Hz,
2 2
4
JHH = 1.2 Hz, 16H, CH CHd), 4.81 (m, 8H, dHCHcis-H); 4.83
2
(m, 8H, dHCH
3
3
trans-H
HH HH
); 5.77 (tdd, J = 16.7 Hz, J = 10.2 Hz,
3
13
1
JHH = 8.2 Hz, 8H, dCH-). C NMR: δ -4.9 (CH Si ); -3.6
3
2
(CH Si ); 17.7 (C ); 18.4 (C ); 18.6 (C ); 18.8 (C ); 19.2 (C ); 19.7
0
1
0
1
0
3
1
29
(C );23.4 (CH CHd); 112.6 (dCH ); 135.2 (dCH-). Si NMR: δ
2
2
0
0.52 (Si ); 0.70 (Si ); 0.99 (Si ). MALDI-TOF MS: m/z 1603.73;
2
1
þ
calcd m/z 1603.95 ((MþAg) ; monoisotopic values).
1
1
3G-Cl . H NMR: δ -0.07 (s, 12H, CH Si ); -0.04 (s, 24H,
1
6
3
2
CH Si ); 0.40 (s, 96H, CH Si ); 0.57 (m, 80H, CH Si
3
0,1,2
1,2
3
3
2
); 0.89
3
0,1
summed over 500 shots with a N
2
laser emitting at 337 nm and
(m, 32H, CH Si ); 1.31 (m, 24H, Si CH CH CH Si ); 1.45
2
2
2
2
2
3
13
1,2
consisted of peaks from singly charged ions only because the
spacing between adjacent isotopic peaks in all isotopic clusters
was ∼1 Da. The values of m/z were assigned using the external
calibration with poly(ethylene glycol); no internal recalibration
was applied.
(m, 32H, Si CH CH CH Si ). C NMR: δ -5.0 (CH Si ); 1.9
2
2
2
3
3 2 2 0 1 0
(CH Si ); 17.7 (C ); 18.1 (C ); 18.4 (C ); 18.5 (C ); 18.6 (C ); 18.8
3
1
1
0
2
29
0
(C ); 18.9 (C ); 19.2 (C ); 23.5 (CH Si ). Si NMR: δ 0.52 (Si );
2
1
0.96 (Si ); 1.27 (Si ); 31.09 (Si ).
2
3
1
1,2
3G-Allyl . H NMR: δ -0.07 (s, 36H, CH Si ); -0.02
1
6
3
3
Si ); 0.57 (m, 112H, CH
0,1,2,3
Preparation of Polyallylcarbosilane Dendrimers. Polyallylcar-
bosilane dendrimers were prepared by standard iterative proce-
dure starting from tetraallylsilane as a core and alternating
hydrosilylation of double bonds with either dichloromethylsi-
lane or chlorodimethylsilane in the presence of Karstedt catalyst
and nucleophilic displacement of chlorine by allylmagnesium
bromide. The procedure was optimized on the basis of pre-
(
s, 96H, CH
3
2
Si
); 1.32 (m, 56H,
=
0
,1,2
1,2,3
3
); 1.51 (ddd, J
4
= 8.2 Hz, J
Si
1.2 Hz, J
CH CH CH Si
2
2
2
HH
HH
4
= 1.2 Hz, 32H, CH CHd), 4.82 (m, 16H, dHC-
2
3
HH
H
cis-H
); 4.84 (m, 16H, dHCH
trans-H
); 5.77 (tdd, J = 16.6 Hz,
HH
3
3
JHH = 10.2 Hz, J
13
HH
= 8.2 Hz, 16H, dCH-). C NMR: δ
1
2 3 2
-5.0 (CH Si ); -4.9 (CH Si ); -3.6 (CH Si ); 18.3 (C ); 18.6
3
3
3
2
1
(C ); 18.8 (C ); 18.9 (C ); 19.0 (C ); 19.7 (C ); 23.4 (CH CHd);
2
1
1
2
1
,5,6
0
viously published methods.
scopic data for only some of the lower generation dendrimers
To our knowledge, spectro-
112.6 (dCH ); 135.2 (dCH-); C carbon peaks are too weak for
proper assignment. Si NMR: δ 0.43 (Si ); 0.70 (Si ); 0.94 (Si );
1.02 (Si ). MALDI-TOF MS: m/z 3418.87; calcd m/z 3419.29
2
2
9
0
3
1
1
2,14
2
of these structural types are available in literature,
so we
þ
include here full characterization for the whole series of struc-
tures prepared.
(MþAg) .
Preparation of 2G-OH . Borane-dimethylsulfide complex
8
1
0
1
G-Cl . H NMR: δ 0.69 (m, 8H, CH Si ); 0.78 (s, 12H,
solution (3.60 mL of 2 M solution in diethylether, 7.20 mmol)
was diluted with diethylether (15 mL) and cooled to -10 °C. A
solution of cyclohexene (1.18 g, 14.4 mmol) in diethylether
(5 mL) was added dropwise, and the reaction mixture was
allowed to warm to room temperature with simultaneous stir-
ring until the white precipitate of dicyclohexylborane appeared
(1 h). The suspension was then cooled again to -10 °C, and a
8
2
1
Si ); 1.20 (t, JHH = 7.9 Hz, 8H, CH
3
1
Si ); 1.55 (m, 8H,
CH
Si CH
3
2
0
1
13
1
Si ); 15.8 (CH
0
Si );
2
CH
2
CH
2
Si ). C NMR: δ 5.4 (CH
3
2
1
29
0
1
3
7.3 (CH CH CH ); 25.9 (CH Si ). Si NMR: δ 1.50 (Si );
2
2
2
2
1
2.04 (Si ).
1
1
Si ); 0.55 (m, 8H,
1
CH Si ); 0.63 (m, 8H, CH CH Si ); 1.33 (m, 8H, Si CH -
G-Allyl
8
. H NMR: δ -0.01 (s, 12H, CH
3
0
1
0
2
2
2
2
1
3
4
CH CH Si ); 1.55 (ddd, JHH = 8.1 Hz, JHH = 1.3 Hz,
8
solution of 2G-allyl (1.00 g, 0.60 mmol) in diethylether (5 mL)
2
2
4
2
3
3
J
J
HH = 0.9 Hz, 16H, CH
2
CHd); 4.84 (tdd, JHH = 10.0 Hz,
was added dropwise. The reaction mixture was stirred for 1 h at
-10 °C, for 1 h at 0 °C, and for additional 40 h at room
temperature. Subsequently, methanol (0.1 mL) was slowly
4
HH = 2.1 Hz, JHH = 0.9 Hz, 8H, dHCHcis-H); 4.86 (tdd,
2
JHH = 16.8 Hz, J
4
= 2.1 Hz, J
= 1.3 Hz, 8H, dHC-
HH
3
HH
3
3
H
8
1
(
trans-H); 5.78 (tdd, JHH = 16.8 Hz, JHH = 10.0 Hz, JHH
=
added to decompose excess Cy BH, the reaction mixture was
2
1
3
1
0
Si ); 17.5 (C );
.1 Hz, 8H, dCH-). C NMR: δ -5.8 (CH
3
stirred for 1 h and the volatiles were removed in vacuo. The
residue was redissolved in fresh diethylether (20 mL) and cooled
to -10 °C. Aqueous NaOH solution (3 mL of 3 M solution) was
0
8.1 (C ); 18.4 (C ); 21.5 (CH
0
2
CHd); 113.1 (dCH ); 134.8
2
29
1
0
dCH-). Si NMR: δ 0.19 (Si ); 0.57 (Si ). MALDI-TOF
þ
MS: m/z 803.16; calcd m/z 803.32 ((MþAg) ; monoisotopic
added upon cooling with subsequent dropwise addition of H O
2 2
values).
2
(2.2 mL of 30% solution). The reaction mixture was stirred for 1
h at -10 °C; then, it was allowed to warm up and was refluxed
for 2 h. The organic layer was separated, and the aqueous layer
was extracted with diethylether. Combined organic extracts
were washed with saturated aqueous NaCl solution and dried
with MgSO . The volatiles were evaporated, and the byproduct
4
(cyclohexanol) was distilled off in vacuo. The product was
obtained as a whitish very viscous oil (0.80 g, 73%). H NMR:
δ -0.07 (s, 12H, CH
CH CH CH OH); 0.56 (m, 48H, CH
CH
Hz, 16H, CH
Hz, 16H, CH
10.9 (CH CH
1
1
G-Cl . H NMR: δ -0.02 (s, 12H, CH Si ); 0.61 (m, 32H,
16
3
0,1 2 2
CH
(
2
Si ); 0.77 (s, 24H, CH
3
Si ); 1.18 (m, 16H, CH
Si ); 1.54 (m, 16H, Si CH
2
Si ); 1.31
CH CH
0
m, 8H, Si CH
1
1
2
CH
2
CH
2
2
2
2
-
2
13
1
2
1
Si ). C NMR: δ -5.1 (CH Si ); 5.5 (CH Si ); 17.3 (C ); 17.5
3
3
1
0
0
0
2
29
(
0
C ); 17.6 (C ); 18.5 (C ); 18.8 (C ); 25.9 (CH Si ). Si NMR: δ
2
0
.53 (Si ); 1.46 (Si ); 32.15 (Si ).
1
2
1
1
1
2
G-Allyl . H NMR: δ -0.07 (s, 12H, CH Si ); -0.01
16
3
2
0,1,2
1
2
(
s, 24H, CH
3
Si ); 0.59 (m, 48H, CH
CH
2
Si
Si ); 1.55 (ddd, JHH = 8.2 Hz, JHH = 1.2
CHd); 4.84 (m, 16H, dHCHcis-H);
); 1.34 (m, 24H,
3
3
Si ); -0.02 (s, 48H, CH Si ); 0.48 (m, 16H,
0
,1
1,2
3
4
0,1,2
Si 4CH
2
CH
2
2
2
2
2
Si
2
); 1.32 (m, 24H,
0,1
1,2
3
3
Hz, JHH = 0.9 Hz, 32H, CH
.85 (m, 16H, dHCH
2
Si CH
2
CH
2
2
Si ); 1.54 (tt, JHH = 8.5 Hz, JHH = 6.7
3
3
3
4
1
(
(
(
); 5.78 (tdd, J = 16.7 Hz, J =
HH
2
2
CH OH); 2.20 (bs, 8H, OH); 3.57 (t, JHH = 6.7
trans-H
HH
3
13
13
1
2
Si );
0.3 Hz, JHH = 8.2 Hz, 16H, dCH-). C NMR: δ -5.7
2
OH). C NMR: δ -4.9 (CH
3
Si ); -3.3 (CH
3
2
1
Si ); 17.7 (C ); 18.0 (C ); 18.2 (C ); 18.6
0
1
1
0
1
0
CH
C ); 18.8 (C ); 19.1 (C ); 21.5 (CH
3
Si ); -5.0 (CH
3
2
2 2
CH OH); 17.7 (C ); 18.4 (C ); 18.6 (C ); 18.8
0
1
0
1
0
1
2
CHd); 113.0 (dCH
2
); 134.8
(C ); 19.2 (C ); 20.0 (C ); 27.1 (CH CH OH); 65.7 (CH OH).
2 2 2
Si NMR: δ 0.98 (Si ); 2.23 (Si ); Si not observed. MALDI-
2
9
0
2
1
29
1
2
0
dCH-). Si NMR: δ 0.46 (Si ); 0.24 (Si ); 0.98 (Si ). MALDI-
þ
þ
TOF MS: m/z 1811.94; calcd m/z 1812.07 ((MþAg) ; mono-
TOF MS: m/z 1663.71; calcd m/z 1664.12 ((MþNa) ; mono-
isotopic values).
isotopic values).

Article
Macromolecules, Vol. 43, No. 10, 2010 4513
Figure 1. Overview of synthesis, structures, and notation of dendrimers.
Preparation of 3G-OH16. This dendrimer was prepared by a
method analogous to that described above for 2G-OH , starting
8
from 3G-allyl16 (1.00 g, 0.30 mmol) and dicyclohexylborane
prepared from borane-dimethylsulfide complex (3.60 mL of
“half-generation” dendrimer because after the subsequent
allylation, its signals would become completely hidden under
the signals of next generation allyl groups. In fact, this kind
of defect was seldom observed in the NMR spectra during
optimization, and its presence was probably due to the use of
partially decomposed catalyst. Much more common was the
defect E, caused by isomerization of the terminal double
bond to the internal one, more sterically hindered and as
such less susceptible to hydrosilylation. These two defects are
characterized by the same loss of molecular weight so they
cannot be distinguished in the mass spectra. It is possible that
part of the defects ascribed by Lorenz et al. to incomplete
2
(
M solution in diethylether, 7.20 mmol) and cyclohexene
1.18 g, 14.4 mmol). The crude product containing cyclohexanol
was dispersed in methanol (4 mL), which was then slowly
distilled off to remove possible traces of boric acid in the form
1
5
of methyl borate (azeotropic distillation). The byproduct
cyclohexanol) was then distilled off in vacuo. The product
was obtained as a yellowish very viscous oil (1.00 g, 93%). H
(
1
1
,2
3
NMR: δ -0.07 (s, 36H, CH
3
3
Si ); -0.02 (s, 96H, CH Si ); 0.48
0
,1,2,3
(
(
m, 32H, CH
2
CH
2
CH
2
OH); 0.56 (m, 112H, CH
2
Si
); 1.55 (tt, JHH = 8.4 Hz,
OH); 2.19 (bs, 16H, OH); 3.57
); 1.32
7
conversion of hydrosilylation is actually due to isomeri-
zation. Signals of internal double bonds are downfield from
0,1,2
1,2,3
3
m, 56H, Si
CH
2
CH
2
CH
CH
t, J = 6.6 Hz, 32H, CH OH). C NMR: δ -4.9 (CH Si );
2
Si
3
J
HH = 6.6 Hz, 32H, CH
2
2
1
those of allyl groups in H NMR spectra, and they can be
3
13
1,2
(
-
1
6
HH
2
3
3
2
1
easily distinguished because of their different multiplicity.
(See Table 1, entry 1-3.) Because both E and Z isomer can be
formed, two sets of peaks are usually observed, the one of E
isomer being predominant. With higher amount of isomer-
ization, the number of different sets of signals increases
because of varying chemical environment of particular de-
fects. The lower limit of detection generally increases with
dendrimer generation, and it is affected by many factors such
as sample concentration and setup of the NMR experiment,
3.3 (CH Si ); 10.9 (CH CH CH OH); 18.4 (C ); 18.5 (C );
3 2 2 2
2
1
1
2
8.8 (C ); 18.9 (C ); 19.0 (C ); 20.0 (C ); 27.1 (CH CH OH);
2 2
0
5.7 (CH OH); C carbon peaks are too weak for proper
2
29
1
2
3
0
assignment. Si NMR: δ 0.94 (Si ); 0.99 (Si ); 2.22 (Si ); Si
not observed. MALDI-TOF MS: m/z 3622.51; calcd m/z
þ
3
622.65 (MþNa) .
Results and Discussion
1
Defect Analysis of Polyallylcarbosilane Skeleton. The
synthesis and structure of the dendrimers under study are
summarized in Figure 1. Polyallylcarbosilane dendrimers
but in typical H spectra, it amounts to several tenths of a
percent. Double-bond isomerization also has notable influ-
ence on the adjacent silicon atom, shifting its signal by nearly
9 ppm to higher field.
1
,5,6
were prepared according to literature procedures.
To
obtain products of sufficient quality; the procedure of hy-
drosilylation had to be optimized. During the optimization,
many batches of dendrimers were prepared that differed in
the type and amount of defects present in their structure;
products enriched in a specific type of defect greatly helped
its identification because the signals of impurity became
Because the coordination ability of internal double bond is
lower than that of allyl group, the stabilization of the active
catalyst is worse, and it tends to decompose, leaving the
internal double bonds unreacted. However, it is possible to
convert the internal double bonds to desired hydrosilylation
product by adding fresh catalyst to the reaction mixture and
prolonged heating. The course of the reaction has to be
monitored by NMR, and additional portions of the catalyst
must be added when necessary. In the presence of THF, only
the desired γ-adduct is formed, which means that the hydro-
silylation is preceded by migration of the double bond back
to the terminal position. Still, the method should be used
only when the amount of isomerization is low because long
heating of the reaction mixture with decomposed catalyst has
a negative influence on the overall purity of the dendrimer.
R-Hydrosilylation was previously reported to accompany
the hydrosilylation step in the synthesis of polyvinylcarbo6-
silanes when the reaction was carried out without a solvent.
We have observed that the regioselectivity of hydrosilylation
of allyl groups by dichloromethylsilane is also solvent-
dependent, and two end-group regioisomers are formed in
1
3
29
clearly visible in C and Si NMR spectra.
Because of the nature of divergent synthesis involving
multiple transformations on a single molecule, even small
overall amount of defects introduced in one reaction step can
cause a serious decrease in dendritic purity. The conversion
of a particular reaction is often checked only by NMR, and
signals corresponding to impurities can easily go unnoticed
because of their low intensity and coincidence with other
signals when they are not expected and searched for.
As can be seen from Figure 2, most defects are formed
during the hydrosilylation step of the iteration cycle. Because
the hydrosilylation product cannot be purified because of
its high moisture sensitivity, the defects are translated into
the next reaction step. Defect A, unreacted allyl group, can
be identified only from the NMR spectra of the chlorosilyl

4
514 Macromolecules, Vol. 43, No. 10, 2010
Krupkov ꢀa et al.
w
Figure 2. Possible pathways for introduction of defect structures in one iteration cycle and the impact of defects on the molecular weight. M (D0)
stands for the molecular weight of perfect dendrimer of the x þ 1 generation, X represents the interior of the dendrimer, and n denotes the branching
multiplicity. (A) incomplete conversion of hydrosilylation; (B) hydrolysis of chlorosilane and condensation of the resulting silanol to either intra- or
intermolecular disiloxane; (C) desired pathway; (D) R-hydrosilylation; (E) isomerization of the double bond (both E and Z isomers are possible);
(F) incomplete conversion of allylation (subsequent aqueous workup causes hydrolysis of unreacted chlorosilyl groups).
Table 1. Overview of NMR Spectra of Regular End Groups and Defects Detected by NMR Spectroscopy
a
end group
nuc.
n
significant signals
a
b
c
d
e
CHd CH
1
1
D- Si( Me)3-n( CH
2
2
)
n
H
2
1
2
1
2
1
b -0.01 (s); c 1.55 (ddd 8.2/1.2/0.9); e 4.84 (m); d 5.78 (tdd 16.7/10.3/8.2)
b -0.02 (s); c 1.51 (ddd 8.2/1.2/1.2); e 4.82 (m); d 5.77 (tdd 16.7/10.2/8.2)
1
2
3
C
b -5.7; c 21.5; e 113.1; d 134.8
b -3.6; c 23.4; e 112.6; d 135.2
9
Si
a 0.2
a 0.7
a
b
c
d
e
1
1
2
1
1
2
1
2
3
4
D- Si( Me)3-n( CHd CH CH
3
)
E-isomer
Z-isomer
H
b
b
n=1 0.01 (s); e 1.82 (dd 6.2/1.6); c 5.60 (qd 18.5/1.6); d 6.05 (qd 18.5/6.2)
n=1 -4.8; e 22.8; c 128.9; d 143.1
3
C
9
Si
a -7.4
e 1.77 (dd 6.8/1.5); c 5.45 (qd 14.1/1.5); d 6.46 (qd 14.1/6.8)
e 22.8; c 127.9; d 143.7
a
b
c
d
e
D- Si( Me)3-n( CHd CH CH
3
)
H
3
C
9
Si
a -9.1
b 0.64 (m); f 0.77 (s); d 1.18 (m); c 1.54 (m)
f 0.40 (s); b 0.57 (m); d 0.89 (m); c 1.45 (m)
f 5.5; b 17.3; c 17.5; d 25.9
f 1.9; b 17.7; c 18.1; d 23.5
a 1.5; e 32.1
a 1.2; e 31.2
b 0.39 (dd 6.4/4.2); f 0.82 (s); d 1.09 (t 7.3); c 1.70 (m)
f 7.6; d 17.6; c 18.5; b 19.6
e 32.3
e 0.79 (s)
e 5.3; a 16.5; b 19.3; c 25.1
d 31.8
b 0.16 (s)
b -2.8; c 23.8; e 114.1; d 133.5
a 10.2
b 0.12 (s)
a
b
c
2
d
e
f
D- Si CH
CH
2
CH
2
Si( Me)3-nCl
n
H
2
1
2
1
2
1
*
*
*
*
*
*
1
3
C
2
9
Si
a
b
c
d
e
) Si( Me)Cl
f
1
1
2
1
1
2
1
1
2
1
1
2
1
1
2
1
1
2
1
1
2
1
1
2
5
6
7
8
9
1
1
1
D- Si CH( CH
2
CH
3
2
H
3
C
9
Si
a
b
CH
c
d
e
Si( Me)Cl]
[D- CH
2
2
CH
2
2
-O
H
3
C
9
Si
a
b
c
d
e
CHd CH
[D- Si( Me)( CH
2
2
)]
2
-O
H
3
C
9
Si
a
b
Si Me
[D-CH
2
CH
2
CH
2
2
]
2
-O
H
3
C
b -2.2
a 9.4
b 0.13 (s)
b -0.1
a 17.7
9
Si
a
b
Si (Me
D-CH
2
CH
2
CH
2
2
)-OH
H
3
C
9
Si
a
b
c
2
d
e f
Si( Me
g
) H
0
1
2
D- Si CH
CH
2
CH
2
2
H
f 0.06 (d 3.7); b 0.57 (m); d 0.65 (m); c 1.38 (m); g 3.89 (sp 3.7)
f -4.3; b 18.2; d 18.9; c 19.0
3
C
9
Si
e -14.1; a 1.2
a
b
c
) CH
d
e
f
D- Si( Me
2
2
CH
2
CH
2
OH
H
b -0.02 (s); c 0.48 (m); d 1.54 (tt 8.5/6.7); f 2.20 (bs); e 3.57 (t 6.7)
3
C
b -3.3; c 10.9; d 27.1; e 65.7
a 2.2
9
Si
a
b
c
d
2
D- Si CH(OH) CH
CH
3
H
d 1.02 (t 7.3); b 3.27 (dd 10.0/4.0)
d 11.7; c 26.8; b 66.9
a 1.5
3
C
9
Si
a
In case of defects, only well-resolved signals, which do not overlap with major structure peaks, are given. Depending on generation, branching
pattern, and chemical environment, H shifts can vary by 0.01 to 0.04 ppm, C shifts by 0.1 to 0.4 ppm, and Si shifts by 0.1 ppm from those reported
here. Values for the second generation are given where possible; the first generation signals are marked with asterisk.
1
13
29

Article
Macromolecules, Vol. 43, No. 10, 2010 4515
the ratio of approximately 10:1 when the reaction is per-
formed in nonpolar media. The presence of donor solvent
such as THF or diethylether in the reaction mixture prevents
the formation of unwanted end-group isomer, which was
identified as 1-(dichloromethylsilyl)propyl group from the
NMR spectra (Table 1, entry 5). Just as in the hydrosilylation
of vinylsilanes, the silyl group adds to the R-carbon, and
compound with two geminal silyl groups is formed, which is
consistent with the large influence of this defect on the NMR
spectra of neighboring dendritic branches. Unlike the vinyl
analog, the formation of R-adduct in the hydrosilylation of
allyl group is a two-step process, migration of the double
bond being its first step. That may be the reason why
R-addition to allylsilanes is less extensive than that to
vinylsilanes.
R-Addition does not change either the molecular weight
or the number of peripheral functions of the defect molecule,
so it does not affect the dendrimer growth in a major way,
but it affects the symmetry of the molecule, and because
of the large influence on the adjacent branches, it spoils
the simplicity of the NMR spectra and impedes their un-
ambiguous interpretation. However, this type of defect was
observed only in the synthesis of the first generation den-
drimer because of the fact that the small amount of THF
retained in the dendrimer structure after the first iteration
cycle can suppress the formation of R-adducts in the follow-
ing hydrosilylation, even when it is performed in nonpolar
media.
mass spectra, exact molecular composition of particular
structures can be deduced.
MALDI-TOF mass spectra of regularly branched dendri-
mers (branching multiplicity n = 2 in all generations) show
a simple pattern illustrated by Figure 3a. The difference
between isotopic multiplets corresponds to the loss of a
repeat unit, diallylmethylsilane (126 Da), which can be a
result of defects A or E. A similar pattern was observed with
related dendrimers of higher branching multiplicity (n = 3),
only the mass difference was respectively higher, correspond-
7
ing to triallylsilane (152 Da). Except for the peak of perfect
dendrimer (m/z 1812), only the peak at m/z 1686 can be
assigned unambiguously; it corresponds to one defect in the
outer layer. Other peaks correspond to molecules with
different combinations of defects, for example, the one at
m/z 1560 to a molecule with two defects in the outer layer or
to a molecule with one defect in the inner layer, which had
reacted in the second iteration cycle.
Dendrimers with lower branching multiplicity in the out-
most layer show more complicated spectra (Figure 3b,c). The
structure contains not only two types of repeat units and so
the defects in the outer and inner layers can be distinguished
but also two other types of defects not observed in the NMR
spectra. The structures proposed for defects are shown in
Figure 3d.
The drop of 58 Da compared with the dominant peak
corresponds to a loss of (CH ) Si fragment and so to the
3
2
replacement of -(CH ) Si(CH ) CH CHdCH sequence in
2
3
3 2
2
2
Considering the susceptibility of the Si-Cl bond to hydro-
lysis, the presence of disiloxane bridges, either intra- or
intermolecular, in the molecule can be expected, at least
when the branching multiplicity n = 2,3. These can be
introduced from the silane used in the hydrosilylation step,
they can appear when handling the moisture sensitive “half-
generation” dendrimers, or eventually they can arise from
incomplete allylation, which is followed by the hydrolysis of
the reaction mixture. Nonchlorinated disiloxanes (and sila-
the peripheral layer of the molecule with a purely aliphatic
chain -(CH ) CHdCH . This defect could be due to hydro-
2
4
2
chlorination of the double bond taking place under hydro-
silylation conditions, utilizing hydrochloric acid inherently
present in chlorosilanes used for hydrosilylation and possi-
bly catalyzed by Karstedt catalyst itself or some product
of its decomposition. Chloropropyl group thus generated
would react in the next step with allylmagnesium bromide to
form the proposed structure. The aliphatic spacer can be
either linear or branched depending on the regioselectivity of
hydrochlorination. However, there is no evidence of the
2
9
nols in case of n = 1) give rise to specific signals in Si NMR
spectra that can hardly be mistaken (Table 1, entries 6-9).
However, the presence of a small amount of these defects is
best detected from proton signals of methyl group adjacent
1
3
branched spacer in the C NMR spectra, in contrast with
the presence of several unassigned methylene peaks.
1
to siloxane because of higher signal-to-noise ratio of H
NMR spectra.
The occurrence of the molecule with higher-than-expected
molecular weight is rather surprising. The excess of 74 Da
corresponds to a (CH ) SiO fragment that could be incor-
Intermolecular siloxane bridges nearly double the mole-
cular weight of affected molecules, so their impact on
dendrimer polydispersity is much higher than that of other
defects. At the same time, however, the large difference in
size between perfect and defect structures makes the separa-
tion of this type of defect molecules possible (at least in
principle), for example, by gel permeation chromatography
3
2
porated into the dendrimer skeleton in case the chlorodi-
methylsilane used in the hydrosilylation step contained
(chlorodimethylsiloxy)dimethylsilane. Considering that hy-
drosilanes can undergo disproportionation and redistribu-
1
6
tion reactions under certain conditions, its presence can be
explained by the following reaction sequence.
(GPC). Intramolecular siloxane bridges create molecular
loops in the structure and introduce “dead end” branches,
whose further growth is inhibited so the defects deepen
in subsequent iteration cycles. Fortunately, the disiloxane
bridge is chemically very stable, so it cannot interfere in a
majority of reactions that can be used for periphery func-
tionalization of carbosilane dendrimers.
2
HSiMe2Cl f H2SiMe2 þ SiMe2Cl2
SiMe2Cl2 þ H2O f SiMe2ClOH þ HCl
The sensitivity of NMR experiment does not allow the
detection of a defect present in amounts smaller than several
tenths of a percent; the limit for identification of unknown
defect is even higher. Because of the multiple functionality
that amplifies the number of defect molecules compared with
the fraction of defect end groups, mass spectrometry is far
more sensitive to defects in divergently synthesized dendri-
mers. Soft ionization technique of MALDI MS ensures that
the molecules stay intact and defect structures cannot be
mistaken for fragment ions. Moreover, from high-resolution
SiMe2ClOH þ HSiMe2Cl f HSiMe2OSiMe2Cl þ HCl
Fortunately, these latter two types of defects have negli-
gible effect on the dendrimer properties, retaining the high
chemical and thermal stability and low polarity of the perfect
structure and not even modifying the number and type of end
groups.
Mass spectra of 2G-allyl and 3G-allyl also confirm the
16
8
ability of isomerized or unreacted allyl groups (defects A
and E) to participate in subsequent iteration cycles, which

4
516 Macromolecules, Vol. 43, No. 10, 2010
Krupkov ꢀa et al.
Figure 3. MALDI-TOF mass spectra of polyallylcarbosilane dendrimers and their interpretation; insets showing isotopic multiplets of dominant
peaks are included in the high-resolution spectra. (a) Mass spectrum of 2G-allyl16; (b) mass spectrum of 2G-allyl ; (c) mass spectrum of 3G-allyl16; and
and 3G-allyl16
8
(
d) structures of affected branches proposed for the defects observed in mass spectra of 2G-allyl
8
.
can decrease the impact of these defects on the structure of
higher dendrimer generations. This is fully consistent with
the results previously obtained from a theoretical model for
model of the formation of defects, quantitative interpreta-
tion of the MALDI-TOF mass spectra of carbosilane den-
drimers gives credible results in limited range of m/z.
11
2
be as high as 83%. In the case of 2G-allyl (Figure 3b), the
peak corresponding to a molecule bearing unreacted defect
from the first generation (loss of 326 Da) is not even apparent
in mass spectrum, whereas there is a rather intensive signal
G-allyl , where the conversion of defects was calculated to
Because of the fact that the molecular composition of parti-
cular defect structures of a given generation is very similar
and the same holds for the width and shape of respective
isotopic multiplets, it is possible with slight inaccuracy to use
the intensity of the most intense peak of each isotopic
multiplet for the calculation instead of multiplet integral.
Using this approximation, aggregate conversion of a given
iteration cycle can be easily obtained from the spectra.
Aggregate conversion of the last iteration cycle, ξ, which is
particularly interesting for the optimization of synthesis, was
obtained from the ratio between the intensity of the perfect
molecule peak, ID0, and the sum of intensities of the peaks
corresponding to structures with just one defect, ID1, which is
given by
1
6
11
8
(
m/z 1378) assigned to a molecule with one defect from the
first iteration cycle, which had reacted in the subsequent
hydrosilylation/allylation sequence.
The presence of dimeric or oligomeric structures was never
detected in MALDI-TOF mass spectra. However, responsive-
ness of the method is known to be a function of molecular
17
weight depending on the experimental conditions. Also, the
broadening of isotopic distribution with increasing molecular
weight results in a decrease in relative intensity of particular
isotopic peaks and flattening of the isotopic multiplet so that
the signal-to-noise ratio decreases. A small amount of high-
molecular-weight impurities can therefore easily go unnoticed.
As previously confirmed through the comparison between
experimental data and data obtained from the theoretical
X
ID1=ID0 ¼ f ð1 - ξÞξf - 1=ξf
ð1Þ
where f stands for the number of groups reacting in the given
iteration cycle, for example, for 1G-allyl f = 4, for both
8

Article
Macromolecules, Vol. 43, No. 10, 2010 4517
a
Table 2. Aggregate Conversions , as Calculated from MALDI-TOF
Mass Spectra Presented in Figure 3
instead of solid hydride. Silanols that prevailed in freshly
synthesized materials slowly condensed to disiloxanes, form-
ing structures identical to defect F (Figure 2, n = 1). Under
hydrosilylation conditions of the next step, silanols can also
undergo O-silylation, the resulting product being identical to
that mentioned above. Finally, any defect structure contain-
ing unsaturated bond present in the starting chlorosilyl
dendrimers (defects A and E) is able to undergo hydrosilyla-
tion.
dendrimer
f
ξ
ξ
A
þ ξ
E
ξ
G
ξ
H
2
2
3
G-allyl16
G-allyl
G-allyl16
8
8
16
91.6%
97.0%
96.2%
8.4%
0.9%
1.0%
8
1.2%
1.5%
0.9%
1.3%
a
ξ: aggregate conversion to desired product; ξ
allyl groups; ξ : conversion of isomerization; ξ
characterized by the loss of 58 Da; ξ
by the excess of 74 Da.
A
: fraction of unreacted
: conversion to defect
H
: conversion to defect characterized
E
G
All three pathways can produce both inter- and intramo-
lecular bridges. The polydispersity of 2G-ArOH , prepared
8
1
according to a published procedure by hydrosilylation of
3
2
G-allyl16 and 2G-allyl f = 8, and for 3G-allyl16 f = 16.
8
Therefore, ξ can be calculated according to eq 2
eugenol with 2G-SiH , was high according to GPC (data not
8
shown here), indicating the presence of intermolecularly
bridged structures. Whereas no peaks with higher-than-
expected molecular weight were detected in MALDI-TOF
mass spectra, intramolecular bridges due to hydrosilylation
of defect E, present in the starting chlorosilane dendrimer
and persisting after the reduction step, were clearly apparent
(Figure 4a).
X
ξ ¼ f =ð
ID1=ID0 þ f Þ
ð2Þ
In the case of 2G-allyl and 3G-allyl , where more than
8
16
one type of defect is apparent in the spectra, the amount of
particular defect is also of interest. Knowing the aggregate
conversion of the iteration cycle, the conversion to particular
defect ξ is obtained from the intensity of corresponding
For hydroboration of carbosilane dendrimers, 9-borabi-
cyclononane is mostly used. However, during the subsequent
oxidation step, 1,5-cyclooctanediol is formed as a byproduct,
of which complete removal from the dendritic polyol is
x
peak ID1x
.
ID1x=ID0 ¼ f ξxξf - 1=ξf
ð3Þ
ð4Þ
7
rather complicated because of its high boiling point. There-
fore, we have chosen another reagent, in situ prepared
dicyclohexylborane, which was also previously applied to
carbosilane dendrimers. The hydroboration of 2G-allyl8
ξx ¼ ðξ=f ÞðID1x=ID0Þ
18
Conversions calculated from the spectra in Figure 3 are
presented in Table 2. Low conversion values found for the
defects characterized by loss of 58 Da and by excess of 74 Da
and 3G-allyl16 using dicyclohexylborane gave satisfactory
results with conversion of allyl groups exceeding 99% ac-
cording to NMR analysis. The hydroboration products were
not isolated, and they were immediately oxidized to desired
dendritic polyols 2G-OH and 3G-OH . The density of
(
ξG and ξ ) are consistent with the fact that these defects
H
were not identified in NMR spectra. Also, the approxima-
tion of integrals by maximum intensities gives aggregate
conversion very close to that previously calculated using
8
16
hydroxyl groups on the periphery of 2G-OH8 and
3G-OH is lower than that of related structures prepared
1
6
1
1
7
integration, as can be seen for 2G-allyl16 (91.6 vs 91.2%).
For comparison, aggregated conversions calculated accord-
ing to eq 2 from the published MALDI-TOF mass spectra
of higher branching polyallylcarbosilane dendrimers, 2G-
previously by Lorenz et al. so they are not soluble in
alcohols but they are well soluble in moderately polar
solvents such at diethylether or chloroform.
As confirmed by NMR, hydroboration proceeded with
high conversion not only on peripheral allyl groups but also
on isomerized interior double bonds (if present in the starting
dendrimer). Considering the substitution of double bond
with electrondonating trialkylsilyl group, we have expected
the formation of 1-hydroxypropyl group, which was consis-
tent with the observed multiplicity of corresponding signals
7
allyl36 and 3G-allyl108, amount to 97 and 96%, respectively.
The conversion of the first iteration cycle can also be esti-
mated from the spectrum of 2G-allyl36 as equal to 95%.
Synthesis and Analysis of Dendritic Polyols. The majority
of carbosilane dendritic polyols is prepared using two stra-
tegies. The first one involves reduction of peripheral chloro-
1
silyl groups, typically with LiAlH , followed by hydrosilyla-
4
in the H NMR spectrum (Table 1, entry 12). No side
tion of suitable unsaturated (and usually protected) alcohol
with dendritic hydrosilane thus obtained. This strategy
enables incorporation of a wide range of alcohols, including
substituted phenols, onto the dendrimer periphery. The
second possibility is the hydroboration/oxidation procedure,
applied to carbosilane dendrimer of any generation with
double bonds on the periphery. This strategy yields simple
structures ending with short hydroxyalkyl groups.
reactions related to hydroboration were detected, but the
subsequent oxidation step was accompanied by basic hydro-
lysis of intermediate alkylboranes to alkanes. This reaction is
well apparent in Si NMR spectrum, where the replacement
of 3-hydroxypropyl group at silicon by propyl group causes a
shift of the corresponding signal by ∼0.9 ppm to higher field.
The products formed under hydroboration-oxidation con-
ditions are summarized in Figure 5.
2
9
The first strategy is more flexible but also more prone to
produce materials of high polydispersity because of many
possible side reactions. Whereas the concurrent O- and
C-silylation can be easily avoided by the protection of
unsaturated alcohol used, other pathways leading to den-
drimer dimers and oligomers are inherently present in the
reaction sequence. Any chlorosilyl group that has not re-
acted in the reduction step would hydrolyze during the
reaction mixture workup to form silanols and disiloxanes.
In fact, we have detected signals of both silanols and
disiloxanes in the NMR spectra of all hydrosilane dendri-
The presence of hydrolytic products in the structure of
dendritic polyols was also confirmed by MALDI-TOF MS
analysis (Figure 4b). Although it is not possible to distin-
guish this defect characterized by loss of 16 Da from un-
reacted allyl group (loss of 18 Da) because of low resolution
in the spectrum of 3G-OH , from the high-resolution spec-
1
6
trum of 2G-OH it is clearly evident that the defect caused by
8
hydrolysis predominates. The same defect, albeit in lower
amount, can be observed in MALDI-TOF mass spectrum of
similar second-generation dendritic polyol synthesized by
Lorenz et al., although the authors claim quantitative con-
7
mers prepared by the common procedure introduced by
Seyferth et al., even when using ether solution of LiAlH
version of the hydroboration-oxidation sequence. In the
6
spectrum of 3G-OH , it is also possible to find defects
16
4

4
518 Macromolecules, Vol. 43, No. 10, 2010
Krupkov ꢀa et al.
Figure 4. MALDI-TOF mass spectra of dendritic polyols and their interpretation including structures of affected branches proposed for the defects.
a) Mass spectrum of 2G-ArOH with inset showing isotopic multiplet of dominant peak and (b) mass spectrum of 3G-OH16, with inset showing
segment of mass spectrum of 2G-OH
(
8
8
.
Conclusions
Combined multinuclear NMR and MALDI-TOF MS analysis
was used for the determination of structure, origin, and amount of
defects present in divergently synthesized polyallylcarbosilane
dendrimers and their polyol derivatives. The main part of defects
characterized by missing repeating units in MS spectra was
ascribed to isomerization of allyl groups under hydrosilylation
conditions. Isomerized double bonds retain some reactivity to-
ward hydrosilylation, which can, on one hand, contribute to
suppression of the impact of defects on the structure of higher
generations; on the other hand, it is a possible source of dendrimer
dimers and oligomers when such defect structure is reduced to
hydrosilyl-functionalized dendrimer and used in hydrosilylation
of an unsaturated substrate. Insensitivity of MALDI-TOFMS for
high-molecular-weight impurities was observed so that their
absence has to be confirmed by complementary NMR and
GPC analyses. The defects incorporated into the structure due
to impurities present in reagents used are of small importance
because they do not alter either polarity or functionality of the
dendritic molecule. The presented summary of significant NMR
signals of a range of impurities as well as simplified equations for
quantitative analysis of MS spectra may help in further optimiza-
tion of synthesis of structurally related dendrimers.
Figure 5. Reactions occurring under hydroboration-oxidation condi-
tions. (A) Desired course of reaction sequence; (B) hydroboration
followed by hydrolysis of intermediate borane; (C) reaction on defect
branch yielding secondary alcohol.
identical to those present in the starting 3G-allyl16 (peaks
þ74, -58, and -100 Da, cf. Figure 3c). The mass difference
from the corresponding dominant peak stays unchanged
after the hydroboration-oxidation sequence, which implies
that these defects do not affect the number of double bonds
on the dendrimer periphery. This fact further supports the
correctness of proposed structures. By contrast, each defect
in the preceding (second) generation lowers the number of
functional groups on the periphery by one, which causes the
increase by 18 Da (one molecule of water) in the difference
between dominant peaks corresponding to lower generation
defects upon going from 3G-allyl16 to 3G-OH16.
Acknowledgment. We thank the Ministry of Education, Youth
and Sport of the Czech Republic (grant LC06070) and to Academy
of Sciences of the Czech Republic (grant AV0Z40500505) for
financial support.
The conversion of allyl to 3-hydroxypropyl groups, as
calculated from the mass spectra in Figure 4b, was 95.9 and
9
the high-resolution spectrum of 2G-OH , the remaining
8.1% for 2G-OH and 3G-OH , respectively. According to
8 16
8
References and Notes
fraction consists mainly of propyl groups formed by hydro-
lysis, which are nonreacrive and would not interfere in
further functionalization of the dendrimer.
(
1) (a) van der Made, A. W.; van Leeuwen, P. W. N. M. J. Chem. Soc.,
Chem. Commun. 1992, 1400–1401. (b) Zhou, L. L.; Hadjichristidis, N.;

Article
Macromolecules, Vol. 43, No. 10, 2010 4519
Toporowski, P. M.; Roovers, J. Rubber Chem. Technol. 1992, 65,
03–314.
(7) Lorenz, K.; M u€ lhaupt, R.; Frey, H.; Rapp, U.; Mayer-Posner, F. J.
Macromolecules 1995, 28, 6657–6661.
3
(
2) (a) Schlenk, C.; Frey, H. Monatsch. Chem. 1999, 130, 3–14. (b) Frey,
H.; Schlenk, C. Top. Curr. Chem. 2000, 210, 69–129. (c) Lukevics, E.;
Arsenyan, P.; Pudova, O. Main Group Met. Chem. 2002, 25, 135–154.
(8) Allgaier, J.; Martin, K.; R €a der, H. J.; M u€ llen, K. Macromolecules
1999, 32, 3190–3194.
(9) Bosman, A. W.; Janssen, H. M.; Meijer, E. W. Chem. Rev. 1999, 99,
1665–1688.
(
d) Silicon-Containing Dendritic Polymers; Dvornic, P., Owen, M. J.,
Eds.; Springer: Dordrecht, The Netherlands, 2009.
(10) (a) Speier, J. L. Adv. Organomet. Chem. 1979, 17, 407–447.
(b) Marciniec, B.; Guli ꢀn ski, J.; Urbaniak; W.; Kornetka, Z. W. In
Comprehensive Handbook on Hydrosilylation; Marciniec, B., Ed.;
Pergamon: Oxford, U.K., 1992; pp 758.
(
3) (a) 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, 659–663. (b) Astruc, D.; Chardac, F. Chem.
Rev. 2001, 101, 2991–3023. (c) Oosterom, G. E.; Reek, J. N. H.;
Kamer, P. C. J.; van Leeuwen, P. W. N. M. Angew. Chem., Int. Ed.
ꢁ
(11) Krupkov ꢀa , A.; Cerm ꢀa k, J.; Walterov ꢀa , Z.; Horsk ꢀy , J. Anal. Chem.
2007, 79, 1639–1645.
2
001, 40, 1828–1849. (d) van Heerbeek, R.; Kamer, P. C. J.;
van Leeuwen, P. W. N. M.; Reek, J. N. H. Chem. Rev. 2002, 102,
717–3756.
(12) Roesler, R.; Har, B. J. N.; Piers, W. E. Organometallics 2002, 21,
4300–4302.
3
(13) Ar ꢀe valo, S.; Benito, J. M.; de Jes uꢀ s, E.; de la Mata, F. J.; Flores,
J. C.; G oꢀ mez, R. J. Organomet. Chem. 2000, 602, 208–210.
(14) (a) Ignateva, G. M.; Rebrov, E. A.; Myakushev, V. D.; Chenskaya,
T. B.; Muzafarov, A. M. Vysokomol. Soedin., Ser. A. 1997, 39,
1271–1280. (b) Ponomarenko, S. A.; Rebrov, E. A.; Boiko, N. I.;
Muzafarov, A. M.; Shibaev, V. P. Vysokomol. Soedin., Ser. A 1998, 40,
1253–1265. (c) Kleij, A. W.; Gossage, R. A.; Klein Gebbink, R. J. M.;
Brinkmann, N.; Reijerse, E. J.; Kragl, U.; Lutz, M.; Spek, A. L.; van
Koten, G. J. Am. Chem. Soc. 2000, 122, 12112–12124.
(15) Chung, T. C.; Raate, M.; Berluche, E.; Schulz, D. N. Macromole-
cules 1988, 21, 1903–1907.
(
4) (a) Matsuoka, K.; Terabatake, M.; Esumi, Y.; Terunuma, D.;
Kuzuhara, H. Tetrahedron Lett. 1999, 40, 7839–7842. (b) Nishikawa,
K.; Matsuoka, K.; Watanabe, M.; Igai, K.; Hino, K.; Hatano, K.; Yamada,
A.; Abe, N.; Terunuma, D.; Kuzuhara, H.; Natori, Y. J. Infect. Dis. 2005,
1
91, 2097–2105. (c) Yamada, A.; Hatano, K.; Matsuoka, K.; Koyama, T.;
Esumi, Y.; Koshino, H.; Hino, K.; Nishikawa, K.; Natori, Y.; Terunuma,
D. Tetrahedron 2006, 62, 5074–5083. (d) Pedziwiatr, E.; Shcharbin, D.;
Chonco, L.; Ortega, P.; de la Mata, F.; G ꢀo mez, R.; Klajnert, B.;
Bryszewska, M.; Mu ~n oz-Fern ꢀa ndez, M. A. J. Fluoresc. 2009, 19,
2
67–275. (e) Rasines, B.; Hern ꢀa ndez-Ros, J. M.; de las Cuevas, N.;
Copa-Pati ~n o, J. L.; Soliveri, J.; Mu ~n oz-Fern ꢀa ndez, M. A.; G ꢀo mez, R.;
dela Mata, F. J. DaltonTrans. 2009, 8704–8713. (f)Ortega, P.; Serramía,
M. J.; Samaniego, R.; de la Mata, F. J.; G ꢀo mez, R.; Mu ~n oz-Fern ꢀa ndez,
M. A. Org. Biomol. Chem. 2009, 7, 3079–3085.
(16) (a) Sheludyakov, V. D.; Zhun, V. I.; Ten, M. K. Zh. Obshch. Khim.
1987, 57, 567. (b) Belyakova, N. R.; Mileshkevich, V. P. Zh. Obshch.
Khim. 1989, 59, 724.
(17) Williams, J. B.; Chapman, T. M.; Hercules, D. M. Anal. Chem.
2003, 75, 3092–3100.
(
(
5) Zhou, L. L.; Roovers, J. Macromolecules 1993, 26, 963–968.
6) Seyferth, D.; Son, D. Y.; Rheingold, A. L.; Ostrander, R. L.
Organometallics 1994, 13, 2682–2690.
(18) Terunuma, D.;Kato, T.;Nishio, R.;Aoki, Y.;Nohira, H.;Matsuoka,
K.; Kuzuhara, H. Bull. Chem. Soc. Jpn. 1999, 72, 2129–2134.