Polycationic (Mixed) Core-Shell Dendrimers for Binding and Delivery of Inorganic/Organic Substrates

Article abstract of DOI:10.1002/1521-3765(20010105)7:1<181::AID-CHEM181>3.0.CO;2-1

The convergent synthesis of a series of polycationic aryl ether dendrimers has been accomplished by a convenient procedure involving quantitative quaternarization of aryl(poly)amine core molecules. The series has been expanded to the preparation of the first polycationic, mixed core-shell dendrimer. All these dendrimers consist of an apolar core with a peripheral ionic layer which is surrounded by a less polar layer of dendritic wedges. These cationic, macromolecular species have been investigated for their ability to form assemblies with (anionic) guest molecules. The results obtained from UV/Vis and NMR spectroscopies, and MALDI-TOF-MS demonstrate that all the cationic sites throughout the dendrimer core are involved in ion pair formation with anionic guests giving predefined guest/host ratios up to 24. The large NMR spectroscopic shifts of resonances correlated with the groupings located in the core of the dendrimers, together with the relaxation time data indicate that the anionic guests are associated with the cationic core of these dendrimers. The X-ray molecular structure of the octacationic, tetra-arylsilane model derivative [Si(C6H3{CH2NMe3}2-3,5)4](8+) * 8I(-) shows that the iodide counterions are primarily located near the polycationic sphere. The new polycationic dendrimers have been investigated for their catalytic phase-transfer behavior and substrate delivery over a nanofiltration membrane.

Full text of DOI:10.1002/1521-3765(20010105)7:1<181::AID-CHEM181>3.0.CO;2-1

FULL PAPER
Polycationic (Mixed) Core ± Shell Dendrimers for Binding and
Delivery of Inorganic/Organic Substrates
Arjan W. Kleij,[a] Rob van de Coevering,[a] Robertus J. M. Klein Gebbink,[a]
Anne-Marie Noordman,[a] Anthony L. Spek,[b] and Gerard van Koten*[a]
Abstract: The convergent synthesis of a
series of polycationic aryl ether den-
drimers has been accomplished by a
convenient procedure involving quanti-
tative quaternarization of aryl(poly)-
amine core molecules. The series has
been expanded to the preparation of the
first polycationic, mixed core ± shell den-
drimer. All these dendrimers consist of
an apolar core with a peripheral ionic
layer which is surrounded by a less polar
layer of dendritic wedges. These cation-
ic, macromolecular species have been
investigated for their ability to form
assemblies with (anionic) guest mole-
cules. The results obtained from UV/Vis
and NMR spectroscopies, and MALDI-
TOF-MS demonstrate that all the cat-
ionic sites throughout the dendrimer
core are involved in ion pair formation
with anionic guests giving predefined
guest/host ratios up to 24. The large
NMR spectroscopic shifts of resonances
correlated with the groupings located in
the core of the dendrimers, together
with the relaxation time data indicate
that the anionic guests are associated
with the cationic core of these dendrim-
ers. The X-ray molecular structure of the
octacationic, tetra-arylsilane model de-
rivative [Si(C6H3{CH2NMe3}2-3,5)4]8‡
´
8Iꢀ shows that the iodide counterions
are primarily located near the polycat-
ionic sphere. The new polycationic den-
drimers have been investigated for their
catalytic phase-transfer behavior and
substrate delivery over a nanofiltration
membrane.
Keywords: dendrimers ´ nanofiltra-
tion
´
noncovalent interactions
´
phase-transfer catalysis
tions
´
polyca-
irradiation or pH changes.[4] Most of the dendritic containers
reported so far use the concept of closing the dendrimer
surface in order to be able to keep the guest molecules inside
the dendrimer host. In this case, high generation dendrimer
species with sufficient loading of peripheral groups are
needed for an effective binding of guest molecules inside
their cavities. However, up to now, few reports exist on well-
defined molecular containers that exhibit flexible binding
properties specifically inside its internal cavities which do not
depend on the peripheral crowding of the system but solely on
the binding properties at the core.[5a±c]
Here, we focus on this particular aspect by using a different
approach of binding guest molecules inside dendritic host
systems. Two types of polycationic species[5d] were designed;
the first one is constructed from polyaryl(di)amine core
molecules and contains an ionic core shell and a less polar
outer shell. The second type is derived from a relatively large,
carbosilane dendritic core producing polycationic mixed
core ± shell dendrimer species with a potentially variable,
but with a distinct distance between the core of the molecule
and the ionic layer (Figure 1).
Introduction
Dendritic polymers[1] with regular and well-defined uni-
molecular architectures currently attract much interest as
soluble supports[2] or as so-called dendritic boxes or molecular
containers.[3] The micro-environment inside these materials is
usually less densely packed compared with the sterically
congested outer shell, which enables the encapsulation of
(small) organic guests in the dendritic host. The latter offers
the possibility to convert these macromolecules into bio-
logically important delivery systems which can release organic
guest molecules by means of external stimulus, for example
[a] Prof. Dr. G. van Koten, A. W. Kleij, R. van de Coevering
Dr. R. J. M. Klein Gebbink, A.-M. Noordman
Debye Institute, Department of Metal-Mediated Synthesis
Utrecht University, Padualaan 8, 3584 CH Utrecht (Netherlands)
Fax : (‡ 31)30-2523615
E-mail: g.vankoten@chem.uu.nl
[b] Prof. Dr. A. L. Spek
Bijvoet Center for Biomolecular Research
Department of Crystal and Structural Chemistry
Utrecht University, Padualaan 8, 3584 CH Utrecht (Netherlands)
E-mail: a.l.spek@chem.uu.nl
These new polycationic dendrimers can potentially be used
for the multiple incorporation of anionic guests inside their
core or cavities. Both types of cationic dendrimers should be
capable of specific binding to a number of different anionic,
Supporting information for this article is available on the WWW under
http://www.wiley-vch.de/home/chemistry/ or from the author.
Chem. Eur. J. 2001, 7, No. 1
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0701-0181 $ 17.50+.50/0
181

G. van Koten et al.
FULL PAPER
organic substrates through electrostatic forces[6] regardless of
the steric features on the surface. Moreover, this behavior is
likely to be fully reversible, which could be relevant for an
eventual reuse of the system. The application of these systems
in phase-transfer catalysis and substrate delivery over a
nanofiltration membrane is presented.
Results and Discussion
Synthesis of model species: In order to investigate whether
arylamine building blocks could be used for the construction
of larger (dendritic) macromolecules, we first carried out
quaternarization of a model tetra-arylsilane compound,
namely Si(C6H3{CH2NMe2}2-3,5)4 (abbreviated as Si(NCN)4)
(1a)[7] with various alkyl halides. In order to be able to vary
the amount of anchoring groups and the final structure of the
polycationic products we have also prepared core molecule
1b (Scheme 1). Core molecules 1a and 1b give rise to eight
and four ammonium centers, respectively, after quaternariza-
tion. Core molecule 1b was obtained as a white solid by in situ
treatment of an excess of [LiC6H4(CH2NMe2)-4] with SiCl4 in
Et2O at ꢀ788C as described earlier for 1a.[7]
Treatment of 1a with an excess of methyl iodide at RT in
CH2Cl2 led to the immediate precipitation of a white solid
material, which was isolated by centrifugation. 1H NMR
spectroscopic analysis performed in D2O pointed to selective
formation of the octatrimethylammonium derivative
[Si(C6H3{CH2NMe3}2-3,5)4]8‡ ´ 8Iꢀ (2) in nearly quantitative
Figure 1. Schematic representation of two types of polycationic macro-
molecular hosts for anionic guest molecules.
Abstract in Dutch: Dit onderzoek beschrijft een serie poly-
kationische arylether-dendrimeren, die in kwantitatieve op-
brengst zijn verkregen door reactie van dendritische benzyl
bromides met organosilanen die vier aryl(di)amine substituen-
ten bevatten. Deze dendrimeren zijn opgebouwd uit een
polykationische kern en een polaire buitenlaag. Tevens wordt
de synthese van het eerste gemengde, polykationische dendri-
meer beschreven. Dit nieuwe type dendrimeer bestaat achter-
eenvolgens uit een [G1] carbosilaandendrimeer kern, een poly-
ionische tussenlaag en een polaire buitenlaag opgebouwd uit
NMe2
Br
Br
NMe2
NMe2
1. 2 equiv tBuLi
2. SiCl4
Et2O, -78 → 0oC
1. 2 equiv tBuLi
2. SiCl4
Et2O, -78 → 0oC
core molecules
NMe2
Si
Si
Â
[G1] Frechet type dendrons. De bindingscapaciteiten van deze
NMe2
4
NMe2
4
dendrimeren ten aanzien van anionische gast-molekulen zijn
onderzocht. Spectroscopische metingen (UV/Vis, NMR en
MALDI-TOF) wijzen erop dat alle kationische centra in de
dendrimeren beschikbaar zijn voor de vorming van ionparen.
Zo is het mogelijk om een maximum aantal van 24 anionische
gasten te binden. De grote NMR spectroscopische verschuivin-
gen gecombineerd met relaxatiemetingen van de groepen die
aanwezig zijn in de kern van de dendrimeren, duiden op een
duidelijke associatie van de anionen met de kationische centra
van de dendrimeren. De kristalstructuur van de octakationi-
sche model-verbinding [Si(C6H3{CH2NMe3}2-3,5)4]8‡ ´ 8Iꢀ laat
zien dat de (jodide) anionen zich voornamelijk in de buurt van
de kationische centra bevinden. De nieuwe polykationische
dendrimeren zijn gebruikt in fase-transfer katalyse en voor het
transport van organische substraten door nanofiltratiemem-
branen.
1b
1a
excess RX
CH2Cl2, RT
X–
X–
+
+
Me2RN
NRMe2
X–
X–
+
+
Me2RN
NRMe2
Si
Me2RN
NRMe2
+
+
X–
X–
Me2RN
+
NRMe2
+
X–
X–
2 R = Me, X = I
3 R = Et, X = Br
4 R = Oct, X = I
Scheme 1. Synthesis of core molecules 1a and b and model polycationic
species 2 ± 4.
182
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0701-0182 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 1

Polycationic Dendrimers
181±192
yield. This procedure proved to be of general use and the octa-
ethyl and octa-octyl derivatives 3 and 4 were also isolated as
white solids in good yields by treatment of 1a with ethyl
bromide and octyl iodide, respectively (Scheme 1).
Compounds 2 ± 4 are hygroscopic solids melting between
73 ± 778C without decomposition, unlike the polycationic
materials prepared by Stoddart et al.[8] The composition of 2 ±
4 was confirmed by NMR spectroscopy, fast atom bombard-
ment (FAB) mass spectroscopy, and elemental microanalysis.
The FAB mass spectra of 2 ± 4 showed characteristic isotopic
distributions at m/z 1801.0, 1585.1, and 2587.0, corresponding
cationic core of the dendrimer species presented here (see
below). Crystals of 2 were obtained from a 2:1 mixture of
MeOH/EtOH at ꢀ208C. The polycationic part of the
molecular structure of 2 is depicted in Figure 2, top and
relevant bond lengths and angles are given in Table 1.
Crystallographic data and other related details are summar-
ized in the Experimental Section.
It should be noted that the refinement of the data was
troublesome because of the large disorder of the solvent
molecules and partial disorder of the iodide anions. A first
impression of the unique solid state structure of octacationic 2
could be obtained emphasizing a fully methylated species with
eight quaternarized amine moieties forming a soft, cationic
sphere at about 5.9 Š from the centroid Si center. Although
partially disordered, the larger part of the iodide anions are
primarily located near/between the cationic centers (Figure 2,
‡
‡
‡
to [2 ꢀ I] , [3 ꢀ Br] , and [4‡H ꢀ I] , respectively. Beside
these easily assignable peaks, the spectra also showed other
complicated isotope patterns that most likely resulted from
the exchange of halide anions with matrix molecules,[9a] as
well as degradation of the quaternary ammonium salts via
[10]
ꢀ
C N bond cleavage.
The solubility characteristics
of these dendritic cationic de-
rivatives varied upon increasing
the length of the alkyl halide:
Octamethylated 2 is only solu-
ble in H2O and MeOH whereas
the octa-octylated derivative 4
is insoluble in water and dis-
solves readily in relatively non-
polar solvents such as CH2Cl2
and CHCl3. This effect can be
rationalized by assuming that
the aliphatic octyl chains in 4
strongly interact with each oth-
er and thus shield the hydro-
philic polycationic core, similar
to
a reversed micelle. The
1H NMR spectra recorded for
4 at different concentrations
showed more than one set of
resonances for the aromatic
protons and CH2N and NMe2
groupings. Similar dependen-
cies of the NMR spectroscopic
patterns on the concentration
of dendritic ammonium salts
have been reported.[9b]
X-ray crystal structure of 2:
Although several groups have
characterized organic mole-
cules crystallographically con-
taining triorganoammonium io-
dide moieties,[11] the structure
of 2 in the solid state is the first
example of a polyammonium
derivative containing more
than four of such groupings
within a single molecule. More-
over, this X-ray crystal struc-
ture can be regarded as a rep-
resentative model for the poly-
Figure 2. Top: Displacement elipsoid plot (ORTEP, 50% probability level) of the polycationic part of the
molecular structure of 2. Hydrogen atoms, iodide anions, and cocrystallized, disordered solvent molecules have
been omitted for clarity. Bottom: Space-filling model of the molecular structure of the polycationic unit of 2
together with the iodide anions.
Chem. Eur. J. 2001, 7, No. 1
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0701-0183 $ 17.50+.50/0
183

G. van Koten et al.
FULL PAPER
Table 1. Selected bond lengths [Š] and angles [8] for 2 with esdꢁs in
parentheses.
bottom). The presence of the relatively large iodide anions
near the cationic core could be responsible for the minor
deviations from ideal geometry found for some bond lengths
and angles (Table 1). The key point of this structure is that it
substantiates earlier implications that anions could be ac-
comodated inside dendritic structures by means of ion pair
formation,[8] though in solution this feature may not be fully
retained.
bond lengths
ꢀ
ꢀ
Si C11
1.896(17)
1.891(18)
1.60(5)
1.42(4)
1.57(3)
Si C21
1.878(19)
1.854(19)
1.47(2)
1.53(2)
1.50(2)
ꢀ
ꢀ
Si C31
Si C41
ꢀ
ꢀ
N11 C111
N11 C112
N11 C113
N12 C121
N12 C122
N12 C123
ꢀ
ꢀ
ꢀ
ꢀ
bond angles
C11-Si-C21
C11-Si-C41
C21-Si-C41
105.7(8)
111.7(8)
108.4(8)
C11-Si-C31
C21-Si-C31
C31-Si-C41
108.0(7)
116.4(8)
106.7(7)
Synthesis of polycationic aryl ether dendrimers: The dendritic
molecules 5 ± 8 (Scheme 2) were constructed from the tetra-
O
O
O
O
O
O
Br –
+
+
NMe2
+
O
O
O
O
Br –
Me2N
N
Me2
Br –
Br –
Br –
Br –
Me2
Me2
O
Me2
N
N
O
O
O
N
O
O
+
+
Br –
Si
Br –
+
Si
+
+
N
+
N
O
N
Br –
Me2
O
Me2
Br –
Me2N
Me2
Br –
Br –
+
+
NMe2
Me N
2
+
Br –
O
O
O
O
O
O
7
5
O
O
[G1]-Br
[G1]-Br
Br
1a
1b
O
O
O
[G2]-Br
[G2]-Br
Br
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
Br –
+
NMe2
Br –
O
O
Me N
N
Me
+
+
2 Br –
Me2
2
Br
–
O
Br
O
O
O
O
Me2
Me2
O
O
N
N
O
N
O
O
O
+
+
O
O
O
O
Br –
Si
Br –
+
+
N
Si
+
N
+
O
O
O
O
O
N
Me2
O
Br –
Me2
Me2
Br –
Br –
Br –
NMe
2
O
O
O
Me N
2
+
Me N
2
+
Br –
+
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
6
8
Scheme 2. Synthesis of the polycationic dendrimers 5 ± 8.
184
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0701-0184 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 1

Polycationic Dendrimers
181±192
Â
arylsilane building blocks 1a and 1b, and Frechet-type
dendrons (abbreviated as [G1]-Br or [G2]-Br)[12a] under mild
reaction conditions (CH2Cl2, RT). These dendrons contain a
benzyl bromide focal point that can be used for the selective
and quantitative quaternarization of the nitrogen atoms
present in the core molecules 1a and 1b.[12b] This gave almost
quantitative yields (95 ± 100%) of the functionalized macro-
molecules 5 ± 8 as white, hygro-
of 5 ± 8 (see below) to give the new dendrimer species
ꢀ
[G1]-SiMe2 NCN ´ 24[G1]-Br (9) as a white solid in quanti-
tatively yield. Dendrimer 9 consists of three distinctive layers:
a rather apolar carbosilane inner shell, a middle ionic shell,
and an less polar outer shell.
1H and 13C{1H} NMR spectroscopic analysis of this new
dendrimer showed that all the tertiary amines had successfully
scopic solids. These dendrimer
species were fully characterized
by NMR spectroscopy (1H and
13C{1H}), electrospray mass
spectrometry (ES-MS), and
combustion analyses, which
confirmed in all cases the iso-
lation of a single product. The
ES-MS spectra recorded for
5 ± 8 showed, for example, the
presence of multiple cationic
fragment ions at m/z 969.3
(calcd for [5 ꢀ 2Br]2‡: 969.1),
1818.6 (calcd for [6 ꢀ 2Br]2‡
:
1818.1), 1849.8 (calcd for
[7 ꢀ 2Br]2‡
:
1849.9),
and
Figure 3. Space-filling models of the calculated energy-minimized polycationic structures of 7 and 8.
2338.7 (calcd for [8 ꢀ 3Br]3‡
:
2338.5), respectively.
NMe2
Compounds 5 ± 8 are soluble
in solvents such as CH2Cl2 and
toluene, but are essentially in-
soluble in water. This behavior
strongly suggests that the local,
ionic environment inside these
dendrimers is well-shielded by
the outer, polyether dendrons
(cf. the position of the iodide
anions in the structure of 2).
Molecular mechanics studies[13]
performed on 7 and 8 support-
ed this assumption and showed
that these dendrimers are able
to adopt structures in which
potential guests can be embed-
ded in the internal cavities (Fig-
ure 3).
Me
Si
Si
Si
Me
NMe2
3
4
[G1]-Br
Ph
Ph
Ph
Ph
Ph Ph
+
24+
N = N(Me2)
O
O
O
O
O
O
Ph
Ph
O
Ph
Ph
O
O
O
Ph
Ph
24 Br–
Ph
Ph
O
·
N
N
O
N
O
O
N
Ph
Ph
Ph
Ph
O
O
N
Me
Me
Si
N
Me
Si
Me
Si
O
O
N
N
Me
Me
Ph
Ph
Ph
Ph
O
O
N
N
Si
Si
Si
Me
Si
O
Me
Si
O
Me
Me
Si
Me
Me
Me
Me
Si
Me
Me
N
N
O
O
O
O
Ph
Ph
Ph
Ph
Si
Si
Preparation and analysis of a
mixed core ± shell dendrimer
species: Beside the preparation
of dendrimers with a central
ionic core, we have also pre-
N
N
Si
Me
Si
Me
O
O
O
Ph
Ph
Ph
N
N
O
Ph
Me
Si
Me
Me
Me
N
N
Si
N
O
O
Si
Me
Ph
Ph
N
Ph
Ph
Me
N
O
O
pared
dendrimer
a
polycationic mixed
species
9
N
O
O
O
N
N
O
(Scheme 3). The synthesis of 9
was realized by treatment of the
known carbosilane dendrimer
Ph
Ph
Ph
Ph
O
O
O
O
Ph
Ph
Ph
Ph
O
O
O
O
O
O
[7]
ꢀ
[G1]-SiMe2 NCN
with
a
Ph
Ph
Ph
Ph
PhPh
slight excess of a first genera-
9
Â
tion Frechet-dendron [G1]-Br
Scheme 3. Synthesis of the mixed dendrimer species 9.
as described for the preparation
Chem. Eur. J. 2001, 7, No. 1
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0701-0185 $ 17.50+.50/0
185

G. van Koten et al.
FULL PAPER
been quaternized with [G1]-Br Frechet dendrons. The 29Si{1H}
NMR spectrum recorded for 9 showed only two distinctive
broad peaks at d ˆ 0.77 and ꢀ2.07. The resonances corre-
sponding to the central silicon and those for the inner layer of
silicon nuclei, most probably coincide at d ˆ 0.77. The increase
in broadness of the Si resonances is indicative of a decrease in
molecular motion upon introduction of [G1] dendrons. The
structure of 9 was further supported by elemental analysis and
ES-MS. The ES mass spectrum recorded for 9 displayed
distinctive peaks at m/z 3171.4 (calcd for [9 ꢀ 4Br]4‡: 3171.7),
2521.5 (calcd for [9 ꢀ 5Br]5‡: 2521.3), and 2088.0 (calcd for
[9 ꢀ 6Br]6‡: 2087.8).
Â
H10
H1
NaSO3
UV/Vis spectroscopy
Na O3S
N
N
NMe2
MO
B
A
H2O
H2O
MO–
Br–
MO–
Br–
CH2Cl2
CH2Cl2
dendrimers 5-10
'mixing'
dendrimer with MO
incorporated
Preparation and analysis of host ± guest systems: We decided
to investigate the exchange of the bromide anions in 5 ± 9 for
anionic organic molecules. In a model study, the CH2Cl2-
insoluble dye methyl orange (abbreviated as MO) was
selected as a diagnostic probe for spectroscopic analysis.
The set-up for the anion exchange in the various polycationic
dendrimers is outlined in Figure 4. In a two-phase layer
system (Figure 4A), stoichiometric amounts of MO were
dissolved in water (orange color) along with one of the
polycationic dendrimers 5 ± 9 in CH2Cl2 (colorless). Thorough
mixing of the two layers gave rise to almost instantaneous
decolorization of the water layer and concomitant coloriza-
tion of the organic phase; this indicated the exchange of
bromide for MO anions. These two-phase mixtures can be
stored for months without any observable colorization of the
water layer. The precise mechanism of formation of these
dye ± dendrimer assemblies is not known. It has been postu-
lated that similar phase-transfer processes are accompanied
by diffusion of water into the dendrimer core.[3h] The dye ±
dendrimer assemblies were studied by UV/Vis spectroscopy
(residual amount of MO in the water layer), 1H NMR
spectroscopy, and MALDI-TOF mass spectrometry (CH2Cl2
layer). From the spectroscopic investigations, the number of
MO molecules present in the dendrimer could be quantified
(Table 2). The spectroscopic analyses pointed to a full
exchange of bromide for MO anions and, therefore, suggest
that all ammonium cations in the dendritic hosts are available
for the formation of dye ± dendrimer assemblies. It should be
noted that the use of both a stoichiometric amount and an
excess of MO in the two-phase system (Figure 4) leads to
quantitative exchange of bromide for MO anions in the
dendrimer species 5 ± 8. Another interesting aspect is
that the number of guests could easily be regulated and
monitored. The addition of less than the stoichiometric
1H NMR spectroscopy
MALDI-TOF-MS
Figure 4. Set-up for the assembly of inorganic/organic anions with the
polycationic dendrimer species 5 ± 10.
amount (1 ± 3 equiv) of MO with respect to the number of
cations present in dendrimer 5, led to the assembly of one,
two, or three MO anions with the cationic dendrimer,
respectively, as was easily recognized in the 1H NMR spectra
by the increase in intensity of new resonances at d ˆ 8.02, 7.78,
7.69, and 6.68 (all doublets, ArH) and 3.02 (singlet, NMe2
group) that were attributed to MO guest anions. The 1H NMR
spectra of these MO-containing species showed a stepwise
increase in Dd for the NMe2 groups of the cationic core unit in
5 upon binding more MO anions, which allowed the (further)
NMR spectroscopic assignment.
It should be mentioned that the assembly of MO with the
polycationic dendrimeric structures 5 ± 8 is fully reversible.
The addition of an excess of bromide anions to the water
phase completely decolorized the organic layer and, simulta-
neously, colorized the water phase. The dye anions could also
be liberated by lowering the pH of the mixture. This gave rise
to the formation of the corresponding acid of MO, which is
soluble in both the organic as well as the water phase.
The dye ± dendrimer assemblies 5 ´ 4MO, 6 ´ 4MO, 7 ´ 8MO,
8 ´ 8MO, and 9 ´ 24MO were isolated as bright orange solids,
which are soluble in CH2Cl2, THF, and benzene similar to the
bromide-containing, polycationic dendrimers.[14] The NMR
spectra of these MO-containing species proved to be very
illustrative. The resonances assigned to the dendrimer core
were broadened, while the signal belonging to the NMe2
grouping of MO was found as a relatively sharp resonance.
Additionally, for the dendrimer core unit typical upfield shifts
were found for the NMe2 grouping, while downfield shifts
were observed for the ArH of the core unit (Table 3 and
Figure 5, Supporting Information).[15]
Table 2. The number of MO molecules incorporated into the dendrimers
1
5 ± 9 and 10 determined by H NMR and UV/Vis spectroscopy.
Dendrimer
1H NMR
UV
The line-broadening in the 1H NMR spectra of 5 ± 8 and the
corresponding MO assemblies was further qualitatively
analyzed by performing spin-lattice (T1) and spin ± spin (T2)
relaxation measurements.[16] The T1 relaxation times for
several groups (especially the ArH and NMe2 groups of the
core unit) showed a steady increase with increasing molecular
weight, that is from 5 to 6 and from 7 to 8 (see also Supporting
5
6
7
8
9
4.1
4.0
8.2[a]
7.5[a]
23.7
4.0
4.0
3.5
8.0
8.0
n.d.[b]
n.d.[b]
10
[a] Integration less accurate because of line broadening. [b] Not deter-
mined.
186
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0701-0186 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 1

Polycationic Dendrimers
181±192
Table 3. Selected 1H NMR spectroscopic data for the core resonances of
the dendrimers 5 ± 10 and their MO-containing derivatives in CDCl3 at
RT.[a]
sponding to the NMe2 grouping of the core and the NMe2
grouping of the MO guest molecules. Nevertheless, an
increase in T1 and further decrease of T2 was clearly noted
when compared with their parent species. These data are in
line with earlier reports on relaxation time data measured for
dendrimer species.[3e, 17]
From the latter results, it seems reasonable to assume that
the MO anions are associated with the polycationic dendri-
mers. As a consequence, the dendrimer core unit will be more
shielded from the lattice (i.e., solvent) compared with the
dendrimers 5 ± 8, giving rise to the observed higher T1 and
lower T2 relaxation times. Molecular modeling[13] has been
Compound
Cationic sites
ArH
NMe2
‡
Si[C6H4(CH2NMe2 )2-4]4-core
5
4
4
7.77 (d)
±
2.98
2.83
2.97
2.88
[b]
5 ´ 4MO
6
6 ´ 4MO
7.76 (br)
[b]
±
‡
Si[C6H3(CH2NMe2 )2-3,5]4-core
7
8
8
8.41, 8.19
8.88, 8.75
8.35, 8.75 (br)
8.92, 8.72 (br)
3.07
2.78
3.00
2.81
7 ´ 8MO
8
8 ´ 8MO
Â
used to estimate the size of the Frechet dendrons [G1]-Br,
‡
[G1] [C6H3(CH2NMe2 )2-3,5]12-core[c]
ꢀ
[G2]-Br as well as of the MO anion itself. These calculated
data showed, in the case of putative close anion ± cation
interactions, that the MO anions are partially sticking out of
the dendrimers 5 and 7, while the MO anions in 6 and 8 could
be completely shielded by the aryl ether dendrons.
9
24
4
8.14 (br)[d]
8.10 (br)[d]
3.06
2.85
9 ´ 24MO
‡
Me2Si[C6H3(CH2NMe2 )2-3,5]2-core
10
8.35, 8.20
8.55, 8.37
3.02
2.92
10 ´ 4MO
MALDI-TOF-MS was carried out for 5 ´ 4MO to further
confirm the proposed composition of the host ± guest system.
The MS-spectrum of 5 ´ 4MO was complicated by the large
number of ion peaks. Nevertheless, all the main peaks could
be readily assigned, and characteristic peaks at m/z 2996.02
and 2692.85 were observed corresponding to the molecular
[a] All chemical shifts are listed in ppm and signals are singlet resonances
unless stated otherwise. Abbreviations used: d ˆ doublet, brˆ broad signal,
MO ˆ methyl orange. [b] The ArHcore of these species are coincident with
ꢀ
ꢀ
other ArH of the dendrimer. [c] [G1] ˆ Si[(CH2)3Si[(CH2)3 SiMe2 ]3]4.
[d] For a possible explanation see ref. [15].
‡
and fragment ions [M‡H] (calcd for: 2996.19) and [M ꢀ
‡
MO] (calcd for: 2691.84), respectively. The large number of
other peaks is most probably due to in situ fragmentation
processes (e.g., loss of multiple anions and the earlier
ꢀ
mentioned C N bond cleavage) and/or exchange with matrix
anions. One of the main fragmentation patterns indicated the
loss of an NMe2 fragment that most likely originated from the
complexed MO anions. For the larger composites (involving
the dendrimers 6 ± 8), however, a clear confirmation of the
proposed structures was more difficult due to an increased
fragmentation behavior and matrix ± MO anion exchange.
Accordingly, no direct molecular ion was observed for 6 ´
4MO. However, illustrative in this respect is the presence of
peaks at m/z 4239.3 (calcd for: 4239.4) and 4390.3 (calcd for:
4389.7), which were attributed to the fragment ions
‡
[M‡DHB ꢀ 2MO]
(DHB ˆ 3,5-dihydroxy-benzoic acid)
‡
and [M ꢀ MO] , respectively. The latter indirectly pointed
to the proposed host ± guest stoichiometry. Under similar
conditions, a fragment ion at m/z 5350.5 (calcd for: 5350.8)
was observed for 7 ´ 8MO (Figure 5A) which was ascribed to
‡
[M ꢀ MO] , while in the case of 8 ´ 8MO a peak at m/z 8746.4
‡
(assigned to [M ꢀ MO] ) could be clearly detected (Fig-
ure 5B).
Mode of interaction between host and guest: To confirm that
electrostatic interactions are the predominant attractive
forces in these structures, we have studied the phase transfer
of MO into CH2Cl2 by a sterically less congested model
compound 10 (Scheme 4). This model compound was pre-
pared analogously to 5 ± 9 and isolated as a white solid in
quantitative yield. Using the set-up depicted in Figure 4, this
tetracationic derivative 10 was also found to form assemblies
with MO anions. However, in contrast to the MO-assemblies
of the core ± shell dendrimeric species 5 ± 9, this particular
MO-containing compound (i.e., 10 ´ 4MO) is sparingly soluble
Figure 5. MALDI-TOF-MS spectra recorded for A) 7 ´ 8MO and B) 8 ´
8MO.
Information). On the other hand, the T2 relaxation times for
the same moieties decreased with higher molecular weights.
This has been followed by line shape analysis of the NMe2
grouping present in the cores of the molecules 5 ± 8. The
1H NMR spectra of the MO-containing dendrimer species
showed increased overlap between the resonances corre-
Chem. Eur. J. 2001, 7, No. 1
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0701-0187 $ 17.50+.50/0
187

G. van Koten et al.
FULL PAPER
Applications: We have investigated the potential use of the
polycationic dendrimers 7 and 8 in phase-transfer catalysis.
For this purpose, we selected the SN2 reaction between
(excess) potassium cyanide and benzyl bromide to afford
benzyl cyanide (Scheme 5) under biphasic conditions (i.e., a
CH2Cl2/H2O system). The formation of the nitrile product was
NMe2
NMe2
Me2N
Me
Si
Me
Me2N
[G1]-Br
CH2Cl2, RT
Br
CN
O
O
KCN
CH2Cl2/H2O
Me2
Me2
N
O
O
O
N
Me
Si
Br–
Br–
Br–
7 or 8
+
+
Br–
Scheme 5. SN2 reaction between benzyl bromide and KCN catalyzed by 7
and 8 in a biphasic experiment.
+
+
N
Me
N
O
Me2
Me2
monitored by 1H NMR spectroscopy. Gradually a singlet
resonance at d ˆ 3.75 appeared while the singlet resonance
corresponding to the CH2Br group of the starting material
decreased simultaneously. In the absence of 7 or 8 as a
catalyst, minor formation of benzyl cyanide was observed
after 20 h (ꢁ6%). However, in the presence of 7 and 8
(0.01 mol% catalytic ammonium groupings) a much higher
conversion of 40 and 59%, respectively, was noted after 20 h.
This clearly shows the suitability of these new polycationic
dendrimers as micro-reactor systems.[18]
Another interesting and rapid developing area in the field
of dendrimer chemistry is the use of membrane technology.
Membrane reactor systems in combination with catalytically
active, nanosized metallodendrimers have recently been used
to bridge the gap between homogeneous and heterogeneous
catalysis.[2d, e] These membrane experiments showed that
recycling of dendrimer catalysts is very promising. We there-
fore anticipated that the newly presented host ± guest systems,
in which MO guests are bonded through electrostatic forces,
would be retained by a suitable membrane filter.
To study this in more detail, a commercially available
dialysis membrane with a cut-off mass of 1000 gmolꢀ1 was
loaded with the strongly orange-colored dendrimer 7 ´ 8MO
(MW ˆ 5655 gmolꢀ1) and placed into a sealed system. After
72 h, no evidence for leaching of the dendrimer 7 ´ 8MO into
the bulk solution (CH2Cl2) could be obtained. Addition of an
excess of [Bu4N]PF6 led to a fast colorization of the bulk
CH2Cl2 solution (<1 h) and a slow, almost complete decol-
orization of the dialysis bag. The presence of the MO anion in
the bulk solution could be confirmed by lowering the pH,
which led to a fast color change from yellow/orange to purple/
red. After repeated dialysis with [Bu4N]PF6 dissolved in
CH2Cl2, the contents of the dialysis membrane were subjected
to NMR spectroscopic analysis (1H, 19F, and 31P). The
spectroscopic data were compared with the NMR spectro-
scopic data obtained for an authentic sample of 7 ´ 8PF6. This
comparison unambiguously revealed the presence of 7 ´ 8PF6
in the membrane vessel after this dialysis procedure and
therefore indicated that dendrimer 7 already has an appro-
priate size for retention in a membrane system.
O
O
10
Scheme 4. Synthesis of the less congested dendritic species 10.
in aromatic solvents. This indicates a less efficient shielding of
the ionic part by the Frechet dendrons in this molecule.
Â
1H NMR spectroscopic quantification by specific signal
integration (Table 2) suggested, just as in the case of 5 and 6,
the presence of four MO anions in 10 ´ 4MO. This provided
further evidence that the MO complexation by these cationic
dendrimers is primarily based on electrostatic interactions.
However, the possibility of p ± p stacking between the
aromatic parts in MO and the attached dendrons leading to
an even larger stabilization of the host ± guest system could
not be excluded. To investigate this in more detail, compound
4 was also used as a potential phase-transfer agent. This
derivative was also shown to be able to ªextractº MO from the
water layer and it seems therefore unlikely that p ± p
interactions play a pivotal role in the binding of MO anions
by the cationic dendrimers 5 ± 9.
Scope of the new host ± guest systems: In a subsequent study,
several other inorganic/organic substrates were tested as
possible guests for the dendritic hosts 5 ± 8 by using similar
procedures as described for the MO anion exchange proce-
dure. The sodium salt of pyrene-1-sulfonic acid (Figure 4) was
successfully extracted by the cationic dendrimer 5 from the
water phase. The presence of anionic pyrene guests in 5 was
1
evident from the recorded H NMR spectrum, in which two
distinctive doublets were found at d ˆ 9.41 and 8.78, which
were attributed to H1 and H10 (see Figure 4). Also in this
case, signal integration pointed to a full (i.e., four pyrene
guests) loading in 5. Interestingly, this material with anionic
pyrene guest showed a much lower solubility in CH2Cl2 when
compared to 5 ´ 4MO.
Likewise, the formation of other guest molecule assemblies
was also demonstrated by 1H NMR spectroscopy. These other
type of molecules include carboxylic acid derivatives such as
methyl red (i.e., 4-(4-dimethylamino-phenylazo)benzoic
acid), sulfates such as p-cresol sulfate, and chromate anions
(Cr2O72ꢀ). The preliminary findings indicate that the poly-
cationic dendrimer species 5 ± 8 and 9 can bind different kinds
of anionic inorganic/organic substrates.
Conclusion
In summary, a synthetic strategy for the formation of stable
assemblies based on a range of inorganic/organic substrates
188
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0701-0188 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 1

Polycationic Dendrimers
181±192
‡
1801 [M ꢀ I] ; elemental analysis calcd (%) for C56H100N8SiI8 (1928.8): C
and novel, cationic dendritic containers is presented. These
polycationic dendrimers are quantitatively prepared with
simple building blocks. The assembly of anionic, organic
molecules with these cationic dendrimers is fully reversible
and can be regulated by addition of other anions or lowering
the pH of the medium. Moreover, the molecular containers
can be fine-tuned with respect to the amount of potential
guests by a predesign of the arylamine silane core or by
changes in stoichiometry between host and guest. These
characteristics make this type of molecular container ex-
tremely versatile for a range of applications. For example,
these polycationic polymers have a great potential as dendri-
meric microreactors while applications in controlled substrate
delivery can be foreseen. In addition, the cavities in these
dendrimers could be of great use in shape-selective (or phase-
transfer) catalysis. Further studies are being carried out to
explore the use of these polycationic dendrimers in a number
of these applications as well as a more detailed investigation
of the location of the anionic guest molecules in these
polycationic dendrimers.
34.87, H 5.23, N 5.81, Si 1.46; found: C 34.89, H 5.26, N 5.74, Si 1.57.
Si(NCN)4 ´ 8EtBr (3): White solid (170 mg, >99%); m.p. 76 ± 778C;
1H NMR (200 MHz, D2O): d ˆ 8.04 (s, 8H, ArH), 7.88 (s, 4H, ArH), 4.61
(s, 16H, ArCH2N), 3.41 (q, 16H, J ˆ 7.0 Hz, CH2CH3), 3.03 (s, 48H,
N(CH3)2), 1.36 (t, 24H, J ˆ 7.0 Hz, CH2CH3); 13C{1H} NMR (50 MHz,
ꢀ
D2O): d ˆ 142.6, 139.5, 134.6, 128.7 (4  ArC), 66.0 (Ar CH2N), 60.2
‡
(CH2), 49.1 (N(CH3)2), 7.8 (CH3); MS (FAB): m/z: 1585.1 [M ꢀ Br] ;
elemental analysis calcd (%) for C64H116N8SiBr8 (1665.0): C 46.17, H 7.02, N
6.73, Si 1.69; found: C 46.28, H 7.08, N 6.73, Si 1.75.
Si(NCN)4 ´ 8octyl-I (4): Off-white solid (0.20 g, 73%); m.p. 75 ± 768C;
1H NMR (200 MHz, CDCl3): d ˆ 8.36 (m, 4H, ArH), 8.32 (m, 8H, ArH),
5.11 (brs, 16H, ArCH2N), 3.65 (m, 16H, CH2CH2N), 3.26 (brs, 48H,
N(CH3)2), 1.79 (m, 16H, CH2CH2N), 1.66 (m, 32H, -CH2-, octyl), 1.35 ± 1.25
(m, 48H, -CH2-, octyl), 0.88 (t, 24H, CH3, octyl); 13C{1H} NMR (50 MHz,
CDCl3): d ˆ 142.9, 140.8, 135.6, 128.9 (4  ArC), 66.4 (ArCH2N), 65.3
(CH2CH2N), 49.8 (N(CH3)2), 31.7, 29.2, 29.1, 26.3, 22.9, 22.6 (6 Â CH2,
‡
octyl), 14.1 (CH3, octyl); MS (FAB): m/z: 2587.0 [M‡H ꢀ I] ; elemental
analysis calcd (%) for C112H212N8SiI8 (2714.3): C 49.56, H 7.87, N 4.13;
found: C 49.83, H 7.78, N 4.87.
Si(CN)4 ´ 4[G1]-Br (5): A mixture of 1b (0.18 g, 0.31 mmol) and [G1]-Br
(0.48 g, 1.25 mmol) in CH2Cl2 (25 mL) was stirred for 3 d at room
temperature under nitrogen. The slightly yellow reaction mixture was then
filtered through a path of Celite, dried over MgSO4, and concentrated in
vacuo. The product was isolated in quantitative yield as a white solid.
1H NMR (300 MHz, CDCl3): d ˆ 7.77 (brd, 8H, 3J ˆ 8.4 Hz, ArH), 7.40 ±
7.10 (m, 48H, ArH), 6.73 (brs, 8H, ArH), 6.71 (brs, 4H, ArH), 5.37 (brs,
8H, CH2N), 5.05 (brs, 8H, OCH2), 4.80 (brs, 8H, NCH2), 2.98 (brs, 24H,
NMe2); 13C{1H} NMR (75 MHz, CDCl3): d ˆ 160.1, 136.8, 136.3, 135.1,
133.4, 130.0, 128.9, 128.7, 127.9, 112.6, 104.1 (ArC), 70.4 (broad, overlapping
CH2N), 67.5 (CH2O), 48.8 (N(CH3)2); MS (ES, THF): m/z: 969.3 [M ꢀ
2Br]2‡, 777.7 [M ꢀ [G1]-Brꢀ 2Br]2‡, 619.7 [M ꢀ 3Br]3‡, 492.0 [M ꢀ [G1]-
Experimental Section
General: All air-sensitive manipulations were carried out in an inert
atmosphere using standard Schlenk techniques. All solvents were carefully
dried and distilled prior to use. Standard chemicals were purchased from
Acros Chimica or Aldrich and used without further purification. The
compounds 1b,[7] [G1]-Br,[12a] and [G2]-Br[12a] were prepared according to
previously reported procedures. 1H, 13C{1H}, 31P, 19F, and 29Si{1H} NMR
spectroscopic measurements were carried out on a Varian Inova/Mercury
200 or 300 MHz spectrometer at 258C and chemical shifts (d) are given in
ppm with TMS or H3PO4 as external standards or relatively to the residual
solvent peak. The MALDI-TOF mass spectra were acquired using a
Voyager-DE BioSpectrometry Workstation mass spectrometer (PerSeptive
Biosystems Inc., Framingham, MA, USA). ES-MS spectra were obtained
from the Analytical Chemistry department, Utrecht University. Elemental
analyses were performed by Dornis und Kolbe, Mülheim/Ruhr, Germany.
Brꢀ 3Br]3‡
,
444.9 [M ꢀ 4Br]4‡
; elemental analysis calcd (%) for
C120H124Br4N4O8Si ´ 8H2O (2242.1): C 64.28, H 6.29, N 2.50; found: C
64.25, H 6.08, N 2.51.
Si(CN)4 ´ 4[G2]-Br (6): A mixture of 1b (79.3 mg, 0.14 mmol) and [G2]-Br
(0.44 g, 0.54 mmol) in freshly distilled CH2Cl2 (25 mL) was stirred for 24 h
at room temperature under nitrogen. The slightly yellow reaction mixture
was concentrated in vacuo, and the residue was washed with Et2O (5 Â
50 mL). The product was isolated as a white pellet by centrifugation,
dissolved in CH2Cl2, dried over MgSO4, filtered, and concentrated in vacuo
to give 6 as a white solid in quantitative yield. 1H NMR (300 MHz, CDCl3):
d ˆ 7.76 (m, 8H, ArH), 7.60 ± 6.90 (brm, 96H, ArH), 6.73, 6.65, 6.62, 6.49
(br m, 36H, ArH), 5.10 ± 4.60 (br signal, 64H, all CH2 groups), 2.97 (brs,
24H, N(CH3)2); 13C{1H} NMR (75 MHz, CDCl3): d ˆ 160.1, 138.7, 136.7,
133.3, 129.9, 128.8 128.5, 128.0, 127.6, 112.3, 106.6, 104.0, 101.6 (ArC), 70.0
(br, overlapping CH2N), 67.5 (br, overlapping CH2O), 48.7 (br, N(CH3)2);
Si(CN)4 (1b): tBuLi (15 mL, 1.5m in pentane, 22.5 mmol) was added at
ꢀ788C to
a solution of 1-bromo-4-[(dimethylamino)methyl]-benzene
(2.79 g, 13.0 mmol) dissolved in THF (40 mL). The resulting white
suspension was stirred for about 10 min and subsequently SiCl4 (0.3 mL,
2.6 mmol) was added. After the mixture reached RT, the suspension was
stirred for an additional period of 19 h. The reaction mixture was quenched
with an excess of H2O. The organic layer was seprated and the H2O layer
extracted with pentane (2 Â 40 mL). The solvent was removed under
MS (ES, THF): m/z: 1818.6 [M ꢀ 2Br]2‡, 1413.9 [M ꢀ [G2]-Brꢀ Br]2‡
,
;
1185.5 [M ꢀ 3Br]3‡
,
916.2 [M ꢀ [G2]-Brꢀ 2Br]2‡
,
869.3 [M ꢀ 4Br]4‡
elemental analysis calcd (%) for C232H220Br4N4O24Si (3796.0): C 73.41, H
5.84, N 1.48, Br 8.42; found: C 73.62, H 5.91, N 1.54, Br 8.34.
reduced pressure to give
a yellow viscous oil. To remove siloxane
impurities, the yellow oily product was filtered through a path of silica
with CH2Cl2/NEt3 120:1 v/v as eluent and recrystallized from pentane to
afford 1b as a white solid material (0.70 g, 1.24 mmol, 47%). 1H NMR
(200 MHz, C6D6): d ˆ 7.80 (d, 8H, 3J ˆ 7.6 Hz, ArH), 7.38 (d, 8H, 3J ˆ
7.6 Hz, ArH), 3.26 (s, 8H, CH2N), 2.07 (s, 24H, N(CH3)2); 13C{1H} NMR
(50 MHz, C6D6): d ˆ 141.5, 137.0, 133.6, 128.3 (ArC), 64.5 (CH2N), 46.6
Si(NCN)4 ´ 8[G1]-Br (7): A mixture of 1a (0.10 g, 0.13 mmol) and [G1]-Br
(0.45 g, 1.40 mmol) in freshly distilled CH2Cl2 (25 mL) was stirred for 96 h
at room temperature under nitrogen. The slightly yellow reaction mixture
was concentrated in vacuo, whereupon Et2O (15 mL) was added to the
residue. The resultant mixture was stirred for 24 h, and the crude product
was isolated by centrifugation. The product was washed with Et2O (15 mL)
to remove the excess of [G1]-Br and isolated in quantitative yield after
drying in vacuo. 1H NMR (300 MHz, CDCl3): d ˆ 8.41 (brs, 8H, ArH), 8.19
(brs, 4H, ArH), 7.29 ± 7.15 (m, 80H, ArH), 7.00 (brs, 16H, ArH), 6.57 (brs,
8H, ArH), 5.19 (brs, 48H, OCH2‡CH2N), 4.84 (brs, 16H, NCH2), 3.07
(brs, 48H, NMe2); 13C{1H} NMR (75 MHz, CDCl3): d ˆ 159.8, 143.3, 140.5,
136.3, 135.4, 129.3, 128.6, 128.4, 128.0, 127.7, 112.5, 104.2 (ArC), 70.2
(OCH2), 67.7, 66.0 (CH2NCH2), 48.8 (N(CH3)2); MS (ES, THF): m/z: 1849.8
‡
(N(CH3)2); MS (MALDI-TOF, DHB): m/z: 565.8 [M‡H] ; elemental
analysis calcd (%) for C36H48N4Si (565.9): C 76.54, H 8.57, N 9.92, Si 4.67;
found: C 76.39, H 8.67, N 9.85, Si 4.85.
Preparation of model compounds: General procedure: The polycationic
compounds 2 ± 4 were prepared by mixing 1a with an excess of MeI, EtBr,
or octyliodide in CH2Cl2 and the initially clear reaction mixtures were
stirred at room temperature (<24 h). After removal of all volatiles, the
products were obtained as white solids by washing with pentane and drying
in vacuo.
[M ꢀ 2Br]2‡
,
1206.5 [M ꢀ 3Br]3‡
,
1078.8 [M ꢀ [G1]-Brꢀ 3Br]3‡
, 884.8
[M ꢀ 4Br]4‡, 789.1 [M ꢀ [G1]-Brꢀ 4Br]4‡, 597.6 [M ꢀ 3[G1]-Brꢀ 4Br]4‡
;
elemental analysis calcd (%) for C216H228N8Br8O16Si (3859.5): C 67.22, H
5.95, N 2.90, Br 16.56; found: C 67.38, H 5.86, N 2.82, Br 16.68.
1
Si(NCN)4 ´ 8MeI (2): White solid (0.41 g, 85%); m.p. 73 ± 748C; H NMR
(200 MHz, D2O): d ˆ 8.19 (s, 8H, ArH), 7.88 (s, 4H, ArH), 4.62 (s, 16H,
CH2N), 3.16 (s, 72H, N(CH3)3); 13C{1H} NMR (75 MHz, D2O): d ˆ 145.3,
142.4, 137.9, 131.8 (4 Â ArC), 71.0 (CH2N), 55.7 (N(CH3)3); MS (FAB): m/z:
Si(NCN)4 ´ 8[G2]-Br (8): A mixture of 1a (58 mg, 73 mol) and [G2]-Br
(0.48 g, 0.59 mmol) in freshly distilled CH2Cl2 (25 mL) was stirred for 24 h
Chem. Eur. J. 2001, 7, No. 1
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0701-0189 $ 17.50+.50/0
189

G. van Koten et al.
FULL PAPER
at room temperature under nitrogen. The slightly yellow reaction mixture
was concentrated in vacuo, and the residue was washed with Et2O (5 Â
50 mL). The product was isolated as a white solid by centrifugation,
dissolved in CH2Cl2, dried over MgSO4, filtered, and concentrated in vacuo
For 5 ´ 4MO (Mw ˆ 2995.8): 1H NMR (300 MHz, CDCl3): d ˆ 8.02 (d, J ˆ
8.1 Hz, 8H, ArH(MO)), 7.78 (d, J ˆ 9.0 Hz, 8H, ArH(MO)), 7.69 (d, J ˆ
7.8 Hz, 8H, ArH(MO)), 7.26 ± 7.18 (m, 56H, ArH), 6.68 (d, J ˆ 9.3 Hz, 8H,
ArH(MO)), 5.20 (brs, 8H, ArCH2), 4.90 (s, 16H, ArOCH2), 4.63 (brs, 8H,
ArCH2), 3.02 (s, 24H, NMe2(MO)), 2.83 (brs, 24H, core NMe2); MS
1
to give 8 in quantitative yield. H NMR (300 MHz, CDCl3): d ˆ 8.35 (brs,
‡
‡
8H, ArH), 7.89 (brs, 4H, ArH), 7.27 ± 6.98 (brm, 160H, ArH), 6.58, 6.54,
6.41 (br overlapping signals, 72H, ArH), 4.95 ± 4.50 (m, 128H, all CH2
groups), 3.00 (brs, 48H, N(CH3)2); 13C{1H} NMR (75 MHz, CDCl3): d ˆ
160.1, 159.9, 139.0, 136.8, 129.4, 128.6, 128.0, 127.8, 112.5, 106.7, 104.5, 101.6
(ArC), 70.0 (br, all CH2O), 67.6 (br, all CH2N), 48.9 (br, N(CH3)2); MS (ES,
(MALDI-TOF, DHB): m/z: 3298.8 [M‡2DHB] , 3147.8 [M‡DHB] ,
‡
‡
‡
2996.0 [M‡H] , 2800.2 [M ꢀ MO ꢀ NMe2‡DHB] , 2692.9 [M ꢀ MO] ,
‡
‡
2648.7 [M ꢀ MO ꢀ NMe2] , 2496.6 [M ꢀ 2MO ꢀ NMe2‡DHB] , 2345.7
‡
‡
[M ꢀ 2MO ꢀ NMe2] , 2194.6 [M ꢀ 3MO ꢀ NMe2‡DHB] , 2150.6 [M ꢀ
‡
3MO ꢀ 2NMe2‡DHB] .
THF): m/z: 2338.7 [M ꢀ 3Br]3‡
;
elemental analysis calcd (%) for
Dendrimer 6 ´ 4MO (Mw ˆ 4693.8): 1H NMR (300 MHz, CDCl3, 608C): d ˆ
8.02 (brd, J ˆ 8.1 Hz, 8H, ArH(MO)), 7.71 (brd, J ˆ 9.0 Hz, 8H, ArH-
(MO)), 7.65 (brd, J ˆ 8.7 Hz, 8H, ArH(MO)), 7.31 ± 7.19 (m, 96H, ArH),
6.71 (brs, 8H, ArH), 6.61 ± 6.58 (br, 24H, overlapping ArH(MO), ArH),
6.46 (brs, 12H, ArH), 4.90, 4.83, 4.66 (brs, 64H, all ArOCH2, ArCH2), 2.95
(s, 24H, NMe2(MO)), 2.88 (brs, 24H, core NMe2); MS (MALDI-TOF,
C
440H420N8Br8O48Si (7255.5): C 72.84, H 5.83, N 1.54; found: C 72.68, H
5.96, N 1.53.
ꢀ
ꢀ
[G1]-SiMe2 NCN ´ 24[G1]-Br (9): A mixture of [G1]-SiMe2 NCN (0.19 g,
0.0499 mmol) and [G1]-Br (0.45 g, 1.17 mmol) in CH2Cl2 (30 mL) was
stirred for 20 h. Subsequently the solvent was removed in vacuo and the
remaining solid washed with Et2O (50 mL) to yield 9 as an off-white solid
(0.64 g, 0.0492 mmol, 99%). 1H NMR (200 MHz, CDCl3): d ˆ 8.14 (m,
36H, ArH), 7.26 (brm, ArH), 6.90 (s, 48H, ArH), 6.60 (s, 24H, ArH), 5.14
(brs, 48H, CH2N), 5.00 (brs, 96H, CH2O), 4.50 (brs, 48H, NCH2), 3.06
(brs, 144H, N(CH3)2), 1.20, 0.84, 0.49 (3 Â m, 96H, SiCH2CH2CH2Si), 0.26
(brs, 72H, Si(CH3)2); 13C{1H} NMR (75 MHz, CDCl3): d ˆ 159.95, 149.87,
143.17, 140.90, 137.78, 136.22, 129.17, 128.52, 128.05, 127.73, 112.43, 104.05
(12 Â ArC), 70.21 (CH2O), 67.65, 66.99 (2 Â CH2N), 48.86 (N(CH3)2), 20.27,
18.65, 17.41 (SiCH2CH2CH2Si, overlapping signals), ꢀ2.55 (Si(CH3)2);
29Si{1H} NMR (60 MHz, CDCl3): d ˆ 0.77 (br signal, Sicore‡Siinner), ꢀ2.07
(br, SiMe2Ph); MS (ES, CH3CN): m/z: 3171.4 [M ꢀ 4Br]4‡, 3075.9 [M ꢀ
‡
‡
DHB): m/z: 4239.3 [M‡DHB ꢀ 2MO] , 4390.3 [M ꢀ MO]
Dendrimer 7 ´ 8MO (Mw ˆ 5655.1): 1H NMR (200 MHz, CDCl3): d ˆ 8.88
(brs, 8H, ArHcore), 8.75 (brs, 4H, ArHcore), 7.89 (d, J ˆ 8.4 Hz, 16H,
ArH(MO)), 7.79 (d, J ˆ 8.8 Hz, 16H, ArH(MO)), 7.69 (d, J ˆ 8.6 Hz, 16H,
ArH(MO)), 7.18 (m, 80H, ArH), 6.69 (d, J ˆ 9.2 Hz, 16H, ArH(MO)), 6.67
(brs, 16H, ArH), 6.52 (brs, 8H, ArH), 4.80 (m, 48H, overlapping ArCH2
and ArOCH2), 4.54 (brs, 16H, ArCH2), 3.04 (s, 48H, NMe2(MO)), 2.78
(brs, 48H, core NMe2); MS (MALDI-TOF, DHB): m/z: 5350.5 [M ꢀ
‡
MO] .
Dendrimer 8 ´ 8MO (Mw ˆ 9051.1): 1H NMR (200 MHz, CDCl3): d ˆ 8.92
(brs, 8H, ArHcore), 8.72 (brs, 4H, ArHcore), 7.87 (brd, J not resolved, 16H,
ArH(MO)), 7.71 (brd, J ˆ 8.1 Hz, 16H, ArH(MO)), 7.66 (brd, J not
resolved, 16H, ArH(MO)), 7.23 (brm, 160H, ArH), 6.76, 6.47, 6.38 (m,
88H, ArH‡ArH(MO)), 4.76, 4.70 (m, 128H, all ArOCH2‡ArCH2), 2.93
(brs, 48H, NMe2(MO)), 2.81 (brs, 48H, core NMe2); MS (MALDI-TOF,
[G1]-Brꢀ 4Br]4‡
,
2521.5 [M ꢀ 5Br]5‡
,
2444.9 [M ꢀ [G1]-Brꢀ 5Br]5‡
,
2088.0 [M ꢀ 6Br]6‡, 2023.93 [M ꢀ [G1]-Brꢀ 6Br]6‡; elemental analysis
calcd (%) for C720H852N24Si17O48Br24 (13006): C 66.49, H 6.60, N 2.58, Br
14.75; found: C 67.06, H 6.39, N 2.44, Br 14.28.
‡
‡
DHB): m/z: 8746.4 [M ꢀ MO] , 8443.1 [M ꢀ 2MO‡H] .
Me2Si(NCN)2 ´ 4[G1]-Br (10): This compound was prepared in a similar
way as described for 5 ± 8. A mixture of Me2Si(NCN)2[19] (0.21 g, 0.48 mmol)
and [G1]-Br (0.70 g, 1.97 mmol) in CH2Cl2 (25 mL) was stirred for 4 h. The
solvent was removed under reduced pressure and the solid residue was
washed with Et2O to give 10 as a white solid (0.75 g, 0.40 mmol, 84%).
1H NMR (200 MHz, CDCl3): d ˆ 8.35 (s, 2H, ArH), 8.20 (s, 4H, ArH),
7.34 ± 7.24 (m, 40H, ArH), 6.91 (s, 6H, ArH), 6.64 (s, 3H, ArH), 5.00 (s,
16H, CH2O), 4.77 (s, 16H, overlapping CH2N), 3.02 (s, 24H, N(CH3)2), 0.63
(s, 6H, Si(CH3)2); 13C{1H} NMR (50 MHz, CDCl3): d ˆ 160.2, 141.8, 141.0,
136.5, 129.2, 128.9, 128.6, 128.4, 128.2, 127.8, 112.8, 112.5, 104.7 (ArC), 70.5
(CH2O), 67.6 (br, overlapping CH2N), 49.2 (br, NCH3)2), ꢀ2.0 (br,
Si(CH3)2); elemental analysis calcd (%) for C102H120N4SiO8Br4 (1973.9): C
65.24, H 6.44, N 2.98; found: C 65.38, H 6.38, N 2.90.
1
Dendrimer 9 ´ 24MO (Mw ˆ 18393): H NMR (200 MHz, CDCl3): d ˆ 8.10
(m, 84H, ArHcore), 7.92 (brd, J ˆ 6.8 Hz, 48H, ArH(MO)), 7.73 (brd, J ˆ
8.8 Hz, 48H, ArH(MO)), 7.65 (brd, J ˆ 8.2 Hz, 48H, ArH(MO)), 7.18 (m,
240H, ArH), 6.72 (brs, 48H, ArH), 6.59 (brd, J ˆ 8.4 Hz, 48H, ArH(MO)),
6.49 (brs, 24H, ArH), 4.81 (m, 192H, all ArOCH2‡ArCH2), 2.95 (s, 144H,
NMe2(MO)), 2.85 (brs, 144H, core NMe2), 1.25 (brm, SiCH2CH2CH2Si),
0.88 (brm, SiCH2CH2CH2Si), 0.68 (brm, SiCH2CH2CH2Si), 0.26 (brs, 72H,
carbosilane dendrimer CH2); MS (MALDI-TOF, DHB): m/z: ꢂ10000 ±
15000 (broad peak).
Dendrimer 10 ´ 4MO (Mw ˆ 2871.7): 1H NMR (200 MHz, CDCl3): d ˆ 8.55
(s, 2H, ArHcore), 8.37 (s, 4H, ArHcore), 7.98 (d, 8H, J ˆ 9.2 Hz, ArH(MO)),
7.85 (d, 8H, J ˆ 8.6 Hz, ArH(MO)), 7.76 (d, 8H, J ˆ 8.0 Hz, ArH(MO)),
7.32 ± 7.25 (m, 40H, ArH), 6.75 ± 6.63 (m, 20H, ArH‡ArH(MO)), 6.64 (s,
3H, ArH), 4.95 (m, 24H, CH2O‡CH2N), 4.53 (s, 8H, CH2N), 3.09 (s, 24H,
NMe2(MO)), 2.92 (brs, 24H, core NMe2), 0.08 (s, 6H, Si(CH3)2); 13C{1H}
NMR (50 MHz, CDCl3): d ˆ 160.19, 152.79, 147.13, 143.60, 141.26, 139.44,
136.46, 129.01, 128.71, 128.17, 127.90, 126.88, 125.57, 122.20, 112.29, 111.77,
104.75 (17 ArC), 70.37 (CH2O), 68.50 (br, overlapping CH2N), 49.08 (br,
NCH3)2), 40.61 (N(CH3)2, MO), ꢀ3.23 (br, Si(CH3)2); MS (MALDI-TOF,
Preparation and characterization of MO-dendrimer assemblies: All
solutions were prepared at room temperature in the presence of air, using
freshly distilled CH2Cl2 and de-ionized water. Two stock solutions were
prepared: A solution of one of the dendrimer species in dichloromethane
(3 Â 10ꢀ4 ± 8 Â ꢀ4 m) and a solution of methyl orange in water. In a typical
procedure, five two-phase systems were prepared in such manner that the
ratio between bromide anions in the dendrimer and methyl orange anions
ranged from 1:4 to 5:4 (in the case of 5 and 6) or from 2:8 to 10:8 (in the
case of 7 and 8). The two-phase systems were mixed for 20 h and then the
layers were separated for analysis. The water layers were diluted to such an
extend that the ultimate amount of methyl orange still present in the water
layers could be determined with UV/Vis absorption spectroscopy using a
calibration curve derived from a series of samples with known concen-
trations of methyl orange in water. The UV/Vis experiments evidenced that
the dendrimer systems are only able to exchange either a maximum of 4 or
8 MO anions, whereas an excess of MO remains in the H2O layer of the
two-phase system. The dichloromethane layers were washed with water,
dried over Na2SO4, and concentrated in vacuo. The isolated orange solids
were analyzed with 1H NMR spectroscopy in CDCl3 and with MALDI-
TOF-MS. The peak assignments were based on the appearance of new
signals in the 1H NMR spectra, which were attributed to the presence of
MO anions and on a stepwise increase in Dd for several groups of the
dendrimers upon assembly with an increasing number of MO anions. From
the obtained data, the maximum amount of MO in the (dendrimeric)
assemblies (i.e., 5 ± 10) was calculated and these results are summarized in
Table 2 (see main text).
‡
‡
DHB): m/z: 2566.1 [M ꢀ MO] , 1958.7 [M ꢀ 3MO] .
Phase-transfer catalysis: A typical experiment is as follows: A mixture of
KCN (0.28 g, 430 m mol) in H2O (5 mL), benzyl bromide (0.19 g,
111 mmol) in CH2Cl2 (5 mL) and
8 (10.6 mg, 0.012 mmol catalytic
ammonium groups) was stirred vigorously at ambient temperature.
1
Samples for H NMR spectroscopic analysis were taken after 3 and 20 h.
Conversions were determined by signal integration and compared with the
noncatalyzed run.
Membrane experiment: A commercially available dialysis membrane with
a cut-off mass of 1000 gmolꢀ1 (SIGMA) was loaded with a solution of
7 ´ 8MO (7.0 mg, 1.24 mmol) in CH2Cl2 (ca. 10 mL). The contents of the
dialysis tubing were placed into a sealed system containing CH2Cl2
(300 mL) and H2O (0.5 mL). After 3 d of gently stirring and replacing
the CH2Cl2/H2O mixture twice, an excess of NBu4PF6 (4.8 g) was added to
the bulk solution. The dialysis tubing slowly started to decolorize with a
simultaneous colorization of the CH2Cl2 bulk solution. The bulk solution
was replaced two times within 3 d with a dialysis solution of NBu4PF6 (4.2 ±
4.5 g) in CH2Cl2. Then the dialysis bag was thoroughly washed with CH2Cl2
to remove excess of NBu4PF6 still present in the tubing and subsequently
190
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0701-0190 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 1

Polycationic Dendrimers
181±192
disconnected from the set-up. After isolation of the contents by extraction
with CH2Cl2 and drying in vacuo, an off-white solid was obtained which was
subjected to NMR spectroscopic analysis. The whole procedure was carried
out twice and confirmed the reproducibility of the membrane experiment.
The NMR spectroscopic data was compared with an authenthic sample of
7 ´ 8PF6, which was synthesized by a similar procedure as described for the
MO-containing dendrimers using the two-phase system depicted in Fig-
ure 4 loaded with 7 (in CH2Cl2) and an excess of sodium hexafluorophos-
phate (in H2O). Spectroscopic data for 7 ´ 8PF6: 1H NMR (300 MHz,
CDCl3): d ˆ 8.01 (brs, 8H, ArH), 7.86 (s, 4H, ArH), 7.33 ± 7.25 (brm, 40H,
ArH), 6.68 (s, 8H, ArH), 6.47 (brs, 16H, ArH), 5.01 (brs, 32H, CH2O), 4.06
(brs, 16H, CH2NCH2), 3.77 (brs, 16H, CH2NCH2), 2.62 ± 2.47 (brs, 48H,
membrane reactors to facilitate catalyst recovery has been demon-
strated. See: d) N. J. Hovestad, E. B. Eggeling, H. J. Heidbüchel,
J. T. B. H. Jastrzebski, U. Kragl, W. Keim, D. Vogt, G. van Koten,
Angew. Chem. 1999, 111, 1763 ± 1765; Angew. Chem. Int. Ed. 1999, 38,
1655 ± 1658; e) N. Brinkmann, D. Giebel, G. Lohmer, M. T. Reetz, U.
Kragl, J. Catal. 1999, 183, 163 ± 168.
[3] For relevant examples of dendrimers acting as (inverted) micelles and/
or molecular containers, see: a) G. Caminati, N. J. Turro, D. A.
Tomalia, J. Am. Chem. Soc. 1990, 112, 8515 ± 8522; b) K. R. Gopidas,
A. R. Leheny, G. Caminati, N. J. Turro, D. A. Tomalia, J. Am. Chem.
Soc. 1991, 113, 7335 ± 7342; c) G. Newkome, C. N. Moorefield, G. R.
Baker, M. J. Saunders, S. H. Grossman, Angew. Chem. 1991, 103,
1207 ± 1209; Angew. Chem. Int. Ed. Engl. 1991, 30, 1178 ± 1180; d) C. J.
1
N(CH3)2); 19F NMR (282 MHz, CDCl3): d ˆ ꢀ70.30 (d, J(F,P) ˆ 714 Hz);
1
31P NMR (81 MHz, CDCl3): d ˆ ꢀ143.14 (m, J(F,P) ˆ 714 Hz).
Hawker, K. L. Wooley, J. M. J. Frechet, J. Chem. Soc. Perkin Trans. 1
Â
1993, 1287 ± 1297; e) J. F. G. A. Jansen, E. M. M. de Brabander-van
den Berg, E. W. Meijer, Science 1994, 266, 1226 ± 1229; f) J. F. G. A.
Jansen, E. M. M. de Brabander-van den Berg, E. W. Meijer, J. Am.
Chem. Soc. 1995, 117, 4417 ± 4418; g) S. Stevelmans, J. C. M. van Hest,
J. F. G. A. Jansen, D. A. F. J. van Boxtel, E. M. M. de Brabander-
van den Berg, E. W. Meijer, J. Am. Chem. Soc. 1996, 118, 7398 ±
7399; h) A. I. Cooper, J. D. Londono, G. Wignall, J. B. McClain,
E. T. Samulski, J. S. Lin, A. Dobrynin, M. Rubinstein, A. L. C. Burke,
Crystal data and data collection
Crystal data for 3: C56H100N8SiI8‡solvent; colorless block-shaped crystal
(0.40  0.25  0.20 mm), Mw ˆ 1928.8 (excluding solvent), triclinic, space
Å
group P1 (no. 2), with a ˆ 13.803(4), b ˆ 15.634(2), c ˆ 22.290(6) Š, a ˆ
98.89(2)8, b ˆ 93.89(2)8, g ˆ 103.16(2)8, Vˆ 4600.7(19) Š3, Z ˆ 2, 1calcd
ꢂ1.51 gcmꢀ3, m(MoKa) ˆ 2.6 mmꢀ1. X-ray data were collected on a Nonius
CAD4T diffractometer (rotating anode, MoKa radiation, Tˆ 150 K, w scan,
qmax ˆ 27.58). Data were corrected for absorption using psi-scan data. The
structure was solved by Direct Methods (SHELXS97) and refined on F2
(SHELXL97). The cation could be refined with anisotropic thermal
parameters for the non-hydrogen atoms. Hydrogen atoms were taken into
account at calculated positions. Several of the iodine anions were found to
be disordered over several sites. In addition, the crystal structure contains
several solvent molecules of crystallisation. Due to their positional disorder
their nature could not be identified satisfactorly. A disorder model with
partially occupied sites was refined to take the scattering matter into
account in the structure factor calculations. Convergence was reached at
R ˆ 0.11 for 7799 reflections with I > 2s(I) and 730 parameters (wR2 ˆ
0.36). A residual density map showed no features apart from absorption
artifacts near iodine.
Â
J. M. J. Frechet, J. M. DeSimone, Nature 1997, 389, 368 ± 371;
i) M. W. P. L. Baars, P. E. Froehling, E. W. Meijer, Chem. Commun.
1997, 1959 ± 1960; j) V. Chechik, M. Zhao, R. M. Crooks, J. Am. Chem.
Soc. 1999, 121, 4910 ± 4911; k) C. Loup, M. A. Zanta, A. M. Caminade,
J. P. Majoral, B. Meunier, Chem. Eur. J. 1999, 5, 3644 ± 3650; l) H.
Stephan, H. Spies, B. Johannsen, L. Klein, F. Vögtle, Chem. Commun.
1999, 1875 ± 1876.
[4] G. Pistolis, A. Malliaris, D. Tsiourvas, C. Paleos, Chem. Eur. J. 1999, 5,
1440 ± 1444.
[5] For examples see: a) G. R. Newkome, B. D. Woosley, E. He, C. N.
Moorefield, R. Güther, G. R. Baker, G. H. Escamilla, J. Merrill, H.
Luftmann, Chem. Commun. 1996, 2737 ± 2738; b) G. R. Newkome,
C. N. Moorefield, G. R. Baker, US Pat. 5650101, 1999; [Chem. Abstr.
1999, 69896]; c) G. R. Newkome, Z. Yao, G. R. Baker, V. K. Gupta, J.
Crystallographic data (excluding structure factors) for the structure of 2
reported in this paper have been deposited with the Cambridge Crystallo-
graphic Data Centre as supplementary publication no. CCDC-102002.
Copies of the data can be obtained free of charge on application to CCDC,
12 Union Road, Cambridge CB21EZ, UK (fax: (‡44)1223-336-033;
e-mail: deposit@ccdc.cam.ac.uk).
Org. Chem. 1985, 50, 2003 ± 2004. For
a relevant example of
polycationic dendrimers see: d) C. Larr, B. Donnadieu, A. M.
Caminade, J. P. Majoral, J. Am. Chem. Soc. 1998, 120, 4029 ± 4030.
[6] For a review on host ± guest complexation including electrostatic
forces see: H. J. Schneider, Angew. Chem. 1991, 103, 1419 ± 1439;
Angew. Chem. Int. Ed. Engl. 1991, 30, 1417 ± 1436.
[7] A. W. Kleij, H. Klein, J. T. B. H. Jastrzebski, W. J. J. Smeets, A. L.
Spek, G. van Koten, Organometallics 1999, 18, 268 ± 276.
[8] P. R. Ashton, K. Shibata, A. N. Shipway, J. F. Stoddart, Angew. Chem.
1997, 109, 2902 ± 2904; Angew. Chem. Int. Ed. Engl. 1997, 36, 2781 ±
2783.
[9] a) Z. Wu, K. Biemann, Int. J. Mass Spectrom. Ion Processes 1997, 165/
166, 349 ± 361. For the synthesis of the studied dendritic cations/anions
see: b) S. W. Krska, D. Seyferth, J. Am. Chem. Soc. 1998, 120, 3604 ±
3612.
[10] See for instance: J. March, Advanced Organic Chemistry, 4th ed.,
Wiley, New York, 1992, pp. 1015 ± 1017, and references therein.
[11] For molecules containg more than two triorganoammonium iodide
groupings see for instance: a) F. P. Schmidtchen, G. Müller, J. Chem.
Soc. Chem. Commun. 1984, 115 ± 166; b) H. W. Schmalle, K. He-
getschweiler, Acta Crystallogr. 1996, C52, 1288 ± 1293.
Acknowledgement
A.W.K. and A.L.S. thank the Council for Chemical Sciences (CW) from the
Dutch Organization for Scientific Research (NWO) for financial support.
Cees Versluis from the Department of Analytical Chemistry of the
University of Utrecht is gratefully acknowledged for the electrospray mass
spectrometric studies. We thank Prof. Dr. Ashok Kakkar for stimulating
suggestions during the preparation of this manuscript.
[1] For recent reviews on dendritic molecules see: a) M. A. Hearshaw,
J. R. Moss, Chem. Commun. 1999, 1 ± 8; b) A. W. Bosman, H. M.
Janssen, E. W. Meijer, Chem. Rev. 1999, 99, 1665 ± 1688; c) G. New-
kome, E. He, C. N. Moorefield, Chem. Rev. 1999, 99, 1689 ± 1746;
d) J. P. Majoral, A. M. Caminade, Chem. Rev. 1999, 99, 845 ± 880;
e) M. Fischer, F. Vögtle, Angew. Chem. 1999, 111, 934 ± 955; Angew.
Â
[12] For the synthesis of Frechet type aryl ether dendrons see for example:
Â
a) C. J. Hawker, J. M. J. Frechet, J. Am. Chem. Soc. 1990, 112, 7638 ±
7647. For an analogous synthetic approach see: b) B. Hong, T. P. S.
Thoms, H. J. Murfee, M. J. Lebrun, Inorg. Chem. 1997, 36, 6146 ± 6147.
[13] CAChe Molecular Modeling package, Molecular Mechanics at the
MM2 parameter level, Oxford Molecular Group, 1998.
[14] The solubility features of the dye ± dendrimer assemblies point to a
clear interaction between the MO anions and the polycationic
dendrimers, particularly because of the relatively low ion-solvating
efficiencies of the solvents CH2Cl2 and benzene with respect to polar
solvents such as DMSO and H2O.
Â
Chem. Int. Ed. 1999, 38, 884 ± 905; f) J. M. J. Frechet, C. J. Hawker,
I. Gitsov, J. W. Leon, J. Macromol. Sci. Pure Appl. Chem. 1996, A33,
1399 ± 1425; g) D. A. Tomalia, H. D. Durst, Top. Curr. Chem. 1993,
165, 193 ± 313.
[2] For examples of dendrimers with catalytic end groups see: a) J. W. J.
Knapen, A. W. van der Made, J. C. De Wilde, P. W. M. N. van Leeu-
wen, P. Wijkens, D. M. Grove, G. van Koten, Nature 1994, 372, 659 ±
663; b) M. T. Reetz, G. Lohmer, R. Schwikardi, Angew. Chem. 1997,
109, 1559 ± 1562; Angew. Chem. Int. Ed. Engl. 1997, 36, 1526 ± 1529;
c) A. W. Kleij, R. A. Gossage, J. T. B. H. Jastrzebski, J. Boersma, G.
van Koten, Angew. Chem. 2000, 112, 182 ± 185; Angew. Chem. Int. Ed.
2000, 39, 179 ± 181. Recently, the use of these catalytic materials in
[15] Note that different cores are present in the dendrimer molecules 5 ±
10. This difference most probably has an influence on the mobility of
the MO anions in the corresponding dye ± dendrimers assemblies and,
1
consequently, on their H NMR spectroscopic features. Interestingly,
Chem. Eur. J. 2001, 7, No. 1
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0701-0191 $ 17.50+.50/0
191

G. van Koten et al.
FULL PAPER
for the cationic model derivative [(C6H5CH2)2NMe2]Br, that does not
contain internal cavities, a similar Dd (namely from d ˆ 3.12 to 3.02)
for the NMe2 group was noted as compared with dendrimers 5 and 6
upon exchange of the bromide for a MO anion. (A. W. Kleij, R.
van de Coevering, R. J. M. Klein Gebbink, G. van Koten, unpublished
results). This latter result suggests a similar proximity (ion pairing)
between the MO anion and the cationic NMe2 groups in these species.
[16] Samples of the dendrimer species 5 ± 8 and their MO-containing
derivatives were prepared in an oxygen-free atmosphere with a
concentration in the range of 1 ± 5 mm in CDCl3. T1 and T2 measure-
ments were carried out on a Varian Inova 300 MHz NMR spectrom-
eter using the implemented dot1 and Carr± Purcell ± Meiboom ± Gill
T2 experiments.
[17] a) A. D. Meltzer, D. A. Tirrell, A. A. Jones, P. T. Inglefield, D. M.
Hedstrand, D. A. Tomalia, J. Am. Chem. Soc. 1992, 25, 4541 ± 4548;
b) A. M. Naylor, W. A. Goddard III, G. E. Kiefer, D. A. Tomalia, J.
Am. Chem. Soc. 1989, 111, 2339 ± 2341.
Â
[18] M. E. Piotti, F. Rivera, R. Bond, C. J. Hawker, J. M. J. Frechet, J. Am.
Chem. Soc. 1999, 121, 9471 ± 9472.
[19] The synthesis and full characterization of this compound will be
reported elsewhere.
Received: May 2, 2000 [F2452]
192
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0701-0192 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 1