Dendrimer disassembly in the gas phase: A cascade fragmentation reaction of frechet dendrons

Article abstract of DOI:10.1002/chem.200900403

The mass spectrometry characterization of Frechet-type dendrons is reported. In order to provide the charges necessary for electrospray ionization, dendrons bearing an OH group at the focal point can be deprotonated and observed in the negative ion mode.

Full text of DOI:10.1002/chem.200900403

FULL PAPER
DOI: 10.1002/chem.200900403
Dendrimer Disassembly in the Gas Phase: A Cascade Fragmentation
Reaction of Frꢀchet Dendrons
Bilge Baytekin,[a] H. Tarik Baytekin,[a] Uwe Hahn,[b] Werner Reckien,[c]
Barbara Kirchner,[c] and Christoph A. Schalley*[a]
Abstract: The mass spectrometric char-
acterization of Frꢀchet-type dendrons
is reported. In order to provide the
charges necessary for electrospray ioni-
zation, dendrons bearing an OH group
at the focal point can be deprotonated
and observed in the negative ion mode.
Alternatively, the corresponding bro-
mides can be converted to quaternary
ammonium ions that can easily be de-
tected in the positive mode. If the
latter ions are subjected to collision-in-
duced dissociation experiments, a frag-
mentation cascade begins with the dis-
sociation of the focal amine. The focal
benzyl cation quickly decomposes in a
fragmentation cascade from the focal
point to the periphery until the periph-
eral benzyl (or naphthylmethyl) cations
are formed. Five different mechanisms
are discussed in detail, three of which
can be excluded based on experimental
evidence. The cascade fragmentation is
reminiscent of self-immolative den-
drimers.
Keywords: dendrimers · fragmenta-
tion mechanisms
·
gas-phase
reactions · mass spectrometry
Introduction
From the periphery, the drug is released at high concentra-
tions because of the many branches to which the prodrug is
attached. This strategy is thus a way to amplify the trigger
signal into the liberation of a multitude of drug molecules.
Dendrimer research has been accelerated significantly by
the development of analytical methods that permitted to
characterize their structures in detail, analyze their purity,
and characterize their properties as a function of increasing
generation number. Among these methods[6] are nuclear
magnetic resonance (NMR), size-exclusion chromatography
(SEC), high-performance liquid chromatography (HPLC)
and, more recently[7] scanning probe methods (SPM). Since
the development of the so-called soft ionization methods,
mass spectrometry has proven to be a very valuable tool,[8]
because it provides detailed information on defects and im-
purities.[9] Dendrimers have thus been ionized by fast atom
bombardment (FAB-MS),[10] matrix-assisted laser desorp-
tion/ionization (MALDI),[11] and electrospray ionization
(ESI).[12] The two latter techniques involve very gentle ioni-
zation that minimizes the fragmentation of the analyte mol-
ecules, although there are examples for artifacts formed
during electrospray ionization[13] and through the destruction
of dendrons and dendrimers during the MALDI process.[14]
The utility of mass spectrometry in dendrimer chemistry
is not limited to the determination of molecular masses of
dendrimers and their purity. Their chemistry in the gas
phase is an interesting and novel area of research.[15] Mass
spectrometry may provide even more information, for ex-
Dendrons and dendrimers[1] are highly branched, ideally
monodisperse and regularly shaped macromolecules with a
significant impact on biomedical[2] and materials sciences.[3]
They are particularly interesting with respect to their nano-
sized structures and their Aufbau principle in generations.[4]
Recently, the idea of self-immolative dendrimers has been
developed, which is particularly promising for the applica-
tion of dendrimers in medicine. Self-immolative dendri-
ACHTUNGTRENNUNG
mers[5] carry multiple copies of a prodrug at their periphery.
At the focal point they bear a disease-specific trigger. Upon
arrival at the location of the disease in the body, a specific
signal activates the trigger, which leads to complete frag-
mentation of the dendrimer scaffold into small subunits.
[a] Dr. B. Baytekin, Dr. H. T. Baytekin, Prof.Dr. C. A. Schalley
Institut fꢁr Chemie und Biochemie, Freie Universitꢂt
Takustr. 3, 14195 Berlin (Germany)
Fax : (+49)30-838-55817
E-mail: schalley@chemie.fu-berlin.de
[b] Dr. U. Hahn
Departamento de Quꢃmica Orgꢄnica
Universidad Autꢅnoma de Madrid, 28049 Madrid (Spain)
[c] Dr. W. Reckien, Prof. Dr. B. Kirchner
Wilhelm-Ostwald Institut fꢁr Physikalische und Theoretische Chemie
Universitꢂt Leipzig, Linnꢀstr. 2, 04103 Leipzig (Germany)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200900403.
Chem. Eur. J. 2009, 15, 7139 – 7149
ꢆ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
7139

ample on the self-assembly of dendrimers,[16] or on weak,
non-covalently bound host-guest complexes of dendritic spe-
cies.[17] These results demonstrate the power of mass spec-
trometry for a detailed characterization of dendrimers with-
out which the fast pace of development in this field would
not have been possible.
final fragmentation products, the peripheral benzyl and
naphthylmethyl groups are detected with the highest intensi-
ties. The suggested mechanism also rationalizes a few side
reactions such as a methyl migration to the peripheral
benzyl or naphthylmethyl groups. A number of control ex-
periments also supports this fragmentation mechanism.
In this contribution, we discuss the mass spectrometric
characterization of Frꢀchet dendrons such as those shown in
Figure 1, which are synthesized in a convergent way.[9a,18]
Figure 2. Dendrimers (G0–G3) with quaternary ammonium ions at the
focal points. With the exception of the triethyl ammonium derivatives,
these dendrimers bear chiral amino moieties (B(n): benzyl-terminated
dendron of nth generation; N(n): naphthylmethyl-terminated G(n) den-
dron)
Results and Discussion
Synthesis of Frꢀchet dendron bromides: All Frꢀchet den-
drons were synthesized by the previously published proce-
dure.[9a,18] Up to generation G2, column chromatography
can easily be used to purify the dendrons. For the prepara-
tion of the G3 dendrons, two literature procedures exist to
convert the benzylic alcohol into the benzyl bromide: One
makes use of PBr3 and creates acid which may rearrange the
benzyl ethers within the already existing dendron scaffold.
The second one uses the Appel reaction and generates the
benzyl bromide under very mild conditions with CBr4 as the
bromide source and Ph3P as the oxygen scavenger. The final
benzyl bromide was attached to the tertiary amine cores by
simple nucleophilic displacement reactions.
Figure 1. G0–G3 Frꢀchet dendrons under study. At the focal points
(FPG) they bear a benzylic hydroxy group or bromide; the periphery
(PG) is decorated either with benzyl or with 2-naphthylmethyl
(“naphme”) groups.
The final dendron carries a benzylic alcohol or a benzyl bro-
mide at the focal point with which it can be attached to a
core molecule. After some brief comments on the analytical
characterization and the analysis of defects, we will focus on
the fragmentation behavior of Frꢀchet dendrons with qua-
ternary ammonium ions at their focal points (Figure 2).
Some of them have a viologen core and are doubly charged
(Figure 2). These dendrons exhibit a fast fragmentation cas-
cade, for which five different mechanisms are discussed.
Three can be ruled out based on experimental evidence
from tandem mass spectrometry. The remaining two are
closely related to each other and also supported by theoreti-
cal calculations. Although the initial step is the loss of an
amine from the focal point, the fragmentation cascade pro-
ceeds so quickly that intermediate fragments are either not
observed at all or appear merely with low intensities. As the
ESI-MS characterization of Frꢀchet dendrons: The Frꢀchet
dendrons terminated with a benzylic hydroxy group at the
focal point can easily be characterized by negative-ion elec-
trospray ionization Fourier-transform ion-cyclotron reso-
nance (ESI-FTICR) mass spectrometry. Deprotonation of
the OH group during the electrospray process occurs, when
for example methanol is used as the spray solvent. In addi-
tion to the deprotonated dendron, a quite intense signal for
the dimeric species bridged by an O···H-O hydrogen bond is
observed. The fact that the dimer survives even quite harsh
ionization conditions can be rationalized by the substantial
strength of this ionic hydrogen bond in the absence of any
competing solvent (ca. 130 kJmolÀ1).[19]
7140
www.chemeurj.org
ꢆ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 7139 – 7149

Dendrimer Disassembly in the Gas Phase
FULL PAPER
The corresponding benzyl bromides, however, are difficult
to ionize with electrospray ionization, because they neither
carry a charge nor bear any functional groups to which a
charge could easily be attached. All attempts to characterize
them directly by ESI-MS yielded more intense signals of
some traces of easy-to-charge impurities. For example, the
sodium adduct of triphenyl phosphine oxide generated as a
side product in the above-mentioned Appel reaction was ob-
served as one of the major signals, although it was not de-
tected in the NMR spectra and thus can only be present in
the sample in trace amounts.
Consequently, 1 equiv of triethyl amine has been added to
the sample solutions of the benzyl bromide dendrons in
methanol to generate TEA-B(n) and TEA-N(n) in situ
before conducting the MS experiment.[20] Each of these solu-
tions gave excellent results with the parent ion being the
most intense signal. Since this approach turned out to be
successful, other, more complex tertiary amines, that is, atro-
pine (A-B(n)), quinine (Q-B(n) and Q-N(n)), and Trçgerꢇs
base (T-B(n) and T-N(n)) were used instead of triethyl
amine (Figure 2). These compounds were isolated before
performing the MS experiments. For all dendrons bearing
an ammonium cation at their focal point, clean mass spectra
with excellent signal-to-noise ratio could be obtained
(Figure 3). Not unusual for the electrospray ionization of
salts, dimers held together by electrostatic forces through
one counter ion (bromide in all cases) can be seen with low
intensity. The exact masses and isotope patterns determined
by experiment are in excellent agreement with those calcu-
lated.
spond to the masses of one branching unit plus a peripheral
end group. For some cases, the defect structures shown in
Figure 4 represent several possible isobaric isomers. Since
the corresponding second-generation dendrons do not show
any substantial impurities, the defects in the third-genera-
tion dendrons must originate from the last steps in the con-
vergent synthesis. During the conversion of the benzyl alco-
hol into the benzyl bromide with PBr3 traces of acids induce
rearrangements of the benzyl ether linkages. While it is easy
to separate the analogous defects from the intact parent ion
for the lower generations, the chromatographic separation
of the intact dendrons becomes increasingly difficult for
higher generations. While the cleavage of benzyl ethers does
not occur in the absence of water, group transfer reactions
are observed in which whole branching units can change
places and even be transferred from one dendron to anoth-
er. Using CBr4/PPh3 for the production of the bromides,
which is also advantageous because of the higher yields ob-
tained, no defects were observed in the mass spectrum of
these compounds (Figure 4, bottom).[18] Consequently, the
defects that prevail in the samples even after chromatogra-
phy are easily seen by ESI mass spectrometry.
Dendritic viologens—The effect of dendron size on dication
stability: Quite different from the ammonium salts discussed
so far are the dendritic viologens shown in Figure 5. These
compounds have been examined by mass spectrometry
before with respect to their host–guest chemistry and they
represent excellent guests for Klꢂrner-type molecular tweez-
ers.[21] The dications are quite stable in solution due to the
presence of stabilizing counterions. However, they decom-
pose slowly over time, when the solvent is nucleophilic. Due
to the short distance between the two charges, significant
charge-repulsion effects can be expected to affect their gas-
phase behavior. Figure 6a shows the ESI- FTICR mass spec-
trum of Viol-G0. Most remarkably, the dication in its bare
form (asterisk in Figure 6a) has never been observed irre-
spective of the ionization conditions applied. Instead, the
sample ions avoid being a dication suffering from charge re-
pulsion by forming singly and doubly charged
À
À
(M2+) (PF6 )2nÀ1 (n = 1–3) and (M2+
AHCTNUGERTGUNNN nACHTUNGERTNNUNG )nCAHTUNGTREN(NUNG PF6 )2nÀ2 (n = 3–6)
clusters. In these clusters, the high positive charge can be
compensated by the counter ions and thus the compounds
are significantly stabilized.
Other signals also speak of a strong tendency to avoid
bare dications: Signals at m/z 203 and m/z 359 are due to
fragmentation of the dication by cleavage of one of the ben-
Figure 3. Two representative examples for positive-ion ESI-FTICR mass
spectra of dendrimers bearing an ammonium cation at the focal point:
Q-N1 (top) and A-N2 (bottom). Not unexpectedly, a bromide-bridged
dimer is observed for Q-N1. Experimental and calculated isotope pat-
terns agree well.
À
zylic C N bonds. A benzyl cation is then created together
with a mono-substituted, singly charged bipyridinium and
charge repulsion is avoided by separating the two charges
on two independent ions. Interestingly, a signal at m/z 561
showed an isotope pattern in the broadband mode of the
FT-ICR instrument, whose relative intensities changed with
the ionization conditions. This pointed to the fact that two
overlapping, non-resolved isotope patterns are observed.
Changing the ionization conditions also changed the relative
amounts of the two species contributing to the overall pat-
When the PBr3 reaction was used for the preparation of
the G3 Br-dendron, the co-generated acid induced the for-
mation of defects in the dendron scaffold. These defects
appear in the mass spectra at mass distances of Dm=212
below and to a minor extent also above the benzyl-terminat-
ed parent ions (Figure 4) and of Dm=264 for the 2-naph-
thylmethyl terminated ones. These mass differences corre-
Chem. Eur. J. 2009, 15, 7139 – 7149
ꢆ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7141

C. A. Schalley et al.
ence between 12C and 13C, both
patterns can be resolved and
independently seen at high res-
olution in Figure 6. Likely, the
proton loss occurs at the ben-
zylic position adjacent to one
of the nitrogen atoms yielding
a zwitterionic structure, which
nevertheless is well stabilized
by conjugation of the anion
with the aromatic ring. Conse-
quently, four different ways
exist for a cation to avoid the
charge repulsion within the di-
cation: Compensation of posi-
tive charges by counter ions,
proton loss, one-electron re-
duction, and fragmentation
leading to the separation of
two singly charged ions.
With the larger dendrimers
Viol-B1, Viol-B2, Viol-N1, and
Viol-N2 the same experiments
gave similar results. However,
one significant difference was
observed: Substitution with
dendrons of the Frꢀchet-type
stabilizes the bare dications so
that signals for them can be
observed in the ESI-FTICR
mass spectra (Figure 6b,c). A
Figure 4. ESI-FTICR mass spectra of the third generation dendrimers A-B3 (top) and TEA-B3 (bottom). Top:
Defects become visible at regular spacings of 212 Da below and to a minor extent above the parent ion of the
intact dendrimer, when the PBr3 reaction is used for dendron preparation. Bottom: Dendrons prepared with
the Appel reaction are essentially defect-free.
clear ranking of stabilities with increasing dendron genera-
tions was observed depending on the harshness of the ioni-
zation conditions. In particular, the capillary exit voltage can
be tuned at our instrument. This accelerates the ions to dif-
ferent velocities with which they then undergo collisions
with residual gas molecules. At higher settings of this volt-
age, only Viol-B2 and Viol-N2 gave clearly observable sig-
nals for the corresponding dications, but no dications of the
smaller generations were observed. By softening the condi-
tions, all four dendritic viologens of first and second genera-
tion gave signals for bare dications, while Viol-B0 did not
yield any bare dications irrespective of the ionization condi-
tions. This ranking was confirmed by other experiments with
the Klꢂrner-type tweezers complexes which showed a pro-
nounced dendritic effect on their gas-phase reactions.[21a]
Collision-induced decomposition of dendrons bearing am-
monium ions at their focal points: The fragmentation pat-
tern of the dendrons under study can be examined in
tandem MS experiments as shown for A-N2 in Figure 7.
First, the monoisotopic parent ion is mass-selected from all
ions present in the ESI mass spectrum (Figure 7a and b).
Then, the dendron ions of interest are subjected to collisions
with Ar as the collision gas. Chemical intuition would cer-
tainly predict that the primary fragment would correspond
to the loss of the tertiary amine giving rise to the corre-
Figure 5. Doubly charged dendrimers (G0–G2) with viologen cores
(“naphme”=2-naphthylmethyl).
tern. A high-resolution mass spectrum confirmed that two
ions differing by only 1 Da correspond to i) a one-electron
reduction occurring during the ionization process (m/z 562)
and ii) the loss of a proton from the dication (m/z 561).
Since the proton has a mass differing from the exact differ-
7142
www.chemeurj.org
ꢆ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 7139 – 7149

Dendrimer Disassembly in the Gas Phase
FULL PAPER
Figure 7. Top: ESI-FTICR mass spectrum of A-N2. Center: Isolation of
the monoisotopic molecular ion. Bottom: Collision-induced decay (CID)
experiment. The most intense fragment ion is the 2-naphthylmethyl
cation at m/z 141. The positions are marked, at which the intermediate
2nd and 1st generation benzyl cations are expected to appear. (*=over-
tone)
we only observe a cation originating from the periphery,
À
when the first reaction step should be C N bond cleavage at
the focal point?
Direct bond cleavage mechanism: The simplest mechanism
that can be imagined involves the direct cleavage of a pe-
ripheral benzyl ether bonds (Scheme 1). However, it is not
likely to take place because of the following reasons: First
of all, it is hard to imagine that it could energetically com-
pete with the expected amine loss, if one considers that it
would require a charge separation through formation of the
benzyl or 2-naphthylmethyl cations and the corresponding
anionic phenolic oxygen, which would compensate for the
charge on the ammonium group of the neutral fragment.
Such a charge separation in the gas phase requires several
hundreds of kJmolÀ1.[22] A second argument against such a
mechanism is the fact that no other dendritic benzyl cations
are observed, although for example the first generation Frꢀ-
chet dendron is connected to the next branching unit in the
G2 dendrons through the same benzyl ether bond. Another
evidence against such a mechanism comes from the CID ex-
periments on viologen-based dendrimers Viol-N1 and Viol-
N2. In the CID spectra of the dications, a bipyridinium frag-
ment is observed (e.g., Bipy-N2 in Scheme 1) which indi-
Figure 6. Positive-ion ESI-FTICR mass spectra of MeOH solutions
(50 mm) of a) Viol-G0, b) Viol-N1, and c) Viol-N2 (the latter two opti-
mized for maximum dication intensities). Insets: High resolution isotope
pattern of the signal at m/z 561 revealing that both electron capture and
proton loss overlap (top left), isotope pattern of the signal at m/z 1559
providing evidence for a superposition of singly and doubly charged clus-
ters (top right), and experimental isotope patterns of the bare Viol-N1
and Viol-N2 dications.
sponding benzyl cation of nth generation. In our example
A-N2, the G2 benzyl cation would thus be expected to
appear at m/z 927. The tertiary amine is a good and stable
leaving group and the benzyl cation is stabilized by conjuga-
tion of the cation with the aromatic ring next to it. Figure 7c
shows the CID mass spectrum of A-N2 as a representative
example. The experimental result differs significantly from
expectation for any of the dendrons under study: Regardless
of the generation number, the most intense (and often the
only) signal in all CID mass spectra corresponds to either
benzyl or 2-naphthylmethyl ions depending on the peripher-
al end group incorporated in the dendrons. Signals for the
expected benzyl cations of 1st to 3rd generation are instead
either not seen at all (G2 and G3) or hardly exceed the
noise (G1). The question arises, which mechanism might ac-
count for this unexpected fragmentation behavior. Why do
À
cates the cleavage of the benzyl N bond. The corresponding
benzyl cation dendron is however missing in the CID mass
spectrum. Instead, the major fragment is again the peripher-
al naphthylmethyl cation. Consequently, there must be a
mechanism connecting an initial amine loss with the cleav-
age of the peripheral benzyl ether bond. The direct bond
cleavage can thus be ruled out as a viable alternative.
Fragmentation cascade involving cyclophanes: A similar
fragmentation behavior is observed for dendrons of all gen-
Chem. Eur. J. 2009, 15, 7139 – 7149
ꢆ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7143

C. A. Schalley et al.
liberates the naphthylmethyl
cation. However, the same
mechanism applied to G1 and
G3 dendrons should generate
G1 benzyl cation, which
cannot
fragment
further
through this same mechanism.
One would therefore expect an
alternating fragmentation pat-
tern: The cascade should ter-
minate at G1 benzyl cations
for G1 and G3 dendrons, while
it forms the peripheral benzyl
or naphthylmethyl cations for
G2 dendrons. This behavior is
not observed. All generations
form predominantly benzyl or
naphthylmethyl cations. Conse-
quently, this mechanism can
also be ruled out.
À
Scheme 1. A direct bond cleavage mechanism (top) is unlikely because of the weak C N bond and unfavora-
ble charge separation. Also, it does not provide an explanation for the fragments observed for viologen-based
dendrimers such as Viol-N2 (bottom).
Benzyl–tropylium rearrange-
ment cascade: A third mecha-
erations under study. Any reaction mechanism connecting
the initial amine loss with cleavage of the peripheral benzyl
ether bond must thus involve a fragmentation cascade which
works itself through the generations: Loss of the tertiary
amine forms a G(n) benzyl cation, which undergoes a quick
nistic scenario avoids this alternation (Scheme 3). It involves
the formation of a G(n) benzyl cation, which rearranges
into a G(n) tropylium ion. In the tropylium ion, the positive
charge can be delocalized over all seven carbon atoms and
the two benzyl ether oxygens attached to the tropylium
core. This creates a good leaving group and is followed by
rearrangement to yield the G
G
next lower generation and so on until the periphery is
reached and no further rearrangements are possible any-
more.
the formation of the G
N
same process occurs until the periphery is reached. Accord-
ing to the literature on the benzyl/tropylium rearrangement,
the tropylium structure is energetically more favorable than
its benzyl analogue.[23] This would provide an explanation,
why the intermediate benzyl cations are not observed with
higher intensities. The ion would gain internal energy from
the reaction energy of each rearrangement step and thus the
individual steps in the fragmentation cascade proceed at
higher and higher rates from step to step.
One candidate for such a fragmentation cascade is depict-
ed in Scheme 2. Amine loss from second generation TEA-
N2 is followed by an electrophilic attack of the benzylic
carbon atom on an oxygen atom from the next shell of
branching units. Consequently, dioxa-metaACTHNUGRTENUNG[2.2]cyclophane
Phane-N2 is generated. The neutral cyclophane is an excel-
À
lent leaving group and cleavage of the oxonium CH2 bond
However, theoretical calculations indicate the rearrange-
ment barrier to be quite substantial (ca. 270 kJmolÀ1).[24]
Consequently, doubts arise that the benzyl–tropylium rear-
rangement scenario holds true. Two control experiments
have been performed. On their basis, we can also rule out
this mechanism (Scheme 4): In control compound C1, the
benzyl ethers at the first branching units have been replaced
by esters. After amine loss and the benzyl–tropylium rear-
rangement, one would expect the fragmentation cascade to
stop, because the tropylium leaving group cannot be
formed. Against expectation, the peripheral benzyl cation is
still the predominant fragment ion. If one of the peripheral
benzyl groups is replaced by methyl as in C2, no benzyl
fragment is observed anymore. Instead, the fragmentation
cascade is terminated directly after the amine loss. After a
benzyl–tropylium rearrangement, both substituents are how-
ever activated and one would expect to observe the remain-
ing benzyl group to be cleaved off, because this cation
Scheme 2. A fragmentation cascade involving the formation of cyclo-
phanes.
7144
www.chemeurj.org
ꢆ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 7139 – 7149

Dendrimer Disassembly in the Gas Phase
FULL PAPER
Scheme 3. A fragmentation cascade involving sequential benzyl–tropyli-
um rearrangements.
Scheme 5. Two similar fragmentation mechanisms involving electrophilic
aromatic substitution reactions. Note that rearomatization of both rings
can occur for both intermediates due to the presence of mobile protons.
step. These G(n) benzyl cations should then be seen in the
mass spectra. Consequently, the concerted reaction is most
likely to happen in all those cases, where the G(n) dendron
is not observed as an intermediate. In those few cases (e.g.,
see below), where the G(n) benzyl cation is indeed observed
as a minor fragment, both mechanisms may well compete.
Rearomatization of the attacked rings occurs and gener-
ates a mobile proton[25] which can easily move across the
whole aromatic ring system. Finally, this proton is anchored
to one of the oxygen atoms. When attached to the remaining
benzyl ether oxygen, again a good leaving group is generat-
ed and the benzyl cation (or 2-naphthylmethyl, respectively)
is liberated. If this benzyl cation is still dendritic, it can react
through the same sequence of steps until the periphery is
reached.
The fact that a quick reaction cascade is observed is not
in agreement with endothermic reaction steps. In line with
this argument, theoretical calculations predict the products
to be more favorable than the precursor benzyl cation by
more than 50 kJmolÀ1 (Figure 8). Since reaction energy is
gained in each step, they proceed faster and faster so that
the intermediates along the fragmentation cascade are not
observed in the mass spectra. The dendron thus immolates
itself in a chain reaction.
Scheme 4. Control experiments ruling out the benzyl–tropylium fragmen-
tation cascade.
would certainly be more stable than the methyl cation from
the second branch. These two control experiments clearly
rule out the benzyl–tropylium rearrangement to be in-
volved.
Fragmentation cascades involving electrophilic aromatic
substitutions: In particular, the latter control experiment
brings us to two additional mechanisms which are quite sim-
ilar and both involve intramolecular electrophilic aromatic
substitution steps (Scheme 5).
The fragmentation starts either with the stepwise loss of
the tertiary amine from the focal point and then creates the
corresponding G(n) benzyl cation, or it commences with a
concerted reaction step, in which the amine is lost simulta-
neous with the electrophilic attack. In a stepwise sequence,
some ions should have sufficient internal energy to accom-
plish the loss of NEt3, while not enough internal energy is
left to overcome the barrier for the electrophilic substitution
Chem. Eur. J. 2009, 15, 7139 – 7149
ꢆ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7145

C. A. Schalley et al.
Figure 8. Calculations at the B3LYP/TZVP level of theory predict the
products of the mechanisms shown in Scheme 5 to be more stable in
energy than the dendron benzyl ion so that each subsequent step in the
cascade should be faster than the preceding one (relative energies in
kJmolÀ1). Note that the dissociation threshold is not shown and the
Figure is just a comparison of the relative reactant and product energies.
The electrophilic substitution mechanism is also in line
with the fragmentation behavior of the control compounds
C1–C5 (Scheme 6). Only the ortho-attack (Scheme 5, right
branch) is considered here. An analogous reaction can of
course occur in the para-position (Scheme 5, left branch).
The ester analogue C1 fragments to yield the benzyl cation,
because the same mechanism can be applied to this com-
pound. The only difference is the formation of a seven- in-
stead of a six-membered ring in the neutral fragment, but
this difference does not interfere with the cascade. It is also
clear, why no benzyl cation is formed from C2. The electro-
philic substitution step attacks the only benzyl group present
and connects it tightly to the core. Since the methyl cation is
too small for the mass range of our instrument, we do not
know whether this cation is formed instead, but it is clear
that no benzyl cation can be liberated anymore. C3 forms
the methoxybenzyl cation as the major product, although
one would expect the electrophilic attack to be faster at the
more electron-rich methoxybenzyl ring. After this reaction,
only the nitrobenzyl would be available for further reaction.
However, the fact that the methoxybenzyl cation is the
major product is in good agreement with the assumption
that all rearrangement steps proceed below any of the avail-
able exit channels. In such a situation, rearrangements are
reversible and the more stable cation is expected to form.
From C3, the methoxybenzyl cation is better stabilized as
compared to the nitrobenzyl cation and thus is the major
product. Finally, symmetrically substituted C4 and C5 have
been synthesized. While no G1 benzyl cation is observed as
an intermediate for C5, the corresponding nitro-substituted
G1 benzyl cation is visible in the CID mass spectrum of C4
as a minor fragment. This indicates the consecutive reaction
steps to be faster for the methoxy derivative as compared to
the nitro compound—in line with the expected rates of the
electrophilic attacks involved in the fragmentation of these
ions.
Scheme 6. Control experiments in support of the intramolecular SE(ar)
reaction pathway.
therefore an intense signal. The corresponding G1 benzyl
cation, however, is only observed with low intensity. Never-
theless, this is one of the rare cases, in which the G1 benzyl
cation intermediate of the fragmentation cascade is indeed
observed. In addition, a few additional fragments appear
which are usually either very low in intensity or absent in
the CID mass spectra of the other dendrimers. A very simi-
lar situation is observed for Viol-N2. Because of charge re-
pulsion, the first dissociation step into the bipyridinium frag-
ment and the G(n) benzyl cation occurs at lower internal
energies. Consequently, the resulting benzyl cation is less ex-
cited, if generated from Viol-N1 or Viol-N2 and becomes
visible as an intermediate. Also, one expects that this inter-
mediate is longer-lived so that more time is available for re-
arrangements that finally lead to the additional fragments at
higher intensities than observed for the singly charged am-
monium dendrimers.
Collision-induced decomposition of viologen-based den-
drimers: Finally, we should take a closer look at the CID
mass spectrum of Viol-N1 as a representative of the doubly
charged dendritic viologens (Figure 9). As discussed above,
the mass-selected dication forms two singly charged frag-
À
ments through a cleavage of one of the benzylic C N bonds.
In the CID mass spectrum, the bipyridinium fragment is
7146
www.chemeurj.org
ꢆ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 7139 – 7149

Dendrimer Disassembly in the Gas Phase
FULL PAPER
Figure 9. For Viol-N1, the G1 benzyl cation is observed as the fragmenta-
tion intermediate. Also, some other side products besides the peripheral
benzyl loss are observed, which are in agreement with the postulated
mechanism.
Scheme 7 provides a plausible mechanistic interpretation
of the additional fragments observed in the CID experi-
ments with Viol-N1. The mobile proton can shift across dif-
ferent positions of the aromatic rings. Similarly, a methyl
cation can be involved. By combining different sequences of
such shifts, different intermediates are formed, which form
the products in energetically quite favorable 1,2-elimination
reactions of simple bond cleavages. Most importantly, the
methyl group can be shifted to the naphthyl ring explaining
the methyl-naphthylmethyl fragment at m/z 155. This shift
also indicates that an intact methyl group is present in the
intermediates and thus lends further support to the forma-
tion of the exocyclic methyl group during the rearomatiza-
tion step in Scheme 5.
Conclusion
ESI-FTICR mass spectrometry proved useful for the ioniza-
tion and characterization of dendrimers and is a sensitive
tool for the detection of defects. While the mass spectra are
quite simple to interpret for the dendritic alcohols (negative
mode) and dendritic ammonium ions (positive mode), the
spectra of the doubly charged dendritic viologens are more
complex. In particular, they show a pronounced stabilization
of the bare dication with increasing dendron size.
Scheme 7. A mechanistic rationalization of the side products observed in
the fragmentation of Viol-N1. All prominent side products can be ex-
plained by invoking a series of simple proton and methyl cation shifts
across the aromatic system. The major naphthyl methyl fragment is
formed in analogy to the mechanism shown in Scheme 5 (right) and not
repeated here.
Most importantly, however, the Frꢀchet dendrimers,
which are certainly one of the most important classes of
dendrimers in the chemical literature, exhibit a fascinating
fragmentation behavior. Collision-induced decomposition
(CID) experiments provided insight into a surprising frag-
mentation cascade initiated by the loss of the neutral amine.
Through several rapid reaction steps of energetically highly
favorable rearrangements of the intermediate benzyl cat-
ions, the peripheral benzyl or 2-naphthylmethyl cations are
generated as the most prominent signals in the CID mass
spectra. This cascade mechanism is reminiscent of the self-
immolative dendrimers mentioned in the beginning of this
article. As well, the fragmentation cascade reminds of the
first name “cascadanes” given to the early dendritic species
synthesized by Vçgtle et al.[26] Based on control compounds,
a number of mechanistic alternatives can be ruled out leav-
ing a mechanism, which involves an electrophilic aromatic
substitution of the benzyl cation core at one of the aromatic
rings in the next dendron shell. This mechanism also pro-
vides a rationalization for minor fragments that are for ex-
ample observed in the fragmentation of the doubly charged
viologen-based dendrimer ions.
Chem. Eur. J. 2009, 15, 7139 – 7149
ꢆ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7147

C. A. Schalley et al.
m) J. M. J. Frꢀchet, D. A. Tomalia, Dendrimers and other dendritic
polymers, Wiley, New York 2001; n) S. M. Grayson, J. M. J. Frꢀchet,
Chem. Rev. 2001, 101, 3819–3868; o) G. R. Newkome, C. D. Schrein-
er, Polymer 2008, 49, 1–173; p) F. Vçgtle, G. Richardt, N. Werner,
Dendritische Molekꢀle: Konzepte, Synthese, Eigenschaften, Anwen-
dungen, Teubner, Wiesbaden 2007; Dendrimer Chemistry: Concepts,
Syntheses, Properties, Applications, Wiley-VCH, Weinheim 2009.
[2] J. Denning, Top. Curr. Chem. 2003, 228, 621–672.
[3] a) M. Venturi, S. Serroni, A. Juris, S. Campagna, V. Balzani, Top.
Curr. Chem. 1998, 197, 193–228; b) V. Balzani, P. Ceroni, M. Maes-
tri, C. Saudan, V. Vicinelli, Top. Curr. Chem. 2003, 228, 159–191.
[4] D. A. Tomalia, Aldrichimica Acta 2004, 37, 39–57.
[5] a) R. J. Amir, N. Pessah, M. Shamis, D. Shabat, Angew. Chem. 2003,
115, 4632–4637; Angew. Chem. Int. Ed. 2003, 42, 4494–4499;
b) F. M. H. de Groot, C. Albrecht, R. Koekkoek, P. H. Beusker,
H. W. Sheeren, Angew. Chem. 2003, 115, 4628–4632, Angew Chem.
Int. Ed. 2003, 42, 4490–4494; c) S. Li, M. L. Szalai, R. M. Kevwitch,
D. V. McGrath, J. Am. Chem. Soc. 2003, 125, 10516–10517. Review:
d) D. Shabat, J. Polym. Sci. A 2006, 44, 1569–1578.
[6] D. Schubert, M. Corda, O. Lukin, B. Brusilowskij, E. Fiskin, C. A.
Schalley, Eur. J. Org. Chem. 2008, 4148–4156, and references there-
in.
[7] Selected examples: a) M. C. Coen, K. Lorenz, J. Kressler, H. Frey,
R. Mꢁlhaupt, Macromolecules 1996, 29, 8069–8076; b) L. Shu, A. D.
Schlꢁter, C. Ecker, N. Severin, J. P. Rabe, Angew. Chem. 2001, 113,
4802–4805; Angew. Chem. Int. Ed. 2001, 40, 4666–4669; c) J. Li,
D. R. Swanson, D. Qin, H. M. Brothers, L. T. Piehler, D. A. Tomalia,
D. J. Meier, Langmuir 1999, 15, 7347–7350; d) H.-J. van Manen, T.
Auletta, B. Dordi, H. Schçnherr, G. J. Vancso, F. C. J. M. van Veggel,
D. N. Reinhoudt, Adv. Funct. Mater. 2002, 12, 811–818; e) F. Wꢁrth-
ner, V. Stepanenko, A. Sautter, Angew. Chem. 2006, 118, 1973–
1976; Angew. Chem. Int. Ed. 2006, 45, 1939–1942; f) R. Bauer, D.
Liu, A. V. Heyen, F. D. Schryver, S. D. Feyter, K. Mꢁllen, Macromo-
lecules 2007, 40, 4753–4761; g) H. T. Baytekin, M. Sahre, A. Rang,
M. Engeser, A. Schulz, C. A. Schalley, Small 2008, 4, 1823–1834;
h) Review: A. D. Schlꢁter, J. P. Rabe, Angew. Chem. 2000, 112, 860–
880; Angew. Chem. Int. Ed. 2000, 39, 864–883.
Experimental Section
Syntheses: All dendrons under study here were synthesized, purified, and
characterized according to well-established procedures published previ-
ously.[20,27] Syntheses and analytical data of the control dendrons C1–C5
are provided in the Supporting Information.
ESI-FTICR mass spectrometry: ESI mass spectra were recorded on a
Bruker APEX IV Fourier-transform ion-cyclotron-resonance (FT-ICR)
mass spectrometer with an Apollo electrospray ion source equipped with
an off-axis 708 spray needle. Typically, acetonitrile served as the spray
solvent and 30–50 mm solutions of the analytes were used. Analyte solu-
tions were introduced into the ion source with a syringe pump (Cole-
Parmers Instruments, Series 74900) at flow rates of about 3–4 mLminÀ1
.
Ion transfer into the first of three differential pumping stages in the ion
source occurred through a glass capillary with 0.5 mm inner diameter and
nickel coatings at both ends. Ionization parameters—some with a signifi-
cant effect on signal intensities—were adjusted as follows: capillary volt-
age: À4.1 to À4.4 kV; endplate voltage: À2.8 to À3.5 kV; capexit voltage:
+200 to +300 V; skimmer voltages: +8 to +12 V; temperature of drying
gas: 2008C. The flows of the drying and nebulizer gases were kept in a
medium range (ca. 10 psi). The ions were accumulated in the instruments
hexapole for 0.5–1 s, introduced into the FT-ICR cell, which was operated
at pressures below 10À10 mbar and detected by a standard excitation and
detection sequence. For each measurement 16 to 512 scans were aver-
aged to improve the signal-to-noise ratio.
Tandem MS experiments: For MS/MS experiments, the whole isotope
patterns of the ions of interest were isolated by applying correlated
sweeps, followed by shots to remove the higher isotopes. After isolation,
argon was introduced into the ICR cell as the collision gas through a
pulsed valve at a pressure of about 10À8 mbar. The ions were accelerated
by a standard excitation protocol and detected after a 2 s pumping delay.
A sequence of several different spectra was recorded at different excita-
tion pulse attenuations in order to get at least a rough and qualitative
idea of the effects of different collision energies on the fragmentation
patterns.
Theoretical calculations: The calculations were performed with the TUR-
BOMOLE program package[28] employing Kohn-Sham density functional
theory (DFT) with the B3LYP functional[29] and the TZVP[30] basis set.
The SCF convergence criterion for these calculations was chosen to be
[8] C. A. Schalley, B. Baytekin, H. T. Baytekin, M. Engeser, T. Felder,
A. Rang, J. Phys. Org. Chem. 2006, 19, 479–490.
[9] Representative examples for defect analysis by mass spectrometry:
a) C. J. Hawker, J. M. J. Frꢀchet, J. Am. Chem. Soc. 1990, 112, 7638–
7647; b) M. Liu, J. M. J. Frꢀchet, Polym. Bull. 1999, 43, 379–386;
c) E. K. Wolter, M. J. Cloninger, Org. Lett. 2002, 4, 7–10; d) G. D.
Engel, L. H. Gade, Chem. Eur. J. 2002, 8, 4319–4329; e) W. Zhang,
S. E. Tichy, L. M. Pꢀrez, G. C. Maria, P. A. Lindahl, E. E. Simanek,
J. Am. Chem. Soc. 2003, 125, 5086–5094; f) S. C. Zimmerman, I.
Zharov, M. S. Wendland, N. A. Rakow, K. S. Suslick, J. Am. Chem.
Soc. 2003, 125, 13504–13518; g) A. Dirksen, U. Hahn, F. Schwanke,
M. Nieger, J. N. H. Reek, F. Vçgtle, L. De Cola, Chem. Eur. J. 2004,
10, 2036–2047.
10À8
.
Acknowledgements
We are grateful to Prof. Fritz Vçgtle for supply with some of the den-
drons examined herein. We thank Prof. Arne Lꢁtzen (Universitꢂt Bonn)
and Prof. Dietmar Kuck (Universitꢂt Bielefeld) for fruitful discussions.
Funding from the Deutsche Forschungsgemeinschaft and the Fond der
Chemischen Industrie (Dozentenstipendium to C.A.S.) is acknowledged.
[10] a) C. G. Juo, L. L. Shiu, C. K. F. Shen, T. Y. Luh, G. R. Her, Rapid
Commun. Mass Spectrom. 1995, 9, 604; b) G. Coullerez, H. J. Ma-
thieu, S. Lundmark, M. Malkoch, H. Magnusson, A. Hult, Surf. In-
terface Anal. 2003, 35, 682–692.
[1] For recent reviews, see: a) G. R. Newkome, C. N. Moorefield, F.
Vçgtle, Dendritic Molecules: Concepts, Synthesis, Perspectives,
Wiley-VCH, Weinheim, 1996; b) F. Zeng, S. C. Zimmerman, Chem.
Rev. 1997, 97, 1681–1712; c) A. Archut, F. Vçgtle, Chem. Soc. Rev.
1998, 27, 233–240; d) D. K. Smith, F. Diederich, Chem. Eur. J. 1998,
4, 1353–1361; e) M. Fischer, F. Vçgtle, Angew. Chem. 1999, 111,
934–955; Angew. Chem. Int. Ed. 1999, 38, 884–905; f) A. W.
Bosman, H. M. Jansen, E. W. Meijer, Chem. Rev. 1999, 99, 1665–
1688; g) G. R. Newkome, E. He, C. N. Moorefield, Chem. Rev. 1999,
99, 1689–1746; h) J.-F. Nierengarten, Chem. Eur. J. 2000, 6, 3667–
3670; i) A. D. Schlꢁter, J. P. Rabe, Angew. Chem. 2000, 112, 860–
880; Angew. Chem. Int. Ed. 2000, 39, 864–883; j) G. R. Newkome,
C. N. Moorefield, F. Vçgtle, Dendrimers and Dendrons. Concepts,
Syntheses, Applications, Wiley-VCH, Weinheim, 2001; k) R. Haag,
Chem. Eur. J. 2001, 7, 327–335; l) S. Hecht, J. M. J. Frꢀchet, Angew.
Chem. 2001, 113, 76–94; Angew. Chem. Int. Ed. 2001, 40, 74–91;
[11] See, for example: a) T. Kawaguchi, K. L. Walker, C. L. Wilkins, J. S.
Moore, J. Am. Chem. Soc. 1995, 117, 2159–2165; b) B. L. Schwartz,
A. L. Rockwood, R. D. Smith, D. A. Tomalia, R. Spindler, Rapid
Commun. Mass Spectrom. 1995, 9, 1552–1555; c) J. W. Leon, M.
Kawa, J. M. J. Frꢀchet, J. Am. Chem. Soc. 1996, 118, 8847–8859;
d) M. S. Wendland, S. C. Zimmerman, J. Am. Chem. Soc. 1999, 121,
1389–1390; e) P. B. Rheiner, D. Seebach, Chem. Eur. J. 1999, 5,
3221–3236; f) L. Bu, W. K. Nonidez, J. W. Mays, N. B. Tan, Macro-
molecules 2000, 33, 4445–4452; g) L. Zhou, D. H. Russell, M. Zhao,
R. M. Crooks, Macromolecules 2001, 34, 3567–3573; h) H. Chen, M.
He, X. Cao, X. Zhou, J. Pei, Rapid Commun. Mass Spectrom. 2004,
18, 367–370.
[12] Selected examples: a) G. J. Kallos, D. A. Tomalia, D. M. Hedstrand,
S. Lewis, J. Zhou, Rapid Commun. Mass Spectrom. 1991, 5, 383–
386; b) J. C. Hummelen, J. L. J. van Dongen, E. W. Meijer, Chem.
7148
www.chemeurj.org
ꢆ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 7139 – 7149

Dendrimer Disassembly in the Gas Phase
FULL PAPER
Eur. J. 1997, 3, 1489–1493; c) L. P. Tolic, G. A. Anderson, R. D.
Smith, H. M. Brothers II, R. Spindler, D. A. Tomalia, Int. J. Mass
Spectrom. 1997, 165, 405–418; d) U. Puapaiboon, R. T. Taylor,
Rapid Commun. Mass Spectrom. 1999, 13, 508–515; e) S. Watanabe,
M. Sato, S. Sakamoto, K. Yamaguchi, M. Iwamura, J. Am. Chem.
Soc. 2000, 122, 12588–12589; f) S. M. Cohen, S. Petoud, K. N. Ray-
mond, Chem. Eur. J. 2001, 7, 272–279; g) Y. Rio, G. Accorsi, H.
Nierengarten, J.-L. Rehspringer, B. Hçnerlage, G. Kopitkovas, A.
Chugreev, A. Van Dorsselaer, N. Armaroli, J.-F. Nierengarten, New
J. Chem. 2002, 26, 1146–1154; h) A. T. Iavarone, E. R. Williams, J.
Am. Chem. Soc. 2003, 125, 2319–2327; i) M. Luostarinen, T. Laiti-
nen, C. A. Schalley, K. Rissanen, Synthesis 2004, 255–262.
[13] B. Baytekin, N. Werner, F. Luppertz, M. Engeser, J. Brꢁggemann, S.
Bitter, R. Henkel, T. Felder, C. A. Schalley, Int. J. Mass Spectrom.
2006, 249, 138–148.
er: Z. Bo, L. Zhang, Z. Wang, X. Zhanh, J. Shen, Mater. Sci. Eng. C
1999, 10, 165–170.
[21] a) C. A. Schalley, C. Verhaelen, F. G. Klꢂrner, U. Hahn, F. Vçgtle,
Angew. Chem. 2005, 117, 481–485; Angew. Chem. Int. Ed. 2005, 44,
477–480; b) V. Balzani, H. Bandmann, P. Ceroni, C. Giansante, U.
Hahn, F.-G. Klꢂrner, U. Mꢁller, W. M. Mꢁller, C. Verhaelen, V. Vici-
nelli, F. Vçgtle, J. Am. Chem. Soc. 2006, 128, 637–648.
[22] Two point charges of opposite signs at a distance of 200 pm attract
each other with nearly 700 kJmolÀ1 according to the Coulomb law.
[23] For reviews, see: a) C. Lifshitz, Acc. Chem. Res. 1994, 27, 138–144.
Also, see: b) I. Howe, F. W. McLafferty, J. Am. Chem. Soc. 1971, 93,
99–105; c) A. Siegel, J. Am. Chem. Soc. 1974, 96, 1251–1252;
d) J. C. Traeger, R. G. McLoughlin, J. Am. Chem. Soc. 1977, 99,
7351–7352; e) C. Cone, M. J. S. Dewar, D. Landman, J. Am. Chem.
Soc. 1977, 99, 372–376; f) F. W. McLafferty, F. M. Bockhoff, Org.
Mass Spectrom. 1979, 14, 181–184; g) J. H. Moon, J. C. Choe, M. S.
Kim, J. Phys. Chem. A 2000, 104, 458–463; h) M. Malow, M. Penno,
K.-M. Weitzel, J. Phys. Chem. A 2003, 107, 10625–10630.
[14] a) J.-C. Blais, C.-O. Turrin, A.-M. Caminade, J.-P. Majoral, Anal.
Chem. 2000, 72, 5097–5105; b) T. Felder, C. A. Schalley, H. Fakhr-
nabavi, O. Lukin, Chem. Eur. J. 2005, 11, 5625–5636.
[15] a) J.-W. Weener, J. L. J. Dongen, E. W. Meijer, J. Am. Chem. Soc.
1999, 121, 10346–10355; b) U. Puapaiboon, R. T. Taylor, J. Jai-
nhuknan, Rapid Commun. Mass Spectrom. 1999, 13, 516–520; c) S.
Koster, M. C. Duursma, X. Guo, R. A. T. M. van Benthem, C. G.
de Koster, J. J. Boon, R. M. A. Heeren, J. Macromol. Chem. 2002,
37, 792–802; d) A. Adhiya, C. Wesdemiotis, Int. J. Mass Spectrom.
2002, 214, 75–88; e) H. Neubert, K. A. Knights, Y. R. de Miguel,
D. A. Cowan, Macromolecules 2003, 36, 8297–8303; f) S. Koster,
M. C. Duursma, J. J. Boon, R. M. A. Heeren, S. Ingemann,
R. A. T. M. van Benthem, C. G. de Koster, J. Am. Soc. Mass Spec-
trom. 2003, 14, 332–341; g) M. He, S. A. McLuckey, Rapid
Commun. Mass Spectrom. 2004, 18, 960–972.
[16] Some selected examples for MS-detected self-assembled dendrim-
ers: a) P. S. Corbin, L. J. Lawless, Z. Li, Y. Ma, M. J. Witmer, S. C.
Zimmerman, Proc. Natl. Acad. Sci. USA 2002, 99, 5099–5104;
b) H.-J. van Manen, F. C. J. M. van Veggel, D. N. Reinhoudt, Top.
Curr. Chem. 2001, 217, 121–162.
[17] a) W. Ong, A. E. Kaifer, Angew. Chem. 2003, 115, 2214–2217;
Angew. Chem. Int. Ed. 2003, 42, 2164–2167; b) T. Yamada, M. Ge,
H. Shinohara, K. Kimura, S. Mashiko, Chem. Phys. Lett. 2003, 379,
458–465; c) M. A. C. Broeren, J. L. J. van Dongen, M. Pittelkow,
J. B. Christensen, M. H. P. van Genderen, E. W. Meijer, Angew.
Chem. 2004, 116, 3641–3646; Angew. Chem. Int. Ed. 2004, 43, 3557–
3562.
[24] T. D. Fridgen, J. Troe, A. A. Viggiono, A. J. Midey, S. Williams, T. B.
McMahon, J. Phys. Chem. A 2004, 108, 5600–5609.
[25] a) D. Kuck, W. Bꢂther, H.-F. Grꢁtzmacher, J. Am. Chem. Soc. 1979,
101, 7154–7157; b) D. Kuck, W. Bꢂther, H.-F. Grꢁtzmacher, Int. J.
Mass Spectrom. Ion Processes 1985, 67, 75–91;
[26] E. Buhleier, W. Wehner, F. Vçgtle, Synthesis 1978, 155–158.
[27] The following articles report the syntheses which have been used:
a) O. Trapp, G. Trapp, J. W. Kong, U. Hahn, F. Vçgtle, V. Schurig,
Chem. Eur. J. 2002, 8, 3629–3634; b) U. Hahn, A. Kaufmann, M.
Nieger, O. Julꢃnek, M. Urbanova, F. Vçgtle, Eur. J. Org. Chem.
2006, 1237–1244; c) U. Herrmann, T. Jonischkeit, J. Bargon, U.
Hahn, Q. Y. Li, C. A. Schalley, E. Vogel, F. Vçgtle, Anal. Bioanal.
Chem. 2002, 372, 611–614; d) P. Ceroni, V. Vicinelli, M. Maestri, V.
Balzani, W. M. Mꢁller, U. Mꢁller, U. Hahn, F. Osswald, F. Vçgtle,
New J. Chem. 2001, 25, 989–993; e) R. Toba, J. M. Quintela, C. Pei-
nador, E. Romꢄn, A. E. Kaifer, Chem. Commun. 2001, 857–858;
f) T. H. Ghaddar, J. F. Wishart, D. W. Thompson, J. K. Whitesell,
M. A. Fox, J. Am. Chem. Soc. 2002, 124, 8285–8289.
[28] R. Ahlrichs, M. Bꢂr, M. Hꢂser, H. Horn, C. Kçlmel, Chem. Phys.
Lett. 1989, 162, 165–169. For the current version, see: http://
www.turbomole.de.
[29] a) A. D. Becke, Phys. Rev. A 1988, 38, 3098–3100; b) A. D. Becke, J.
Chem. Phys. 1993, 98, 5648–5652.
[30] The turbomole basis set library is available via anonymous ftp from
ftp://ftp.chemie.uni-karlsruhe.de/pub/basen.
[18] C. J. Hawker, J. M. J. Frꢀchet, J. Chem. Soc. Chem. Commun. 1990,
1010–1013.
[19] M. Meot-Ner ACHTUNGTRENNUNG(Mautner), Chem. Rev. 2005, 105, 213–284.
[20] The synthesis, isolation and characterization of Frꢀchet dendrimers
with triethylammonium at the focal point has been described earli-
Received: February 13, 2009
Published online: June 16, 2009
Chem. Eur. J. 2009, 15, 7139 – 7149
ꢆ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7149