Activation of aryl ethers and aryl sulfides by the Fe(η5-C5H5)+ group for the synthesis of phenol dendrons and arene-centered poly-olefin dendrimers

Article abstract of DOI:10.1039/B000370K

Organo-iron mediated syntheses of dendrons and dendrimers are described. The reaction of [FeCp(η⁶-p-EtOC₆H₄Me)][PF₆], 1, with Bu(t)OK or KOH and allyl bromide in THF gives both the triallylation of the acidic Me

Full text of DOI:10.1039/B000370K

View Article Online / Journal Homepage / Table of Contents for this issue
Activation of aryl ethers and aryl sulÐdes by the Fe(g5-C H )‘ group
5 5
for the synthesis of phenol dendrons and arene-centered poly-oleÐn
dendrimers¤
Valerie Sartor,a Sylvain Nlate,a Jean-Luc Fillaut,a Laurent Djakovitch,a Francoise Moulines,a
Valerie Marvaud,a Frederic Neveu,a Jean-Claude Blais,b Jean-Francois Letardc and Didier
Astruc*a
a Groupe de Chimie Supramoleculaire des Metaux de T ransition, L COO (CNRS UMR 5802),
Universite Bordeaux I, 33405 T alence cedex, France. E-mail: d.astruc=lcoo.u-bordeaux.fr
b L aboratoire de Chimie Structurale Organique et Biologique (CNRS EP 103), Universite Paris
V I, 4 Place Jussieu, 75252 Paris, France
c L aboratoire des Sciences Moleculaires, ICMCB (CNRS UPR 9048), Universite Bordeaux I,
33608 Pessac, France
Received (in Montpellier, France) 12th January 2000, Accepted 22nd February 2000
Organo-iron mediated syntheses of dendrons and dendrimers are described. The reaction of [FeCp(g6-p-
EtOC H Me)][PF ], 1, with ButOK or KOH and allyl bromide in THF gives both the triallylation of the acidic
6
4
6
Me substituent and the heterolytic cleavage of the OÈMe bond, which leads to the triallyl oxocyclohexadienyl
complex FeCp[g5-p-(allyl) CC H 2O], 2. If excess ButOK is added, the major compound obtained in this one-pot,
3
6 4
eight-step reaction is the iron-free phenol derivative p-(allyl) CC H OH, 3, in 50% yield. Surprisingly, the
3
6 4
analogous aryl sulÐde complexes [FeCp(g6-p-RSC H Me)][PF ], 4a (R \ Me) and 4b (R \ Et), also give 3
6
4
6
[together with (allyl) CC H SR, 5] by reaction with ButOK and allyl bromide. The dendron 3 reacts with 1,3,5-
3
6 4
C H [C(CH CH CH X) ] , 6, giving the 27-allyl dendrimer 7 (Ðrst generation) whose purity, monitored by
6
3
2
2
2 3 3
MALDI-TOF mass spectroscopy, is not as good when the reaction is performed with X \ I (6a) as with
X \ mesylate (6b). Iteration of this divergent synthesis gives the 81-allyl dendrimer 10 (second generation) and
ultimately the dendrimer 13 (third generation) with a theoretical number of 243 allyl groups. Protection and
functionalization of 3 also allows to carry out convergent dendritic syntheses: reaction of 3 with the protected
functional dendron C H C(O)OC H C(CH CH CH I) , 17, followed by deprotection gives the phenol nona-allyl
2
5
6
4
2
2
2 3
dendron OHC H C[CH CH CH O-p-C H C(CH CH2CH ) ] , 18. Compound 18 reacts with the 1,3,5-
2 3 3
C Me (CH Br) and C (CH Br) cores to cleanly give the corresponding arene-centered 27-allyl and 54-allyl
dendrimers 20 and 21, as shown by the dominant molecular peaks in the MALDI-TOF mass spectra. Reaction of
17 with 18 followed by deprotection does not give the expected 27-allyl dendron, but (very selectively) a 19-allyl
dendron 19 resulting from selective double branching and single dehydroiodation, which shows the limits of the
convergent strategy.
6
4
2
2
2
6
4
2
6
3
2
3
6
2
6
Activation des ethers dÏaryles et des sulfures dÏaryle par le gre†on Fe(g5-C H )‘ pour la synthese de dendrons
5
5
phenoliques et de dendrimeres a peripherie poly-oleÐnique et possedant un coeur aromatique. Dans ce memoire, nous
decrivons la synthese de dendrons et de dendrimeres. La reaction de [FeCp(g6-p-EtOC H Me)][PF ], 1, avec
6
4
6
ButOK ou KOH et le bromure dÏallyle produit a la fois la tri-allylation du substituant methyle acide et la coupure
heterolytique de la liaison OÈMe, conduisant au complexe tri-allyle FeCpg5-p -(allyl) CC H 2O], 2. Si on ajoute
3
6 4
un exces de ButOK, le principal compose obtenu, dans cette reaction en un seul pot pour huit etapes, est le derive
phenolique decomplexe p-(allyl) CC H OH, 3, avec un rendement de 50%. De facon apparemment surprenante,
3
6 4
lÏanalogue soufre [FeCp(g6-p-RSC H Me)][PF ], 4a (R \ Me) and 4b (R \ Et), donne aussi 3 [en meme temps
6
4
6
que (allyl) CC H SR, 5] par reaction entre 1, ButOK et le bromure dÏallyle. Le dendron 3 reagit avec 1,3,5-
3
6 4
C H [C(CH CH CH X) ] , 6, pour donner le dendrimere 27-allyle 7 (1e5re generation). La purete de 7, determinee
6
3
2
2
2 3 3
par spectrometrie de masse MALDI-TOF, nÏest pas aussi bonne quand la reaction est conduite avec X \ I (6a) que
quand elle est avec X \ mesylate (6b). LÏiteration de cette synthese divergente donne le dendrimere 81-allyle 10 (2nde
generation), et Ðnalement le dendrimere 13 (3e5me generation) comprenant un nombre theorique de 243 groupements
allyles terminaux. La protection et la fonctionnalisation de 3 permet aussi de mener a bien des synthese
dendritiques convergentes: la reaction de 3 avec le dendron fonctionnel protege
C H C(O)OC H C(CH CH CH I) , 17, suivie de deprotection, conduit au dendron nona-allyl phenol
2
5
6
4
2
2
2 3
OHC H C[CH CH CH O-p-C H C(CH CH2CH ) ] , 18. Ce compose 18 reagit avec les coeurs 1,3,5-
6
4
2
2
2
6
4
2
2 3 3
C Me (CH Br) et C (CH Br) pour donner proprement les dendrimeres 27-allyle 20 et 54-allyle 21,
6
3
2
3
6
2
6
respectivement, comme lÏindiquent les pics moleculaires dominant dans les spectres de masse MALDI-TOF. La
reaction de 17 avec 18 suivie de deprotection ne conduit pas au dendron 27-allyle attendu, mais (tres selectivement)
au dendron 19-allyle 19 resultant du double branchage selectif et dÏune dehydroiodation, ceci montrant la limite de
la strategie convergente.
¤ Dedicated to Oliver Kahn, fully living in our memories as a generous friend and as an outstanding lecturer and scientist who pioneered the
conceptual approach of molecular magnetism.
Electronic supplementary information (ESI) available: MALDI-TOF mass spectra of the 19-allyl dendron 19, 27-allyl dendrimer 20, 54-allyl
dendrimer 21 and 81-allyl dendrimer 10. See http://www.rsc.org/suppdata/nj/b0/b000370k/
DOI: 10.1039/b000370k
New J. Chem., 2000, 24, 351È370
351
This journal is ( The Royal Society of Chemistry and the Centre National de la Recherche ScientiÐque 2000

View Article Online
The temporary complexation of aromatics by 12-electron
organo transition-metal fragments in robust 18-electron com-
plexes such as [Cr(arene)(CO) ], [Mn(arene)(CO) ]` and
the same arene ligand in a one-pot reaction, directly leading
to a dendron. This goal has been successfully achieved and the
results published in preliminary form;10 it is the subject of the
present article. Our new and apparently surprising Ðnding of
the reaction of the corresponding aryl sulÐdes is now included,
together with the synthesis of new dendrimers such as a 54-
allyl dendrimer resulting from the convergent method. These
results are key steps in the context of our ongoing synthesis11
and applications12 of metallodendrimers.
3
3
[MCp(arene)]` (M \ Fe or Ru) provides a powerful means to
activate aromatics towards nucleophilic and deprotonation
reactions.1h5 This trend has been known and exploited for a
long time, especially in the [Cr(arene)(CO) ] series.1h3 In the
3
[MCp(arene)]` series, the activation is stronger than in the
[Cr(arene)(CO) ] one, however, which has led us to deproton-
3
ate such complexes bearing methyl substituents on the arene
ligands.6 Recent thermodynamics studies have indeed shown
Results
that the pK of these 18-electron complexes are lower by
a
about 15 units than those of the free aromatics in DMSO.7
Heterolytic cleavage of the exo-cyclic O–C bond of aryl ether
complexes in [FeCp(p-CH C H OEt)], 1
Although stepwise deprotonation is known, we have recently
been more concerned with the perfunctionalization of these
complexes using a base (KOH or ButOK) and a halide RX
[R \ alkyl, allyl, benzyl, etc.; X \ Br or I, eqn. (1)].6 This
reaction consists in the iteration of deprotonation-alkylation
sequences, yet it is di†erent from catalysis, and represent an
original way to activate ligands and synthesize dendritic cores.
3 6
4
The yellow complex 1 is readily available13 and can be syn-
thesized in large quantities. It reacts with ButOK and allyl
bromide in THF to give the orange oxocyclohexadienyl
complex 2 as a red oil in 50% yield after puriÐcation on a
silica column. This reaction combines the heterolytic cleavage
of the exocyclic OÈC bond9 and triallylation of the methyl
group located in the para position14 [eqn. (3)].
(3)
(1)
Other products found in minor amounts in this reaction are
the decomplexation products resulting from partial reaction
and/or decomplexation. Formation of 2 has been optimized
by variation of the reaction temperature and time (see
experimental section). First, ButOK deprotonates the methyl
group to give a deep red cyclohexadienyl complex, which is
followed by the Ðrst allylation, then the two other
deprotonation-allylation sequences follow before cleavage of
the exocyclic OÈC bond. This can be veriÐed by protonation
of the red intermediates, giving a yellow mixture of aryl ether
The advantage of dendrons over dendrimers is that den-
drons have a functional trunk whereas dendrimers do not,
contrary to what etymology suggests.8 Thus, dendrons can be
attached to a variety of functional molecular and nanoscale
materials. Therefore, it was of great interest for us to investi-
gate an extension of the CpFe` induced aromatic per-
functionalization strategy to functional aromatics bearing
methyl groups in order to synthesize dendrons, in addition to
dendritic cores. Sometime ago, we disclosed the CpFe`
induced heterolytic cleavage of the OÈalkyl bond of pheny-
lalkyl and diphenyl ether complexes, leading to oxocyclohexa-
dienyl complexes [eqn. (2)], the deprotonated form of
phenols.9
allylated
complexes
of
the
type
[FeCpMg6-p-
EtOC H CH (allyl) N][PF ].
6
4
3~n
n
6
Synthesis of the triallyphenol dendron p-OHC H C(allyl) , 3
6
4
3
Compound 2 is decomplexed in acetonitrile using a Hg
lamp,15 which leads to the substitution of the arene ligand by
three acetonitrile ligands in the cation; thus separation is
facile, giving a 70% yield of p-OHC H C(allyl) , 3 a light
6
4
3
brown solid [eqn. (4)].
(2)
(4)
The next challenge was to perform both the per-
functionalization and heterolytic CÈO cleavage reactions on
An alternative synthesis of 3, which turns out to be more
convenient than the above procedure, is the direct synthesis
352
New J. Chem., 2000, 24, 351È370

View Article Online
using the precursor 1 in a one-pot reaction. The procedure
consists in using the same reactants as for the synthesis of 2
except that another portion of ButOK is added at the end of
the reaction in order to achieve decomplexation. This one-pot
reaction produces 3 from 1 in 50% yield after puriÐcation
using column chromatography [eqn. (5)]. Compound 3 is air-
stable for a day but long-term storage must be carried out in
the dark under an inert atmosphere, best in a refrigerator.
analog [eqn. (6)]. The same reaction proceeded with the ethyl
aryl ether complex, although the yields were lower.
Divergent dendrimer syntheses using the dendron 3
The nonafunctional dendritic cores 6a (X \ I) and 6b
(X \ mesylate) are available by nona-allylation of [FeCp(g6-
mesitylene)][PF ] with KOH and allyl bromide,14 followed
6
by photolytic decomplexation,15 regioselective hydro-
boration18 and oxidation of the nonaborane18 to the nona-
ol14 and subsequent functionalization. First, 6a is synthesized
as follows. Reaction of the nona-ol dendritic core with
SiMe Cl in THF in the presence of Et N provided the nona-
3
3
trimethylsilyl dendritic core, which was converted in situ to
the nona-iodo dendrimer 6a by reaction with NaI [eqn.
(7)].19,20
The MALDI-TOF mass spectrum of 6a shows the suc-
cessive loss of all the iodo atoms, and is therefore of little
value to determine the purity of 6a. In 1H NMR, the terminal
methylene carbon peak undergoes only a slight shift from 3.4
ppm for the nona-ol (CD OD) to 3.1 ppm for 6a (CDCl ). The
reaction is best characterized with 13C NMR by the replace-
ment of the signal of the primary alcohol carbon at 63.7 ppm
in the nona-ol by the upÐeld signal of the carbon attached to
the iodo atom at 8.0 ppm in 6a (see experimental section),
which shows that it is quantitative.
(5)
3
3
Williamson-type coupling reaction21 of the nona-iodo core
6a with the dendron 3 was Ðrst tested using tertiobutylphenol,
which has the advantages of being commercially available and
having no oleÐn group, unlike 3. In this way, the side dehy-
droiodation reaction leading to the formation of oleÐn termini
can be monitored by 1H NMR, because the downÐeld oleÐn
signals are easily observable. The parameters potentially inÑu-
encing the substitution-elimination competition21,22 were
varied using this model reaction. The results were that room
temperature was preferred to higher temperatures, that the
nature of the base (CsF vs. K CO ) had little inÑuence,23 and
Cleavage reactions of thio aryl ethers using ButOK
Aryl thiolate ligands are useful in bio-inorganic chemistry16
and we wished to examine the possible extension of the reac-
tions of eqns. (3) and (4) to the heterolytic cleavage of aryl
sulÐde p complexes analogous to 1. These aryl sulfphide com-
plexes, 4a (R \ CH ) and 4b (R \ C H ), are as easily avail-
3
2 5
able as the ether complexes by reactions of the corresponding
alkyl aryl thiols with [FeCp(g6-p-ClC H CH )] in the pres-
6
4
3
ence of Na CO .6 The tentative SÈC cleavage in 4 was inves-
2
3
2
3
tigated with both the methyl and ethyl aryl sulphide
complexes. Optimization of the reaction time and temperature
and search for the best alkyl group on sulfur led to the forma-
that small amounts of elimination products were always
tion of
2 in 20% yield and the free aryl sulÐde
CH SC H C(allyl) , 5a, in 45% yield. The formation of 5a
3
6
4
3
was expected, but that of 2 was not; it was checked twice by
MALDI-TOF (matrix-assisted laser desorption ionizationÈ
time of Ñight)17 mass spectroscopy, which showed its very
clean formation by the presence of the molecular peak as the
dominant peak; no peak was observed for the awaited sulfur
(7)
(6)
New J. Chem., 2000, 24, 351È370
353

View Article Online
detected by 1H NMR. Reaction of 6a with 3 in DMF at room
temperature in the presence of K CO yielded the 27-allyl
2
3
dendrimer 7 in 44% yield after chromatographic puriÐcation.
The 1H and 13C NMR spectra of 7 are very clean, and the
MALDI-TOF mass spectrum shows that the molecular peak
[M ] Na]` at 2559 dominates; an important peak corre-
sponding to the octa-substituted dendrimer with dehydroioda-
tion of the non-substituted branch and
a small peak
corresponding to the hepta-substituted dendrimer are also
found. Thus, although the major dendrimer formed is the 27-
allyl dendrimer, the 25-allyl dendrimer is also formed in good
amounts and the 23-allyl dendrimer is formed in small
amounts [eqn. (8)]. If the intensity of the peaks in the mass
spectrum reÑects the respective amounts, the Williamson
coupling is achieved to an extent of 95% whereas elimination
reaches about 5%, which is too much for further dendritic
construction.
In order to prepare a purer 27-allyl dendrimer, we then
switched to the nona-mesylate dendritic core 6b in order to
eliminate the side dehydrohalogenation reaction as much as
possible. Indeed, Kellog et al. reported that the use of mesy-
late as a leaving group for substitution by propionic acid with
cesium carboxylate in DMF did not give any elimination
product.24a Such a preference was already noted in the con-
struction of polyether dendrimers.24b,c Reaction of the nona-ol
with chloromethylsulfonate in pyridine25 gives a 95% yield of
the nona-mesylate dendritic core 6b as a colorless oil that is
best stored at [25 ¡C. The completion of the reaction is char-
acterized by the shift of the signals of the terminal methylene
in CDCl from 3.4 (CH OH) to 4.1 (CH OSO Me) in 1H
3
2
2
2
NMR and from 63.7 to 70.1 in 13C NMR. Reaction of 6b with
the model tertiobutylphenol led to the total absence of signals
due to oleÐnic protons in the 1H NMR observations, indicat-
ing that elimination, if any, is occurring in lower amounts
than the observation threshold in 1H NMR. Reaction of 6b
with 3 in DMF in the presence of CsF produces the 27-allyl
dendrimer as a colorless oil in 27% yield after chromato-
graphic puriÐcation [eqn. (9)].
(8)
The 1H and 13C NMR spectra are exactly identical to those
of 7 obtained from 6a as above. The MALDI-TOF mass spec-
trum shows the molecular peaks [M ] Na]` at 2559 and
[M ] K]` at 2575 with only a tiny peak at 2331 correspond-
ing to octa-substitution and dehydromesylation of one
branch. The proportion of dehydromesylation can be esti-
mated to be of the order of 1%. In the crude reaction product,
some hydroxymethyl termini were detected. Indeed, the very
facile hydrolysis of the mesylate group to the hydroxy group is
known.26 Dendrimers containing this impurity were removed
during the chromatographic puriÐcation, but their formation
reaches 30%, which consequently leads to lower yields of 7
than when starting from the polyiodoalkyl precursor.
Attempts to mesylate these residual hydroxy groups and con-
dense them again with 3 were not successful, possibly because
of steric inhibition and hydrophobicity of the dendritic cavity
around the reaction site.
The dendrimer 7 synthesized from 6b was used for further
syntheses. Thus, regioselective hydroboration of 7 was carried
out using disiamylborane (HBsiam ) and the 27-borane den-
2
drimer was oxidized to the 27-alcohol dendrimer 8, a white
solid, using H O in basic medium [eqn. (10)]. The MALDI-
2
2
TOF mass spectrum of 8 shows only the molecular peaks
[M ] Na]` at 3045.11 and [M ] K]` at 3061.06, as well as a
trace of 25-ol at ([M ] Na]` at 2779.74 and [M ] K]` at
2795.88) as expected from the purity of the precursor (Fig. 1).
Reaction of 8 with SiMe Cl in the presence of Et N was
yellow solid, was obtained in 43% yield. Reaction of 8 with
chloromethylsulfonate gave the 27-mesylate dendrimer 9b as a
colorless oil in 92% yield. Reaction of 9b with 3 proceeds as
for the Ðrst generation and gives the 81-allyl dendrimer 10 as
a colorless oil in 71% yield after chromatographic separation
[eqn. (11)]. The MALDI-TOF mass spectrum does not show
the molecular peak, but many peaks corresponding to lack of
substitution and deshydroiodation are observed. On the other
hand, reaction of 9b with 3 as for the Ðrst generation gives 10
3
3
carried out in THF instead of MeCN, because 8 is not soluble
in MeCN, leading to the 27-alkoxytrimethylsilane interme-
diate,27 which was converted in situ to the 27-iodo dendrimer
by reaction with NaI in the presence of SiMe Cl in MeCN
3
after removing THF.28 The 27-iodo dendrimer 9a, a pale
354
New J. Chem., 2000, 24, 351È370

View Article Online
(9)
Fig. 1 MALDI-TOF mass spectrum of the 27-alcohol dendrimer 8 (molecular peaks M ] Na` at 3047.73 and M ] K` at 3063.74).
New J. Chem., 2000, 24, 351È370
355

View Article Online
(10)
356
New J. Chem., 2000, 24, 351È370

View Article Online
(11)
New J. Chem., 2000, 24, 351È370
357

View Article Online
Scheme 1
in 6.4% yield after repeated chomatographic separation. The
Convergent syntheses of dendrons and dendrimers using the
mass spectrum of this sample of 10 shows the molecular peak
as the dominant peak [M ] Na]` at 8723.38. Less intense
peaks corresponding to the lack of one to nine dendrons and
deshydromesylation are also observed.
dendron 3
Heterobifunctionalization and protection of the triallyl phenol
dendron 3 has been achieved in order to carry out convergent
dendrimer syntheses. First, the phenol group of 3 is protected
by methylation using methyl iodide in DMSO29 giving 14,
then the methyl aryl ether is functionalized by regioselective
hydroboration using disiamylborane and oxidation of the tri-
borane by H O in basic aqueous medium,18 which gives the
Finally, hydroboration of this sample of 10 followed by oxi-
dation of the 81-borane dendrimer as for the preceding gener-
ation gave the 81-ol dendrimer 11 in 53% yield as a white
powder, and reaction of chloromethylsulfonate with 11 as
above gave the 81-mesylate dendrimer 12 as a colorless oil in
85% yield. Reaction of 12 with 3 gave the 243-allyl dendrimer
13 (Scheme 1) in 11% yield after chromatographic separation.
This reaction was monitored in 1H NMR by the disap-
pearance of the signals due to the CH OSO Me group at 3
2
2
protected triol 15. Reaction of 15 with Me SiCl and NaI19,31
leads to both iodation of the branch termini and deprotection
3
of the phenol, giving 16. Protection of the phenol group of 16
is achieved by propionylation using propionyl iodide in the
presence of diisopropylethylamine,31 giving the protected
functional dendron 17 (Scheme 2). Tentative propionylation
using propionyl chloride leads to partial chlorination of the
branches and must be discarded.32 Reaction of 17 with 3 fol-
lowed by deprotection using K CO successfully leads to the
2
2
and 2.3 ppm and the appearance of the signals of the oleÐn
protons downÐeld near 5.6 and 5 ppm. The mass spectrum of
13 could not be obtained, possibly due to polydispersity.
However, a clear 13C NMR spectrum was recorded with the
expected signals corresponding to the surface groups, those of
the mesylate groups having totally disappeared (Fig. 2).
2
3
nona-allyl dendron 18 [eqn. (12)].
358
New J. Chem., 2000, 24, 351È370

View Article Online
Fig. 2 [1H]13C NMR spectrum of the 243-allyl dendrimer 13 in CDCl solution (Bruker AC 250 FT, 62.9 MHz). The solvent triplet (CD) is
3
noted around d \ 77 vs. SiMe . For assignments, see Experimental section.
4
Continuation of the convergent strategy was attempted by
reaction of the dendron 18 with the protected phenol triodo
derivative 17 in the presence of K CO , followed by deprotec-
The dendron 18 also reacts with 1,3,5-tribromomesitylene in
the presence of K CO to give the 27-allyl dendrimer 20 as a
light brown oil in 30% yield (non-optimized) after chromato-
graphic puriÐcation. The reaction is selective and complete as
indicated by the MALDI-TOF mass spectrum showing the
2
3
2
3
tion using K CO at higher temperature. The clean 1H and
2
3
13C NMR spectra are in full agreement with the formation of
the expected phenol 27-allyl dendron. The MALDI-TOF mass
spectrum shows a unique molecular peak [M ] Na]` at
2077.7, however, corresponding to the very selective substitut-
ion of two iodides by two dendrons and dehydroiodation of
the third branch. Thus, the phenol derivative formed, 19, has
only 19 allyl branches (Scheme 3); it is obtained as a light
brown oil in 25% yield after chromatographic separation.
(12)
Scheme 2
New J. Chem., 2000, 24, 351È370
359

View Article Online
Scheme 3
molecular peak [M ] K]` as the only signiÐcant peak at
2936.95. Finally, 18 reacts with hexakis(bromomethyl)benzene
in the same way to give the 54-allyl dendrimer 21 as a white
crystalline solid in 83% yield after chromatography (Scheme
4); 21 was characterized by its molecular peak appearing as
the main peak in the MALDI-TOF mass spectrum.
group on the oxygen atom protects it during the three
deprotonation-allylation sequences. Yet its bond to the
oxygen atom is fragile enough to be cleaved thereafter, while
the ironÈarene bond is still resistant. Synthesis of the tri-
dendate oxocyclohexadienyl complex 2 thus requires these
seven steps to proceed in a precise order in a one-pot reaction.
Deprotonation of a methyl arene substituent37,38 by ButOK
is followed by allylation of the deep red methy-
lenecyclohexadienyl intermediate to generate a yellow cationic
arene complex, whose structure only di†ers from that of the
starting complex by the substitution of an allyl group for a
benzylic hydrogen atom. The monosubstituted allyl complex
cannot be isolated in pure form because it is deprotonated in
situ by the methylenecyclohexadienyl intermediate, and
double substitution by allyl groups is always observed in
minor amounts, even if only one equivalent of ButOK is used.
Such methylenecyclohexadienyl intermediates have been iso-
lated and their structure is known.37,38 The structure of the
deprotonated species can be described by two mesomers: a
dominant methylenecyclohexadienyl form and a zwitterionic
form with a minor contribution9 (Chart 1).
Discussion
Formation of the triallylphenol dendrons 2 and 3
Heterolytic cleavage of exocyclic carbonÈoxygen bonds is
known from a previous study with aryl ether ligands lacking
methyl groups,9 and is also known in other ligands and
metals33,34 and even in the absence of metal under drastic
conditions.35 Oxocyclohexadienyl or phenate complexes
resulting from exocyclic OÈC cleavage have been known for a
long time, being easily formed by deprotonation of phenol
complexes.36 The use of such phenol p complexes would not
be appropriate here because their deprotonation to phenate p
complexes would be a dead end. On the other hand, the alkyl
360
New J. Chem., 2000, 24, 351È370

View Article Online
Scheme 4
The fact that two tautomers rather than two mesomers
should be considered cannot be discarded, although we prefer
the mesomer description in the absence of evidence for two
species on the infrared time scale of 1013 s~1. Two other ana-
logous deprotonation-allylation sequences must proceed
before heterolytic cleavage of the exocyclic OÈC of the aryl
ether complex by ButOK. Therefore, the reaction mixture is
kept at low temperature each of the three times that ButOK is
added in the course of the synthesis. Once the OÈC bond has
been cleaved, it is no longer possible to deprotonate the ligand
because the oxocyclohexadienyl complex formed is neutral,
thus no longer acidic. Minor amounts of reaction products
analogous to 2 and containing only two allyl substituents are
sometimes formed, and can be eliminated during the chro-
matographic puriÐcation. Their formation results from the
fact that the third deprotonation-allylation sequence is much
slower than the Ðrst two because of steric inhibition due to the
allyl substituents. The synthesis of 3 proceeds analogously to
that of 2, but also requires one more step: the cleavage of the
ironÈarene bond by ButOK. This last step does not occur at
low temperature, but only more slowly at room temperature
in the presence of excess ButOK added at the very end of the
reaction. We know that electron-rich anions such as ButOK
can transfer an electron to 18-electron organo iron cations of
the [FeCp(g6-arene)]` type to produce unstable 19-electron
intermediates, which decompose. The decomposition of such
19-electron species occurs because the 19th electron occupies
an antibonding e* orbital, which destabilizes the metalÈarene
1
bond,38 especially when the arene bears a heteroatom in an
exocyclic position and long alkyl chains,39 which is the case
for 2 (Scheme 5).
Although 2 is not strictly an arene complex, the arene con-
tribution to the mesomers is indicated by the reduction
potential9b ([2.3 V vs. FeCp `@0 in DMF), which is accessible
2
although it is more negative that those of the regular cationic
arene complexes38 ([1.7 V vs. FeCp `@0 in DMF). We believe
2
that, after electron transfer, the reduced form 2~ is a true,
although very unstable, 19-electron complex of phenate.40 The
reaction conditions of eqn. (4) and Scheme 5 avoid this
reduction before the seven other steps. Thus the eight steps of
this one-pot synthesis occur in the right order as indicated in
the overall mechanism represented in Scheme 6. The ion-
pairing and salt e†ects (i.e., counteranion in Scheme 5) are
New J. Chem., 2000, 24, 351È370
361

View Article Online
Scheme 5
very inÑuential factors, which have been investigated in the
case of the same reaction with the ethoxybenzene complex9b,c
[eqn. (2)].
It is also probable that, in the reaction of 1 (Scheme 5), allyla-
tion of 2 by allyl bromide occurs in situ and is followed by
cleavage by ButOK to re-form 2. Thus, 1 is presumably a cata-
lyst for the alkylation of allyl bromide by ButOK, although
investigation of turnovers have not been carried out (Scheme
8).
In summary, the best route to 3 is the one-pot eight-step
procedure from 1, ButOK, allyl bromide and THF, which is
very convenient.
The formation of 3 from the aryl sulÐde complexes 4a and
4b may appear as a rather surprising feature. The formation of
5 in large amounts in this reaction indicates that the exocyclic
SÈC bond is resistant to cleavage in the iron complex, thus the
7th step of Scheme 6 is skipped. This is conÐrmed by the
absence of any aryl thiol formation and, instead, by the forma-
tion of 3. It is known that thioalkyl groups are excellent
leaving groups in the nucleophilic aromatic substitution of
electron-poor aromatics.41 Thus ButO~ plays the role of the
incoming nucleophile in the reaction with 4, then the classic
OÈC cleavage by ButOK can occur as for 1 (Scheme 7).
We know that the OÈC cleavage reaction is general with
aryl alkyl ether complexes and even in diaryl ether complexes.
Divergent and convergent dendrimer syntheses
The methods of dendrimer synthesis in molecular chemistry
are quite varied.8,42h54 They are divergent42h46 or
convergent42,47,48 or can result from mixing both types. The
divergent technique is most often used because it can lead to
Scheme 6
362
New J. Chem., 2000, 24, 351È370

View Article Online
to 10 (2nd generation, 81-allyl) and 13 (3rd generation, 223-
allyl), and the failure to synthesize the 27-allyl phenol dendron
(Scheme 3) in the convergent strategy. The interest in the
divergent strategy lies both in the synthesis of pure dendrimer
of modest size (7) and in the relatively fast synthesis of the
dendrimer 13 with a theoretical number of 243 branches. This
construction is summarized in Scheme 9.
Although this strategy gives the 81- and 243-allyl dendri-
mers 10 and 13 in low yields if the mesylate route is required,
the related routes in which dehydroiodation is not available,
such as the synthesis of the 54-allyl dendrimer 21, give dendri-
mers as crystalline compounds with excellent purity and high
yields. E†orts along this line to synthesize large quantities of
dendrimers are ongoing in our laboratories. Except perhaps
for 21, the elemental analyses were not correct for the large
dendrimers from 27- to 243-allyl. This problem is classic: it is
now well known that solvent molecules and various guest
molecules are very difficult to remove from dendrimer
samples, even after several days under vacuum, which often
leads to frustrating attempts to carry out successful elemental
analyses.55 This feature emphasizes even more the consider-
able assistance of MALDI-TOF mass spectroscopy in their
characterization (see also the spectra in the Electronic Supple-
mentary Information). The lack of success in obtaining a
MALDI-TOF mass spectrum for 13 argues in favor of a poly-
disperse dendrimer56 in spite of its satisfactory 1H and 13C
NMR spectra (see Fig. 2). The increasing bulk around the
reaction sites at the termini of the branches as the generation
number increases is probably a limiting factor to synthesizing
the 243-allyl dendrimer by the divergent route. An indication
of this problem is represented by the molecular models of the
27-, 81- and 243-allyl dendrimers (Fig. 3). These models do not
give conformational information. In particular, they do not
address the problem of the location of the termini (at the
periphery57 or turning inside58). The overall bulk is clearly
apparent, however.
The design and practical synthesis of the tridentate dendron
3 has several advantages: (i) it allows one to carry out both
the divergent and convergent strategies; (ii) it involves a rapid
dendritic growth since the number of branches is multiplied
by three, whereas most dendrimer syntheses only double the
number of branches at each generation; (iii) the branches of
the dendron are relatively long, which avoids steric problems
at least for the Ðrst generations; (iv) it can be functionalized in
di†erent ways on both the phenol and oleÐn sides; (v) its syn-
thesis in reasonable yields from a cheap starting material,
which can be easily synthesized on a scale of several hundred
grams, is now available from the present study in a one-pot
procedure; (vi) the convergent route leads to a clean nona-
allyl phenol dendron 18, which is used as a building block for
the one-step clean synthesis of dendrimers of moderate size, as
exempliÐed by the synthesis of the 54-allyl dendrimer 21.59
Scheme 7
the formation of large dendrimers, but it is marred by impu-
rities, which become more and more important as the gener-
ation number increases. The convergent method provides
clean dendrimers for the low generations, but cannot be
applied to the synthesis of large dendrimers because of the
increasing steric bulk at the branching points. After Achar and
PuddephattÏs dendrimer syntheses astutely using oxidative
addition on platinum centers,54 the present dendrimer synthe-
sis strategy is the second one to use organo transition-metal
chemistry. The growth of the dendrimer is fast and large den-
drimers are synthesized in only a very few generations. All the
problems noted above are typically met in the present work
however, as exempliÐed by the decreased dendrimer purity
met in the divergent strategy from 7 (1st generation, 27-allyl)
Conclusions
1. The Ðrst organo iron strategy for the construction of den-
drimers and dendrons has been successfully developed using
an original synergy between two modes of activation of aro-
matics by the electron-withdrawing cation CpFe`: the
enhanced acidity of the benzylic protons and the nucleophilic
cleavage of the OÈC bond of aryl ether complexes.
2. The triallyl phenol dendron 3 is synthesized in 50%
overall yield in a one-pot eight-step synthesis from [FeCp(g6-
p-MeC H OEt)][PF ], 1, a cheap starting material easily
6
4
6
available on a scale of several hundred grams.
3. The cationic aryl sulÐde iron complexes 4 analogous to 1
do not undergo SÈC cleavage by reaction with ButOK and
allyl bromide, but 3 and the metal-free triallyl aryl sulÐde 5
are formed; the apparently surprising formation of 3 is
accounted for by a nucleophilic aromatic substitution of the
Scheme 8
New J. Chem., 2000, 24, 351È370
363

View Article Online
Scheme 9
364
New J. Chem., 2000, 24, 351È370

View Article Online
to have a good level of purity, their molecular peak being
always dominant (see the Electronic Supplementary
Information).
7. This study opens practical and powerful routes to further
studies in dendritic and nanosized materials chemistry and
their applications based on the dendrons 3 and 18. Various
strategies along these lines are presently ongoing in our labor-
atories.
Experimental
General
All reactions were carried out in an inert atmosphere in oven-
dried or Ñamed glassware using Schlenk techniques and a
Vacuum Atmosphere Company dry-lab. Reagent grade sol-
vents used for the syntheses are dried and distilled prior to
use: tetrahydrofuran (THF) is distilled over NaÈbenzo-
phenone ketyl, then over CaH ; 1,2-dimethoxyethane (DME)
2
is distilled over NaÈbenzophenone ketyl (caution: THF and
DME are not distilled to dryness); N,N-dimethylformamide
(DMF) and dimethylsulfoxide (DMSO) are distilled over
CaH immediately before use; pyridine is distilled over KOH
2
immediately before use. All other chemicals are used as
received.
1H NMR spectra were recorded with a Bruker AC 250 FT
(250 MHz) and a Bruker AC 200 FT (200 MHz), and 13C
NMR spectra were recorded with a Bruker AC 250 FT (62.9
MHz). All chemical shifts are reported in ppm with internal
reference to tetramethylsilane or to the solvent. The elemental
analyses were carried out at the CNRS center in Vernaison
(France). All mass spectra were recorded using the MALDI-
TOF technique on a Voyager Elite TOF mass spectrometer
(Perspective Biosystems, Framingham, USA). One lL of an
acetone solution of the matrix (2,5-dihydroxybenzoic acid)
and analyte at concentrations of 0.1 M and 0.1 mM, respec-
tively, was deposited on the target for MALDI analysis.
[FeCp(g6-p-ClC H Me)]-[PF ] and [FeCp(g6-mesitylene)]-
6
4
6
[PF ], precursors of 2 and [FeCp(g6-1,3,5-triallyl-methyl-
6
benzene)][PF ],14 respectively, are synthesized by ligand
exchange between ferrocene and p-ClC H Me in the presence
of AlCl and Al powder according to refs. 60 and 61, respec-
tively, modiÐed with improved yields according to ref. 62 (1
6
6
4
3
equiv. H O per ferrocene is added; the arene is used as solvent
2
at reÑux overnight). [FeCp(g6-p-EtOC H Me)][PF ], 1, is
6
4
6
synthesized in essentially quantitative yield by reaction
between [FeCp(g6-p-ClC H Me)][PF ], Na CO
and
6
4
6
2
3
ethanol (used as solvent at reÑux overnight) according to ref.
13.
Syntheses
[FeCp(g6-p-triallylmethylphenate)], 2 [eqn. (3)]. [FeCp(g6-
p-EtOC H Me)][PF ], 1 (see above),6c,7 (5 g, 12.4 mmol) and
Fig. 3 Molecular modeling (Cerius program) of the (a) 27-allyl den-
drimer 7, (b) 81-allyl dendrimer 10 and (c) 243-allyl dendrimer 13.
6
4
6
ButOK (12.6 g, 112.5 mmol) are introduced into a Ñamed
three-necked Ñask under nitrogen, then 100 mL of freshly dis-
tilled THF is introduced by canula into this reaction mixture
at [50 ¡C. After stirring 15 min., 20 mL (0.232 mmol) of cold
allyl bromide is added. The reaction medium is stirred at
ambient temperature in the dark for 2 days, ButOK (12.6 g,
112.5 mmol) and allyl bromide (10 mL, 0.116 mmol) in 50 mL
THF are introduced at [50 ¡C and the reaction mixture is
stirred for 2 more days at ambient temperature in the dark.
ButOK (12.6 g, 112.5 mmol) and allyl bromide (5 mL, 0.058
mmol) in 10 mL THF are again added at [50 ¡C and the
reaction mixture is stirred for an additional 2 days. Finally,
ButOK (9 g, 80 mmol) and allyl bromide (5 mL, 0.058 mmol)
in 10 mL THF are added at [50 ¡C and the reaction mixture
is allowed to stir for 2 more days under the same conditions.
The solid residue is removed from the reaction mixture by
Ðltration, the solvent of the Ðltrate is removed under vacuum,
thioalkyl group by ButO~, followed by the classic OÈC cleav-
age.
4. Functionalization of a nona-allyl dendritic core has led to
the use of 3 as a building block for the rapid divergent dendri-
mer synthesis up to a theoretical number of 243 branches after
only three generations.
5. A convergent strategy based on 3 has led to the clean
synthesis of a nona-allylphenol dendron 18, a very useful
building block for clean, one-step synthesis of larger dendri-
mers such as the 54-allyl dendrimer 21.
6. MALDI-TOF mass spectroscopy has been systematically
used all along this work. It has helped to considerably
improve the syntheses whereas 1H and 13C NMR brought
comparatively much less accurate information. Finally, the
dendrons or dendrimers 3 to 18 (except 13) have been shown
New J. Chem., 2000, 24, 351È370
365

View Article Online
and the resulting solid residue is puriÐed by chromatography
on a neutral alumina column using methanol as the eluent,
which gives a red oil (2.2 g, 6.2 mmol, 50% yield). 1H NMR
d 7.30È7.11 (m, CH
arom
, 4H); 5.66È5.45 (m, CHCH , 3H);
2
5.05È4.97 99 (m, CHCH , 6H); 2.93 (q, SCH CH , 2H), 2.43
2
2
3
(d, CH CHCH , 6H); 1.31 (t, SCH CH , 3H,); 13C NMR
2
(CDCl )
3
2
2
3
(CD COCD ) d 5.75 (m, CH
arom
5.06È4.97 (m, CHCH and C H , 8H); 4.57 (m, CHCH , 6H);
, 2H); 5.49 (m, CH
, 2H);
d
143.50 (SC
6 quat 4 2
H ); 128.83 (CH ); 127.29 (CH ); 117.74
H ); 135.40 (CH CH); 134.38
3
3
arom
(C
2
5
5
2
6 quat
4
arom
arom
2.44 (m, CH , 6H). 13C NMR (CD COCD ) d 154.89
(CHCH ); 43.00 (C CH ); 41.79 (CH CHCH ); 27.74
quat
(SCH CH ); 14.45 (SCH CH ). Anal. calcd for C S: C
2
3
3
2
2
2
2
H
(C
(C
O); 135.26 (CHCH ); 119.24 (CHCH ); 101.89
C); 84.18 (CH CO); 74.93 (C H ); 73.22
quat arom
quat arom
2
2
2
3
2
3
18 24
79.37; H 8.89. Found C 80.00; H 8.91. MS (EI) m/z (rel. int.):
arom
5
5
(CH C); 43.56 (CH ); 42.54 (C CH ). Anal. calcd for
272 (23) [M]`, 231 (100), 202 (6), 190 (33).
arom
2
quat
2
C
H
OFe, H O: C 68.83; H 7.16; C 8.74; Fe 15.28. Found
21 24
2
C 68.82; H 7.20.
Nona-iodo core 6a [eqn. (7)]. The nona-ol core synthesized
according to ref. 14 (1.17 g, 1.82 mmol) in 200 mL MeCN,
p-Triallylmethylphenol, 3, by photochemical decomplexation
of 2 [eqn. (4)]. The complex 2 (2 g, 0.575 mmol) and 200 mL
degassed MeCN are introduced under nitrogen into a Schlenk
Ñask. The solution is irradiated for 12 h with a Hg vapor
lamp, the solvent is removed under vacuum and the residue is
dissolved in 200 mL ether. This solution is washed with 100
mL 6 N HCl, dried over sodium sulfate, Ðltered and the
solvent is removed under vacuum. The residue is puriÐed by
then dry NaI (10 g, 66 mmol) and freshly distilled SiMe Cl (5
mL, 39 mmol) are successively introduced into a Ñamed
3
Schlenk Ñask. This reaction mixture is stirred under nitrogen
in the dark for 12 h at 80 ¡C. The solvent is then removed
under vacuum, the residue is dissolved in 200 mL ether, this
solution is washed with a saturated aqueous solution of
Na S O and Ðltered, and the solvent is removed under
2 2
3
vacuum. This gives 6a as a pale yellow solid (2.64 g, 1.62
chromatography on
a silica column with pentaneÈether
mmol, 89%). 1H NMR (CDCl ) d 7.04 (s, C H , 3H); 3.14 (m,
3
6 3
(95 : 5) as the eluent, which yields 3 as a light brown solid
CH I, 18H); 1.81 (m, CH CH I, 18H); 1.56 (m, CH C
,
2
2
2
2 quat
(0.920 g, 0.40 mmol, 70% yield). 1H NMR (CDCl ) d 7.18 and
18H). 13C NMR (CDCl ) d 145.51 (C
); 121.98 (CH );
3
3
quat arom
arom
6.83 (m, CH
, 4H); 6.58 (m, OH, 1H); 5.68È5.55 (m,
42.8 (C CH ); 38.56 (CH CH I); 27.79 (CH C ); 8.5
arom
quat
2
2
2
2 quat
CHCH , 9H); 5.07È5 (m, CHCH , 18H); 2.44 (d, CH , 18H).
(CH I). Anal. calcd for C
H
I : C 28.48; H 3.52; I 70.
2
2
2
OH); 137.60
2
36 57 9
13C NMR (CDCl )
d
153.40 (C
Found C 26.55; H 3.49.
3
quat arom
(C
C); 134.56 (CHCH ); 127.73 (CH COH); 117.42
(CHCH ); 114.89 (CH C); 42.59 (C CH ); 41.87 (CH ).
quat arom
2
arom
Nona-mesylate core 6b [eqn. (7)]. The nona-ol (0.6 g, 0.934
mmol) in 7 mL anhydrous pyridine is introduced in a Ñamed
Schlenk tube under an inert atmosphere. The solution is
cooled to [5 ¡C, and chloromethylsulfonate (13 mL, 16
mmol) is slowly added to this solution over 1 h. The solution
is stirred for 4 h at 0È2 ¡C. Ice (50 g) is added to this solution
and iced 6 N HCl is added until pH 3 is reached. Extraction
with 3 ] 50 mL dichloromethane is followed by drying over
sodium sulfate. After Ðltration and removal of the solvent
under vacuum, the residue is washed with 50 mL cold
pentane. The nona-mesylate 6b is obtained in this way as a
colorless oil and is stored at [25 ¡C (1.2 g, 0.89 mmol, 95%
yield). 1H NMR (CDCl ) d 7.02 (s, C H , 3H); 4.09 (m,
2
arom
quat
2
2
Anal. calcd for C
H
O: C 84.16; H 8.83; O 7.01. Found C
16 20
84.14; H 8.82.
p-Triallylmethylphenol, 3, by one-pot synthesis from
[FeCp(g6-p-EtOC H Me)] [PF ], 1 [eqn. (5)]. [FeCp(g6-p-
6
6
4
6
EtOC H Me)][PF ], 1 (5 g, 12.4 mmol) and ButOK (12.6 g,
6
4
112.5 mmol), are introduced into a Ñamed three-necked Ñask
under Ar, then 100 mL of freshly distilled THF is introduced
by canula into this reaction mixture at [50 ¡C. After stirring
15 min., cold allyl bromide (20 mL, 0.232 mmol) is added. The
reaction medium is stirred at ambient temperature in the dark
for 1 day, then ButOK (12.6 g, 112.5 mmol) and allyl bromide
(10 mL, 0.116 mmol) in 50 mL THF are introduced at
[50 ¡C, and the reaction mixture is warmed to 25 ¡C over 3 h
and stirred 1 more day at ambient temperature in the dark.
ButOK (12.6 g, 112.5 mmol) and allyl bromide (8 mL, 0.092
mmol) are again added at [50 ¡C and the reaction mixture is
stirred for 1 more day. Finally, ButOK (18.3 g, 163 mmol) is
added in 80 mL THF at [50 ¡C and the reaction mixture is
allowed to stir for 1 more day under the same conditions. The
solvent is removed under vacuum and the solid residue is dis-
solved in 500 mL ether, this solution is washed with 100 mL 6
N HCl, dried over sodium sulfate and Ðltered. The solvent is
removed under vacuum and the residue is chromatographed
on a silica column using pentaneÈether (95 : 5) as the eluent,
which gives 3 as a light brown solid (1.71 g, 0.75 mmol, 60%
yield).
3
6 3
CH O, 18H); 2.93 (s, CH , 27H); 1.72 (m, CH CH O, 18H);
2
3
2
2
1.42 (m, CH C , 18H). 13C NMR (CDCl ) d 145.12
2 quat
); 121.99 (CH ); 70.73 (CH O); 42.36 (C CH );
quat arom arom quat
36.91 (CH ); 33.62 (CH CH O); 23.56 (CH C ).
2 quat
3
(C
2
2
3
2
2
27-Allyl dendrimer 7 from the nona-iodo core 6a [eqn. (8)].
Triallylmethylphenol 3 (1.25 g, 5.5 mmol) in 10 mL of freshly
distilled DMF and potassium carbonate (3.85 g, 27 mmol) are
introduced into a Ñamed Schlenk tube under nitrogen and the
reaction mixture is stirred for 30 min., then the nona-iodo
core 6a (0.5 g, 3.06 mmol) in 20 mL anhydrous DMF is pro-
gressively added. The reaction mixture is stirred for 3 days at
room temperature, the organic derivative is extracted with
3 ] 100 mL ether, the solution is dried over sodium sulfate,
Ðltered, and the solvent is removed under vacuum. The
product is puriÐed by chromatography on a neutral alumina
column using a pentaneÈether mixture (98 : 2) as the eluent.
This gives the 27-allyl dendrimer 7 as a colorless oil (0.342 g,
44%).
Reaction of ButOK and allyl bromide with [FeCp(g6-p-
MeC H SR)] [PF ], R = Me (4a) or Et (4b) eqn. (6)]. The
6
4
6
reaction has been carried out as from 1, and the organic deriv-
atives 2 and 5 are separated by column chromatography. 2 is
obtained in 20% yield with R \ Me and 15% yield with
R \ Et. 5a and 5b are characterized as above. 5a p-
MeSC H C(CH ÈCH2CH ) ]: 45% yield. 1H NMR (CDCl )
27-Allyl dendrimer 7 from the nona-mesylate core 6b [eqn.
(9)]. Cesium Ñuoride (5.4 g, 36 mmol) is heated at 160 ¡C
under vacuum for 3 h in a Ñamed Schlenk tube. Then, 3 (5.49
g, 24 mmol) is dissolved in 15 mL of freshly distilled DMF
and this solution is added into the Schlenk tube under nitro-
gen. The mixture is stirred for 30 min., then the nona-mesylate
6b (1.2 g, 0.89 mmol) in 15 mL anhydrous DMF is progres-
sively added over 30 min. This reaction mixture is stirred for 3
days at room temperature under nitrogen. After extraction
with 3 ] 50 mL pentane, the organic layer is dried over
6
4
2
2 3
3
d 7.23 (m, CH
, 4H); 5.64È5.50 (m, CHCH , 3H); 5.07È4.99
arom
2
(m, CHCH , 6H); 2.50 (s, SCH , 3H); 2.43 (d, CH CHCH ,
6H). 13C NMR (CDCl )
(CH CH); 134.42 (C
(CH ); 117.73 (CHCH ); 43.30 (C CH ); 41.81
arom quat
(CH CHCH ); 15.89 (SCH ). Anal. calcd for C
2
3
2
2
d
143.42 (SC
H ); 135.40
3
6 quat
4
H ); 127.29 (CH ); 126.38
2
6 quat
4
arom
2
3
2
H
S: C
17 22
8.76. 5b [p-
2
2
79.01;
H
8.58. Found C
79.45;
H
EtSC H C(CH ÈCH2CH ) ]: 58% yield. 1H NMR (CDCl )
6
4
2
2 3
3
366
New J. Chem., 2000, 24, 351È370

View Article Online
sodium sulfate and Ðltered. The solvent is removed under
vacuum and the residue is chromatographed on a neutral
alumina column using a pentaneÈether mixture (98 : 2), which
gives 7 as a colorless oil (0.61 g, 0.24 mmol, 27%). 1H NMR
(CDCl ) d 7.32 (s, C H , 3H); 7.2 (d, C H , 18H); 6.8 (d,
27-Mesylate dendrimer 9b. The synthesis of 9b from 0.800 g
(0.264 mmol) 8 is e†ected according to the same procedure as
that described above for the synthesis of 6b, and yields 9b as a
colorless oil (1.2 g, 0.234 mmol, 92% yield). 1H NMR (CDCl )
3
d 7.38 (s, C H , 3H); 7.20 and 6.77 (m, C H , 36H); 3.80 (m,
3
6
3
6
4
6
3
6 4
C H , 18H); 5.66È5.49 (m, CHCH , 27H); 5.06È4.99 (m,
CH OC
, 18H); 3.77 (m, CH OSO , 54H); 2.83 (m,
SO CH , 81H); 1.94 (m, CCH CH , 72H); 1.63 (m,
6
4
2
2
arom
2
2
CHCH , 54H); 3.9 (m, CH O, 18H); 2.43 (d, CH CHCH ,
2
2
2
2
2
3
2
2
54H); 1.97 (m, CCH CH , 18H); 1.67 (m, CH CH O, 18H).
13C NMR (CDCl ) d 156.86 (OC
137.48 (C
(CH ); 117.59 (CHCH ); 113.94 (C H ); 68.24 (CH O);
arom
43.17 (C CH ); 42.71 (C CH ); 42.03 (CH CHCH );
quat quat
34.02 (CH CH O); 24.17 (CCH CH ). MALDI-TOF MS:
CCH CH , 72H). 13C NMR (CDCl ) d 156.79 (C
O);
2
2
2
2
2
2
3
quat arom
H ); 145.74 (C
H ); 134.71 (CH CH); 127.67 (CH ); 122.44
H );
144.59 (C
); 136.42 (C
); 126.8 (CH ); 121.89
(CH ); 113.9 (CH ); 70.6 (CH SO ); 67.87 (CH OC);
41.22 (C CH ); 36.76 (SO CH ); 32.43 (CCH CH ); 23.15
quat
(CH CH O).
3
6 quat
4
6 quat
3
quat arom
quat arom
arom
6 quat
4
2
arom
arom
arom
2
2
2
2
6
4
2
2
2
3
2
2
2
2
2
2
2
2
2
2
2
2
[M ] Na]` 2558.88 (calcd 2558.80).
81-Allyl dendrimer 10 [eqn. (11)]. The 81-allyl dendrimer 10
is synthesized from the 27-iodo dendrimer 9a (0.330 g, 0.0551
mmol) using 1.2 g (5.26 mmol) of triallylmethylphenol accord-
ing to the same procedure as that described above for the
synthesis of 7 from 6a. The product is puriÐed by chromatog-
raphy on a neutral alumina column using a pentaneÈether
mixture (3 : 1) as the eluent, which gives 10 as a colorless oil
(0.342 g, 0.0393 mmol, 71% yield of crude 10, see text). Alter-
natively, 10 is synthesized from 9b (0.343 g, 0.0668 mmol)
according to the same procedure as that described above for
the synthesis of 7 from 6b. The product is puriÐed by chroma-
tography on neutral alumina columns using a pentaneÈether
mixture (3 : 1) as the eluent, which gives 10 as a colorless oil
27-Alcohol dendrimer 8 [eqn. (10)]. The 27-alcohol is syn-
thesized from the 27-allyl precursor 7 in the same way as the
nona-ol according to ref. 6a. This procedure follows the
hydroboration-oxidation by Brown et al.63 A molar THF
solution of BH (127 mL, 127 mmol) is introduced into a
3
Ñamed Schlenk tube under nitrogen and cooled to [15 ¡C. A
2 M THF solution of 2-methyl-2-butene (127 mL, 254 mmol)
is cooled to [15 ¡C, then added into the Schlenk tube con-
taining the BH solution and the reaction mixture is kept at
3
[12 ¡C for 2 h. The 27-allyl dendrimer 7 (0.800 g, 0.315
mmol) dissolved in 40 mL of freshly distilled THF is then
added and the reaction mixture is stirred for one 1 day in the
dark. Then, 30 mL degassed water, 30 mL 3 M NaOH and 45
mL 30% H O are added. The reaction mixture is left for 15 h
(0.030 g, 0.0035 mmol, 6.4% yield). 1H NMR (CDCl ) d 7.2
3
and 6.8 (m, C H , 144H); 5.66È5.49 (m, CHCH , 81H); 5.06È
6
4
2
4.99 (m, CHCH , 162H); 3.9 (m, CH O, 72H); 2.43 (d,
CH CHCH , 162H); 1.97 (m, CCH CH , 72H); 1.67 (m,
2
2
2
2
at 50 ¡C, the two phases are separated, the aqueous phase is
saturated with potassium carbonate and extracted with 5 ] 50
mL THF. The organic phases are gathered collected and dried
over sodium sulfate, Ðltered, and concentrated to 30 mL under
reduced pressure. The 27-alcohol dendrimer 8 is obtained by
precipitation upon addition of 100 mL ether (0.800 g, 0.265
mmol, 84% yield). 1H NMR (CD OD) d 7.39 (s, C H , 3H);
2
2
2
2
CH CH O, 72H). 13C NMR (CDCl ) d 156.81 (OC
H );
2
2
3
6 quat
4
145.79 (C
H ); 137.46 (C
H ); 134.62 (CH CH); 127.52
6 quat
3
6 quat
4
2
(C H ); 122.42 (C H ); 117.44 (CHCH ); 113.81 (C H ); 68.12
6
4
6
3
2
6 4
(CH O); 43.17 (C CH ); 42.60 (C CH ); 41.91
quat quat
(CH CHCH ); 33.79 (CH CH O); 23.72 (CCH CH ).
2
2
2
2
2
2
2
2
2
MALDI-TOF MS: [M ] Na]` 8723.38 (calcd 8723.91).
3
6
3
, 18H);
7.21 and 6.77 (m, C H , 36H); 3.78 (m, CH OC
81-Alcohol dendrimer 11 (Scheme 1). The dendrimer 11 is
synthesized from 10 (obtained from the 27-mesylate
dendrimer) according the same procedure as that used for the
syntheses of the nona-ol and 27-alcohol 8 (see above). From
0.18 g (0.020 mmol) of 10, 11 is obtained as a white powder
6
4
2
arom
3.44 (m, CH OH, 54H); 1.95 (m, CCH CH , 72H); 1.64 (m,
2
2
2
CCH CH , 72H). 13C NMR (CD OD) d 158.3 (C
O);
2
2
3
quat arom
148.24 (C
); 140.25 (C
); 128.63 (CH ); 124.7
quat arom
quat arom
arom
(CH ); 115.25 (CH ); 69.50 (CH OC); 63.71 (CH OH);
arom
arom
2
2
42.95 (C CH ); 42.88 (C CH ); 34.89 (CCH CH ); 27.95
quat quat
(CH CH O); 25.59 (CH CH O). MALDI-TOF MS
(108 mg, 0.0106 mmol, 53% yield). 1H NMR (CD OD) d 7.21
2
2
2
2
2
3
and 6.77 (m, C H , 144H); 3.78 (m, CH OC
, 72H); 3.44
2
2
2
[M ] Na]` 3045.11; [M ] K]` 3063.7 (calcd 3061.06).
6
4
2
arom
(m, CH OH, 162H); 1.95 (m, CCH CH , 234H); 1.64 (m,
2
2
2
CCH CH , 234H). 13C NMR (CD OD)
d
157.99
2
2
3
27-Iodo dendrimer 9a. The 27-alcohol dendrimer 8 (0.338 g,
0.128 mmol) is dissolved in 30 mL of freshly distilled THF;
this solution is introduced into a Schlenk tube under an inert
(C
O); 140.04 (C
); 128.43 (CH ); 115.04
quat arom
quat arom
arom
(CH ); 69.26 (CH OC); 63.5 (CH OH); 42.89 (C CH );
arom
2
2
quat
2
2
42.66 (C CH ); 34.69 (CCH CH ); 27.7 (CH CH O); 24.5
quat
(CH CH O).
2
2
2
2
atmosphere, then freshly distilled SiMe Cl (4.4 mL, 22.7
3
mmol) and Et N (2.4 mL, 38.5 mmol) are added. The reaction
3
2
2
mixture is stirred for 2 days at room temperature, then the
81-Mesylate dendrimer 12 (Scheme 1). 12 is synthesized
from 11 (0.60 g, 0.0059 mmol) according to the same pro-
cedure as that described above for 6b and 9b, which gives 12
as a colorless oil (0.082 g, 0.0050 mmol, 85% yield). 1H NMR
solvent is removed under vacuum, the residue is dissolved in
10 mL of anhydrous MeCN; SiMe Cl (2.2 mL, 11.3 mmol)
3
and NaI (5.7 g, 38 mmol) are added. The solution is kept for
one 1 day at 40 ¡C under nitrogen, then the solvent is removed
under vacuum and the residue is extracted with 100 mL
dichloromethane, this solution is washed with 3 ] 50 mL of
an aqueous solution saturated with Na S O and dried over
(CDCl ) d 7.3 and 6.7 (m, C H , 144H); 4.9 (m, CH OC
arom
72H); 3 (m, CH OSO , 162H); 2.3 (m, SO CH , 243H); 1.7
,
3
6
4
2
2
2
2
3
(m, CCH CH , 234H); 1.3 (m, CCH CH , 234H). 13C NMR
2
2
2
2
(CDCl ) d 157.9 (C
O); 137.8 (C
); 127.6 (CH );
2 2
3
3
quat arom
quat arom
arom
sodium sulfate. After Ðltration, the 27-iodo dendrimer 9a is
obtained as a pale yellow solid by precipitation from this solu-
tion upon addition of 200 mL ether (0.330 g, 0.055 mmol, 43%
yield). 1H NMR (CDCl ) d 7.40 (s, C H , 3H); 7.20 and 6.75
114.1 (CH ); 70.6 (CH SO ); 68.3 (CH OC); 41.6
(C CH ); 37.7 (SO CH ); 32.4 (CCH CH ); 23.9
quat
(CH CH O).
arom
2
3
2
2
2
2
2
2
2
2
3
6
3
(m, C H , 36H); 3.78 (m, CH OC
, 18H); 3.1 (m, CH I,
243-Allyl dendrimer 13 (Scheme 1). 13 is synthesized from
12 (0.082 g, 0.005 mmol) and triallylmethyl phenol (0.327 g,
1.43 mmol) using the same procedure as that described above
for the synthesis of the 27-allyl dendrimer 7. 13 is separated by
chromatography on a neutral alumina columns using a
pentaneÈether mixture (70 : 30) as the eluent, which gives 13
as a light brown oil (18 mg, 0.000 66 mmol, 11% yield). 1H
6
4
2
arom
2
54H); 1.92 (m, CCH CH , 72H); 1.60 (m, CCH CH , 72H).
2
2
2
2
13C NMR (CDCl ) d 158.3 (C
O); 148.25 (C
);
3
quat arom
quat arom
140.27 (C
); 128.60 (CH ); 124.69 (CH ); 115.24
quat arom
arom
arom
(CH ); 69.50 (CH OC); 42.94 (C CH ); 42.9 (C CH );
arom quat quat
34.9 (CCH CH ); 27.96 (CH CH O); 25.28 (CH CH O); 7.6
2
2
2
2
2
2
2
2
2
(CH I).
2
New J. Chem., 2000, 24, 351È370
367

View Article Online
NMR (CDCl ) d 7.22 and 6.79 (m, C H , 144H); 5.7È5.51 (m,
orange oil (2.85 g, 4.27 mmol, 94%). 1H NMR (CDCl ) d 7.21
3
6
4
3
CH CH , 81H); 5.04È4.99 (m, CHCH , 162H); 3.91 (m,
and 7.01 (m, C H , 4H); 3.06 (t, CH I, 6H); 2.42 (q,
2
2
2
6
4
2
CH O, 72H); 2.39 (d, CH CHCH , 162H); 1.95 (m,
CH CH CO, 2H); 1.71 (m, CH CH I, 6H); 1.58 (m,
2
2
2
3
2
2
2
CCH CH , 72H); 1.65 (m, CH CH O, 72H). 13C NMR
(CDCl ) d 156.71 (OC
CH C , 6H); 1.11 (t, CH CH CO, 3H). 13C NMR (CDCl )
d 172.83 (CO); 148.93 (C
127.15 (CH ); 121.33 (CH ); 42.17 (C CH ); 38.58
(CH CH I); 27.85 (CH CH CO); 27.5 (CH C ); 9.2
2
2
2
2
2 quat
3
2
3
C);
H ); 137.56 (C
H ); 134.61
CO); 143.13 (C
3
6 quat
4
4
6 quat
4
quat arom
quat arom
2
(CH CH); 127.47 (C H ); 117.30 (CHCH ); 113.70 (C H );
2
6
2
6
4
arom
arom
quat
68.09 (CH O); 43.15 (C CH ); 42.50 (C CH ); 42
quat quat
(CH CHCH ); 33.77 (CH CH O); 23.69 (CCH CH ).
2
2
2
2
2
3
2
2 quat
(CH CH CO); 7.84 (CH I).
2
2
2
2
2
2
3
2
2
p-[1,1-Di(2-propenyl)-3-butenyl]methoxybenzene (or p-
triallylmethyl-methoxybenzene), 14 (Scheme 2). Tri-
allylmethylphenol 3 (1.9 g, 8.32 mmol), methyl iodide (4 mL,
30 mmol), KOH (3.6 g, 64 mmol) and 7 mL DMSO are intro-
duced under nitrogen into a Schlenk tube. The reaction
mixture is stirred for 2 h at room temperature in the dark,
then the excess of methyl iodide is removed under vacuum, the
residue is extracted with 3 ] 20 mL ether, this solution is
washed with Na S O and dried over sodium sulfate. This
9-Allylphenol dendron 18 [eqn. (12)]. Triallylmethylphenol 3
(4.09 g, 25 mmol) dissolved in 25 mL of freshly distilled DMF,
then potassium carbonate (3.5 g, 25 mmol) are introduced into
a Ñamed Schlenk tube, and the mixture is stirred 30 min.
under nitrogen at room temperature. The protected tri-iodo
dendron 17 (2.85 g, 4.26 mmol), dissolved in 10 mL of anhy-
drous DMF, is slowly added, and the reaction mixture is
stirred for 2 days at room temperature, then 3 g potassium
carbonate and 5 mL water are added, and the reaction
mixture is stirred for 12 h at 40 ¡C. The product is extracted
with 100 mL ether, this solution is dried over sodium sulfate,
then Ðltered and concentrated to 5 mL. The product is puri-
Ðed by chromatography on a silica-gel column using pentaneÈ
ether (4 : 1) as the eluent. The 9-allyl phenate dendron 18 is
obtained in this way as a light brown oil (1.2 g, 1.31 mmol,
2 2
3
gives 14 as a light brown oil (1.80 g, 7.44 mmol, 90% yield).
1H NMR (CDCl ) d 7.20 and 6.85 (m, CH
, 4H); 5.68È5.55
3
arom
(m, CHCH , 9H); 5.07È5 (m, CHCH , 18H); 3.78 (s, CH O,
2
2
3
3H); 2.44 (d, CH , 18H). 13C NMR (CDCl ) d 156.10
2
3
(C
O); 137.29 (C
C); 134.22 (CHCH ); 127.23
quat arom
quat arom
2
(CH CO); 117.08 (CHCH ); 112.21 (CH C); 54.78 (CH );
arom
42.25 (C CH ); 41.54 (CH ). Anal. calcd for C
quat
84.25; H 9.15. Found C 83.84; H 9.05.
2
arom
3
H
O: C
30% yield). 1H NMR (CDCl ) d 7.18 and 6.83 (m, CH
,
2
2
17 22
3
arom
16H); 5.68È5.55 (m, CHCH , 9H); 5.07È5 (m, CHCH , 18H);
2
2
3.9 (m, CH O, 6H); 2.44 (d, CH , 18H); 1.9 (m, C CH ,
2
2
quat
2
p-[1,1-Di(3-hydroxypropyl)-4-hydroxybutyl]methoxybenzene,
15 (Scheme 2). The hydroboration-oxidation of 14 (1.8 g, 7.43
mmol) is carried out as described above for the synthesis of
the 27-alcohol dendrimer 8 from the oleÐnic precursor. This
gives 15 as a colorless oil (1.9 g, 6.4 mmol, 86% yield). 1H
NMR (CDCl ) d 7.22 and 6.9 (m, C H , 4H); 3.74 (s, CH O,
6H); 1.65 (m, CH CH , 6H). 13C NMR (CDCl ) d 156.87
2
2
3
(C
(C
O); 153.55 (C
OH); 138.52 (C
C); 137.77
quat arom
quat arom
arom
quat arom
quat arom
C); 134.76 (CHCH ); 127.91 (CH COH); 127.73
2
arom
(CH COH); 117.63 (CHCH ); 115.19 (CH C); 114.35
(CH C); 68.40 (CH O); 42.77 (C CH ); 42.15 (CH CH),
arom quat
33.87 (CH CH O); 23.81 (CCH CH ). MALDI-TOF MS:
2
arom
2
2
2
3
6
4
3
2
2
2
2
3H); 3.47 (m, CH OH, 6H); 1.70 (m, CH CH OH, 6H); 1.3
[M ] Na]` 935.60 (calcd 936.33).
2
2
2
(m, CH C
,
6H). 13C NMR (CDCl )
d
156.76
2 quat
3
(C
OH); 137.32 (C
C); 127.75 (CH ); 114.80
19-Allylphenol dendron 19 (Scheme 3). The dendron 19 is
prepared from the nona-allyl phenol dendron 18 (0.750 g, 0.8
mmol), the protected tri-iodo dendron 17 (0.100 g, 0.15 mmol),
potassium carbonate (0.120 g, 0.8 mmol) and 6 mL DMF
using the procedure described above for the synthesis of the
nona-allylphen dendron 18. The reaction product 19 is puri-
Ðed by chromatography on a silica gel column using a
pentaneÈether mixture (4 : 1) as the eluent, which gives 19 as a
quat arom
quat arom
arom
(CH ); 63.24 (CH ); 53.41 (CH O); 42.37 (C H ); 34.58
arom
(C CH ); 27.89 (CH CH O). Anal. calcd for C
quat 17 28
68.89; H 9.52. Found C 68.9; H 8.67.
3
2
quat
2
H
O : C
2
2
2
4
p-[1,1-Di(3-iodopropyl)-4-iodobutyl]phenol, 16 (Scheme 2).
The triol dendron 15 (1.5 g, 5.06 mmol), 5 mL of anhydrous
MeCN, then dry NaI (6 g, 40 mmol) and freshly distilled
SiMe Cl (5 mL, 39 mmol) are successively introduced into a
light brown oil (0.110 g, 0.037 mmol, 25%). 1H NMR (CDCl )
3
3
Schlenk tube under nitrogen. The closed reaction medium is
d 7.25 and 6.84 (m, CH
, 36H); 5.57 (m, CHCH , 19H);
arom
2
left for 2 days at 80 ¡C in the dark, then the solvent is removed
under vacuum and the residue is extracted with 2 ] 100 mL
ether, the ether solution is washed with an aqueous solution
saturated with Na S O , dried over sodium sulfate and Ðl-
5.06È4.99 (m, CHCH , 38H); 3.92 (m, CH O, 16H); 2.46 (d,
2
2
CH , 38H); 1.87 (m, C CH , 16H); 1.65 (m, CH CH ,
2
quat
3
2
2
2
16H). 13C NMR (CDCl ) d 156.88 (C
O); 153.48
quat arom
C); 134.71
(C
OH); 138.61 (C
C); 137.77 (C
2 2
3
quat arom
quat arom
quat arom
tered. The solvent is removed under vacuum, which gives 16
(CHCH ); 127.63 (CH COH); 117.50 (CHCH ); 115.04
arom
(CH C); 113.91 (CH C); 68.24 (CH O); 42.7 (C CH );
arom arom quat
42.15 (CH CH), 33.83 (CH CH O); 23.78 (CCH CH ).
2
2
as a pale yellow oil (3 g, 4.90 mmol, 97%). 1H NMR (CDCl )
d 7.10 and 6.81 (m, C H , 4H); 3.06 (t, CH I, 6H); 1.64 (m,
3
2
2
2
6
4
2
2
2
2
2
CH CH I, 6H); 1.54 (m, CH C , 6H). 13C NMR (CDCl ) d
MALDI-TOF MS [M ] Na]` 2077.70 (calcd 2078.02).
2
2
2 quat
OH); 137.59 (C
quat arom quat arom
3
C); 127.40 (CH );
arom
153.76 (C
115.40 (CH ); 41.8 (C CH ); 38.58 (CH CH I); 27.69
27-Allyl dendrimer 20 (Scheme 4). The synthesis of 20 from
the nona-allylphenol dendron 18 (0.200 g, 0.218 mmol), 2,4,6-
tris(bromomethyl)mesitylene (0.018 g, 0.004 56 mmol), pot-
assium carbonate (0.045 g, 0.325 mmol) in 10 mL acetone is
achieved according to the procedure described above for the
synthesis of the 27-allyl dendrimer 7 from the nona-iodo core
6a. The reaction product is puriÐed by chromatography on a
silica gel column using dichloromethane as the eluent, which
gives 20 as a light brown oil (40 mg, 0.0138 mmol, 30% yield).
1H NMR (CDCl ) d 7.24 and 6.82 (m, C H , 36H); 5.4 (m,
arom
(CH C ); 8.45 (CH I). Anal. calcd for C
2 quat
H 3.31. Found C 30.85; H 3.03.
quat
2
2
2
H
I O: C 31.55;
2
16 23 3
sp-[1,1-Di(3-iodopropyl)-4-iodobutyl]phenolpropionate, 17
(Scheme 2). A solution of C H COI is prepared by slowly
2
5
adding 2.4 mL C H COCl to 5 g of Me SiI, stirring 15 min.,
2
5
3
then adding 22.6 mL dichloromethane. This solution is added
into a Schlenk Ñask containing 16 (2.7 g, 4.52 mmol) in 30 mL
dichloromethane, then NEt Pri (0.84 mL, 4.83 mmol) is also
2
3
6 4
added. The reaction mixture is stirred under nitrogen for 16 h,
CHCH , 27H); 4.98 (m, CHCH , 54H); 3.8 (m, CH O, 18H);
2
2
2
then extracted using 50 mL dichloromethane. The organic
solution is washed with a saturated aqueous solution of
sodium bicarbonate, then with an aqueous solution saturated
with Na S O and dried over sodium sulfate. After removing
2.41 (s, CH , 9H); 2.38 (d, CH CHCH , 54H); 1.87 (m,
CCH CH , 18H); 1.58 (m, CH CH O, 18H). 13C NMR
(CDCl ) d 156.87 (OC
(CH CH); 127.63 (C H ); 117.5 (CHCH ); 113.85 (C H );
68.23 (CH O); 42.69 (C CH ); 41.97 (CH CHCH ); 33.82
quat
3
2
2
2
2
2
2
H ); 137.56 (C
H ); 134.7
3
6 quat
4
4
6 quat
4
2 2
3
2
6
2
6 4
the solvent under vacuum, the product 17 is obtained as an
2
2
2
2
368
New J. Chem., 2000, 24, 351È370

View Article Online
8
9
(a) G. R. Newkome, C. N. MooreÐeld and F. Vogtle, Dendritic
Molecules: Concepts, Syntheses and Perspectives, VCH, New
York, 1996; (b) Advances in Dendritic Macromolecules, ed. G.
Newkome, JAI, Greenwich, CT, 1994, vol. 1; 1995, vol. 2; 1996,
vol. 3; 1999, vol. 4; (c) T op. Curr. Chem., ed. F. Vogtle, SpringerÈ
Verlag, Berlin, 1998, vol. 197; 2000, vol. 210.
(a) F. Moulines, L. Djakovitch, M.-H. Delville, F. Robert, P.
Gouzerh and D. Astruc, J. Chem. Soc., Chem. Commun., 1995,
463; (b) F. Moulines, L. Djakovitch and D. Astruc, New J. Chem.,
1996, 20, 1071; (c) for ion-pairing aspects and salt e†ects, see A.
Loupy, B. Tchoubar and D. Astruc, Chem. Rev., 1992, 92, 1141.
(CH CH O); 29.73 (CH ); 23.77 (CCH CH ). MALDI-TOF
MS [M ] Na]` 2919.05; [M ] K]` 2935.96 (calcd
[M ] Na]` 2919.9; [M ] K]` 2935.89).
2
2
3
2
2
54-Allyl dendrimer 21 (Scheme 4). An ethanol solution (10
mL) of the nona-allyl dendron 18 (0.380 g, 0.416 mmol) and
K CO (0.060 g, 0.428 mmol) was stirred at room temperature
2
3
for 1 h in a Ñamed and deaerated Schlenk tube. Then, hexa-
kishexabromomethylbenzene (0.040 g, 0.063 mmol) was added
to the mixture, and the latter was heated at reÑux for 2 days.
After evaporation of the solvent under vacuum, the residue
was dissolved in a Et OÈH O mixture (50 mL : 50 mL). The
10 This article mostly overlaps with the thesis of Dr Valerie Sartor,
Universite Bordeaux I, 1998. For a preliminary communication,
see: V. Sartor, L. Djakovitch, J.-L. Fillaut, F. Moulines, F.
Neveu, V. Marvaud, J. Guittard, J.-C. Blais and D. Astruc, J. Am.
Chem. Soc., 1999, 121, 2929.
11 (a) J.-L. Fillaut and D. Astruc, New J. Chem., 1996, 20, 945; (b) V.
Marvaud, D. Astruc, E. Leize, A. Van Dorsselaer, J. Guittard and
J.-C. Blais, New J. Chem., 1997, 21, 1309; (c) E. Alonso, C.
Valerio, J. Ruiz and D. Astruc, New J. Chem., 1997, 21, 1139.
12 (a) C. Valerio, J.-L. Fillaut, J. Ruiz, J. Guittard, J.-C. Blais and D.
Astruc, J. Am. Chem. Soc., 1997, 119, 2588; (b) C. Valerio, E.
Alonso, J. Ruiz, J.-C. Blais and D. Astruc, Angew. Chem., Int. Ed.,
1999, 38, 1747.
13 (a) N. A. Nesmeyanov, N. A. VolÏkenau and I. N. Bolesova, Dokl.
Akad. Nauk SSSR, 1967, 175, 606; (b) D. Astruc, T etrahedron,
1983, 39, 4027.
14 (a) F. Moulines, L. Djakovitch, R. Boese, B. Gloaguen, W. Thiel,
J.-L. Fillaut, M.-H. Delville and D. Astruc, Angew. Chem., Int.
Ed. Engl., 1993, 32, 1075; (b) D. Astruc, F. Moulines, B. Gloaguen
and L. Toupet, in preparation.
2
2
organic phase was dried over Na SO , Ðltered, and the
2
4
solvent was removed under vacuum. The crude product was
puriÐed by chromatography and obtained as a white solid in
83% yield (0.295 g, 0.052 mmol). 1H NMR (CDCl ) d 7.18 (m,
3
CH
, 48H); 6.80 (m, CH
, 48H); 5.52 (m, CHCH , 54H);
arom
arom
2
5.24 (s, PhCH O, 108H); 4.99 (m, CHCH , 108H); 3.80 (m,
CH O, 36H); 2.41 (d, CH CHCH , 108H); 1.77 (m,
2
2
2
2
2
2
CCH CH , 36H), 1.56 (m, CH CH O, 36H). 13C NMR
2
2
2
(CDCl ): d 156.87 (OC
H ); 156. (OC
H ); 137.55 (C
H ); 139.22
H ); 134.66
3
6 quat
4
6 quat
4
(C
H ); 137.98 (C
6 quat
4
6 quat
4
6 quat
4
(CH CH); 127.58 (C H ); 117.49 (CHCH ); 114.49 (C H );
2
6
4
2
6 4
113.64 (C H ); 68.15 (CH O); 63.68 (CH O); 42.65
6
4
2
2
(C CH ); 42.06 (C CH ), 41.95 (CH CHCH ); 33.84
quat quat
(CH CH O); 23.77 (CCH CH ). MALDI-TOF MS:
2
2
2
2
2
2
2
2
[M ] Na]` 5654.1 (calcd 5653.3). Anal. calcd for
15 D. Catheline and D. Astruc, J. Organomet. Chem., 1984, 272, 417.
16 I. Bertini, H. B. Gray, S. J. Lippard and J. S. Valentine, Bio-
inorganic Chemistry, University Science Books, Mill Valley, 1994.
17 MALDI-TOF MS: S. A. Hofstadler, R. Bakhtiar and R. D.
Smith, J. Chem. Educ., 1979, 57, 1887.
C
H
O
: C 84.47; H 8.71. Found C 83.47; H 8.66.
396 486 24
Acknowledgements
18 (a) A. Pelter, K. Smith and H. C. Brown, Borane Reagents. Best
Synthetic Methods, Academic Press, New York, 1988; (b) using
the large disiamylborane and 9-BBN (9-borabicyclo[3.3.1]
nonane), Brown and colleagues have shown that the anti-
Markovnikov regioselectivity of the hydroboration reaction of
terminal oleÐns is superior to 99%, see ref. 18(a), p. 182. We have
therefore followed this procedure with disiamylborane since the
oxidation products (a primary and a secondary alcohol, which
are light) are easier to remove than with 9-BBN; 18c (c) H. C.
Brown, C. Snyder, B. C. S. Rao and J. Zweifel, T etrahedron, 1986,
42, 5505.
The Institut Universitaire de France (IUF), the Centre
National de la Recherche ScientiÐque (CNRS), the Uni-
versities Bordeaux I and Paris VI, the Region Aquitaine,
Rhone-Poulenc (thesis grant to L. D.) and the Ministere de la
Recherche et de la Technologie (MRT, thesis grant to J. G.)
are gratefully acknowledged for Ðnancial support. J.-C. B.
undertook the mass spectrometric studies and F. N. is a Ðnal-
year student from Polytechnique, ““promotionÏÏ 1998, Palai-
seau, France.
19 G. Olah, S. C. Narang, B. G. B. Gupta and R. Malhotra, J. Org.
Chem., 1979, 44, 1247.
Notes and references
20 Hexa(iodobutyl)benzene was previously prepared by the more
tedious hydrozirconation and hydroboration routes: F. Mou-
lines, L. Djakovitch, J.-L. Fillaut and D. Astruc, Synlett, 1992, 57.
21 J. March, Advanced Organic Chemistry, Wiley, New York, 4th
edn., 1992, pp. 386 and 1003.
22 H. Feuer and J. Hooz, in T he Chemistry of the Ether L inkage, ed.
S. Patai, Wiley, New York, 1967, ch. 10, p. 445; D. V. Banthorpe,
Reaction Mechanisms in Organic Chemistry. Substitution Reac-
tions, Elsevier, Amsterdam, 1963, p. 49.
23 Inorganic Ñuoride salts as bases have been reported to give the
best results in the competition between substitution and elimi-
nation for the alkylation of phenols: (a) J. M. Miller, K. H. So
and J. H. Clark, Can. J. Chem., 1979, 57, 1887; (b) T. Ando, J.
Yamawaki, T. Kawate, S. Sumi and T. Hanafusa, Bull. Chem. Soc.
Jpn., 1982, 55, 2504.
24 (a) W. H. Kruizinga, B. Strijveen and R. M. Kellog, J. Org.
Chem., 1981, 46, 4321; (b) A. B. Padias, H. K. Hall, D. A. Tomalia
and J. R. McConnell, J. Org. Chem., 1987, 52, 5305; (c) M. K.
LothianÈTomalia, D. M. Hedstrand, D. A. Tomalia, A. B. Padias
and H. K. Hall, T etrahedron, 1997, 53, 15495.
1
J. P. Collman, L. S. Hegedus, J. R. Norton and R. G. Finke,
Principles and Applications of Organotransition-Metal Chemistry,
University Science Books, Mill Valley, CA, 1987, ch. 8.
2
(a) M. F. Semmelhack, J. Organomet. Chem. L ibrary, 1976, 1, 361;
(b) Pure Appl. Chem., 1981, 53, 2379; (c) T etrahedron, 1981, 37,
3957; (d) Ann. N.Y . Acad. Sci., 1977, 295, 36; (e) in Comprehensive
Organic Chemistry, ed. B. M. Trost, Pergamon Press, Oxford,
1991, vol. 4, p. 517.
3
(a) G. Jaouen, Ann. N.Y . Acad. Sci., 1977, 295, 59; (b) H. Des
Abbayes and M.-A. Boudeville, J. Org. Chem., 1977, 42, 4104; (c)
E. P. Kundig, Pure Appl. Chem., 1985, 57, 1855; (d) F.
RoseÈMunch and E. Rose, J. Organomet. Chem., 1989, 377, C9;
(e) M.-C. SenechalÈTocquer, D. Senechal, J.-Y. Le Bihan, D.
Gentric and B. Caro, Bull. Soc. Chim. Fr., 1992, 114, 8288; ( f ) F.
RoseÈMunch, E. Rose, V. Gagliardini and C. Renard, Coord.
Chem. Rev., 1988, 178–180, 249; (g) F. RoseÈMunch and E. Rose,
Curr. Org. Chem., 1999, 3, 445.
4
5
For a review comparing the di†erent activating organo
transition-metal groups and purely organic reactions, see: L.
Balas, D. Jhurry, L. Latxague, S. Grelier, Y. Morel, M.
Hamdami, N. Ardoin and D. Astruc, Bull. Soc. Chim. Fr., 1990,
127, 401.
Theoretical aspects: (a) S. G. Davies, M. Green and D. M. P.
Mingos, T etrahedron, 1978, 34, 20; (b) O. Eisenstein and R. H.
Ho†man, J. Am. Chem. Soc., 1981, 103, 4308; (c) D. Astruc, P.
Michaud, M. Madonik, J.-Y. Saillard and R. H. Ho†man, Nouv.
J. Chim., 1985, 9, 41.
(a) D. Astruc, New J. Chem., 1992, 16, 305; (b) T op. Curr. Chem.,
1991, 160, 47; (c) Acc. Chem. Res., in press.
H. J. Trujillo, C. M. Casado, J. Ruiz and D. Astruc, J. Am. Chem.
Soc., 1999, 121, 5674.
25 R. K. Muller, R. Joos, D. Felix, J. Schreiber, C. Wintner and A.
Eschenmoser, Organic Synthesis, Wiley, 1988, vol. 6, p. 56.
26 K. K. Andersen, in Comprehensive Organic Chemistry, ed. B. M.
Trost, Pergamon, Oxford, 1991, vol. 3, p. 341.
27 E. J. Corey and B. B. Snider, J. Am. Chem. Soc., 1972, 94, 2549.
28 The conversion of alkoxytrimethylsilanes into iodoalkyl deriv-
atives must be carried out at 40 ¡C, not at higher temperatures.
At 80 ¡C, both this reaction and the cleavage of aryl ethers in the
6
7
presence of NaI and Me SiCl have been reported: T. Morita, S.
3
Yoshida, H. Okamoto and H. Sakurai, Synthesis, 1979, 379.
29 R. A. W. Johnstone and M. E. Rose, T etrahedron, 1979, 35,
2169.
New J. Chem., 2000, 24, 351È370
369

View Article Online
30 T. W. Greene, Protective Groups in Organic Chemistry, Wiley,
disperse: S. M. Aharoni III, C. R. Crosby III and E. K. Walsh,
Macromolecules, 1982, 15, 1093.
New York, 1979, ch. 3, p. 87.
31 Other known phenol-protecting groups using PhCH Br,31a
45 (a) D. A. Tomalia, H. Baker, J. Dewald, M. Hall, G. Kallos, S.
Martin, J. Roeck, J. Ryder and P. Smith, Polym. J., 1985, 17, 117;
(b) for superb and seminal reviews, see: D. A. Tomalia, A. N.
Naylor III and W. A. Goddard, Angew. Chem., Int. Ed. Engl.,
1990, 29, 138; D. A. Tomalia and H. Dupont Durst, T op. Curr.
Chem., 1993, 165, 193.
2
PPh Li,31b CF CO H,31c and H /Pd/C31d did not allow a regio-
2
3
2
2
selective deprotection of the phenol function in the present case:
(a) M. C. Venuti, B. E. Loc, G. H. Jones and J. M. Young, J.
Med. Chem., 1988, 31, 2132; (b) G. Veriot and A. Collet, Acros
Org. Acta, 1995, 1, 40; (c) M. W. Bhatt and S. U. Kulkarni, Syn-
thesis, 1983, 249; (d) K. L. Wooley, C. J. Hawker and J. M. J.
Frechet, J. Am. Chem. Soc., 1991, 113, 4252.
46 G. R. Newkome, Z. Q. Yao, G. R. Baker and V. K. Gupta, J.
Org. Chem., 1985, 50, 2003; (b) for a seminal review, see: G. R.
Newkome, C. N. MooreÐeld and G. R.Baker, Aldrichimica Acta,
1992, 25, 31.
47 (a) C. J. Hawker and J. M. J. Frechet, J. Chem. Soc., Chem.
Commun., 1990, 1010; (b) for a seminal review, see: J. M. J.
Frechet, Science, 1994, 263, 1710.
48 (a) T. M. Miller and T. X. Neenan, Chem. Mater., 1990, 2, 346; (b)
another pioneering work in convergent dendritic construction is:
J. S. Moore and Z. Xu, Macromolecules, 1991, 24, 5893.
49 (a) S. Serroni, G. Denti, S. Campagna, A. Juris, M. Ciano and
V. Balzani, Angew. Chem., Int. Ed. Engl., 1992, 31, 1493; (b)
for a concise review, see: V. Balzani, S. Campagna, G. Denti,
A. Juris, S. Serroni and M. Venturi, Acc. Chem. Res., 1998, 31,
26.
32 Propionyl iodide is synthesized by reaction of propionyl chloride
with trimethylsilyl iodide: A. H. Schmidt, M. Russ and D.
Grosse, Synthesis, 1981, 216.
33 (a) F. RoseÈMunch, J.-P. Djukic and E. Rose, T etrahedron L ett.,
1990, 31, 2589; (b) J.-P. Djukic, F. RoseÈMunch, E. Rose, F.
Simon and Y. Dromzee, Organometallics, 1995, 14, 2027.
34 M. Kimura, M. Morita, H. Mitani, H. Okamoto, K. Satake and
S. Morosawa, Bull. Soc. Chim. Jpn., 1992, 62, 2557.
35 (a) E. Staude and F. Patat, in T he Chemistry of the Ether L inkage,
ed. S. Patai, Wiley, New York, 1967, p. 46; (b) A. Maercher,
Angew. Chem., Int. Ed. Engl., 1987, 26, 972; (c) E. Bunsel, J. M.
Dust, A. Jonczyk, R. A. Manderville and I. Onyido, J. Am. Chem.
Soc., 1992, 114, 5610.
36 (a) D. J. ColeÈHamilton, R. J. Young and G. Wilkinson, J. Chem.
Soc., Dalton T rans., 1976, 1995; (b) J. F. Helling and W. A. Hen-
drickson, J. Organomet. Chem., 1979, 168, 87; (c) U. Koelle, M.
Hong Wang and G. Raabe, Organometallics, 1991, 10, 2573; (d) B.
Chaudret, Z. X. He and Y. Huang, J. Chem. Soc., Chem.
Commun., 1989, 1844; (e) H. El Amouri, M. Gruselle, P. A.
Jackson, G. Jaouen and J. Vaisserman, Organometallics, 1990, 9,
2871; ( f ) J. Le Bras, H. El Amouri, Y. Besace, J. Vaisserman and
G. Jaouen, Bull. Soc. Chim. Fr., 1995, 132, 1073.
50 (a) G. R. Newkome, E. He and C. N. MooreÐeld, Chem. Rev.,
1999, 99, 1689; (b) M. A. Hearshaw and J. R. Moss, Chem.
Commun., 1999, 1.
51 (a) C. M. Casado, I. Cuadrado, M. Moran, B. Alonso, B. Garcia,
B. Gonzales and J. Losada, Coord. Chem. Rev., 1999, 185–186,
53; (b) I. Cuadrado, M. Moran, C. M. Casado, B. Alonso and J.
Losada, Coord. Chem. Rev., 1999, 189, 123.
52 J.-P. Majoral and A.-M. Caminade, Chem. Rev., 1999, 99, 845.
53 N. Ardoin and D. Astruc, Bull. Soc. Chim. Fr., 1995, 132, 875.
54 (a) S. Achar and R. J. Puddephatt, Angew. Chem., Int. Ed. Engl.,
1994, 33, 847; (b) S. Achar and R. J. Puddephatt, J. Chem. Soc.,
Chem. Commun., 1994, 1895.
55 (a) A. Archut, F. Vogtle, L. De Cola, G. C. Azzellini, V. Balzani,
P. S. Ramanujam and R. H. Berg, Chem. Eur. J., 1998, 4, 699; (b)
P. Lange, H. Schier and H. Schmidbauer, Inorg. Chem., 1996, 35,
637.
56 (a) Tomalia45b and Newkome56b have introduced the notion of
monodisperse vs. polydisperse dendrimer. Indeed, there is a con-
tinuum of situations; (b) G. R. Newkome, C. D. Weiss and C. N.
MooreÐeld, Macromolecules, 1997, 30, 2300.
57 P.-G. De Gennes and H. Hervet, J. Phys. L ett., 1983, 44, L351.
58 (a) R. L. Lescanec and M. Muthukumar, Macromolecules, 1991,
24, 4892; (b) M. L. MansÐeld and L. I. Klushin, Macromolecules,
1993, 26, 4262.
59 A dendron of closely related structure, 4-triallylsilylphenol, has
recently been independently reported: (a) R. A. Gossage, E.
MunozÈMartinez and G. van Koten, T etrahedron L ett., 1998, 39,
2397; (b) for the Ðrst reports of the triallylphenol dendrons 2 and
3, see ref. 9(a) and L. Djakovitch, Ph.D. Thesis, Universite Bord-
eaux I, 1995.
60 (a) I. U. Khand, P. L. Pauson and W. E. Watts, J. Chem. Soc., C,
1968, 2261; (b) for a review on the synthesis of [FeCp(g6-
arene)]` complexes, see ref. 13(b).
37 D. Astruc, J.-R. Hamon, E. Roman and P. Michaud, J. Am.
Chem. Soc., 1981, 103, 7502.
38 D. Astruc, Electron-T ransfer and Radical Processes in T ransition-
Metal Chemistry, VCH, New York, ch. 3.
39 J.-R. Hamon, P. Hamon, K. Boukheddaden, J. Linares, F. Varret
and D. Astruc, Inorg. Chim. Acta, 1995, 240, 105.
40 The 19-electron FeI complex of a phenate is more stable than the
19-electron complex of an oxocyclohexadienyl species having its
negative charge on the metal center: for comparison with methy-
lenecyclohexadienyl complexes, see ref. 7, which discusses the
thermodynamic data.
41 J. H. R. Tucker, M. Gingras, H. Brand and J.-M. Lehn, J. Chem.
Soc., Perkin T rans. 2, 1997, 1303., .
42 The divergent method was pioneered by the groups of Vogtle,43
Denkewalter,44 Tomalia45 and Newkome.46 Later, the con-
vergent route was reported independently by the groups of
Frechet47 and Miller (and presumably also by Moore and Xu).48
The Ðrst inorganic dendrimers were reported by BalzaniÏs
group.49 For recent reviews by pioneers on organometallic den-
drimers, see refs. 50 and 51. For a recent review on heteroatom-
containing dendrimers, see ref. 52. For a comprehensive review
on dendrimers, see ref. 53.
43 (a) E. Buhleier, W. Wehner and F. Vogtle, Synthesis, 1978, 155;
(b) for an early ““highlightÏÏ, see: H.-B. Mekelburger, W. Jaworek
and F. Vogtle, Angew. Chem., Int. Ed. Engl., 1992, 31, 1571.
44 R. G. Denkewalter, J. F. Kolc and W. J. Lukasavage, US Pat.
4 410 688, 1983; Chem. Abstr., 1984, 100, 103907. The dendrimers
reported in this patent involve a 1 ] 2 asymmetric branching
pattern and the incorporation of multiple chiral centers at each
tier. No characterization support the concept in the patent, but
these lysine dendrimers were studied and reported to be mono-
61 (a) A. N. Nesmeyanov, N. A. VolÏkenau and I. N. Bolesova, Dokl.
Akad. Nauk SSSR, 1963, 149, 615; (b) see ref. 60(b).
62 D. Astruc, J.-R. Hamon, M. Lacoste, M.-H. Desbois and E.
Roman, in Organometallic Synthesis, ed. R. B. King, Elsevier,
Amsterdam, 1988, vol. IV, p. 172.
63 H. C. Brown, M. W. Rathke, M. M. Rogic and N. R. De Luc,
T etrahedron, 1988, 44, 2751.
370
New J. Chem., 2000, 24, 351È370