Towards a biomimetic poly-aminoketone foldamer: synthesis of a triply protected monomer and its coupling to a dimer, trimer and tetramer

Article abstract of DOI:10.1016/j.tet.2006.12.081

The design of a new biomimetic foldamer, relying on the weak amine-carbonyl interaction for secondary structure formation, is presented. The efficient synthesis of a triply protected monomer starting from glycidol was developed. This monomer contains a dioxolane-protected keto group that will allow liberation of the ketone functionality in the backbone once construction of the oligomeric backbone is complete. This monomer contains two additional orthogonal protecting groups at its two termini, the Fmoc and the TBDMS groups. The Fmoc group in particular permits oligomerisation towards the N terminus as seen in Fmoc solid phase peptide synthesis. Construction and full characterisation of a ketone-protected dimer, trimer and tetramer are reported.

Full text of DOI:10.1016/j.tet.2006.12.081

Tetrahedron 63 (2007) 2199–2207
Towards a biomimetic poly-aminoketone foldamer: synthesis
of a triply protected monomer and its coupling to a
dimer, trimer and tetramer
*
Romain Barbe and Jens Hasserodt
´
Laboratoire de Chimie, UMR CNRS 5182, Ecole Normale Superieure de Lyon, 46 Allee d’Italie, 69364 Lyon, France
´
Received 2 November 2006; revised 18 December 2006; accepted 22 December 2006
Available online 4 January 2007
Abstract—The design of a new biomimetic foldamer, relying on the weak amine–carbonyl interaction for secondary structure formation, is
presented. The efficient synthesis of a triply protected monomer starting from glycidol was developed. This monomer contains a dioxolane-
protected keto group that will allow liberation of the ketone functionality in the backbone once construction of the oligomeric backbone is
complete. This monomer contains two additional orthogonal protecting groups at its two termini, the Fmoc and the TBDMS groups. The Fmoc
group in particular permits oligomerisation towards the N terminus as seen in Fmoc solid phase peptide synthesis. Construction and full
characterisation of a ketone-protected dimer, trimer and tetramer are reported.
Ó 2007 Elsevier Ltd. All rights reserved.
1. Introduction
concentration of the metal ion. Even so, in an exceptional
case nature has relied on this phenomenon, namely the struc-
turally reinforcing zinc finger motif.2
The field of bio-inspired polymeric materials has yielded an
astonishing variety of backbone constitutions,1 largely start-
ing from a modification of the peptide backbone. In most of
these studies, the underlying interest is the creation of solu-
ble oligomers that adopt a defined secondary structure in
a given medium, thus showing a particular folding pattern,
which led to the proposal of the term foldamer. Since its in-
ception, this field has yielded many new insights into the way
that the backbone architecture is determining the three-
dimensional structure in solution and in the solid state.
Other weak contacts may not have been looked at for reasons
of stability of interacting functional groups or simply be-
cause of their non-existence under desired environmental
conditions, i.e., aqueous media. The latter constitutes also
a recurrant problem in the field of peptidomimetic oligomers
or even small to medium-sized natural peptides; in fact, the
secondary structure of these compounds is often studied in
chlorinated solvents such as chloroform or in methanol.3
In most of the cases, structural integrity is lost when passing
on to pure aqueous media.4
Foldamers can be subdivided into two distinct classes: those
that exploit high pre-organisation on the level of their mono-
meric units, and those that rely on weak donor–acceptor
interactions between one repetitive unit and another being
distant on the backbone, or on another strand as observed
in peptides or peptidomimetics. Of course, foldamers incor-
porating both strategies have also surfaced. However, while
the structural diversity of the monomers has been enormous
in all of these endeavours, this is not true for the number of
underlying weak bonds. In effect, the choice has been largely
limited to the all-important hydrogen bond. Metal-to-ligand
coordinative bonds, though considered weak and kinetically
labile to a certain extent, cannot principally serve as a re-
placement for H bonds; in their case, the resulting secondary
structure is not determined solely by the properties of the
monomers, but necessitates the presence of a sufficient
We propose here to create a new type of oligomer that does
not rely on the cooperative effect of hydrogen bonding for
folding.5 Rather, for secondary structure formation, our
oligomer will make use of a rarely observed natural weak
bond: the tertiary amine–carbonyl interaction. The d+N/
C]Odꢀ interaction has been described as a through-space
homoconjugation of the n and p molecular orbitals of the
nitrogen and C]O double bond, respectively, where elec-
tron density is shifted from the n orbital to the p* orbital,
giving rise to an enhanced dipole moment and a hypsochro-
mic shift of the carbonyl UVabsorption.6,7 What makes this
weak bond particularly attractive for the purpose of rivalling
hydrogen bonds, and consequently for practical applica-
tions, is its favoured formation in water.8 As a direct con-
sequence the envisaged oligomers may actively adopt their
intrinsic energy-minimised secondary structure upon water
solvation rather than loosing it.9
* Corresponding author. Tel.: +33 472 72 83 94; fax: +33 472 72 88 60;
e-mail: jens.hasserodt@ens-lyon.fr
0040–4020/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2006.12.081

2200
R. Barbe, J. Hasserodt / Tetrahedron 63 (2007) 2199–2207
O
O
N
O
N
O
N
O
N
O
N
O
N
N
O
N
O
N
O
N
O
N
O
N
O
O
δ-
O
δ-
H
δ+
O
O
Ph
Ph
δ−
O
N
δ+
O
δ-
6
5
O
O
N
N
δ+
O
O
H
N
δ+
O
O
O
O
O
O
O
O
N
N
N
N
1
2
3
4
n
7
Figure 1. Examples of compounds displaying the tertiary amine–carbonyl
interaction.
The d+N/C]Odꢀ interaction has been observed for nearly
80 years in a class of alkaloids10 (Fig. 1, 1) and in artificially
made derivatives thereof. Researchers in the 1950s have
performed solution-phase studies on these interactions in
tropane derivatives (Fig. 1, 2),11 before they served as ob-
jects for the estimation of collision trajectories in nucleo-
philic attack on carbonyl groups by way of X-ray structure
analysis in the 1970s.12 Later, sporadic occurrences in the
synthetic literature (Fig. 1, 3 and 4)13 proved that these inter-
actions were by no means limited to alkaloid systems. More
recently we demonstrated that the d+N/C]Odꢀ inter-
action can be incorporated into a peptidomimetic aimed to
inhibit a particular protease.6
O
O
O
O
+
O
N
HN
O
PG1
PG2
8b
8a
O
O
N
O
PG1
PG2
8
Figure 2. Design of retrosynthesis of a new biomimetic poly-aminoketone
foldamer.
obtain as quickly as possible a first molecule for structural
studies. Alternatively and in analogy to the Merrifield pep-
tide synthesis, a polymer-supported monomer-wise built-
up of an oligomer chain may be envisaged. Such an approach
should be particularly fruitful when employing the superior
and already optimised Fmoc methodology15 in peptide syn-
thesis. This approach then allows for the very flexible incor-
poration of a range of different monomers to expand the
technology. During our synthesis development down below,
we thus eventually decided to focus on the application of the
Fmoc group to our synthetic targets. We present here the
efficient synthesis of a triply protected monomer, and its
coupling to a dimer, trimer and tetramer.
2. Results and discussion
In line with the theory of cooperativity as applied to folding
in biopolymers, it is now assumed that the repetition of a mo-
tive composed of a ketone functionality and a tertiary amine
group in a polymer backbone will lead to interactions either
between sites that are distant on the backbone (analogous to
beta-sheet formation in proteins) or in the vicinity of the
backbone (alpha helix). In analogy to the peptide bond
where the amide nitrogen functions as a hydrogen donor
and the amide carbonyl as acceptor, our monomer was cho-
sen so as to possess at the same time a tertiary amine moiety
as donor and a ketone functionality as acceptor. The struc-
ture of our monomer was also conceived so as to allow for
the formation of chair-like six-membered ring elements to
form upon folding in accord with the (pseudo)tetrahedral
geometries that the carbonyl carbon and the nitrogen centre
will adopt (Fig. 2, 5).
2.1. Monomer synthesis
To obtain the target monomer 8 (Fig. 2), we first explored the
use of ethyl bromopyruvate, which is available in large quan-
tities (technical grade), and is triply functionalised (Scheme
1). These advantages appeared ideal for a rapid construction
of the monomer. However, direct protection of the ketone
moiety by a dioxolane moiety turned out to be impossible.
Introduction of the protected secondary nitrogen found in
8 required the use of N-benzyl-N-methylamine, despite the
inherent disadvantages of the benzyl group as a protecting
group on nitrogen. Reaction of 9 with this reactant led to
10a that was found to be sensitive to moisture leading to
the formation of N-benzyl-N-methylformamide. Optimisa-
tion of the conditions (under argon) gave a yield of 63%. Ac-
cording to NMR data, this compound existed in a tautomeric
equilibrium, with the major isomer being the enolic form
10b. Any attempt for its protection, either on the level of
its carbonyl form (leading to 12) or its enol form (leading
to 13), only led to decomposition, so that nucleophilic sub-
stitution as the first step could be ruled out.
In loose analogy to peptide coupling, we then decided to ex-
ploit reductive amination as the principal coupling reaction
to build up the oligomer. We were guided by the well-known
efficacy of this reaction, its applicability to the synthesis on
solid support and the mild conditions it requires so that
protecting groups are not endangered.14 In order to avoid
the risk of precipitation of the oligomer by aggregation we
envisaged grafting two oligomeric chains onto a suitable
moiety acting as beta-turn, thus forcing the backbone to
fold back upon itself (Fig. 2, 6).
In view of the unknown properties of a highly functionalised
monomer in its fully protected or partially protected form,
we anticipated a major challenge in synthetic optimisation
of the underlying chemistry and opted for a convergent syn-
thetic scheme (Fig. 2). A convergent solution-phase synthe-
sis as proposed in Figure 2 is the starting point of choice to
Ketalisation of 9 according to the literature-described proce-
dure16 using methanol instead of ethylene glycol was found
to be a surprisingly effective reaction with a 76% yield,

R. Barbe, J. Hasserodt / Tetrahedron 63 (2007) 2199–2207
2201
H
O
ii
iv
v
O
O
O
OR
RN
iii
OBn
FmocN
OBn
NHMeBn
Bn
Bn
O
i
EtO
N
N
EtO
Br
EtO
10b
OH
O
18a R = H
18b R = Bn
19 R = H
O
21
O
Bn
9
10a
N
20 R = Fmoc
TBSMSCl
base
vii FmocN
FmocN
OH
O
O
O
O
O
O
vi
viii
O
O
FmocN
OBn
TBDMS
O
24
O
23
O
O
EtO
11
Br
Br
EtO
EtO
13
HN
OBn
O
O
O
O
O
22
12
Bn
N
25
O
O
iii
i
ii
Scheme 2. Synthesis of compounds 24 and 25. Reagents and conditions: (i)
BnBr, NaH, DMF, 91%; (ii) NH2Me (40 wt % solution in water), 100%; (iii)
Fmoc Cl, NaHCO3, dioxane/water, 91%; (iv) oxalyl chloride, DMSO,
ET3N, CH2Cl2, ꢀ78 ꢁC, 93%; (v) ethylene glycol, benzene, APTS, reflux,
96%; (vi) BBr3, CH2Cl2, ꢀ78 ꢁC, 76%; (vii) Dess–Martin periodinane,
NaHCO3, CH2Cl2, 40–56%; (viii) Et3N, acetonitrile, 72%.
MeO
Br
HO
Br
EtO
O
O
O
O
O
9
14
15
iv
Bn
HO
Br
HO
N
O
O
16
17
debenzylation leading to alcohol 23 and its subsequent oxi-
dation to aldehyde 24, and the N terminus by Fmoc cleavage
leading to secondary amine 25. Immediately, we were forced
to realise that the high steric hindrance found in 22 signifi-
cantly inhibited palladium-catalysed hydrogenation of the
benzyl group. In fact, classic conditions at 1 bar hydrogen
and 10% palladium on carbon caused no reaction. All at-
tempts under numerous conditions with catalysts such as
Pd/C, PtO2, or Raney nickel were to no avail. When the pres-
sure was raised above 1 bar, the Fmoc group started to get
cleaved. We thus turned to tribromoborane as an alternative
deprotection agent for benzyl-protected alcohols. Usually,
the use of this reagent can be regarded as much more stren-
uous to any given substrate than catalytic hydrogenation.
However, in this particular case the substrate 22 presents
an ethylenglycoldiether motive, and the associated chelating
properties17 are conducive to mild cleavage of the benzyl
group as depicted in Scheme 3. Nonetheless, the reaction ne-
cessitated rigorous optimisation before giving satisfactory
yields: after treatment with BBr3 at ꢀ70 ꢁC and complete
consumption of starting material, the aggressive reactant
had to be destroyed by transferring the entire mixture at
ꢀ78 ꢁC via a cannula into a solution of sodium bicarbonate
at 0 ꢁC. This protocol turned out to be remarkably effective.
The reaction should be stopped before 100% consumption is
achieved, and the desired alcohol 23 (76% yield) can be sep-
arated by column chromatography from remaining 22 (19%
yield), which, in turn, can easily be reintroduced into a sub-
sequent batch ready for deprotection. Thereby achieved
yields of 23 are significantly higher than those reported in
the literature.18
Scheme 1. Synthesis of amino alcohol 17. Reagents and conditions: (i)
MeOH, HC(OMe)3, H2SO4, 76%; (ii) LiAlH4, Et2O, 88%; (iii) acetone,
Amberlyst-15Ò, H2O (0.3 equiv); (iv) HNMeBn, K2CO3 (3 equiv), 43%
(two steps).
and the subsequent reduction of 14 by LiAlH4 gave the
bromoalcohol 15 in 88% yield. However, either flash column
chromatography or distillation led to significant decomposi-
tion of 14 or 15, and the latter had to be introduced into the
nucleophilic substitution as is. As feared, compound 15 was
inert to N-benzyl-N-methylamine, likely because of steric
reasons, and a deprotection prior to nucleophilic substitution
was explored. All classic conditions of ketal cleavage led to
decomposition of 15, but the use of thoroughly washed sul-
fonic acid-based resin Amberlyst-15Ò in acetone was found
to be an exceptionally mild procedure. Ketone 16, being too
delicate for purification, was introduced directly into the re-
action with N-benzyl-N-methylamine and gave 17 in 43%
over two steps. Once again, the obtained compound showed
a pronounced instability during silica gel chromatography,
and numerous attempts to ketalize it in order to reach the
target synthon 8 were in vain. These results illustrate the
synthetic challenge associated with a densely functionalized
low-molecular weight compound.
We thus turned to an entirely independent synthetic strategy
starting from glycidol (18a), another tri-functionalised and
easily affordable synthon (Scheme 2). Our initial scheme in-
cluded the use of silyl-protected glycidol, but its subsequent
reaction with N-benzyl-N-methylamine led to poor regio-
selectivity during epoxide opening (repartition 68/32).
However, benzyl-protected glycidol (18b) could be reacted
with methylamine in quantitative yields, and subsequent
protection with the largely preferred protecting group
Fmoc gave alcohol 20 in 91% yield. Swern oxidation and
protection of the resulting keto functionality by dioxolane
formation furnished the desired triply protected monomer
22 (a manifestation of 8) in an overall six-step synthesis
with excellent yields throughout.
Br
C2
C1
δ+
Fmoc
Fmoc
N
O
N
OH
O
O
O
BBr3
O
Scheme 3. Mechanism of debenzylation.
In spite of the literature precedence of a Swern oxidation
carried out on a molecule with the same hydroxymethyl-
dioxolane moiety as found in 23,19 the latter did not give
satisfactory results upon exposure to the same conditions.
Using Pfitzner–Moffat conditions20 (DCC instead of oxalyl
chloride) did not improve the yield. Compound 23 also
2.2. Preparation for monomer coupling
In order to prepare the protected monomer 22 for coupling
via reductive amination, it had to be selectively deprotected
on both termini, namely the O-terminal end by

2202
R. Barbe, J. Hasserodt / Tetrahedron 63 (2007) 2199–2207
resisted oxidation by PCC or PDC. Finally, the reactant of
Dess and Martin21 was applied, albeit with mediocre results
(40–56% yield according to the batch). The most likely
reason for the reduced yield must be seen in the electron-
withdrawing effect of the ketal moiety in the alpha position
to the aldehyde group; the latter thus tends to get hydrated as
is found in alpha-fluorinated or -chlorinated aldehydes and
ketones. These hydrated forms are often found to have
much higher water solubility and may accordingly be lost
during work-up. However, the application of the Dess–
Martin reagent has been retained for its ease of execution
and because no purification step is needed.
RN
N
OBn
i
iii
24
25
O
O
O
O
+
26 R = Fmoc
27 R = H
ii
FmocN
N
N
OBn
O
O
O
O
O
O
28
˚
Scheme 4. Synthesis of trimer 28. Reagents and conditions: (i) BH3/Py, 4 A
molecular sieves, MeOH, 60%; (ii) Et3N, acetonitrile, 92%; (iii) 24, BH3/Py,
˚
4 A molecular sieves, MeOH, 57%.
acid.24 However, the latter may produce conflicting results
depending on the system to which it is applied. In fact, the
acid also activates aldehyde groups for reduction, and it
was observed that NaBH3CN, while not acting on aldehydes
at pH 6–8, becomes an efficient reductant at pH 3–4.23 In the
majority of cases, the formation and subsequent reduction of
the immonium ion is much more rapid than aldehyde reduc-
tion. Yet in the case of sterically hindered aldehydes, compe-
tition by the aldehyde reduction path becomes a problem.24
2.3. Coupling by reductive amination
Reductive aminations can be achieved by use of two general
types of independent reductants: molecular hydrogen and
complex hydrides.22 In view of the presence of a benzyl
and an Fmoc group, application of hydrogen can be ruled
out. Among the complex hydrides only those with a suffi-
ciently low reactivity can be considered so as to avoid the
possible reduction of the aldehyde component. We focussed
on testing cyanoborohydride (NaBH3CN),23 triacetoxyboro-
hydride (NaBH(OAc)3)24 and the borane/pyridine adduct
(BH3/Py).25 Whatever the nature of the complex hydride,
reductive aminations are usually carried out at room temper-
ature. Parameters that can be modified for optimisation are
(a) the nature of the solvent (the above-mentioned complex
hydrides are stable in protic solvents, even when acidified),
(b) the potential presence or rigorous absence of water and
(c) the acidification of the medium and the potential addition
of a Lewis acid.23 While certain authors chose to shift the
equilibrium towards imine or immonium formation by use
of zeolites (for its dehydrating26 or catalytic27 properties),
others acidified their medium by adding glacial acetic
Inview of the considerable steric hindrance in our case, a sig-
nificant amount of work had to go into the search for condi-
tions minimising reduction of aldehyde 24 (Scheme 4). This
study was greatly facilitated by LCMS analysis (Fig. 3).
Table 1 illustrates the effect of variation of reactant and con-
ditions. It becomes easily apparent that the reductant BH3/Py
in MeOH does not cause any aldehyde reduction. Its Lewis-
acid character constitutes an additional advantage in that it
€
makes addition of a Bronstedt-acid catalyst obsolete. While
this reaction is particularly slow (48 h), the reaction compo-
nents do not appear to degrade during this period.
Thus obtained dimer 26 was subsequently deprotected on the
amino terminus by use of triethylamine (92%) and coupled
+
+
24
25 (1.35 eq)
26
23
NaBH3CN (2eq)
THF/H2O/AcOH (90/5/5)
FmocN
O
OH
FmocN
O
Ph
HN
O
Ph
N
O
O
O
O
O
O
O
Exact Mass: 237.1
Exact Mass: 369.2
Exact Mass: 588.3
0
5
10
15
temps (min)
Figure 3. LCMS chromatogram of dimer formation using reaction conditions different from those finally retained (Column Zorbax SB-CN; eluant: H2O+
HCO2H (0, 1%)/acetonitrile; 1:0–0:1 in 15 min).

R. Barbe, J. Hasserodt / Tetrahedron 63 (2007) 2199–2207
2203
Table 1. Optimisation of coupling conditions: minimisation of aldehyde
reduction
Surprisingly, the presence of a silyl group in the reactant 30
as compared to the benzyl group in 25 led to mediocre cou-
pling yields towards dimer 31. The efficient conditions elab-
orated for the coupling to dimer 26 were thus of no great use
in this new reaction. No side product explaining this reduc-
tion in yield could be detected by LCMS analysis. Finally,
coupling conditions employing the Lewis acid tri-isoprop-
oxytitanium chloride were retained (44% yield). The thus
obtained dimer was coupled to a tetramer by first deprotect-
ing 31 using BF3 etherate, oxidising the resulting alcohol 33
to aldehyde 34 under Swern conditions, and finally coupling
it to the deprotected secondary amine 32 using the condi-
tions that proved successful in the coupling to 26. The tetra-
mer 35 was obtained in 42% yield.
Reductive
agent
Solvent
AcOH
5%
H2O
5%
Ratio of peak
areas 26/23
NaBH3CN
NaBH3CN
BH3/Py
BH3/Py
BH3/Py
BH3/Py
THF
2.27
0.38
0.66
No alcohol
6.36
18.16
1.42
1.57
CH2Cl2
CH2Cl2
MeOH
THF
5%
5%
THF
5%
5%
NaHB(OAc)3
NaHB(OAc)3
NaHB(OAc)3
NaHB(OAc)3
CH2Cl2
MeOH
THF
5%
5%
3.39
0.75
THF
with aldehyde 24 to furnish trimer 28 in 57% yield. Such
an experiment should demonstrate the feasibility of step-
by-step construction of the oligomer as is found in solid-
phase synthesis of heteropolymers, but does not allow for
rapid synthesis of an oligomer of appreciable size. A conver-
gent synthesis using two dimers of the type 26 turned out to
be impossible because the benzyl group was surprisingly
resistant to the use of BBr3. In view of the limited access to
dimer 26, it appeared prudent to protect alcohol 23 with a
more readily cleavable protectinggroup, namely theTBDMS
group. The resulting silyl ether 29 (Scheme 5) was depro-
tected with piperidine instead of triethylamine since the latter
reacts much more slowly and thus causes multiple products
of decomposition. However, use of piperidine has an incon-
venient consequence, in fact, it is well-known that the fluo-
rene formed during deprotection reacts with piperidine to
give fluorenylmethylpiperidine28 that unfortunately cannot
be removed by chromatography since the desired secondary
amine 30 proved to be very labile on silica gel. On the other
hand, the presence of fluorenylmethylpiperidine in the subse-
quent coupling step causes formation of a large proportion of
the alkylation product of piperidine by aldehyde 24. In the
end, two to three cycles of precipitation/filtration of fluore-
nylmethylpiperidine in methanol at ꢀ78 ꢁC were sufficient
to completely remove the side product as evidenced by NMR.
3. Conclusion
We have presented here the design of an original polymer
backbone on the basis of a weak functional group interaction
between a tertiary amine and a ketone moiety. This interac-
tion has been amply characterised in the course of the last 50
years, but never for its capacity to show cooperative effects
when present in multiple copies on the same molecule. Right
from the start, the heavily functionalised structure of the
initially chosen monomers and their corresponding triply
protected forms made the need for thorough synthetic opti-
misation likely. After initially failing in developing a synthe-
sis starting from ethyl bromopyruvate, we have succeeded in
discovering an efficient synthetic path towards a protected
monomer using glycidol. This building block is stable over
extended periods of time, lends itself to perfectly selective
deprotection, and gives mediocre (42%) to satisfactory
(60%) yields during equimolar solution-phase coupling de-
pending on the identity and the complexity of the reaction
partners. We have thus prepared a dimer, a trimer and a tetra-
mer. The deprotection of each reaction partner in these
syntheses and the coupling behaviour vary considerably in
each case. The proximity of the functional groups on the
backbone caused complications such as difficulty during
debenzylation and reductive amination, effects that can be
attributed at the same time to electronic and steric influ-
ences. However, this same proximity was chosen precisely
for the eventual observation of a folding pattern that involves
favourable formation of chair-like six-membered rings in the
secondary structure (Fig. 2) and that is congruent with beta-
sheet formation in proteins. Grafting of these oligomers onto
beta-turn mimics, deprotection of the ketone groups on the
backbone and structural characterisation are in progress.
FmocN
OTBDMS
OTBDMS
HN
OTBDMS iii
i
ii
23
O
O
O
O
29
30
FmocN
N
O
FmocN
FmocN
N
OH
O
v
O
O
O
O
O
O
O
O
O
31
33
iv
vi
N
HN
N
OTBDMS
O
O
O
O
O
O
+
34
32
4. Experimental
4.1. General
vii
FmocN
N
N
N
O
OTBDMS
O
O
O
O
O
O
O
All reactions were carried out in anhydrous solvents in dried
glassware. CH2Cl2 and Et3N were distilled under argon on
CaH2. Compounds 14 and 15 were prepared according to
the literature protocols.16 1H and 13C NMR spectra were re-
corded on a Bruker DPX 200 spectrometer (200 MHz) and
a Varian Unity 500 spectrometer (500 MHz). Coupling pat-
terns in the 1H NMR spectra are designated as s, singlet; d,
35
Scheme 5. Synthesis of tetramer 35. Reagents and conditions: (i)
TBDMSCl, Et3N, DMAP, CH2Cl2, 97%; (ii) piperidine, acetonitrile, 93%;
(iii) 24, NaBH(AcO)3, TiCl(Oi-Pr)3, CH2Cl2, 44%; (iv) piperidine, aceto-
nitrile, 45%; (v) BF3/OEt2, acetonitrile, 73%; (vi) oxalyl chloride, DMSO,
ET3N, CH2Cl2, ꢀ78 ꢁC, 78%; (vii) BH3/Py, zeolite 4A, MeOH, 42%.

2204
R. Barbe, J. Hasserodt / Tetrahedron 63 (2007) 2199–2207
doublet; dd, double doublet; t, triplet; m, multiplet. HRMS
4.1.4. Fmoc-protected amino alcohol 20. To a stirred solu-
tion of compound 19 (35.203 g, 180 mmol) in 1,4-dioxane
(750 mL) was added a solution of NaHCO3 (10 wt % solu-
tion in water, 750 mL) and 9-fluorenylmethyl chloroformate
(48.901 g, 0.189 mol, 1.05 equiv). The mixture was stirred
overnight, then diluted with water (1.8 L) and extracted
with CH2Cl2 (3ꢂ700 mL). The combined organic layers
were washed with brine (500 mL), dried over sodium sulfate
and the solvent was removed in vacuo. Compound 20 was
obtained as a colourless oil after flash chromatography using
ꢀ
were recorded by the Centre de Spectrometrie de Masse,
Universite de Lyon, France. LRMS were recorded on an
ꢀ
Agilent 1100 Series LC/MSD apparatus. For those com-
pounds existing as a mixture of rotamers, the signal corre-
sponding to equivalent protons or carbons on different
rotamers is separated by ‘and’.
4.1.1. Aminoketone 17. A solution of 15 (500 mg,
2.51 mmol) and water (15 mL, 0.83 mmol, 0.3 equiv) in ace-
tone (10 mL) was stirred for 16 h with an Amberlyst-15Ò
resin (400 mg). The resin was removed by filtration. After
cooling the mixture at 0 ꢁC, K2CO3 (1.04 g, 7.53 mmol,
3 equiv) and N-methylbenzylamine (364 mg, 3.01 mmol,
1.2 equiv) were added. The mixture was stirred for 20 min,
allowed to warm to room temperature and stirred for
20 min. The solvent was removed in vacuo and the residue
was taken in water (15 mL). The aqueous layer was ex-
tracted with ethyl acetate (3ꢂ15 mL). The combined organic
layers were washed with brine (10 mL), dried over sodium
sulfate and the solvent was removed in vacuo. Compound
17 was obtained as a pale yellow oil after flash chromato-
graphy using CH2Cl2/ethyl acetate 65:35 as an eluant
1
pentane/ethyl acetate 2:1 as an eluant (68.5 g, 91%). H
NMR (CDCl3, 500 MHz): d 2.91 and 2.98 (2s, 3H), 3.06–
3.24 (m, 2H), 3.43–3.47 (m, 2H), 4.02 (m, 1H), 4.20–4.25
(m, 1H), 4.41–4.82 (m, 2H), 4.55 (s, 2H), 7.32–7.41 (m,
9H), 7.59 (m, 2H), 7.7–7.8 (m, 2H); 13C NMR (CDCl3,
50 MHz): d 36.01, 47.25, 51.50 and 52.63, 66.68 and
67.50, 69.23 and 69.77, 71.87, 73.38, 119.80, 124.90,
126.99, 127.59, 127.67, 128.37, 137.77, 141.26, 143.91,
156.12 and 157.55; HRMS (CI) m/z: calcd for C26H28NO4:
418.2018; found: 418.2019.
4.1.5. Ketone 21. To a stirred solution of oxalyl chloride
(5.30 mL, 0.063 mol, 1.7 equiv) in CH2Cl2 (190 mL) under
argon atmosphere and at ꢀ78 ꢁC was slowly added DMSO
(9.2 mL, 0.130 mol, 3.5 equiv). After 15 min a solution of
20 (15.51 g, 0.037 mol, 1 equiv) in CH2Cl2 (140 mL) in
a separate round bottom flask under argon atmosphere and
at ꢀ78 ꢁC was slowly transferred via a cannula to the first
flask. The mixture was stirred for 2 h at ꢀ78 ꢁC. Triethyl-
amine (37 mL, 0.265 mol, 7.2 equiv) was added and the so-
lution was stirred for 30 min at ꢀ78 ꢁC and 30 min at 0 ꢁC.
CH2Cl2 (750 mL) was added and the solution was washed
with saturated NH4Cl solution (100 mL) and brine
(100 mL), dried over sodium sulfate and evaporated to dry-
ness. Compound 21 was obtained as a colourless oil after
flash chromatography using pentane/ethyl acetate 4:1 then
2:1 as an eluant (14.05 g, 93%). 1H NMR (CDCl3,
500 MHz): d 2.90 and 3.00 (2s, 3H), 3.86 and 4.02 (2s,
2H), 4.14 (s, 1H), 4.28–4.30 (m, 2H), 3.90 and 4.75 (2d,
J¼6.5 Hz), 4.51 and 4.60 (2s, 2H), 7.30–7.40 (m, 9H),
7.49 (m, 1H), 7.62 (m, 1H), 7.71 (m, 1H), 7.77 (m, 1H);
13C NMR (CDCl3, 50 MHz): d 35.52 and 36.00, 47.15,
55.81 and 56.25, 66.87 and 67.77, 73.61, 74.05 and 74.21,
119.82, 124.64 and 125.02, 127.01, 127.61, 127.85,
128.09, 128.52, 136.83, 141.23, 143.89, 155.95 and
156.60, 203.62 and 204.13; HRMS (CI) m/z: calcd for
C26H26NO4: 416.1862; found: 416.1865.
1
(210 mg, 43%). H NMR (CDCl3, 200 MHz): d 2.30 (s,
3H), 3.24 (s, 2H), 3.56 (s, 2H), 4.35 (s, 2H), 7.24–7.31 (m,
5H); 13C NMR (CDCl3, 50 MHz): d 42.9, 62.0, 63.3, 67.0,
126.6, 127.2, 128.5, 137.2, 209.4.
4.1.2. Benzyl ether 18b. To a stirred solution of glycidol
(17.2 mL, 0.255 mol) and benzylbromide (40.4 mL,
0.338 mol, 1.3 equiv) in dry DMF (650 mL) at 0 ꢁC under
argon atmosphere was added sodium hydride (60% disper-
sion in mineral oil, 10.19 g, 0.255 mol, 1 equiv). The mix-
ture was allowed to warm to room temperature and stirred
overnight. The solvent was removed in vacuo and the residue
was taken up with water (500 mL). The aqueous layer was
extracted with CH2Cl2 (3ꢂ250 mL). The combined organic
layers were washed with brine (60 mL), dried over sodium
sulfate and the solvent was removed in vacuo. Compound
18b was obtained as a colourless oil after flash chromato-
graphy using cyclohexane/ethyl acetate 85:15 as an eluant
1
(63.7 g, 91%). H NMR (CDCl3, 500 MHz): d 2.61 (m,
1H), 2.80 (m, 1H), 3.19 (m, 1H), 3.44 (dd, J¼6 Hz, J0¼
11 Hz, 1H), 3.78 (dd, J¼3 Hz, J0¼11 Hz, 1H), 4.56 (d, J¼
12 Hz, 1H), 4.62 (d, J¼12 Hz, 1H), 7.3–7.4 (m, 5H); 13C
NMR (CDCl3, 50 MHz): d 44.28, 50.87, 70.84, 73.33,
127.77 and 128.45, 137.94; HRMS (EI) m/z: calcd for
C10H12O2: 164.0837; found: 164.0837.
4.1.6. Ketal 22. A solution of compound 21 (18.74 g,
45 mmol), p-toluenesulfonic acid (6.60 g, 38 mmol,
0.85 equiv) and ethylene glycol (50 mL, 900 mmol,
20 equiv) in benzene (400 mL) was stirred under reflux for
3.5 h. During this time, the water was removed with
a Dean–Stark apparatus. The solution was washed with
water (100 mL) and brine (2ꢂ100 mL). The aqueous layer
was then extracted with ethyl acetate (2ꢂ100 mL). The com-
bined organic layers were washed with brine (100 mL),
dried over sodium sulfate and the solvent was removed
in vacuo. Compound 22 was obtained as a colourless oil after
flash chromatography using pentane/diethylether 4:6 as an
eluant (19.69 g, 96%). 1H NMR (CDCl3, 500 MHz):
d 3.03 (s, 3H), 3.37 and 3.51 (2s, 2H), 3.55 and 3.60 (2s,
2H), 3.89–4.01 (m, 4H), 4.22–4.26 (m, 1H), 4.41–4.43 (m,
4.1.3. Amino alcohol 19. To a stirred solution of methyl-
amine (40 wt % solution in water, 650 mL) at 0 ꢁC was
slowly added a solution of compound 18b (26.18 g,
0.159 mol) in a minimum of CH2Cl2. The mixture was al-
lowed to warm to room temperature and stirred overnight.
The solvent was removed in vacuo. Compound 19 was ob-
tained as a colourless oil (31.05 g, 100%). For the character-
isation, the ammonium salt can be obtained with a solution
1
of HCl (gas) in diethylether. H NMR (CDCl3, 500 MHz):
d 2.67 (s, 3H), 3.05 (m, 2H), 3.53 (m, 2H), 4.38 (m, 1H),
5.49 (m, 2H), 7.30–7.34 (m, 5H), 8.66 (s (L), 1H), 9.31 (s
(L), 1H); 13C NMR (CDCl3, 50 MHz): d 33.70, 52.38,
65.66, 71.79, 73.46, 127.85, 128.39, 137.51; HRMS (CI)
m/z: calcd for C11H18NO2: 196.1338; found: 196.1339.

R. Barbe, J. Hasserodt / Tetrahedron 63 (2007) 2199–2207
2205
2H), 4.57 and 4.60 (2s, 2H), 7.29–7.34 (m, 7H), 7.38–7.42
(m, 2H), 7.61–7.67 (m, 2H), 7.76–7.79 (m, 2H); 13C NMR
(CDCl3, 50 MHz): d 36.01 and 36.62, 47.36, 51.37 and
51.77, 65.41 and 65.50, 67.38, 71.59 and 71.71, 73.63,
108.91 and 109.21, 119.86, 124.98, 126.95, 127.54,
128.29, 138.01, 141.28 and 144.13, 156.86; HRMS (CI)
m/z: calcd for C28H30NO5: 460.2124; found: 460.2121.
HRMS (CI) m/z: calcd for C13H20NO3: 238.1443; found:
238.1442.
4.1.10. Dimer 26. To a stirred solution of aldehyde 24
(473 mg, 1.29 mmol) and amine 25 (366 mg, 1.54 mmol,
1.2 equiv) in methanol (15 mL) under argon atmosphere
was added dropwise BH3/Py (0.143 mL, 1.42 mmol,
˚
1.1 equiv) and 4 A molecular sieves. After 24 h, an addi-
4.1.7. Alcohol 23. To a stirred solution of 22 (11.61 g,
31 mmol) in CH2Cl2 (230 mL) at ꢀ78 ꢁC under argon atmo-
sphere was added dropwise a solution of boron tribromide
in CH2Cl2 (31 mL, C¼1 mol/L, 1 equiv). After 4 h, the solu-
tion was transferred to saturated NaHCO3 solution (500 mL)
at 0 ꢁC. The layers were separated and the aqueous layer was
extracted with CH2Cl2 (3ꢂ100 mL). The combined organic
layers were washed with brine (60 mL), dried over sodium
sulfate and the solvent was removed in vacuo. Compound
23 was obtained as a colourless oil after flash chromato-
graphy using cyclohexane/ethyl acetate 2:1, then 1:1 as an
tional sample of BH3/Py (0.090 mL, 0.90 mmol, 0.7 equiv)
was added. After another 24 h the mixture was filtered, the
filtrate was diluted with saturated NaHCO3 (15 mL), stirred
for 30 min and extracted with diethylether (3ꢂ30 mL). The
combined organic layers were washed with brine (20 mL),
dried over sodium sulfate and the solvent was removed
in vacuo. Compound 26 was obtained as a colourless oil after
flash chromatography using pentane/ethyl acetate 10:3 as an
eluant (455 mg, 60%). 1H NMR (CDCl3, 500 MHz): d 2.41
(s, 3H), 2.55and2.59(s, 2H), 2.64(m, 2H), 2.98(s, 3H), 3.56–
3.59 (m, 4H), 3.82–3.86 (m, 4H), 3.962 (m, 4H), 4.22–4.26
(m, 1H), 4.38–4.42 (m, 2H), 4.58 (d, J¼22 Hz, 2H), 7.30–
7.40 (m, 9H), 7.61–7.76 (m, 4H); 13C NMR (CDCl3,
50 MHz): d 35.82 and 36.59, 45.07 and 47.11, 51.94,
53.23, 61.41, 62.32 and 62.59, 64.40, 64.62, 64.84, 64.99,
66.93 and 67.17, 71.15, 73.20, 109.74, 110.42 and 110.72,
119.58, 124.70, 124.94, 126.69, 127.20, 127.26, 127.92,
138.15, 140.95, 143.94, 156.56 and 156.72; HRMS (ESI)
m/z: calcd for C34H41N2O7: 589.2914; found: 589.2916.
1
eluant (7.09 g, 76%). H NMR (CDCl3, 500 MHz): d 2.99
(s, 3H), 3.36 (d, J¼7 Hz, 2H), 3.46 (s, 2H), 3.81 (t, J¼
7 Hz, 1H), 3.93–3.99 (m, 4H), 4.25 (t, J¼6.5 Hz, 1H), 4.47
(d, J¼6.5 Hz, 2H), 7.32 (m, 2H), 7.41 (m, 2H), 7.59 (m,
2H), 7.78 (m, 2H); 13C NMR (CDCl3, 50 MHz): d 35.63
and 36.15, 46.80, 50.53, 61.96, 64.63 and 64.99, 66.85 and
67.28, 108.82 and 109.17, 119.53, 124.45, 126.63, 127.27,
140.84, 143.35, 156.25 and 157.23; HRMS (CI) m/z: calcd
for C21H24NO5: 370.1654; found: 370.1657.
4.1.11. Amine 27. To a stirred solution of 26 (227 mg,
0.39 mmol) in acetonitrile (4 mL) was added dropwise
3 mL of triethylamine. The mixture was stirred for 48 h be-
fore the solvent was removed in vacuo. Compound 27 was
obtained as a yellow oil after purification by neutral-alumina
chromatography using ethyl acetate, then methanol as an
eluant (130 mg, 92%). 1H NMR (CDCl3, 500 MHz): d 2.42
(s, 3H), 2.54 (s, 3H), 2.69 (s, 2H), 2.70 (s, 2H), 3.02 (s, 2H),
3.46 (s, 2H), 3.92–4.00 (m, 8H), 4.59 (s, 2H), 7.20–7.35 (m,
5H); 13C NMR (CDCl3, 50 MHz): d 36.40, 44.94, 55.19,
61.53, 62.30, 64.51, 64.88, 71.05, 73.28, 109.76, 110.21,
127.18, 127.23, 127.99, 138.08; HRMS (ESI) m/z: calcd
for C19H31N2O5: 367.2233; found: 367.2234.
4.1.8. Aldehyde 24. To a stirred solution of NaHCO3
(4.80 g, 57.1 mmol, 3.5 equiv) and 23 (6.05 g, 16.3 mmol)
in CH2Cl2 (150 mL) at 0 ꢁC under argon atmosphere was
added Dess–Martin periodinane (12.27 g, 28.9 mmol,
1.8 equiv). After 1.5 h, aqueous Na2S2O3 (75 mL) was
added followed by saturated NaHCO3 solution (75 mL).
The mixture was stirred for 30 min. The layers were sepa-
rated and the aqueous layer was extracted with diethylether
(3ꢂ150 mL). The combined organic layers were washed
with brine (60 mL), dried over sodium sulfate and the sol-
vent was removed in vacuo (3.37 g, 56%). The crude product
was used without further purification. Its NMR spectra are
congruent with the presence of a minor amount of the hy-
1
drated form. H NMR (CDCl3, 500 MHz): d 2.97 and 3.02
4.1.12. Trimer 28. To a stirred solution of amine 27 (87 mg,
0.24 mmol) and aldehyde 24 (105 mg, 0.29 mmol,
1.2 equiv) in methanol (3 mL) under argon atmosphere
was added dropwise BH3/Py (0.027 mL, 0.27 mmol,
(2s, 3H), 3.51 and 3.70 (2s, 2H), 3.94–4.07 (m, 4H), 4.25
(m, 1H), 4.39–4.48 (m, 2H), 7.32 (t, J¼7.5 Hz, 2H), 7.40
(t, J¼7.0 Hz, 2H), 7.58 (d, J¼7.5 Hz, 2H), 7.76 (d,
J¼7.5 Hz, 2H), 9.03 and 9.42 (2s, 0.7H); 13C NMR
(CDCl3, 50 MHz): d 36.37, 47.13, 50.34, 65.85, 67.61,
106.35, 119.84, 124.84, 126.95, 127.56, 141.20, 143.82,
156.69, 195.19; HRMS (CI) m/z: calcd for C21H22NO5:
368.1498; found, 368.1497.
˚
1.1 equiv) and 4 A molecular sieves. After 48 h the mixture
was filtered, diluted with saturated NaHCO3 (7 mL), stirred
for 30 min and extracted with diethylether (3ꢂ10 mL). The
combined organic layers were washed with brine (20 mL),
dried over sodium sulfate and the solvent was removed
in vacuo. Compound 28 was obtained as a colourless oil after
flash chromatography (97 mg, 57%). 1H NMR (CDCl3,
500 MHz): d 2.39 (s, 3H), 2.40 and 2.41 (2s, 3H), 2.56–
2.60 (m, 4H), 2.64 (s, 2H), 2.66 (s, 2H), 3.02 and 3.03 (2s,
3H), 3.58 and 3.60 (2s, 2H), 3.64 and 3.68 (2s, 2H), 3.85–
3.96 (m, 12H), 4.22–4.27 (m, 1H), 4.37–4.41 (m, 2H),
4.59 (d, J¼13.6 HZ, 2H), 7.28–7.33 (m, 6H), 7.36–7.40
(m, 2H), 7.60–7.61 (m, 1H), 7.69–7.77 (m, 4H); 13C NMR
(CDCl3, 50 MHz): d 36.16 and 36.93, 45.45, 47.38, 52.33,
61.79, 62.15, 62.26, 62.74 and 62.92, 64.65, 64.89, 65.17,
67.20 and 67.57, 71.89, 73.52, 110.16, 110.82 and 111.10,
111.81, 119.83, 124.97 and 125.25, 126.92 and 127.30,
4.1.9. Amine 25. Triethylamine (3 mL, 21.5 mmol) was
added dropwise to a stirred solution of 22 (435 mg,
0.95 mmol) in acetonitrile (10 mL). The mixture was stirred
for 48 h before removing the solvent in vacuo. Compound
25 was obtained as a yellow oil after purification by neutral-
alumina chromatography using CH2Cl2, then CH2Cl2/
MeOH 100:20 as an eluant (162 mg, 72%). 1H NMR
(CDCl3, 500 MHz): d 2.43 (s, 3H), 2.79 (s, 2H), 3.46
(s, 2H), 3.96 (s, 4H), 4.54 (s, 2H), 7.20–7.34 (m, 5H);
13C NMR (CDCl3, 50 MHz): d 36.70, 54.87, 65.26,
71.12, 73.26, 108.82, 127.31, 127.48, 128.06, 137.96;

2206
R. Barbe, J. Hasserodt / Tetrahedron 63 (2007) 2199–2207
127.40, 127.49, 128.16, 138.51, 141.23, 144.25, 156.90 and
157.11; HRMS (ESI) m/z: calcd for C40H52N3O9: 718.3704;
found: 718.3706.
2H), 7.65 (d, J¼14.5 Hz, 1H), 7.73 (d, J¼14.2 Hz, 1H),
7.75–7.77 (m, 2H); 13C NMR (CDCl3, 50 MHz): d ꢀ5.54,
18.12, 25.72, 35.92 and 36.704, 45.17, 47.23, 52.07, 61.31,
62.52 and 62.79, 64.73, 65.04 (peak with shoulder), 67.03
and 67.31, 110.23, 110.58 and 110.85, 119.64, 124.78
and 125.03, 126.76, 127.33, 141.07, 144.07, 156.69 and
156.88; HRMS (CI) m/z: calcd for C33H49N2O7Si:
613.3309; found: 613.3307.
4.1.13. Silyl ether 29. To a stirred solution of 23 (5.59 g,
15.1 mmol), triethylamine (2.3 mL, 16.4 mmol, 1.1 equiv)
and DMAP (100 mg, 0.82 mmol, 0.05 equiv) in CH2Cl2
at 0 ꢁC under argon atmosphere was added dropwise a solu-
tion of tert-butylchlorodimethylsilane (4.1 g, 27 mmol,
1.8 equiv) in CH2Cl2 (6 mL). The mixture was allowed to
warm to room temperature and stirred for 3 h. The preci-
pitated ammonium salts were filtrated and washed with
diethylether (4ꢂ15 mL). The combined organic layers
were washed with water (2ꢂ25 mL) and brine (25 mL),
dried over sodium sulfate and the solvent was removed
in vacuo. Compound 29 was obtained as a colourless oil after
flash chromatography using cyclohexane/ethyl acetate 10:1,
then 10:4 (7.12 g, 97%). 1H NMR (CDCl3, 500 MHz): d 0.06
(s, 6H), 0.89 (s, 9H), 3.03 (s, 3H), 3.52 (s, 1H), 3.53 (s, 2H),
3.60 (s, 1H), 3.87–3.99 (m, 4H), 4.22–4.27 (m, 1H), 4.41 (d,
J¼7.0 Hz, 2H), 7.31 (t, J¼7.3 Hz, 2H), 7.40 (t, J¼7.4 Hz,
2H), 7.61 (d, J¼7.5 Hz, 1H), 7.67 (d, J¼7.4 Hz, 1H), 7.76
(d, J¼7.3 Hz, 2H), 13C NMR (CDCl3, 50 MHz): d ꢀ5.34,
18.35, 25.91, 36.07 and 36.73, 47.42, 51.10 and 51.45,
65.45 and 65.60, 67.47, 67.60, 109.43 and 109.76, 119.93,
125.09, 127.03, 127.62, 141.33, 144.22, 156.97; HRMS
(CI) m/z: calcd for C26H38NO5Si: 484.2519; found:
484.2516.
4.1.16. Amine 32. To a stirred solution of 31 (323 mg,
0.53 mmol) in acetonitrile (2 mL) was added dropwise
0.5 mL of piperidine. After 3 h, the solvent was removed
in vacuo and the solid was mixed with cold methanol
(ꢀ78 ꢁC) in a mortar. The mixture was filtered and the solid
was washed with cold methanol (ꢀ78 ꢁC). The filtrate was
taken up and the cycle of evaporation/mixing/filtration was
repeated two or three times until there was no signal of
9-fluorenylmethylpiperidine in NMR spectra (91 mg,
1
45%). H NMR (CDCl3, 500 MHz): d 0.07 (s, 6H), 0.90
(s, 9H), 2.42 (s, 3H), 2.49 (s, 3H), 2.62 (s, 2H), 2.63 (s,
2H), 2.83 (s, 2H), 3.64 (s, 2H), 3.951–3.969 (m, 8H); 13C
NMR (CDCl3, 50 MHz): d ꢀ5.45, 18.22, 25.81, 36.73,
45.12, 55.54, 61.52, 62.54, 64.69, 65.07, 65.16, 110.30,
110.38; HRMS (CI) m/z: calcd for C18H39N2O5Si:
391.2628; found: 391.2627.
4.1.17. Alcohol 33. To a stirred solution of 31 (286 mg,
0.47 mmol) in acetonitrile (5.5 mL) at 0 ꢁC under argon
atmosphere was added dropwise BF3/etherate (48 wt %,
0.197 mL, 1.6 equiv). The mixture was allowed to warm to
room temperature, stirred for 2 h, subsequently diluted
with ethyl acetate (30 mL) and treated with 10 mL of water.
The layers were separated and the organic layer was washed
with brine (2ꢂ10 mL), dried over sodium sulfate and the
solvent was removed in vacuo. Compound 33 was obtained
as a colourless oil after flash chromatography using CH2Cl2,
4.1.14. Amine 30. To a stirred solution of 29 (3.92 g,
8.1 mmol) in acetonitrile (43 mL) was added dropwise
4.3 mL of piperidine. After 3 h, the solvent was removed
in vacuo and the solid was mixed with cold methanol at
ꢀ78 ꢁC, transferred to a mortar, and well homogenised.
The mixture was filtered and the solid was washed with
cold methanol at ꢀ78 ꢁC. The filtrate was taken up and the
cycle of evaporation/mixing/filtration was repeated two or
three times until the disappearance of the signal of 9-fluo-
renylmethylpiperidine in the corresponding NMR spectra
(1.96 g, 93%). 1H NMR (CDCl3, 500 MHz): d 0.07 (s,
6H), 0.90 (s, 9H), 2.47 (s, 3H), 2.78 (s, 2H), 3.62 (s, 2H),
4.01 (s, 4H); 13C NMR (CDCl3, 50 MHz): d ꢀ5.51, 18.20,
25.77, 36.66, 54.55, 65.16, 65.48, 109.245; HRMS (CI)
m/z: calcd for C12H28NO3Si: 262.1838; found: 262.1840.
1
then CH2Cl2/MeOH 100:2 as an eluant (170 mg, 73%). H
NMR (CDCl3, 500 MHz): d 2.42 (s, 3H), 2.47 and 2.60
(2s, 2H), 2.67 and 2.73 (2s, 2H), 2.99 (s, 3H), 3.50 and
3.52 (2s, 2H), 3.60 and 3.64 (2s, 2H), 3.87–4.00 (m, 8H),
4.24–4.27 (m, 1H), 4.42–4.46 (m, 2H), 7.31 (t, J¼7.5 Hz,
2H), 7.39 (t, J¼7.5 Hz, 2H), 7.59–7.65 (m, 2H), 7.76 (d,
J¼7.5 Hz, 2H); 13C NMR (CDCl3, 50 MHz): d 35.94 and
36.57, 45.39, 47.22, 52.05, 62.31, 62.31, 62.90, 64.48,
65.00, 67.27, 109.23, 110.78, 119.79, 124.85, 126.88,
127.48, 141.18, 143.99, 156.72; HRMS (CI) m/z: calcd for
C27H35N2O7: 499.2443; found: 499.2443.
4.1.15. Dimer 31. To a solution of 30 (494 mg, 1.89 mmol)
in CH2Cl2 (10 mL) under argon atmosphere was added a
solution of 24 (833 mg, 2.27 mmol, 1.2 equiv) in CH2Cl2
(3 mL),
tri-isopropoxytitanium
chloride
(1.1 mL,
4.1.18. Aldehyde 34. To a stirred solution of oxalyl chloride
(0.153 mL, 1.78 mmol, 1.5 equiv) in CH2Cl2 (20 mL) under
argon atmosphere and at ꢀ78 ꢁC was slowly added DMSO
(0.292 mL, 4.11 mmol, 3.5 equiv). After 5 min a solution
of 33 (594 mg, 1.19 mmol) in CH2Cl2 (12 mL) was
added dropwise. After 20 min, triethylamine (1.15 mL,
8.53 mmol, 7.2 equiv) was added and the solution was
stirred for 30 min at ꢀ78 ꢁC and 30 min at 0 ꢁC. CH2Cl2
(750 mL) was added and the solution was washed with
saturated NH4Cl solution (2ꢂ25 mL), dried over sodium
sulfate and evaporated to dryness. Compound 34 was ob-
tained as a colourless oil after flash chromatography using
CH2Cl2, then CH2Cl2/MeOH 100:1 as an eluant (460 mg,
78%). Its NMR spectra are congruent with the presence of
3.27 mmol, 1.7 equiv) and sodium triacetoxyborohydride
(2.0 g, 9.44 mmol, 5 equiv). The mixture was stirred for
2 h then diluted with diethylether (45 mL) and treated with
saturated NaHCO3 solution (30 mL). The layers were sepa-
rated and the aqueous layer was extracted with diethylether
(2ꢂ30 mL). The combined organic layers were washed with
brine (20 mL), dried over sodium sulfate and the solvent was
removed in vacuo. Compound 30 was obtained as a colour-
less oil after flash chromatography using pentane/diethyl-
1
ether 90:10, then 50:50 as an eluant (503 mg, 44%). H
NMR (CDCl3, 500 MHz): d 0.05 and 0.07 (2s, 6H), 0.88
and 0.90 (2s, 9H), 2.42 (s, 3H), 2.56–2.61 (m, 4H), 3.03 (s,
3H), 3.63–3.69 (m, 4H), 3.89–3.95 (m, 8H), 4.22–4.28 (m,
1H), 4.40 (m, 2H), 7.30 (t, J¼7.4 Hz, 2H), 7.38–7.39 (m,
1
a minor amount of the hydrated form. H NMR (CDCl3,

R. Barbe, J. Hasserodt / Tetrahedron 63 (2007) 2199–2207
2207
Jaun, B.; Matthews, J. L.; Schreiber, J. V.; Oberer, L.; Hommel,
U.; Widmer, H. Helv. Chim. Acta 1998, 81, 932–982; Appella,
D. H.; Christianson, L. A.; Klein, D. A.; Richards, M. R.;
Powell, D. R.; Gellman, S. H. J. Am. Chem. Soc. 1999, 121,
7574–7581.
500 MHz): d 2.39–2.49 (m, 3H), 2.59 and 2.93 (2s, 2H), 2.93
and 2.97 (2s, 2H), 2.99–3.01 (m, 3H), 3.44–3.48 (m, 2H),
3.83–4.01 (m, 8H), 4.24 (m, 1H), 4.41–4.44 (m, 2H), 7.32
(m, 2H), 7.39 (m, 2H), 7.60 (d, J¼7.5 Hz, 1H), 7.67 (d,
J¼7.5 Hz, 1H), 7.76 (d, J¼7.5 Hz, 2H), 9.49 and 9.53 (2s,
0.5H); 13C NMR (CDCl3, 50 MHz): d 36.09 and 36.72,
44.88, 47.33, 52.17 and 52.34, 61.38, 62.39, 65.06, 65.31,
67.32, 106.68 and 107.24, 110.53 and 110.70, 119.84,
124.95, 126.93, 127.52, 141.24, 144.12, 156.87, 197.34;
HRMS (CI) m/z: calcd for C27H33N2O7: 497.2288; found:
497.2287.
4. Gung, B. W.; Zou, D.; Stalcup, A. M.; Cottrell, C. E. J. Org.
Chem. 1999, 64, 2176–2177.
ꢀ
5. Barbe, R. Ph.D. Thesis, Ecole Normale Superieure de Lyon
(France), 2006.
6. Gautier, A.; Pitrat, D.; Hasserodt, J. Bioorg. Med. Chem. 2006,
14, 3835–3847.
7. Leonard, N. J.; Oki, M. J. Am. Chem. Soc. 1955, 77, 6239–
6240.
8. Leonard, N. J. California Institute of Technology, personal
communication.
4.1.19. Tetramer 35. To a solution of 34 (66 mg,
0.133 mmol) and 32 (62 mg, 156 mmol, 1.2 equiv) in meth-
anol (5 mL) under argon atmosphere was added dropwise
˚
9. Cheng, R. P.; Gellman, S. H.; DeGrado, W. F. Chem. Rev. 2001,
101, 3219–3232; Venkatraman, J.; Shankaramma, S. C.;
Balaram, P. Chem. Rev. 2001, 101, 3131–3152.
10. That is the pyrrolizidines and related systems. See for instance:
Lee, J.; Ha, J. D.; Cha, J. K. J. Am. Chem. Soc. 1997, 119,
8127–8128; Leonard, N. Rec. Chem. Prog. 1956, 17, 243–257.
11. Bell, M. R.; Archer, S. J. Am. Chem. Soc. 1960, 82, 151–155.
12. Burgi, H. B.; Dunitz, J. D. Acc. Chem. Res. 1983, 16, 153–161;
Rademacher, P. Chem. Soc. Rev. 1995, 24, 143–150.
13. Carroll, J. D.; Jones, P. R.; Ball, R. G. J. Org. Chem. 1991, 56,
4208–4213; McCrindle, R.; McAlees, A. J. J. Chem. Soc.,
Chem. Commun. 1983, 61–62.
14. (a) Devraj, R.; Cushman, M. J. Org. Chem. 1996, 61, 9368–
9373; (b) Bomann, M. D.; Guch, I. C.; DiMare, M. J. Org.
Chem. 1995, 60, 5995–5996.
15. Chan, W. C.; White, P. D. Fmoc Solid Phase Peptide Synthesis:
A Practical Approach; Oxford University Press: Oxford, 2000.
16. Chari, R. V. J.; Kozarich, J. W. J. Org. Chem. 1982, 47, 2355–
2358.
BH3/Py (0.015 mL, 0.143 mmol, 1.1 equiv) and 4 A molec-
ular sieves. After 48 h the mixture was filtered, treated with
saturated NaHCO3 solution (5 mL), stirred for 30 min and
extracted with diethylether (3ꢂ15 mL). The combined
organic layers were washed with brine (10 mL), dried over
sodium sulfate and the solvent was removed in vacuo. Com-
pound 35 was obtained as a colourless oil after flash chroma-
tography using CH2Cl2, methanol/CH2Cl2 100:2, and then
100:10 as an eluant (49 mg, 42%). 1H NMR (CDCl3,
500 MHz): d 0.06 (s, 6H), 0.89 (s, 9H), 2.404–2.428 (m,
9H), 2.520 (s, 1H), 2.581–2.660 (m, 12H), 2.782 (s, 1H),
3.030–3.040 (m, 3H), 3.650–3.706 (m, 4H), 3.850–3.967
(m, 16H), 4.230–4.272 (m, 1H), 4.382–4.415 (m, 2H),
7.28–7.31 (m, 2H), 7.37–7.40 (m, 2H), 7.60–7.62 (m, 1H),
7.70–7.72 (m, 1H), 7.74–7.76 (m, 2H); 13C NMR (CDCl3,
50 MHz): d ꢀ5.42, 18.24, 25.84, 36.09 and 36.86, 45.27,
45.41, 45.52, 47.29, 52.21, 61.83, 62.25, 62.91, 64.57,
64.81, 64.99, 65.26, 65.72, 67.26, 109.23, 110.38, 110.76
and 111.01, 111.82, 119.78, 124.85, 126.86, 127.43,
141.17, 144.14, 156.73 and 157.03; HRMS (ESI) m/z: calcd
for C45H71N4O11Si: 871.4889; found: 871.4884.
17. Also observed during use of other Lewis acids: Hori, H.;
Nishida, Y.; Ohrui, H.; Meguro, H. J. Org. Chem. 1989, 54,
1346–1353.
18. Ward, D. E.; Gai, Y.; Kaller, B. F. J. Org. Chem. 1995, 60,
7830–7836; Yamada, H.; Aoyagi, S.; Kibayashi, C. J. Am.
Chem. Soc. 1996, 118, 1054–1059.
Acknowledgements
19. Aubert, C.; Gotteland, J. P.; Malacria, M. J. Org. Chem. 1993,
58, 4298–4305.
Financial support from the French Research Ministry and
from the C.N.R.S. is gratefully acknowledged.
20. Epstein, W. W.; Sweat, F. W. Chem. Rev. 1967, 67, 247–260.
21. Dess, D. B.; Martin, J. C. J. Am. Chem. Soc. 1991, 113, 7277–
7287.
Supplementary data
22. Johnson, H. E.; Crosby, D. G. J. Org. Chem. 1962, 27, 2205;
Borch, R. F.; Bernstein, M. D.; Durst, H. D. J. Am. Chem.
Soc. 1971, 93, 2897–2904.
23. Borch, R. F.; Durst, H. D. J. Am. Chem. Soc. 1969, 91, 3996–
3997.
1H, 13C NMR and mass spectral data are available in the
electronic supplementary information (ESI). Supplementary
data associated with this article can be found in the online
version, at doi:10.1016/j.tet.2006.12.081.
24. Abdel-Magid, A. F.; Carson, K. G.; Harris, B. D.; Maryanoff,
C. A.; Shah, R. D. J. Org. Chem. 1996, 61, 3849–3862.
25. (a) See Ref. 14b; (b) Pelter, A.; Rosser, R. M.; Mills, S.
J. Chem. Soc., Perkin Trans. 1 1984, 717–720.
26. Breitenbucher, J. G.; Hui, H. C. Tetrahedron Lett. 1998, 39,
8207–8210; Brussee, J.; Van Benthem, R. A. T. M.; Kruse,
C. G.; Van der Gen, A. Tetrahedron: Asymmetry 1990, 1,
163–166; Mattson, R. J.; Pham, K. M.; Leuck, D. J.; Cowen,
K. A. J. Org. Chem. 1990, 55, 2552–2554.
27. Taguchi, K.; Westheimer, F. H. J. Org. Chem. 1971, 36, 1570–
1572.
28. Carpino, L. A.; Han, G. Y. J. Org. Chem. 1972, 37, 3404–
3409.
References and notes
1. Stigers, K. D.; Soth, M. J.; Nowick, J. S. Curr. Opin. Chem.
Biol. 1999, 3, 714–723; Hill, D. J.; Mio, M. J.; Prince, R. B.;
Hughes, T. S.; Moore, J. S. Chem. Rev. 2001, 101, 3893–4011.
2. Cox, E. H.; McLendon, G. L. Curr. Opin. Chem. Biol. 2000, 4,
162–165; Laity, J. H.; Lee, B. M.; Wright, P. E. Curr. Opin.
Struct. Biol. 2001, 11, 39–46.
3. Seebach, D.; Abele, S.; Sifferlen, T.; Haenggi, M.; Gruner, S.;
Seiler, P. Helv. Chim. Acta 1998, 81, 2218–2243; Seebach,
D.; Abele, S.; Gademann, K.; Guichard, G.; Hintermann, T.;