Synthesis of chiral ligands from hydrindene diols with a consecutive heterodomino transformation as the key reaction step

Article abstract of DOI:10.1002/ejoc.200600730

We report here the elaboration of a bicyclo[2.2.2]octane system, synthesized by a consecutive domino process, towards highly functionalized stereopure derivatives, which could be used as multidentate ligands for the generation of asymmetric catalysts. Wil

Full text of DOI:10.1002/ejoc.200600730

FULL PAPER
DOI: 10.1002/ejoc.200600730
Synthesis of Chiral Ligands from Hydrindene Diols with a Consecutive
Heterodomino Transformation as the Key Reaction Step
Laure Finet,[a] Mohamed Dakir,[a] Isabel Castellote,[a] Talbi Kaoudi,[a] Loïc Toupet,[b] and
Siméon Arseniyadis*[a]
Keywords: Chirality / Ligand design / Domino reactions / Bicyclo[2.2.2]octane
We report here the elaboration of a bicyclo[2.2.2]octane sys-
tem, synthesized by a consecutive domino process, towards
highly functionalized stereopure derivatives, which could be
used as multidentate ligands for the generation of asymmet-
ric catalysts.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2007)
Introduction
Ever since the introduction by Kagan of the C2-symmet-
ric ligand DIOP in 1971,[1] a plethora of chiral ligands were
prepared and tested and they demonstrated the beneficial
effects of the C2 symmetry compared with nonsymmetrical
ligands. Even though the greater part of chiral ligands[2]
possess C2 symmetry, efficient nonsymmetrical ligands that
allow for high enantiocontrol are also known in the litera-
ture: a desymmetrized DIOP ligand gave higher optical
(and chemical) yields.[3] Waldmann et al. introduced the C2-
symmetric bicyclo[2.2.2]octane framework as a TADDOL
analogue (1,4-diols).[4] More recently, Frejd et al.[5] intro-
duced a new class of 1,3-diols that are based upon 2-
substituted bicyclo[2.2.2]octane-2,6-diols (BODOLs, 3,
Scheme 1). These bicyclic 1,3-diols were prepared by baker’s
yeast monoreduction of bicyclo[2.2.2]octane-2,6-dione, fol-
lowed by protection of the alcohol, introduction of a side-
arm in the 2-position by nucleophilic addition to the car-
bonyl group, and subsequent deprotection. A number of
chiral bidentate ligands were synthesized and evaluated as
ligands for asymmetric catalysis by this research team.
There are several routes in the literature for the prepara-
tion of bicyclo[2.2.2]octane derivatives.[6] For the current
methodology to be useful in chiral ligand preparation,
straightforward methods for the synthesis of the requisite
bicyclic framework are important. A new class of domino
transformations has been developed in our laboratories
Scheme 1. a) 1.2 equiv. PhI(OAc)2, 24 h, then 1.2 equiv. Pb(OAc)4,
PhMe, 15 h; then K2CO3/MeOH, H2O, 25 °C. b) TBSCl, DMF-
Imidazole, 0 °C, 1 h 30 min.
through the reaction of hydrindene diols of type 1 with two
oxidants and one base in a consecutive way.[7] An advantage
of this method, the combination of two heterodomino reac-
tions[8] (ring expansion/ring system interchange), lies in that
the whole transformation can be carried out in a one-pot
operation. The purpose of this work is to put forward con-
cise routes, which would allow for the stereoclean construc-
tion of optically pure hydroxylated bicyclic frameworks
starting from easily available type 1 hydrindene diols[9] as
portrayed in Scheme 1. To this aim, it was sought to exploit
the rigid, stereochemically well-defined bicyclo[2.2.2]octane
skeleton of 2 with appropriate substituents that could first
serve as control elements and subsequently as chelating
arms for several metals.
[a] Institut de Chimie des Substances Naturelles, CNRS,
Avenue de la Terrasse 91198 Gif-sur Yvette, France
Fax: +33-1-69823029
E-mail: simeon.arseniyadis@icsn.cnrs-gif.fr
[b] GMCM-UMR 6626, Université de Rennes I, Campus de
Beaulieu,
Results and Discussion
Elaboration of Bidentate Ligands
35042 Rennes, France
Supporting information for this article is available on the
WWW under http://www.eurjoc.org or from the author.
Our parent target system was thus 2, prepared in two
steps from hydrindene diols of type 1 as shown in Scheme 1.
Eur. J. Org. Chem. 2007, 335–341
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
335

L. Finet, M. Dakir, I. Castellote, T. Kaoudi, L. Toupet, S. Arseniyadis
FULL PAPER
Insofar as C-2 carbinol stereoselectivity is concerned, it was tion (TBAF, 1 h, 60 °C) afforded a 64% isolated yield of 7.
likely that increased selectivity could be acquired by in- Reaction of 2b with propynyllithium (prepared in situ from
creasing the steric hindrance in the carbonyl substrate, as propyne, nBuLi, HMPA), in dry THF at –78 °C to 25 °C,
exemplified by 1,3-control of the bulky TBS group in 2a for 1 h directly afforded desired diol 8 (70%); no TBS-pro-
and 2b. The synthesis of bidentate ligands involves direct tected diol was isolated under these conditions. Addition of
nucleophilic addition at the carbonyl (Scheme 2), whereas α-alkoxyorganolithium, derived from Bu3SnCH2OMOM[12]
for the synthesis of tridentate ligands we favored acyloin by transmetalation with nBuLi (nBuLi, Bu3SnCH2OMOM,
formation by ozonolysis of silyl enol ether 11 (Scheme 3) in dry THF, –78 °C, 20 min),[13] to 2b gave a mixture of two
followed by nucleophilic addition (Scheme 4).
products: 9a and recovered starting material 2b (87% in
The very promising results obtained by Frejd et al. with 1.9:1 ratio), which were separated by SiO2 flash chromatog-
BODOLs of type 3 (Scheme 1) encouraged us to begin with raphy and isolated in their pure forms. Fluoride deprotec-
2a. Thus, our first target was bidentate ligand 4b, and ad- tion as above gave desired bidentate ligand 9b in 78% com-
dition of commercially available allylmagnesium bromide bined yield. Finally, vinylmagnesium bromide (1  in dry
seemed to be an efficient route. Upon subjection of 2a to THF, –78 °C for 1 h, then overnight at 25 °C) addition to
standard Grignard addition conditions (AllylMgBr, THF, 2b afforded bidentate ligand 10 in 73% isolated yield.
–78 °C, 30 min), a mixture of adducts 4a (50%) and 5
(12%) was obtained along with 4b (36%) where the TBS
protecting group was cleaved. Even though the isolated
chemical yield was excellent, the selectivity for the desired
isomer was poor; unwanted 5 was present in 12%.[10]
We thus switched to 2b, where the TBS protecting group
was used as a steric director, in combination with the OtBu
substituent, in a stereodefined construction of both bi- and
tridentate ligands. Contrary to sterically unbiased 2a, and
much to our pleasure, our parent target systems 2b for bi-
dentate and 12b for tridentate ligands (Schemes 2 and 4,
respectively) underwent nucleophilic additions with com-
plete selectivity in the required direction and showed a com-
plete syn preference with respect to the OtBu substituent.
Initial efforts were devoted to optimize the yield of carbi-
Scheme 2. a) AllylMgBr, Et2O, –78 °C. b) TBAF, THF, 60 °C. c)
MeLi, THF, –78 °C. d) Propyne, nBuLi, HMPA, –78 °C, to 25 °C.
e) Bu3SnCH2OMOM, nBuLi, –78 °C. f) VinylMgBr, THF, –78 °C.
nols 6–10 (either in its TBS-protected form or as a diol).
A four- to fivefold excess of the nucleophile was found
to give good results. Additional reaction parameters such
as temperature and time as well as choice of the quenching
agent were also looked at (optimal conditions are described
in detail in the Experimental Section).
Elaboration of Tridentate Ligands
Preparation of bicyclo[2.2.2]octane derived 1,3-biden-
tates was straightforward. Starting from 2b, nucleophilic
Because there is potential for novel tridentate ligands,[14]
addition (Grignard or RLi, THF, –78 °C) and subsequent routes to conformationally restricted triols may be of inter-
fluoride deprotection (nBu4NF, THF, 60 °C) afforded the est. The three functional groups (free hydroxy, tBu-pro-
targets in high chemical yields and complete selectivity. tected hydroxy, and carbonyl) on each ring, combined with
Upon prolonged reaction times, removal of the TBS group the efficient steric control ensured by the rigid bicyclic
was observed; in some cases, such as 8, the fluoride depro- framework of 2b, offer easy chemodifferentiation for further
tection step was unnecessary. By means of these operation- selective transformations. Inspired by the work published
ally simple reactions, enantiomerically pure bicyclic 1,3-di- on BODOLs,[5] we decided to synthesize more oxygenated
ols 6–10, potential bidentate ligands, were readily available derivatives starting from 2b, targeting 1,3,4-triols, which
in multigram scale; their synthesis is portrayed in Scheme 2. could be derived from 12 (Scheme 3). For the synthesis of
When bicyclic aldol 2b was treated with excess allylmagne- the targeted tridentate ligands we needed to determine the
sium bromide (1  in Et2O, 5 equiv.) at –78 °C for 1 h, TBS- stereochemistry of the added hydroxy group at C-3 and to
protected carbinol 6a was furnished in 45% isolated yield obtain a high as possible ratio of 12/13 (or 15/16) and there-
along with 35% yield of target bidentate ligand 6b. When fore alternative methods of α-hydroxylation of 2b were in-
the reaction was left overnight at room temperature, a 56% vestigated. Direct acetoxylation of 2b by treatment with
yield of 6b was obtained along with 22% yield of 6a. Fluo- Pb(OAc)4 in benzene (90 °C, 2 d) afforded a modest 35%
ride deprotection (TBAF, THF)[11] then removed the silyl isolated yield of 15 along with 0.5% of 16 and unreacted
group, which resisted direct deprotection in the reaction starting material (40%). While the diastereoselectivity of
medium to afford bidentate ligand 6b in 90% yield. Expo- the key hydroxylation reaction to provide 15, a precursor
sure of 2b to an excess of MeLi (1.6  in Et2O, 7 equiv.), in to desired acyloin 12, was excellent (70:1 in favor of 15),
dry THF cooled to –78 °C, followed by fluoride deprotec- the yield remained poor; therefore, an alternative method
336
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 335–341

Synthesis of Chiral Ligands from Hydrindene Diols
FULL PAPER
Scheme 3. a) TMSOTf, collidine, CH2Cl2, 0 °C. b) O3, MeOH, –78 °C, then HCl, THF, 25 °C. c) m-CPBA, NaHCO3, CH2Cl2, 0 °C. d)
Pb(OAc)4, PhH, 90 °C. e) TMSCHN2, Et2O, MeOH, 0 °C. f) Bu3SnCH2OMOM, nBuLi, –78 °C.
of α-hydroxylation was investigated. Attempted next, the α- was observed in the signals corresponding to 7α-H and
hydroxylation of 2b by the oxidation of its corresponding 6α,β-H. On the other hand, enhancement of the signal due
silyl enol ether with a peracid proved less effective (Rubot- to proton 8α-H (δ = 1.90 ppm) and 6α-H (δ = 1.99 ppm)
tom hydroxylation).[15] Thus, generation of silyl enol ether upon irradiation of 5α-H (δ = 3.49 ppm) identified the α-
11 by sequential treatment of 2b with TMSOTf in the pres- face proton signals, which established that 5α-H, 6α-H, 7α-
ence of collidine, in CH2Cl2 at 0 °C, followed by the ad- H, 8α-H are cis to each other. This confirmed that minor
dition of m-CPBA and NaHCO3 in dry CH2Cl2 at 0 °C for acyloin 13 was the undesired isomer, while major acyloin
20 min gave desired α-acetoxy ketone 12 in only 29% iso- 12 was the one hydroxylated from the face opposite to C-
lated yield, together with 39% of 13 and 25% of recovered 5-OtBu.
starting material, as portrayed in Scheme 3. Replacement of
The stereochemistry at C-3 was also confirmed to be as
the oxidant by MTO–H2O2[16] failed to give the desired acy- shown by conversion to orthoester 22 (Scheme 5), to aceto-
loin and the starting material was recovered intact. In a nide 28 (Scheme 7), and by X-ray crystallographic analysis
better approach, 2b was transformed by a three step se- of 18b (vide infra). The latter was obtained by direct alk-
quence (TMSOTf, collidine, CH2Cl2, 0 °C, 2 h, followed by oxylithiation of 15 (16% isolated yield), along with deacety-
ozonolysis in CH2Cl2 at –78 °C, and subsequent desilylation lated counterpart 18a (37%) and 28% of recovered starting
with HCl in THF, at 25 °C for 20 min) to monoprotected material. Although of limited preparative interest, this reac-
α-ketol 12 in 58% yield, along with epimer 13 (12%), and tion proved beneficial and allowed for the creation of crys-
the normal ozonolysis product 14a (8%). The latter was talline 18b which was studied by single-crystal X-ray dif-
characterized as its methyl ester by subsequent esterification fraction analysis (Figure 1); thus, unequivocal stereochemi-
(TMSCHN2, MeOH/Et2O, 0 °C) affording aldehyde ester cal assignment and further confirmation of the correctness
14b in 93% isolated yield. The best conversion was achieved of our previous assignments by NMR techniques (spatial
by carrying out ozonization in MeOH, at –78 °C. This re- proximity measurements) was enabled.
sulted in a 77.4% combined yield of a 21.7:1 mixture of
endo-12 and exo-13 adduct, separated by chromatography,
while the diastereoselectivity of the key hydroxylation reac-
tion was improved and favored 12 (a fivefold raise in 12:13
ratio). The abnormal route of oxidation observed on hin-
dered silyl enol ether 11 derived from the bicyclo[2.2.2]oc-
tane framework parallels the anomalous ozonization on
camphor derived hindered silyloxyalkene reported by
Heathcock et al.[17]
The individual isomers could be characterized by NMR
techniques; the connectivity in the NMR spectrum was de-
duced from COSY, HMQC and HMBC experiments,
whereas the stereochemical assignments of the resultant bi-
cyclic frameworks of 12 and 13 were based on spatial prox-
imity effects, measured by the 1D NOEDIFF technique on
the latter product, which allowed for diagnostic NOEs. Par-
Figure 1. Perspective drawing of the X-ray structure of 18b (Chem.
3D output from X-ray coordinates; gray, carbon; red, oxygen; cyan,
ticularly revealing in 13 was the enhancement of the signals
due to proton 8β-H (d, δ = 1.55 ppm) upon irradiation of
3-H (d, δ = 3.81 ppm), and vice versa, which indicates the
proximity of 3-H and 8β-H. Enhancements were observed
hydrogen; purple, silicon). TBS and tBu protecting groups are sim-
plified for clarity.
On the basis of these results, the stereochemistries of 12,
in signals corresponding to 1-H (δ = 2.56 ppm) and 8α-H 13, 15, and 16 were assigned as shown in Scheme 3. A
(δ = 1.90 ppm) when 7α-H (δ = 3.98 ppm) was irradiated. stereodefined construction of tridentate ligands was then
Similarly, when 1-H (δ = 2.56) was irradiated enhancement carried out with the use of the same nucleophilic partners
Eur. J. Org. Chem. 2007, 335–341
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
337

L. Finet, M. Dakir, I. Castellote, T. Kaoudi, L. Toupet, S. Arseniyadis
FULL PAPER
as for the bidentate ligands. The operationally simple and
straightforward reaction conditions used with 2b, proved to
be successful with 12 as well as with 12b and 15. Highly
functionalized derivatives 17–21 were prepared by nucleo-
philic addition, which occurred exclusively on the sterically
less crowded face, in good to high yields and stereopure
(Scheme 4). Reaction of unprotected acyloin 12 with pro-
pynyllithium (prepared as above, in situ from propyne,
nBuLi, HMPA), in dry THF (–78 °C to 25 °C, for 1 h) was
completely stereoselective and cleanly afforded desired triol
17 in 72% isolated yield after silica gel chromatography; no
detectable amounts of the corresponding TBS-protected
diol was observed. Proceeding as above, the reaction be-
tween TBS-protected bicyclic aldol 12b and the α-alkoxy-
organolithium derived from Bu3SnCH2OMOM by trans-
metalation with nBuLi (nBuLi, Bu3SnCH2OMOM, in dry
THF, –78 °C, 20 min) gave a mixture of two products, 18a
and 18 (70% in 2.9:1 ratio) together with unreacted starting
material (10%), which were separated by SiO2 flash
chromatography and isolated in their pure forms. Fluoride
deprotection, as above, gave desired tridentate ligand 18 in
89% yield. Exposure of 12b to an excess of MeLi (2.2  in
THF, 5 equiv.) directly afforded 19 in a 90% yield. Again,
the ratio 19:19a was variable depending upon the reaction
time and temperature; with longer reaction times, more de-
protection occurred in the reaction medium, whereas under
the optimal conditions (2.2 equiv. of MeLi, dry THF,
–78 °C), either bis- or monoprotected triols could be ob-
tained and further transformed by fluoride deprotection
(TBAF, 1 h, 60 °C). Vinylmagnesium bromide addition (1 
in dry THF, –78 °C) of either free acyloin 12 (for 1 h at
–78 °C, then 2 h at 0 °C; 20a/20, 90% combined yield, 4:1
ratio) or its TMS-protected counterpart 12b (for 1 h at
–78 °C, then overnight at 25 °C; 91% combined of 20a/20,
1:35 ratio) afforded tridentate ligand 20 in high isolated
yields following fluoride deprotection. Reaction of bicyclic
aldol 12b with excess allylmagnesium bromide (1  in Et2O,
6 equiv.) at –78 °C for 1 h furnished a 95% combined yield
of TBS-protected carbinol 21a along with target tridentate
ligand 21. When the reaction was left overnight at room
temperature, a 56% yield of 21a was obtained along with
22% of 21. Fluoride deprotection (TBAF, THF) then re-
moved the silyl group, which resisted direct deprotection in
the reaction medium to provide 21 in 75% yield. All of the
substrates synthesized so far possess a tBu-protected hy-
Scheme 4. a) Propyne, nBuLi, HMPA, –78 °C, to 25 °C. b) TBAF,
THF, 60 °C. c) TMS-Im, CH2Cl2, 0 °C. d) Bu3SnCH2OMOM,
nBuLi, –78 °C. e) MeLi, THF, –78 °C. f) AllylMgBr, THF, –78 °C.
g) VinylMgBr, THF, –78 °C.
droxy group at C-5 (arbitrary numbering) or attached to
chains that modify their lipohilic/lipophobic character. We
thus set out to develop a synthetic route to 23, by an ortho-
ester-type temporary protection, which allows for such link-
ing endeavors. Targeted orthoester 22 was initially isolated
as an unexpected product[19] during the chemical manipula-
tion of 12b and 25 (Scheme 5). At the outset, the procedure
involved the treatment of 12b with MeLi and the trapping
of the resulting alcoholate anion with freshly distilled acetic
anhydride to get 24 in a one-pot operation.
Scheme 5. a) MeLi, THF, –78 °C to 0 °C, then Ac2O, then 35%
HCl, 0 °C. b) BF3·Et2O, PhMe, 0 °C. c) MeLi, THF, –78 °C to
0 °C. d) Ac2O, py, DMAP, 0 °C. e) TBAF, THF, 60 °C (85%) or
THF/HCl/H2O, 0 °C, 50 min (83%). f) TBAF, THF, 0 °C. g)
(CO)2Cl2, DMSO, Et3N, CH2Cl2,–78 °C to 0 °C (80%).
When TMS (acyloin)/TBS (aldol) protected bicyclic
droxy group at C-5 in order to afford face selectivity for the framework 12b was treated with MeLi in THF with the aim
C-3 hydroxylation and further chemoselective elaboration to produce 19 after TMS/TBS deprotection (Scheme 5), a
for other synthetic uses. In all cases investigated, the nucleo- mixture of mono and bis-deprotected carbinols was ob-
philic addition only occurred on the less hindered face of served before the fluoride deprotection step. This observa-
the carbonyl, and only exo adducts were formed. A conve- tion was suggestive of an acetic anhydride mediated
nient route to these substrates is shown in Scheme 4.
quenching, which indeed straightforwardly afforded corre-
Suitably protected bicyclic tetraols 17–21 with their chal- sponding orthoester 22 in high yield. We then discovered
lenging array of six asymmetric centers, might also serve as that carbinol acetate 24 is prone to undergo both acid- or
attractive building blocks for several synthetic endeavors. fluoride-catalyzed orthoester formation. This was exem-
Further manipulation of the title compounds should enable plified by the experiments of 24 (TBAF conditions) and 25
the construction of even more elaborated chelating systems (Swern conditions). The one-pot transformation of 12b into
since the ligands synthesized in this work could be easily 22 and hence to target 23 is summarized in Scheme 5. Com-
fastened to polymeric supports[18] through the extra hy- pound 12b was thus exposed to MeLi in THF at –78 °C for
338
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 335–341

Synthesis of Chiral Ligands from Hydrindene Diols
FULL PAPER
20 min, and the reaction mixture was warmed to 0 °C. After could be easily prepared by appending the suitable func-
an additional 20 min stirring at this temperature, the reac- tionalities to the pre-existing chelating arms; this leads to a
tion mixture was diluted with THF, Ac2O was added, and tridimensional framework differing from the one offered by
the mixture was quenched with aqueous HCl to afford an supertripodal ligands.[24]
80% isolated yield of the target compound. Orthoester was
also formed if 24 was allowed to sit at room temperature in
THF in the presence of TBAF overnight (85% isolated
yield) or if the latter was treated with HCl/H2O in THF
(83% isolated yield). Thus, fluoride deprotection of 24 fur-
nished 25 when performed at 0 °C for a short time while
longer exposure and heating directly gave orthoester 22 in
high yield. A single crystal of the latter was grown in
Scheme 6. a) MOMCl, iPr2NEt, 25 °C.
The high propensity for orthoester formation[25] of the
tridentate ligand precursors constitutes additional proof of
the stereochemistry and confirms that all OH groups point
towards the same space of the bicyclic framework, which
leaves the entering nucleophile and the tBu-protected hy-
droxy group on the opposite site. Yet, additional proof lies
in the ease with which acetonide 28 is formed upon treat-
ment of triol 20 with acetone in the presence of p-TsOH at
25 °C; only one out of the two possible isopropylidene
alcohols was obtained in 75% isolated yield (Scheme 7).
Furthermore, the total 1,2- versus 1,3-diol selectivity ob-
served in acetonide formation could be useful in synthetic
endeavors.
CH2Cl2/heptane, and an X-ray diffraction study corrobo-
rated the structure depicted in Figure 2.
Figure 2. Perspective drawing of X-ray structure of 22 (Chem. 3D
output from X-ray coordinates; gray, carbon; red, oxygen; cyan,
hydrogen). tBu protecting group is simplified for clarity.
Scheme 7. a) p-TsOH, 4 Å MS, 0 °C to 25 °C, 6 h (75%).
Finally, the latter was also obtained, accidentally though
reproducibly, through an attempted Swern oxidation of 25,
aimed at carbonyl formation for use in other synthetic pro-
jects. The yield of this transformation is nearly quantitative;
triethylamine, although systematically added in the early
runs, is not needed. Orthoester 22 proved resistant to
Conclusions
A central objective associated with the present work was
AcOH–H2O, 60 °C for 48 h (only traces of the correspond- to capitalize on our earlier experience with converting 1,2-
ing diol–acetate was detected), as well as to treatment with unsaturated bicyclic diols of type 1 into the bicyclo[2.2.2]-
p-TsOH in MeOH/H2O for 24 h at 0 °C, reduction with Li- octane framework of type 2 (Scheme 1). A series of bicy-
AlH4 in THF at 0 °C to room temperature, and reaction clo[2.2.2]octane derivatives has been prepared for which
with MeLi in THF at –78 °C for 1 h then at room tempera- further applications can be easily imagined. The reported
ture overnight; in those three attempts, unreacted orthoes- approach combines the advantages of domino reactions
ter was recovered.[20] The tBu group was easily cleaved by with the effortlessness of stereoselective elaboration of bicy-
exposure to BF3·Et2O in toluene at 0 °C for 45 min, a clic chiral ligands,[26] in gram quantities. The efficient prep-
method developed earlier in our laboratory.[21]
aration of 1,3-diols (6–10) and 1,3,4-triols (17–21) contain-
Optically pure tridentate ligands 17 and 19, obtained in ing bicyclo[2.2.2]octane frameworks exemplifies this syn-
a five step process and high yield from the Hajos–Parrish thetic strategy. The potential of this route to form the bicy-
ketone[22] derived unsaturated diol 1b, could also serve as clo[2.2.2]octane framework compares favorably with other
precursors for an efficient construction of multidentate li- reported methods by its use of readily available optically
gands, that could be considered as close relatives of super- pure precursors,[27] mild reaction conditions, high selectiv-
tripodal ligands.[23] Straightforward conversion of 17 and ity, and a reasonable number of synthetic steps. We believe
19 into tris-MOM (methoxymethyl ether) derivatives 26 the current approach will also be applicable to purposes
(75%) and 27 (90% based on recovered starting material), other than chiral ligand construction, such as stereodefined
respectively, was accomplished by protection of the triols construction of medium sized rings and hence in the total
with the use of chloromethyl methyl ether in the presence of synthesis of biologically important targets (following paper,
Hünig’s base (Scheme 6). Several other multidentate ligands this journal).
Eur. J. Org. Chem. 2007, 335–341
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
339

L. Finet, M. Dakir, I. Castellote, T. Kaoudi, L. Toupet, S. Arseniyadis
FULL PAPER
0.88 (s, 3 H), 1.10 (s, 9 H), 1.32 (s, 3 H), 1.44 (dt, J = 4.2, 15.9 Hz,
1 H), 1.56 (ddd, J = 1.8, 9.6, 14.3 Hz, 1 H), 1.65 (dd, J = 6.0,
14.3 Hz, 1 H), 1.80 (m, 1 H), 1.81 (ddd, J = 3.8, 9.0, 15.9 Hz, 1
H), 2.95 (br. s, 1 H), 3.25 (dd, J = 4.3, 9.0 Hz, 1 H), 3.56 (br. s, 1
H), 3.64 (br. s, 1 H), 3.83 (m, 1 H), 4.51 (br. s, 1 H) ppm. 13C NMR
(75 MHz, CDCl3): δ = 19.9, 27.9, 28.7, 34.3, 35.6, 39.1, 42.9, 70.7,
Experimental Section
General: The term “usual work up” refers to washing the organic
layer with brine, drying with anhydrous MgSO4, and evaporating
the solvent in vacuo with a rotary evaporator at aspirator pressure.
Melting points are uncorrected. IR spectra were recorded with an
FTIR instrument through NaCl cell windows. Experimental evi-
dence favoring the structures investigated came from a comprehen-
sive range of 1H- and 13C NMR spectroscopic data (500/125 and
300/75 MHz respectively, 1D and 2D experiments) and corrobo-
rated by spatial proximity (n.O.e) studies using mainly the 1D
NOEDIFF technique.[28] For all compounds investigated, multi-
plicities of the 13C NMR spectroscopic resonances were assigned
by the SEFT technique.[29] Chemical shifts in the 1H NMR spectra
are expressed downfield from TMS with the use of the residual
nondeuterated solvent as an internal standard (CDCl3 1H NMR,
7.26 ppm; C6D6 1H NMR, 7.15 ppm). 13C NMR chemical shifts
are reported relative to CDCl3 or the C6D6 triplet centered at δ =
77.0 ppm and 128.0 ppm, respectively. Mass spectra acquired in the
positive ion mode under electron spray ionization (ES+) using a
mobile phase of methanol, will be abbreviated as ESIMS (MeOH).
HR will be added for the high resolution mass measurements
(HRESIMS).
70.8, 72.0, 73.0, 73.1 ppm. ESIMS (MeOH): m/z
= 281.1
[M + Na]+. HRESIMS: calcd. for C14H26O4Na 281.1729; found
281.1731. C14H26O4 (258.18): calcd. C 65.09, H 10.14; found C
64.78, H 9.83.
Preparation of Multidentate Ligands in the Form of Tris-MOM Pro-
tected Triols - 7-tert-Butoxy-2,3,5-trimethoxymethoxy-1-methyl-3-
prop-1-ynylbicyclo[2.2.2]octane (26): To a stirred solution of 17
(40 mg, 0.14 mmol) in dry dichloromethane (1 mL) was added
iPr2Net (0.24 mL, 0.15 mmol, 6.2 equiv.) followed by the addition
of MOMCl (0.06 mL, 0.84 mmol, 6 equiv.). The mixture was stirred
at room temperature overnight, quenched with water, and extracted
with dichloromethane. The combined extracts were washed with
hydrochloric acid (1 ), worked up as usual, and chromatographed
on silica gel (heptane/EtOAc, 2:1) to yield 45 mg (75%) of 26. [α]
20
= +166 (c 1.20, CHCl ). IR (film): ν = 2966, 2928, 2838, 2816,
˜
D
3
1
1387, 1361, 1191, 1149, 1107, 1072, 1038, 995, 917 cm–1. H NMR
(500 MHz, CDCl3): δ = 0.90 (s, 3 H), 1.13 (s, 9 H), 1.50 (ddd, J =
1.8, 9.8, 13.2 Hz, 1 H), 1.70 (ddd, J = 1.7, 6.8, 15.7 Hz, 1 H), 1.88
(s, 3 H), 1.86–1.96 (m, 2 H), 2.05 (td, J = 1.8, 4.7 Hz, 1 H), 3.28
(dd, J = 5.8, 9.3 Hz, 1 H), 3.33 (s, 3 H), 3.43 (s, 3 H), 3.44 (s, 3 H),
3.73 (t, J = 8.5 Hz, 1 H), 4.09 (d, J = 1.7 Hz, 1 H), 4.56 and 4.59
(ABquart, J = 6.8 Hz, 2 H), 4.71 and 5.06 (ABquart, J = 6.4 Hz,
2 H), 4.94 and 4.96 (ABquart, J = 5.9 Hz, 2 H) ppm. 13C NMR
(125 MHz, CDCl3): δ = 3.7, 20.4, 28.8, 33.1, 34.8, 38.9, 42.0, 55.2
55.9, 56.2, 70.4, 72.8, 72.9, 73.7, 77.7, 81.4, 81.7, 93.0, 94.7,
95.9 ppm. ESIMS (MeOH): m/z (%) = 437.2 (100) [M + Na]+.
HRESIMS: calcd. for C18H38O7Na 437.2515; found 437.2509.
C22H38O7 (414.26): calcd. C 60.80, H 8.70; found C 60.61,
H 8.35.
CCDC-610968 (18b), and CCDC-611342 (22) contain the supple-
mentary crystallographic data for this paper. These data can be
obtained free of charge from The Cambridge Crystallographic
Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
7-tert-Butoxy-1-methyl-3-prop-1-ynylbicyclo[2.2.2]octane-2,3,5-triol
(17): To a solution of propyne in dry THF (2 mL) at –78 °C under
an argon atmosphere, nBuLi (1.6 , 1.46 mL, 2.34 mmol, 8 equiv.)
was added dropwise. The mixture was warmed to –15 °C and stir-
ring was continued at this temperature. After 2 h, HMPA (0.3 mL)
was added at –20 °C, and the resulting mixture was stirred for
10 min. A solution of 12 (100 mg, 0.29 mmol, 1 equiv.) in THF
(1 mL) and HMPA (0.3 mL) was then added at –30 °C, and the
mixture was left to stir overnight at room temperature, diluted with
diethyl ether, quenched with sat NH4Cl and worked up as usual.
The residue was purified by chromatography (heptane/EtOAc, 1:1)
7-tert-Butoxy-2,3,5-trimethoxymethoxy-1,3-dimethylbicyclo[2.2.2]-
octane (27): Proceeding as above, to 19 (60 mg, 0.23 mmol) in dry
dichloromethane (2 mL) was added iPr2Net (0.24 mL, 1.42 mmol,
6.2 equiv.) followed by the addition of MOMCl (0.1 mL,
1.38 mmol, 6 equiv.). The mixture was stirred at room temperature
overnight after addition of excess MOMCl, quenched with water,
and extracted with dichloromethane. The combined extracts were
washed with hydrochloric acid (1 ⁾, worked up as usual and chro-
matographed on silica gel (heptane/EtOAc, 2:1) to yield quantita-
tively 27. [α]2D0 = +201 (c 0.35, CHCl ). IR (film): ν = 2971, 2928,
to give 17 (58 mg, 72%). [α]2D0 = 88 (c 0.9, CHCl ). IR (film): ν =
˜
3
3324, 2972, 2360, 2335, 1449, 1363, 1192, 1094, 1065, 994, 971,
1
668 cm–1. H NMR (500 MHz, CDCl3): δ = 0.86 (s, 3 H), 1.12 (s,
9 H), 1.64 (m, 3 H), 1.86 (s, 3 H), 1.90 (m, 1 H), 2.15 (m, 1 H),
3.26 (dd, J = 5.6, 9.4 Hz, 1 H), 3.78 (td, J = 1.9, 8.5 Hz, 1 H), 3.98
(s, 1 H) ppm. 13C NMR (100 MHz, CDCl3): δ = 3.6, 19.5, 28.7,
33.7, 35.8, 38.7, 43.7, 69.4, 70.4, 71.0, 72.6, 73.0, 79.1, 82.9 ppm.
ESIMS (MeOH): m/z (%) = 305.1 (100) [M + Na]+. HRESIMS:
calcd. for C16H26O4Na 305.1729; found 305.1730. C16H26O4
(282.18): calcd. C 68.06, H 9.28; found C 67.66, H 9.22.
˜
3
1727, 1389, 1364, 1145, 1110, 1093, 1039, 919 cm–1. 1H NMR
(500 MHz, CDCl3): δ = 0.89 (s, 3 H), 1.10 (s, 9 H), 1.40 (ddd, J =
2.3, 4.8, 14.3 Hz, 1 H), 1.48 (s, 3 H), 1.48 (m, 1 H), 1.88 (ddd, J =
4.5, 9.4, 14.0 Hz, 1 H), 1.97 (m, 2 H), 3.26 (dd, J = 4.8, 9.3 Hz, 1
H), 3.33 (s, 3 H), 3.41 (s, 3 H), 3.45 (s, 3 H), 3.64 (s, 1 H), 3.69 (t,
J = 8.5 Hz, 1 H), 4.55 and 4.58 (ABquart, J = 6.7 Hz, 2 H), 4.59
and 4.92 (ABquart, J = 8.0 Hz, 2 H), 4.74 and 4.85 (ABquart, J =
6.6 Hz, 2 H) ppm. 13C NMR (100 MHz, CDCl3): δ = 20.5, 26.0,
28.7, 33.3, 35.4, 39.6, 39.8, 55.1, 55.5, 55.8, 70.7, 72.9, 73.9, 76.4,
78.8, 92.0, 94.6, 97.0 ppm. ESIMS (MeOH): m/z (%) = 413.2 (100)
[M + Na]+. HRESIMS: calcd. for C20H38O7Na 413.2515; found
413.2508. C20H38O7 (390.5): calcd. C 61.51, H 9.81; found C 61.96,
H 9.61.
7-tert-Butoxy-1,3-dimethylbicyclo[2.2.2]octane-2,3,5-triol (19): To a
stirred solution of 12b (200 mg, 0.46 mmol) in THF (6 mL) at
–78 °C was added methyl lithium (2.2  in THF, 0.46 mL,
2.2 equiv.). The reaction was kept at –78 °C for 30 min, warmed
up to 0 °C, and left to react for a further 45 min. After complete
consumption of the starting material, a solution of HCl (4 ) was
added, the mixture was left to react at room temperature for an-
other 20 min, neutralized by the addition of a saturated solution
of NaHCO3, extracted 3 times with EtOAc, and worked up as us-
ual. The crude residue was purified by chromatography over silica
gel (heptane/EtOAc, 1:1) to provide tridentate ligand 19 (110 mg,
90%) as a white solid. [α]2D0 = +123 (c = 1.05, CHCl3). M.p. 134–
Supporting Information (see footnote on the first page of this arti-
135 °C. IR (film): ν = 3324, 2964, 2927, 1451, 1366, 1208, 1136,
cle): Procedures for the synthesis and spectral characterization for
˜
1074, 1033, 1000, 926, 815 cm–1. 1H NMR (300 MHz, CDCl3): δ = all compounds including structural analysis of 18b and 22.
340
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 335–341

Synthesis of Chiral Ligands from Hydrindene Diols
FULL PAPER
tane-2-one into the requisite [2.2.2]octane-2,5-diol. F. Alm-
qvist, N. Ekman, T. Frejd, J. Org. Chem. 1996, 61, 3794–3798.
[11] Commercially available nBu4NF (from Aldrich) effects the de-
protection in a large array of temperatures, depending on its
origin; the recently bought reagent needs 60 °C heating while
the older lots bought in the years 2002–2003 used to effect the
same transformations at temperatures as low as –70 °C.
[12] R. L. Danheiser, K. R. Romines, H. Koyama, S. F. Gee, C. R.
Johnson, J. R. Medich, Org. Synth. 1992, 71, 133–139.
[13] W. C. Still, J. Am. Chem. Soc. 1978, 100, 1481–1486.
[14] Y.-J. Cherng, J.-M. Fang, T.-J. Lu, J. Org. Chem. 1999, 64,
3207–3212.
[15] G. M. Rubottom, M. A. Vazquez, D. R. Pelegrina, Tetrahedron
Lett. 1974, 49, 4319–4322; G. M. Rubottom, R. Marrero, J.
Org. Chem. 1975, 40, 3783–3784; A. Hassner, R. H. Reuss,
H. W. Pinnick, J. Org. Chem. 1975, 40, 3427–3429.
[16] S. Stankovic, J. H. Espenson, J. Org. Chem. 1998, 63, 4129.
[17] R. D. Clark, C. H. Heathcock, J. Org. Chem. 1976, 41, 1396–
1403.
[18] For a solid phase synthesis of bicyclo[2.2.2]octane derivatives,
see:S. V. Ley, D. M. Mynett, W.-J. Koot, Synlett 1995, 1017–
1020 and references cited therein.
[19] Orthoesters have been isolated as unexpected products, for ex-
ample during the synthesis of muscarine: W. M. Daniewski, M.
Chmielevski, B. Guzik, P. Hintze, J. Org. Chem. 1985, 50, 5360–
5362 and chemical modifications of taxinine: M. Hosoyama,
H. Shigemori, J. Kobayashi, J. Chem. Soc., Perkin Trans. 1
2000, 449–451.
Acknowledgments
The authors thank Professor Jean-Yves Lallemand for his kind
interest and constant encouragements.
[1] H. B. Kagan, T. P. Dang, J. Chem. Soc. Chem. Commun. 1971,
481.
[2] Stereochemically pure 1,n-diols have established synthetic ap-
plications in catalytic asymmetric transformations: construc-
tion of titanium diolates used extensively in organic synthesis,
such as Sharpless’ epoxidation catalyst with Ti-tartrates (1,2-
diols): K. B. Sharpless, K. Katsuki, J. Am. Chem. Soc. 1980,
102, 5974–5976; Seebach’ TADDOLates (1,4-diols): D. Seeb-
ach, A. K. Beck, A. Heckel, Angew. Chem. Int. Ed. Engl. 2001,
40, 92–138 and references cited therein; BINOLates introduced
by Nakai: M. Mori, T. Nakai, Tetrahedron Lett. 1997, 38,
6233–6236; C. F.-Y. Zhang, C.-W. Yip, R. Cao, A. S. C. Chan,
Tetrahedron: Asymmetry 1997, 8, 585–589; Matteson used pin-
anediol derived bidentate ligands (1,2-diols) to construct chiral
boronates for his homologation sequence and obtained excel-
lent diastereoselectivities; D. S. Matteson, R. Ray, J. Am.
Chem. Soc. 1980, 102, 7590–7591.
[3] K. Inoguchi, S. Sakuraba, K. Achiwa, Synlett 1992, 169–178.
[4] H. Waldmann, M. Weigerding, C. Dreisbach, C. Wandrey,
Helv. Chim. Acta 1994, 77, 2111–2116.
[5] Asymmetric reduction of ketones with catecholborane with the
use of 2,6-BODOL complexes of titanium(IV) as catalysts: F. [20] Although orthoester 22 proved surprisingly resistant to typical
Almqvist, L. Torstensson, A. Gudmundsson, T. Frejd, Angew.
Chem. Int. Ed. Engl. 1997, 36, 376–377; I. Sarvary, F. Almqvist,
T. Frejd, Chem. Eur. J. 2001, 7, 2158–2166; 1,3-diols as cata-
lysts for the diethylzinc addition to aldehydes: I. Sarvary, Y.
Wan, T. Frejd, J. Chem. Soc., Perkin Trans. 1 2002, 645–651;
V. Thornqvist, S. Manner, M. Wingstrand, T. Frejd, J. Org.
Chem. 2005, 70, 8609–8612.
hydrolysis conditions, efficient conversion to 25 was realized by
heating for 15 h at reflux with 5% HCl in tetrahydrofuran.
[21] S. Arseniyadis, R. Rodriguez, E. Cabrera, A. Thompson, G.
Ourisson, Tetrahedron 1991, 47, 7045–7058.
[22] Z. G. Hajos, D. R. Parrish, J. Org. Chem. 1973, 38, 3239–3243;
Z. G. Hajos, D. R. W. Parrish, German Patent 2, 102, 623
(Hoffmann-La Roche, July 29, 1971); (Chem. Abstr. 1971, 75,
129414r); U. Eder, G. Sauer, R. Wiechert, Angew. Chem. Int.
Ed. 1971, 10, 496–497.
[23] L. R. Gahan, T. W. Hambley, A. M. Sargeson, M. R. Snow,
Inorg. Chem. 1982, 21, 2699–2706.
[24] S. R. Cooper, Pure Appl. Chem. 1990, 62, 1123–1125.
[25] This trend could be promising insofar as complexation with a
metal is concerned.
[26] The bicyclic aldol framework of 2b offered chemoselectivity
because of the threefold different oxygenation pattern, and
stereoselectivity, because of the fact that chemical modifica-
tions of bridged polycyclic compounds are highly stereoselec-
tive.
[27] Tietze’s requirements (among others): enantiomeric purity in-
stalled at the outset catalytically.
[6] The most commonly employed ways of building up the bicy-
clo[2.2.2]octane system, which have a wide range of applica-
tions in the synthesis of natural products, are described in re-
view articles: V. Singh, Acc. Chem. Res. 1999, 32, 324–333; C.-
C. Liao, R. K. Peddinti, Acc. Chem. Res. 2002, 35, 856–866.
[7] According to Tietze’s classification,[8] the process is a consecu-
tive reaction, as another reagent is added after the first trans-
formation without isolation of the first formed product.L. Fi-
net, J. I. Candela, T. Kaoudi, N. Birlirakis, S. Arseniyadis,
Chem. Eur. J. 2003, 9, 3813–3820.
[8] L. F. Tietze, U. Beifuss, Angew. Chem. Int. Ed. 1993, 32, 131–
163; L. F. Tietze, Chem. Rev. 1996, 96, 115–136; L. F. Tietze,
A. Modi, Med. Research Reviews 2000, 20, 304–322.
[9] 1a: J. I. Candela Lena, M. Rico Ferreira, J. I. Martín Her-
nando, S. Arseniyadis, Tetrahedron: Asymmetry 2001, 12, 3281–
3291; 1b: S. Arseniyadis, D. V. Yashunsky, R. Pereira de Freitas,
M. Muñoz-Dorado, P. Potier, L. Toupet, Tetrahedron 1996, 52,
12443–12458.
[28] I. D. Sanders, J. K. M. Sanders, J. Am. Chem. Soc. 1980, 102,
5703–5711.
[29] C. LeCocq, J.-Y. Lallemand, J. Chem. Soc., Chem. Commun.
1981, 150–152.
[10] It should be pointed out that Frejd et al. experienced a similar
lack of selectivity in the reduction of hydroxybicyclo[2.2.2]oc-
Received: August 17, 2006
Published Online: November 6, 2006
Eur. J. Org. Chem. 2007, 335–341
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
341