Pd-catalyzed asymmetric intermolecular hydroalkoxylation of allene: An entry to cyclic acetals with activating group-free and flexible anomeric control

Article abstract of DOI:10.1021/ja508587f

A ligand-directed metal-catalyzed asymmetric intermolecular hydroalkoxylation of alkoxyallene is reported. Combined with ring-closing-metathesis, this reaction offers a new atom-efficient synthetic method toward various cyclic acetals with elaborate anomeric control. Synthetic utility of the reaction was demonstrated by the atom-efficient and stereodivergent access to various mono- and disaccharides.

Full text of DOI:10.1021/ja508587f

Communication
pubs.acs.org/JACS
Pd-Catalyzed Asymmetric Intermolecular Hydroalkoxylation of
Allene: An Entry to Cyclic Acetals with Activating Group-Free
and Flexible Anomeric Control
Wontaeck Lim, Jungjoon Kim, and Young Ho Rhee*
Department of Chemistry, Pohang University of Science and Technology, Hyoja-dong San 31, Pohang, Kyungbook,
Republic of Korea 790-784
S
* Supporting Information
on the nature of the pre-existing substituents. The proposed
ABSTRACT: A ligand-directed metal-catalyzed asymmet-
ric intermolecular hydroalkoxylation of alkoxyallene is
reported. Combined with ring-closing-metathesis, this
reaction offers a new atom-efficient synthetic method
toward various cyclic acetals with elaborate anomeric
control. Synthetic utility of the reaction was demonstrated
by the atom-efficient and stereodivergent access to various
mono- and disaccharides.
method is distinguished from the conventional ones in that the
activating groups are not needed. In addition, the diversity can
be easily pursued with regard to the ring size and the anomeric
configuration of the cyclic acetal. Thus, this strategy offers a
highly flexible and chemically efficient approach toward cyclic
acetals, and can open up new possibilities for the structurally
and stereochemically divergent synthesis of mono- and
oligosaccharides.6 For the efficient flexible anomeric control,
the stereoinduction in the hydroalkoxylation step should be
independent of the pre-existing substituents in the alcohol
moiety. Upon the basis of our own experience in the
hydroamination7 and related studies,8 we envisioned that the
metal-catalyzed hydroalkoxylation invoking π-allyl intermedi-
ates would be highly suitable for this transformation.
Unlike various carbon- and nitrogen nucleophiles, however,
unactivated aliphatic alcohols are known as a poor nucleo-
phile in asymmetric π-allyl chemistry.9−12 In fact, this concern
was justified in our initial optimization (Table 1). As shown in
entry 1, the reaction of enantioenriched alcohol 1 with
n-pentoxyallene 3 (10 equiv) in the presence of chiral Trost
ligand 2 (6.25 mol %) and Pd(OAc)2 (5 mol %) in CH2Cl2
showed no conversion (entry 1). Quite interestingly, the
reaction employing sterically bulkier cyclohexyloxyallene 4
slowly proceeded to give a mixture of acyclic acetal 5a and 5b
in notable 67% yield after 4 days.13 The subsequent RCM
proceeded smoothly to give cyclic acetals 6a and 6b in 97%
yield with 10:1 ratio (65% yield over two steps).14 Upon the
basis of the 1H NMR analysis, the absolute configuration of the
anomeric carbon in 6a center was tentatively assigned to be (S)
(entry 2, for the rigorous determination of the anomeric
configuration, see below). Encouraged by this preliminary
result, we continued the optimization process. Concentration of
the solution (entry 3) as well as employing Pd2(dba)3 as the
precatalyst increased the yield (entry 4), even though the
reactions were still very slow. After extensive studies to speed
up the reaction, we discovered a remarkable solvent effect. For
example, the use of polar DMF almost stopped the reaction
(entry 5), while using nonpolar toluene dramatically reduced
the reaction time without harming the selectivity (entry 6). In
the latter case, the hydroalkoxylation reaction was completed
within 4 h to give 5a as the exclusive diastereomer in 90% yield
symmetric formation of carbon−oxygen bond represents
A
one of the most fundamental transformations in synthetic
organic chemistry. Metal-catalyzed addition of alcohols to
allenes (hydroalkoxylation) is considered as a most powerful
method, due to the excellent chemical efficiency and the rich
functional groups installed in the product.1 In most cases, the
scope has been discussed within the realm of oxacycle synthesis
which exploits intramolecular reaction.2 However, the more
challenging intermolecular asymmetric hydroalkoxylation of
allene, to our best knowledge, still remains unknown. In this
context, we envisioned that the chiral ligand-directed addition
of olefinic alcohol to alkoxyallene would generate enantioen-
riched acyclic mixed acetal A (Method A, Scheme 1). The
Scheme 1. Basic Concept
subsequent RCM (ring-closing-metathesis) reaction3,4 would
give the cyclic acetal B. This unprecedented pathway proposes
a unique way to control the anomeric information on the cyclic
acetals, which is of crucial importance in delineating the
structure and function of structurally complex oligosaccharides
and glycoconjugates.5 In conventional approaches for the
anomeric control, synthesis of the oxacyclic framework pre-
cedes formation of the anomeric center (Method B, Scheme 1).
This method generally requires stoichiometric amount of
activating groups and extensive protective group strategies.
Moreover, controlling the anomeric configuration relies heavily
Received: August 20, 2014
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja508587f | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
a
a,b
Table 1. Optimization
Table 2. Scope of the Alcohol
b
c
,
d
e
entry
allene
(eq)
method
solvent
conc. time yield
ratio
(M)
(h)
(%)
(6a:6b)
1
2
3
4
5
6
7
8
3 (10)
4 (10)
4 (10)
4 (10)
4 (10)
4 (10)
4 (2)
A
A
A
B
B
B
B
B
B
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
DMF
0.05
0.05
0.5
0.5
0.5
0.5
0.5
0.5
0.5
96 No Rxn
96 65
-
10:1
10:1
10:1
2.4:1
>20:1
>20:1
>20:1
1:15
96 86
96 92
96 10
Toluene
Toluene
Toluene
Toluene
4
4
6
4
90
93
82
95
4 (1.2)
4 (2)
f
9
a
For more detailed information on the optimization, see the SI.
Method A: Pd(OAc)2 (5 mol %) and ligand 2 (6.25 mol %) were
b
a
used. Method B: Pd2(dba)3 (2.5 mol %) and ligand 2 (5 mol %) were
Typical procedure: (i) first step, a mixture of Pd2(dba)3, ligand 2,
Et3N (1.5 equiv), alkoxyallene 4 (2 equiv), and the starting material (1
equiv) was reacted in toluene (0.5 M) at 40 °C. (ii) Second step: a
mixture of acyclic acetal and 1st generation Grubbs catalyst (5 mol %)
c
d
used. Combined yield for two steps. In all cases, the RCM reaction
e
proceeded in >95% yield. Determined by the integration of 1H NMR
f
spectrum. ligand (S,S)-2 was used.
b
c
was reacted at rt. R′ = c-C6H11. Method A: Pd (2.5 mol %)/(R,R)-
2 (5 mol %). Method B: Pd (2.5 mol %)/(S,S)-2 (5 mol %). Method
C: Pd (3.5 mol %)/(R,R)-2 (7 mol %). Method D: Pd (3.5 mol %)/
(over two steps). Notably, reducing the amount of alkoxyallene
4 to 2 equiv under this condition still showed full conversion
within 4 h (entry 7). Further reduction of allene 4 to 1.2 equiv
gave the product 6a in 82% yield (entry 8). As we expected, the
use of enantiomeric ligand (S,S)-2 had no adverse effect in the
yield and selectivity to give the diastereomeric acyclic acetal 5b,
which could be uneventfully transformed into cyclic acetal 6b
by RCM (entry 9).15 This result corroborates our hypothesis
that the stereoselectivity in the hydroalkoxylation step is
controlled by the chiral information on the ligand.
With the optimized conditions (entries 7 and 9 in Table 1) in
hand, we explored an array of substrates for the two-step
synthesis of diverse anomerically well-defined cyclic acetals
(Table 2). As shown in entries 1 and 2, the substrate 7
possessing benzyloxy group at the neighboring carbon gave
diastereomeric cyclic acetal products (8a and 8b) in good yield
with high selectivity (entries 1 and 2). In addition, the scope of
the reaction was successfully extended to the synthesis of
various 7-membered cyclic acetals (entries 3−6). In some cases,
higher loading of Pd2(dba)3 catalyst (3.5 mol %) was required
for the complete conversion in the hydroalkoxylation step
(entries 3−4 and 6). Nevertheless, the subsequent RCM
reaction using 5 mol % Grubbs catalyst proceeded consistently
to give the cyclic acetal products (10 and 12) in good yield with
no decrease in the diastereoselectivity. Then, we examined
substrates 13 and 15 possessing additional alkoxy substituent at
the allylic position to see if the current method can be used for
the synthesis of cyclohexyl glycoside forms of 2,3,6-trideox-
ysugars such as amicetose and rhodinose. As described in en-
tries 7 and 8, employing the two-step protocol with 3.5 mol %
Pd2(dba)3 and the subsequent RCM converted the alcohol 1316
into the dehydrogenated forms of amicetose glycosides 14a and
14b in good yield and stereoselectivity. It should be noted that
d
e
(S,S)-2 (7 mol %). Combined yield for two steps. Determined by
f
1
the integration of H NMR spectrum after the RCM reaction. 0.10
g
equiv of Et3N was used. 4 equiv of allene was used.
both anomeric configurations could be accessed with
comparable chemical efficiency.17 Additionally, the reaction
was successfully expanded to the stereodivergent synthesis of
dehydrogenated rhodinose glycosides 16a and 16b, which
proceeded from the alcohol 15 in good yield with >10:1
stereoselectivity (entries 9 and 10). In these cases, complete
conversion in the hydroalkoxylation required the use of 4 equiv
of alkoxyallene 4.
Next, we investigated achiral olefinic alcohols 17−19 to
rigorously determine the enantioselectivity of the ligand-directed
hydroalkoxylation reaction. As shown in eq 1, 5- to 7-membered
cyclic acetals 23−25 were accessed in high yields and
enantioselectivity by using the two-step protocol shown in
Table 1, via the corresponding acyclic acetal intermediates 20−
22 (For the determination of the ee and absolute configuration
of compounds 23−25, see the Supporting Information (SI)).
The yield and enantioselectivity of the cyclic acetals did not
vary significantly. Thus, this result firmly establishes the
generality of the asymmetric hydroalkoxylation reaction.
B
dx.doi.org/10.1021/ja508587f | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
From a mechanistic viewpoint, the current result is par-
ticularly noteworthy because unactivated aliphatic alcohols have
been considered as a poor nucleophile in asymmetric π-ally
chemistry. As depicted in Scheme 2, the initially formed Pd−H
Scheme 3. Stereoselective Dihydroxylation: Monosaccharide
Synthesis
Scheme 2. Mechanism
Scheme 4. Synthesis of Disaccharides Containing 2,3,6-
Tri-deoxysugar
species will generate kinetically favored anti π-ally complex
associated with alkoxide counteranion.7,8 High level of
enantioselectivity for (S)-configuration with the use of (R,R)-2
ligand can be reasonably explained by the fast outer-sphere
addition of the alkoxide anion to the anti π-ally complex relative
to the π−σ−π equilibration process. Key observations presented
in the optimization are consistent with this hypothesis. For
example, using nonpolar solvent toluene may render a tighter ion
pair between kinetic π-allyl complex and the alkoxide anion, thus
fastening the attack of the alkoxide ion.18Particularly noteworthy
is the size effect of the alkoxy moiety in the allene species. Higher
yield and selectivity with bulkier alkoxyallene 4 seems apparently
confusing because increasing the size of alkoxy moiety should
shift the π−σ−π equilibrium toward the more stable syn π-allyl
complex. However, we point out that the bulky alkoxy moiety
may kinetically slow the equilibration process because the σ-allyl
intermediate should be more sterically congested than the
π-ally complexes.
From a synthetic viewpoint, the current method produces
highly valuable building blocks that can be potentially used for the
synthesis of various monosaccharides. To address this issue, we
examined the stereoselective dihydroxylation of various cyclic
acetals obtained in this work (Scheme 3). As summarized in eq 2,
compounds 23−25 obtained from achiral olefinic alcohols gave
the trans-dihydroxylated products 26−28 in high yield with
excellent (>25:1) stereoselectivity.19,20 Furthermore, dihydroxyla-
tion of the cyclic pyran acetal compounds 6a and 8a showed
excellent diastereoselectivity to give cyclohexyl 4,6-dideoxyallo-
glycoside 29 and 4-deoxyalloglycoside 30 (eq 3). The acetal-
directed diastereoselectivity is also demonstrated by the reaction
of substrates 6b and 8b, which generated the corresponding 4,6-
dideoxymannoglycoside 31 and 4-deoxymannoglycoside 32 with
high diastereoselectivity.21
In addition to the flexible synthesis of monosaccharides
presented above, the current method could be expanded for the
synthesis of oligosaccharides containing 2,3,6-trideoxysugars,
which are found in numerous bioactive natural products22
(Scheme 4). They represent an attractive synthetic challenge,
due to the stereochemical diversity of the anomeric configura-
tion as well as the instability of the 2,3,6-trideoxysugar moiety.
However, it is well-known that the stereodivergent anomeric
control based upon the conventional (substrate-controlled)
reactions is extremely difficult.6a,23
To our delight, the coupling reaction of the easily available
alcohol 3316 and the deoxysugar-derived alkoxyallene 3416 in
the presence of ligand (R,R)-2 gave the disaccharide 35a in
95% yield with >25:1 selectivity (Scheme 4). Switching the
ligand to enantiomeric (S,S)-2 gave the diastereomeric
disaccharide 35b again in comparable yield with 11:1
selectivity.21 These examples firmly establishes the potential
utility of the conceptually new chiral ligand-driven stereo-
induction in anomerically flexible oligosaccharide synthesis.
In summary, we developed a highly efficient synthesis of
cyclic allylic acetal using the Pd-catalyzed asymmetric
intermolecular hydroalkoxylation as the key method for the
unique anomeric control. The synthetic utility of the proposed
method was demonstrated by the anomerically flexible
synthesis of various mono- and disaccharides. Currently, we
are working on expanding the scope of the proposed reaction in
the flexible and chemoselective synthesis of various oligosac-
charides. The result of these efforts will be reported in a due
course.
C
dx.doi.org/10.1021/ja508587f | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
H.; Ye, K.-Y.; Wu, Q.-F.; Dai, L.-X.; You, S.-L. Adv. Synth. Catal. 2012,
354, 1084. (d) Kiener, C. A.; Shu, C.; Incarvito, C.; Hartwig, J. F. J.
Am. Chem. Soc. 2003, 125, 14272.
ASSOCIATED CONTENT
■
S
* Supporting Information
Experimental data and copies of 1H and 13C spectra for all new
compounds. This material is available free of charge via the
Internet at http://pubs.acs.org.
(11) For sporadic examples describing metal-catalyzed asymmetric
allylic substitution reaction using aliphatic alcohols as a nucleophile,
see: (a) Roggen, M.; Carreira, E. M. Angew. Chem., Int. Ed. 2011, 50,
5568. (b) Lam, F. L.; Au-Yeung, T. T.-L.; Kwong, F. Y.; Zhou, Z.;
Wong, K. Y.; Chan, A. S. C. Angew. Chem., Int. Ed. 2008, 47, 1280.
(c) Evans, P. A.; Leahy, D. K.; Andrews, W. J.; Uraguchi, D. Angew.
Chem., Int. Ed. 2004, 43, 4788. (d) Shu, C.; Hartwig, J. F. Angew.
Chem., Int. Ed. 2004, 43, 4794.
AUTHOR INFORMATION
Corresponding Author
yhrhee@postech.ac.kr
■
(12) For selected examples on the metal-catalyzed nonenantiose-
lective hydroalkoxylation of allene, see: (a) Cui, D. M.; Zheng, Z.-L.;
Zhang, C. J. Org. Chem. 2009, 74, 1426. (b) Kinderman, S. S.;
Doodeman, R.; Van Beijma, J. W.; Russcher, J. C.; Tjen, K. C. M. F.;
Kooistra, T. M.; Mohaselzadeh, H.; Van Maarseveen, J. H.; Hiemstra,
H.; Schoemaker, H. E.; Rutjes, F. P. J. T. Adv. Syn. Catal. 2002, 344,
736. (c) Doodeman, R.; Rutjes, F. P. J. T.; Hiemstra, H. Tetrahedron
Lett. 2000, 41, 5979.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by the National Research Founda-
tion funded by the Korean government (NRF-2013R1A2A2A
01068684).
(13) Unlike the acylic acetals 5a and 5b, cyclic acetals 6a and 6b
1
could be easily resolved by H NMR. The characteristic anomeric
REFERENCES
■
methine hydrogen peak appears at 3.85 ppm (for compound 6a) and
at 4.07 ppm (for compound 6b).
(14) The reaction at high temperature (above 50 °C) gave the
(1) For some reviews on the metal-catalyzed hydrofunctionalization,
see: (a) Le Bras, J.; Muzart, J. Chem. Soc. Rev. 2014, 43, 3003.
(b) Munoz, M. P. Chem. Soc. Rev. 2014, 43, 3164. (c) Krause, N.;
Winter, C. Chem. Rev. 2011, 111, 1994. (e) Patil, N. T.; Yamamoto, Y.
Chem. Rev. 2008, 108, 3395.
product in significantly lower selectivity.
(15) The high selectivity confirms that the RCM step does not harm
the selectivity obtained in the hydroalkoxylation step.
(16) For the detailed procedure for the preparation of 13, 15, 33, and
34, see the SI.
(17) The structure of compound 14a was unambiguously confirmed
by an alternative synthesis starting from known the acetyl derivative of
ent-14a (R = Ac) by deacetylation and the subsequent MEM
protection (see the SI). For the reference, see: Chen, P.; Lin, L.
Tetrahedron 2013, 69, 10045.
(18) For an early example describing a similar effect of toluene in the
Pd-catalyzed asymmetric allylic substitution with phenol, see: Trost, B.
M.; Toste, F. D. J. Am. Chem. Soc. 1999, 121, 4545.
(19) The ee of compounds 23−25 was determined by the chiral
HPLC analysis of the corresponding bis-benzyl ethers. For the detailed
information, see the SI.
(20) For a related report on the racemic synthesis of dihydroxylated
cyclic acetals, see Yu, X. M.; Han, H.; Blagg, B. S. J. J. Org. Chem. 2005,
70, 5599.
(2) For some recent examples on the intramolecular metal-catalyzed
asymmetric hydroalkoxylation of allenes, see: (a) Wang, Y. M.;
Lackner, A. D.; Toste, F. D. Acc. Chem. Res. 2014, 47, 889. (b) Yu, S.;
Ma, S. Angew. Chem., Int. Ed. 2012, 51, 3074. (c) Zhang, Z.;
Widenhoefer, R. A. Angew. Chem., Int. Ed. 2007, 46, 283. (d) Hamilton,
G. L.; Eun, J. K.; Mba, M.; Toste, F. D. Science 2007, 317, 496.
(3) For selected recent reviews on the RCM reaction, see: (a) Deiters,
A.; Martin, S. F. Chem. Rev. 2004, 104, 2199. (b) Grubbs, R. H.; Miller,
S. J.; Fu, G. C. Acc. Chem. Res. 1995, 28, 446.
(4) For a seminal study describing the oxacycle synthesis by way of
the ring closing metathesis of cyclic acetals, see: Fu, G. C.; Grubbs, R.
H. J. Am. Chem. Soc. 1992, 114, 5426.
(5) For some recent reviews on the oligosaccharide synthesis, see:
(a) Satoh, H.; Manabe, S. Chem. Soc. Rev. 2013, 42, 4297. (b) Crich,
D. Acc. Chem. Res. 2010, 43, 1144. (c) Zhu, X.; Schmidt, R. R. Angew.
Chem., Int. Ed. 2009, 48, 1900. (d) Smoot, J. T.; Demchenko, A. V.
Advances in Carbohydrate Chemistry and Biochemistry; Derek, H., Ed.;
Academic Press: 2009; Vol. 62, pp 161.
(6) (a) For some examples on the use of Pd-catalyzed allylic
substitution in stereodivergent glycosylation, see: O’Douherty, G. A.
Advances in Carbohydrate Chemistry and Biochemistry; Derek, H., Ed.;
Academic Press: New York, 2013; Vol. 69, pp 55. Also see: Kim, H.;
Men, H.; Lee, C. J. Am. Chem. Soc. 2004, 126, 1336. (b) For a recent
example describing stereoselective ring-closing glycosylation, see: Liu,
H.; Li, X. J. Org. Chem. 2014, 79, 5834.
(7) (a) Kim, H.; Rhee, Y. H. J. Am. Chem. Soc. 2012, 134, 4011.
(b) Kim, H.; Lim, W.; Im, D.; Kim, D.; Rhee, Y. H. Angew. Chem., Int.
Ed. 2012, 51, 12055. (c) Kim, H.; Rhee, Y. H. Synlett 2012, 23, 2875.
(8) (a) Trost, B. M.; Xie, J.; Sieber, J. D. J. Am. Chem. Soc. 2011, 133,
(21) For the structural determination of 29, 31, 35a, and 35b via the
1
analysis of 2-D H NMR spectra, see the SI.
(22) (a) Herold, K.; Gollmick, F. A.; Groth, I.; Roth, M.; Menzel, K.
D.; Mollmann, U.; Grafe, U.; Hertweck, C. Chem.Eur. J. 2005, 11,
̈
̈
5523. (b) Chakraborty, T. K.; Goswami, R. K.; Sreekanth, M.
Tetrahedron Lett. 2007, 48, 4075. (c) Kawamura, N.; Sawa, R.;
Takahashi, Y.; Sawa, T.; Kinoshita, N.; Naganawa, H.; Hamada, M.;
Takeuchi, T. J. Antibiot. 1995, 48, 1521.
(23) For a recent report describing this issue, see: Chen, J.-H.; Ruei,
J.-H.; Mong, K.-K. T. Eur. J. Org. Chem. 2014, 1827.
20611. (b) Trost, B. M.; Simas, A. B. C.; Plietker, B.; Jakel, C.; Xie, J.
̈
Chem.–Eur. J. 2005, 11, 7075. (c) Trost, B. M.; Jakel, C.; Plietker, B. J.
̈
Am. Chem. Soc. 2003, 125, 4438. (d) Krautwald, S.; Schafroth, M. A.;
Sarlah, D.; Carreira, E. M. J. Am. Chem. Soc. 2014, 136, 3020.
(9) For some recent reviews on the metal-catalyzed asymmetric
allylic substitution; (a) Takemoto, Y.; Miyabe, H. In Catalytic
Asymmetric Synthesis; John Wiley & Sons, Inc.: Hoboken, NJ, 2010;
p 227. (b) Trost, B. M. Org. Process Res. Dev. 2012, 16, 185. (c) Trost,
B. M.; Zhang, T.; Sieber, J. D. Chem. Sci. 2010, 1, 427. (d) Mohr, J. T.;
Stoltz, B. M. Chem.Asian J. 2007, 2, 1476.
(10) For some recent examples on the asymmetric allylic
etherification of phenol, see: (a) Trost, B. M.; Rao, M.; Dieskau, A.
P. J. Am. Chem. Soc. 2013, 135, 18697. (b) Kim, D.; Reddy, S.; Singh,
O. V.; Lee, J. S.; Kong, S. B.; Han, H. Org. Lett. 2013, 15, 512. (c) He,
D
dx.doi.org/10.1021/ja508587f | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX