One-pot synthesis of chiral dihydrobenzofuran framework via Rh/Pd catalysis

Article abstract of DOI:10.1021/ol302646a

A one-pot synthesis of the chiral dihydrobenzofuran framework is described. The method utilizes Rh-catalyzed asymmetric ring opening (ARO) and Pd-catalyzed C-O coupling to furnish the product in excellent enantioselectivity without isolation of intermedia

Full text of DOI:10.1021/ol302646a

ORGANIC
LETTERS
2012
Vol. 14, No. 21
5542–5545
One-Pot Synthesis of Chiral
Dihydrobenzofuran Framework via
Rh/Pd Catalysis
Gavin Chit Tsui, Jennifer Tsoung, Patrick Dougan, and Mark Lautens*
Davenport Research Laboratories, Department of Chemistry, University of Toronto,
80 St. George Street, Toronto, Ontario, Canada M5S 3H6
mlautens@chem.utoronto.ca
Received September 25, 2012
ABSTRACT
A one-pot synthesis of the chiral dihydrobenzofuran framework is described. The method utilizes Rh-catalyzed asymmetric ring opening (ARO)
and Pd-catalyzed CꢀO coupling to furnish the product in excellent enantioselectivity without isolation of intermediates. Systematic metalꢀligand
studies were carried out to investigate the compatibility of each catalytic system using product enantiopurity as an indicator.
Traditional organic syntheses usually adopt a “stop-
and-go” approach, which involves the purification of
intermediates from each step to ensure that the consecutive
reactions proceed smoothly and avoid various compatibility
issues. The purification processes, however, are often time
consuming, costly, and wasteful. Therefore, a more efficient
“one-pot” approach, in which multiple chemical reactions
are carried out sequentially in a single reaction vessel with-
out the purification of intermediates, has emerged as a
desirable and practical method.1 A report of the total
synthesis of (ꢀ)-oseltamivir (Tamiflu) by three “one-pot”
operations highlights the power of this approach.2 Of course
process chemistry research groups have routinely adopted
this strategy to streamline productions. However, the reali-
zation of the strategy is rare when multiple metal catalysts
and ligands exist within a single vessel.
active neolignans (obtusafuran and kadsurenone)3a and
lithospermic acids,3b which possess stereogenic centers at
the C2 and C3 positions (Figure 1). Although methods
exist in synthesizing dihydrobenzofuran frameworks, their
asymmetric syntheses remain rare.4 Until recently, only
CꢀH activation approaches have been employed in the
total syntheses of (þ)-lithospermic acid to construct the
dihydrobenzofuran core asymmetrically.3b,5
Recently, our group reported examples of a one-pot/
domino synthesis of dihydroquinolines using two transi-
tion-metal catalysts (Rh/Pd and BINAP/X-Phos).6 In a
continuing effort to develop efficient and selective one-pot
reactions using two metal catalytic systems, we envisioned
the use of the Rh-catalyzed asymmetric ring-opening (ARO)
(4) For recent examples of dihydrobenzofuran frameworks, see: (a)
Kurono, N.; Honda, E.; Komatsu, F.; Orito, K.; Tokuda, M. Tetra-
hedron 2004, 60, 1791. (b) Grant, V. H.; Liu, B. Tetrahedron Lett. 2005,
46, 1237. (c) Takeda, N.; Miyata, O.; Naito, T. Eur. J. Org. Chem. 2007,
1491. (d) Torraca, K. E.; Kuwabe, S.-I.; Buchwald, S. L. J. Am. Chem.
Soc. 2000, 122, 12907. (e) Nguyen, R.-V.; Yao, X.; Li, C.-J. Org. Lett.
2006, 8, 2397. (f) Liu, Q.; Han, B.; Zhang, W.; Yang, L.; Liu, Z.-L.; Yu,
W. Synlett 2005, 2248. (g) Bertolini, F.; Pineschi, M. Org. Prep. Proc. Int.
2009, 41, 385. (h) John, J.; Indu, U.; Suresh, E.; Radhakrishnan, K. V.
J. Am. Chem. Soc. 2009, 131, 5042.
(5) O’Malley, S. J.; Tan, K. L.; Watzke, A.; Bergman, R. G.; Ellman,
J. A. J. Am. Chem. Soc. 2005, 127, 13496.
(6) Panteleev, P.; Zhang, L.; Lautens, M. Angew. Chem., Int. Ed.
2011, 50, 9089.
Dihydrobenzofurans are common motifs in natural
products and are the key structural elements in biologically
(1) For reviews on one-pot reactions, see: (a) Albrecht, Ł.; Jiang, H.;
Jørgensen, K. A. Angew. Chem., Int. Ed. 2011, 50, 8492. (b) Climent,
M. J.; Corma, A.; Iborra, S. Chem. Rev. 2011, 111, 1072. (c) Walji, A. M.;
MacMillan, D. W. C. Synlett 2007, 1477.
(2) Ishikawa, H.; Suzuki, T.; Hayashi, Y. Angew. Chem., Int. Ed.
2009, 48, 1304.
(3) (a) Benbow, J. W.; Katoch-Rouse, R. J. Org. Chem. 2001, 66,
4965. (b) Wang, D.-H.; Yu, J.-Q. J. Am. Chem. Soc. 2011, 133, 5767 and
the references therein.
r
10.1021/ol302646a
Published on Web 10/12/2012
2012 American Chemical Society

reaction of oxabicyclic alkenes with 2-halo-arylboron
reagents combined with subsequent Pd-catalyzed intramo-
lecular CꢀO coupling to form chiral dihydrobenzofuran
frameworks in a “one-pot” operation.
chiral dihydrobenzofuran product 5 in 84% yield without
loss of ee (Scheme 1).10,14
Scheme 1. Optimized Conditions for Rh-Catalyzed Asymmetric
Ring-Opening Reaction of Oxabenzonorbornadiene 1 and Pd-
Catalyzed Intramolecular CꢀO Coupling of 4a
Figure 1. Biologically active natural products containing dihy-
drobenzofuran frameworks.
Inanefforttodevelop a one-pot/domino reaction (ARO
followed by CꢀO coupling) for synthesizing the dihydro-
benzofuran product 5 directly from oxabicycle 1 and
boronic ester 3a, we combined the optimal Rh-catalyzed
ARO conditions with Pd-catalyzed intramolecular CꢀO
coupling conditions all in the same vessel at the same time
(Scheme 2).15
We have previously reported Rh(I)- and Pd(II)-cata-
lyzed ring-opening reactions of oxa- and azabicyclic al-
kenes using boronic acids.7 They offer powerful methods
for constructing dihydronaphthalene cores with well-defined
stereogenic centers.8 Furthermore, as Pd- and Cu-cata-
lyzed CꢀO couplings are well-studied processes,9 it is
therefore plausible to develop a “one-pot” protocol by
combining both catalytic systems toward the synthesis of
the chiral dihydrobenzofuran framework.
Scheme 2. Rh-Catalyzed ARO/Pd-Catalyzed CꢀO Coupling
One-Pot/Domino Reaction
Extensive studies were carried out to optimize the ring-
opening reaction using 2-chloro-/bromophenylboronic acids
11
and their ethyleneglycol esters.10 The Rh/(R,S)-PPF-PtBu2
catalyst gave complete conversions,12 and the best yield
(74%) was obtained with ethylene glycol boronic ester 3a
in the presence of catalytic [Rh(COD)Cl]2/(R,S)-PPF-
PtBu2 to afford 4a in 96% ee (Scheme 1). Pd(OAc)2/
X-Phos13 was found to be the optimal catalytic system
for the intramolecular CꢀO coupling step which led to the
(7) (a) Rh(I): Lautens, M.; Dockendorff, C.; Fagnou, K.; Malicki, A.
Org. Lett. 2002, 4, 1311. (b) Pd(II): Lautens, M.; Dockendorff, C. Org.
Lett. 2003, 5, 3695.
(8) For a leading reference and review on Rh-catalyzed ARO reac-
tions, see: (a) Lautens, M.; Fagnou, K. Proc. Natl. Acad. Sci. U.S.A.
2004, 101, 5455. (b) Fagnou, K.; Lautens, M. Chem. Rev. 2003, 103, 169.
(9) For reviews on metal-catalyzed CꢀO/CꢀN couplings, see: (a) Pd:
Surry, D. S.; Buchwald, S. L. Chem. Sci. 2011, 2, 27. (b) Pd: Schlummer,
B.; Scholz, U. Adv. Synth. Catal. 2004, 346, 1599. (c) Cu: Evano, G.;
Blanchard, N.; Toumi, M. Chem. Rev. 2008, 108, 3054.
The substrate was completely consumed, and three
major products were isolated: the desired product 5, the
aromatized product 6, and the ring-opened product 4a.
The presence of 4a showed that the CꢀO coupling step was
impededunder the multiple metal/ligand system. A modest
27% of the desired product 5 was obtained by this domino
process, together with 12% of the aromatized product 6
(a total of 39% of products arising from the domino reaction).
On the other hand, the ARO step was not affected, as the
reaction was complete. However, a significant deterioration of
ee was observed in product 5 (44% vs 96% ee; cf. Scheme 1).
(10) See Supporting Information for full details.
(11) (R,S)-PPF-PtBu2 = (R)-1-[(S)-2-(Diphenylphosphino)ferrocenyl]-
ethyl-di-tert-butylphosphine. It belongs to the Josiphos ligand family.
(12) The use of 2-halo-arylboron reagents has not been successfully
demonstrated in the earlier report using Rh catalyst; see ref 7a.
(13) X-Phos = 2-Dicyclohexylphosphino-20,40,60-triisopropylbiphenyl.
(14) CuI/Phen (1,10-phenanthroline) was an effective catalyst system
for the CꢀO coupling step of the bromo-substituted intermediate (72%
yield), but the chloro-substituted intermediate 4a gave no reaction (see
Supporting Information). Since 2-bromo-arylboron reagents were not
effective in the Rh-catalyzed ring-opening step, and we did not further
pursue the use of a Cu catalyst in combination with Rh in the one-pot
process.
(15) See Supporting Information for details of the one-pot/domino
procedure. Other combinations of metals and ligands including Pd-
(MeCN)2Cl2/dppp/Pd(OAc)2/X-Phos, Pd(MeCN)2Cl2/dppp/X-Phos,
Pd(MeCN)2Cl2/dppp/CuI/Phen, [Rh(COD)Cl]2/Josiphos/palladacycle,
and [Rh(CO)2Cl]2/Josiphos/Pd(OAc)2/X-Phos gave no reaction or only
decomposition.
Org. Lett., Vol. 14, No. 21, 2012
5543

To investigate the origin for this ee loss and the factors
hinderingtheCꢀO couplingstep, competition studieswere
carried out to probe catalyst/ligand combinations that
were interfering with the domino sequence (Tables 1 and 2).
In the Rh-catalyzed ARO reaction (Table 1), added
X-Phos ligand did not affect the ee (entries 1ꢀ2) since
Rh binds selectively to Josiphos over X-Phos. When Pd-
(OAc)2 was added, a severe decrease in ee was observed
(entry 3). A control experiment showed that Pd(OAc)2
alone did not catalyze the reaction (entry 6); thus the ee
deterioration was likely due to the formation of a compet-
ing catalytic species such as Pd/Josiphos. In fact, Pd/
Josiphos was efficient at catalyzing the ring-opening reac-
tion (61% yield) but the ee was much lower (28% ee)
(entry 4). Furthermore, Pd/X-Phos also catalyzed the
reaction to a significant extent, which gave 43% of racemic
product (entry 5). Therefore, the ee loss in the one-pot,
domino process (Scheme 2) was a result of catalytic
activities from three metal complexes: Rh/Josiphos, Pd/
Josiphos, and Pd/X-Phos; the latter two contributed to the
lowering of the ee. The ratio of each species and their
relative catalytic reactivity (rate) is unclear at the moment.
material was consumed (entry 3). The formation of Pd/
Josiphos species has already been demonstrated in the
ARO reaction (cf. Table 1, entry 4). Interestingly, when
both Rh and Josiphos were added, the reaction yield
increased back to previous levels (entry 4). This experiment
showed that Pd/X-Phos and Rh/Josiphos were the pre-
dominant metalꢀligand complexes (may also exist in
equilibrium) where Pd/X-Phos served as an active catalyst
for the CꢀO coupling and Rh/Josiphos was a noninterfer-
ing bystander. Conversion of 4a to 5 was completely
inhibited in the presence of oxabicycle 1 (entries 5ꢀ6). A
plausible explanation may be that the oxabicycle acted as a
ligand which bound to Pd (via the oxygen and alkene
moiety) and created an inactive complex. Finally, adding
boronic ester 3a also reduced the yield (entry 7). This can
be explained if 3a undergoes oxidative addition with Pd(0)
and ties up the catalyst. Or Pd(OAc)2 undergoes transme-
talation with 3a to create a Pd(II)-aryl species that cannot
be reductively eliminated to generate Pd(0), which is
necessary for the CꢀO coupling catalytic cycle. Therefore,
the intramolecular CꢀO coupling step of the domino
sequence (Scheme 2) was severely impeded by the compo-
nents from the ARO step: free (R,S)-PPF-PtBu2, oxabi-
cycle 1, and boronic ester 3a.
Table 1. Competition Studies of the Ring-Opening Reaction
Table 2. Competition Studies of the Intramolecular CꢀO Cou-
pling Reaction
entry
catalyst
ligand
yield (%)a ee (%)
1b [Rh(COD)Cl]2 (5%) (R,S)-PPF-PtBu2 (10%)
2
74
68
96
96
[Rh(COD)Cl]2 (5%) (R,S)-PPF-PtBu2 (10%)
X-Phos (15%)
entry
catalyst
ligand
additive
yield (%)
[Rh(COD)Cl]2 (2.5%) (R,S)-PPF-PtBu2 (5%)
Pd(OAc)2 (7%)
65
24
1b
2
Pd(OAc)2 (7%)
Pd(OAc)2 (7%)
[Rh(COD)Cl]2
(5%)
X-Phos (15%)
X-Phos (15%)
ꢀ
ꢀ
84a
71c
3
4
5
6
Pd(OAc)2 (7%)
Pd(OAc)2 (7%)
Pd(OAc)2 (7%)
(R,S)-PPF-PtBu2 (5%)
61
43
<5c
28
ꢀ
X-Phos (15%)
ꢀ
3
4
5
Pd(OAc)2 (7%)
X-Phos (15%)
(R,S)-PPF-
ꢀ
ꢀ
12c
77c
<5
ꢀ
a Isolated yield. b cf. Scheme 1. c Starting material recovery.
PtBu2 (10%)
X-Phos (15%)
(R,S)-PPF-
Pd(OAc)2 (7%)
[Rh(COD)Cl]2
(5%)
Pd(OAc)2 (7%)
[Rh(COD)Cl]2
(5%)
PtBu2 (10%)
X-Phos (15%)
(R,S)-PPF-
PtBu2 (10%)
X-Phos (15%)
Similar competition experiments were also conducted
for the intramolecular CꢀO coupling reaction (Table 2).
The added Rh-catalyst had a small effect on the yield
(entries 1ꢀ2). Therefore, the only species that catalyzed the
CꢀO coupling was Pd/X-Phos. However, when Josiphos
was added, a poor yield was obtained, probably because
both ligands competed to bind with Pd and Pd/Josiphos
was an inactive species which reduced the rate the starting
oxabicycle 1
(1.0 equiv)
6
7
Pd(OAc)2 (7%)
oxabicycle 1
(1.0 equiv)
boronic ester
3a (1.0 equiv)
<5
Pd(OAc)2 (7%)
X-Phos (15%)
30c
a Isolated yield. b cf. Scheme 1. c Determined by H NMR spectro-
scopy of the crude material.
1
(16) Rh-catalyzed ARO reaction of 6: [Rh(COD)Cl]2 (2.5 mol %),
(R,S)-PPF-PtBu2 (5 mol %), Cs2CO3 (5 M in H2O) (1 equiv), THF,
75 °C, 15 h; 17% yield of 7 (34% S.M. rec.), 6% ee.
If the ARO step gives complete conversion and equal
molar amounts of substrate and reagent are used, then
oxabicycle 1 and boronic ester 3a should not be present
after the ring-opening reaction. And because Rh/Josiphos
did not interfere in the CꢀO coupling step, a one-pot, two-
step protocol that bypasses the workup and purification of
(17) For a recent review on the synthetic methods of indoline scaf-
folds, see: Liu, D.; Zhao, G.; Xiang, L. Eur. J. Org. Chem. 2010, 3975.
(18) For examples of synthesizing chiral indolines, see: (a) Kinetic
resolution: Arp, F. O.; Fu, G. C. J. Am. Chem. Soc. 2006, 128, 14264. (b)
Pd-catalyzed asymmetric hydrogenation: Wang, D.-S.; Chen, Q.-A.; Li,
W.; Yu, C.-B.; Zhou, Y.-G.; Zhang, X. J. Am. Chem. Soc. 2010, 132,
8909. (c) Direct asymmetric hydrosilylation: Xiao, Y.-C.; Wang, C.;
Yao, Y.; Sun, J.; Chen, Y.-C. Angew. Chem., Int. Ed. 2011, 50, 10661.
5544
Org. Lett., Vol. 14, No. 21, 2012

the intermediate would be possible. To this end, we carried
out the Rh-catalyzed ARO reaction, let it reach completion,
and then added the Pd-catalyst and ligand dissolved in the
same solvent, and the reaction was continued. We were
able to isolate the final product 5 in 67% yield (over two
steps) and in 96% ee (Scheme 3). This result was equally
efficient as the two-pot, two-step process (ARO: 74%
yield, 96% ee; CꢀO coupling: 84% yield), but without
the extra workup and purification step. The structure of 5
was unambiguously confirmed by X-ray crystallography
(see Supporting Information). In fact, even with X-Phos
present intheARO step, intermediate4awasstill produced
with 100% conversion, and subsequent addition of Pd-
(OAc)2 to the same reaction vessel started the CꢀO
coupling reaction to give product 5 in 33% yield.
in high enantioselectivity (96% ee) by a one-pot procedure
utilizing a Rh-catalyzed asymmetric ring-opening reaction
(ARO) followed by a Pd-catalyzed intramolecular CꢀO
coupling reaction. This protocol was equally efficient
compared to the combined individual reactions, but with-
out any workup or isolation of the intermediate. The ring-
opening/CꢀN coupling sequencecan also be applied to the
azabicycle substrate 6 for synthesizing indoline product 8.
Furthermore, systematic competition studies were carried
out to investigate the interfering components in the one-
pot/domino process utilizing the enantioselectivity of the
ARO reactionasanindicator. Thisnovel approach toward
metalꢀligand studies could serve as a new model for
studying different types of multimetal catalyzed domino
reactions. Further expansion of the substrate scope and
derivatization of the dihydrobenzofuran product toward
valuable structural motifs are currently being investigated.
Scheme 3. Asymmetric One-Pot Synthesis of Dihydrobenzo-
furan 5 via Rh-Catalyzed ARO/Pd-Catalyzed Intramolecular
CꢀO Coupling Sequence
Scheme 4. Racemic Synthesis of Indoline 8 via Pd-Catalyzed
Ring-Opening Reaction/Pd-Catalyzed Intramolecular CꢀN
Coupling
Acknowledgment. We gratefully thank NSERC, Merck
Frosst Canada and Merck for an Industrial Research
Chair. We also thank the University of Toronto for
support of our program and Solvias AG for the generous
gift of chiral ligands.
We also attempted to synthesize indoline 8 from azabi-
cycle 6 using similar protocols, but the Rh-catalyzed ARO
step was low yielding and not enantioselective.16 However,
the racemic ring-opened product 7 was obtained in good
yield with the Pd/dppp catalyst. Intramolecular CꢀN
coupling using Pd/X-Phos afforded the desired indoline
product 8 (Scheme 4). Indolines are common structural
motifs in natural alkaloids and many biologically active
molecules.17 Our approach offered a new method to con-
struct the indoline scaffold in a diastereoselective manner.18
In conclusion, we have successfully demonstrated that
chiral dihydrobenzofuran framework 5 can be synthesized
Supporting Information Available. Experimental pro-
cedures and full characterization for all compounds. This
material is available free of charge via the Internet at
http://pubs.acs.org.
The authors declare no competing financial interest.
Org. Lett., Vol. 14, No. 21, 2012
5545