Catalytic functionalization of unactivated sp3 C-H bonds via exo -directing groups: Synthesis of chemically differentiated 1,2-Diols

Article abstract of DOI:10.1021/ja3082186

We describe a Pd-catalyzed site-selective functionalization of unactivated aliphatic C-H bonds, providing chemically differentiated 1,2-diols from monoalcohol derivatives. The oxime was employed as both a directing group (DG) and an alcohol surrogate for this transformation. As demonstrated in a range of substrates, the C-H bonds β to the oxime group are selectively oxidized. Besides activation of the methyl groups, methylene groups (CH₂) in cyclic substrates and methine groups (CH) at bridge-head positions can also be functionalized. In addition, an intriguing oxidative skeleton rearrangement was observed using the menthol-derived substrate. The use of exo-directing groups in C-H activation, as illustrated in this work, would potentially open doors for the discovery of new transformations and new cleavable DGs.

Full text of DOI:10.1021/ja3082186

Communication
pubs.acs.org/JACS
Catalytic Functionalization of Unactivated sp3 C−H Bonds via exo-
Directing Groups: Synthesis of Chemically Differentiated 1,2-Diols
Zhi Ren,† Fanyang Mo,† and Guangbin Dong*
Department of Chemistry and Biochemistry, University of Texas at Austin, Austin, Texas 78712, United States
S
* Supporting Information
Scheme 1. Alcohol or Masked-Alcohol Directed C−H Bond
Functionalization (Ac, Acetate)
ABSTRACT: We describe a Pd-catalyzed site-selective
functionalization of unactivated aliphatic C−H bonds,
providing chemically differentiated 1,2-diols from mono-
alcohol derivatives. The oxime was employed as both a
directing group (DG) and an alcohol surrogate for this
transformation. As demonstrated in a range of substrates,
the C−H bonds β to the oxime group are selectively
oxidized. Besides activation of the methyl groups,
methylene groups (CH2) in cyclic substrates and methine
groups (CH) at bridge-head positions can also be
functionalized. In addition, an intriguing oxidative skeleton
rearrangement was observed using the menthol-derived
substrate. The use of exo-directing groups in C−H
activation, as illustrated in this work, would potentially
open doors for the discovery of new transformations and
new cleavable DGs.
ite selectivity represents one of the grand challenges for
1
Scatalytic functionalization of aliphatic C−H bonds. Elegant
approaches have been developed for selective activation of
specific C−H bonds based on intrinsic reactivity;2 to date,
directing groups (DGs), broadly defined, still serve as highly
effective tools for controlling which sp3 C−H bond to activate.3
Among various DGs, those derived from alcohols or masked
alcohols are particularly attractive due to the ubiquity of hydroxyl
groups. Using alcohols or masked alcohols as DGs for activation
of more reactive aryl and allylic C−H bonds has been
demonstrated by Yu,4 Hartwig,5 and White;6 however, limited
success has been made with unactivated sp3 C−H bonds.7 It is
well documented that the C−H bonds at the δ-position of
alcohols can be functionalized through a 1,5-H shift using an
oxygen-centered radical approach.8 Du Bois et al. have developed
carbamate- and sulfamate-directed nitrene insertions to access
1,2 or 1,3 amino alcohol moieties (Scheme 1A).9 Recently, Baran
disclosed an efficient four-step one-pot sequence to convert a
C−H activation (such as metal-oxo insertion);2 (2) designing
DGs for such a site-selective reaction is nontrivial because, aside
from serving as a good ligand for transition-metal catalysts, the
DG should be easily removable to restore the alcohol. To address
these challenges, here, we report our development of a Pd-
catalyzed functionalization of the unactivated β sp3 C−H bonds,
where oximes are employed as unusual exo-DGs.
We postulated that properly designed oximes would likely
serve as both an alcohol surrogate and a DG for metal-catalyzed
site-selective functionalization of β C−H bonds. Oxime-directed
catalytic C−H activation was first reported by Sanford, where
oximes serve as a carbonyl equivalent directing oxidation of the
sp3 C−H bonds on the nitrogen side (eq 1).12In this case, an
endo-palladacycle is formed through coordination with the oxime
nitrogen. In contrast to the endo-metallocycles (π-bond of the
DG inside the metallocycle), formation of the exo-metallocycles
(π-bond of the DG outside the metallocycle) is rare in catalytic
reactions13 and often less competitive when formation of an
endo-cycle is possible (Figure 1).14 However, we hypothesized
̋
trifluoromethyl carbamate to a 1,3-diol using a Hofmann−Loffler
reaction-like approach (Scheme 1B).10 Most recently, Hartwig
reported a powerful approach to convert monoalcohols to 1,3-
diols using Ir-catalysts (Scheme 1C).1c
We have been interested in 1,2-diol synthesis directly from
monoalcohols or masked alcohols via catalytic activation of
unactivated sp3 C−H bonds, which, to our knowledge, has not
been realized previously.11 The challenge is twofold: (1)
compared to the γ-position that gives 1,3-diols, the β-position
of an alcohol is relatively electron-deficient due to the inductive
effect of the oxygen and is, thus, less reactive toward electrophilic
Received: August 18, 2012
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja3082186 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
a
Table 1. Selected Optimization of Reaction Conditions
bd
,
entry
variation from the “standard” conditions
yield
c
Figure 1. endo-Metalation vs exo-Metalation.
1
none
67% (61% )
0%
2
no Pd(OAc)2
that, by blocking or obviating potential reaction sites on the nitrogen
site of oximes, selective activation of the C−H bonds on the alcohol
site should become feasible (Scheme 2). Thus, using such an oxime
3
5 mol % Pd(OAc)2
no PhI(OAc)2
50%
0%
4
5
no Ac2O
61%
50%
7%
6
H2O instead of Ac2O
chlorobenzene instead of HOAc
toluene instead of HOAc
1,2-dichloroethene instead of HOAc
K2S2O8 instead of PhI(OAc)2
add 0.1 equiv of LiOAc
0.25 mL of HOAc
1 mL of HOAc
Scheme 2. Our Strategy Design
7
8
7%
9
10%
24%
67%
53%
52%
0%
10
11
12
13
14
15
use “DG A”
use “DG B”
31%
a
Reaction conditions: all the reactions were run on a 0.1 mmol scale
b
with 0.5 mL of solvents in 2 h. NMR yield determined using 1,1,2,2-
tetrachloroethane as internal standard. Isolated yield. Product 1b
c
d
as a DG, a five-membered exo-palladacycle (intermediate B) is
expected to be generated via metalation of the β C−H bond.
Subsequent oxidation of intermediate B would provide the
desired 1,2-diol motif (C). Note that the resulting diols would be
chemically differentiated, because removal of the acetate group
and cleavage of the oxime N−O bond can be operated under
orthogonal reaction conditions (vide infra).
To test our hypothesis, oxime-derived 2-butanol (1a) was
employed as the initial substrate. Having optimized the reaction
conditions, we were delighted to find that the desired protected
1,2-diol (1b) was obtained in 67% yield with excellent site
selectivity; oxidation at the γ-position was not observed (entry 1,
Table 1). A number of control experiments were conducted to
provide further insight of this reaction. Not surprisingly, without
the Pd catalyst or the oxidant, no desired product was formed
(entries 2 and 4). Use of less catalyst (5 mol %) gave incomplete
conversion on a small scale (0.1 mmol, vide infra, eq 2) (entry 3).
The yield was slightly diminished in the absence of Ac2O (entry
5),15 while addition of a small amount of water led to a significant
reduction of the yield, suggesting water may hamper the desired
reaction pathway (entry 6).
Use of other solvents, such as chlorobenzene, toluene, and 1,2-
dichloroethane, proved to be less efficient (entries 7−9). When
K2S2O8 was employed as the oxidant, a significant amount of the
elimination product (2,6-dimethoxybenzonitrile) was formed
(entry 10). Addition of LiOAc did not affect the reaction yield
(entry 11). Control of the reaction concentration seems to be
critical: at higher concentration decomposition products were
formed, while at lower concentration incomplete conversion was
observed (entries 12 and 13). Finally, 2,6-dimethoxybenzyl
oxime proved an optimal DG because use of either the 2-pyridyl
(DG A) or 2,6-dichlorobenzyl (DG B) analogue is much less
efficient (entries 14 and 15).16 It is likely that the pyridine-
derived oxime chelates strongly with the Pd and consequently
inhibits catalyst turnover, while the 2,6-dichlorobenzyl oxime has
relatively weak coordination ability compared to the 2,6-
dimethoxybenzyl oxime.
consisted of a mixture of oxime E/Z stereoisomers; see SI.
The substrate scope is depicted in Table 2. All the substrates
were prepared from the corresponding alcohols via first
conversion to the corresponding N-alkoxyphthalimides using a
Mitsunobu (for 1° and 2° alcohols) or SN1 (for 3° alcohols)-type
reaction, followed by a one-pot deprotection/condensation with
2,6-dimethoxybenzaldehyde (for a summary and details, see
Supporting Information (SI)). Substrates derived from primary,
secondary, and tertiary alcohols all reacted smoothly under the
optimized conditions, and the corresponding vicinal diols were
afforded in good yields (entries 1−9). Substrates with different
steric properties were examined, and it seems that increasing
sterics on one side of the alcohol does not have a significant
impact on the overall reactivity (entries 5−7); however, it seems
clear that the secondary alcohol-derived substrates generally give
higher reactivity (conversion) than the tertiary ones (entries 8
and 9). The oxidation occurred site selectively on the β-position
(X-ray crystal structure of product 9b was obtained, see SI), and
in the presence of a methyl and a nonmethyl alkyl group at the β-
position, reaction with the methyl group is generally preferred
(entries 3−7).
In addition to methyl groups, cyclic methylene groups (CH2)
also react. In the case of a cyclopentanol-derived substrate, both
the mono- and bis-oxidation products (diol 10b and triol 10c)
were obtained (entry 10).17 It is interesting to note that the
reaction gave a trans-isomer for diol 10b but a cis/trans-isomer
for triol 10c. This suggests the first metalation occurs selectively
on the opposite side of the DG, but the second takes place on the
same side. Such selectivity is likely dictated by the inherently
preferred conformation of these substrates in the transition states
(for details, see SI).
B
dx.doi.org/10.1021/ja3082186 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
functional group compatibility is not a focus of this
communication, we envision this reaction should tolerate a
range of functionalities, such as ether, ester, oxime, and aryl
groups given that the DG contains aryl methyl ether and oxime
groups and that the products contain acetate groups.20
Table 2. Summary of the Substrate Scope
Reactions on different scales have also been examined with
substrate 1a (eq 2). On larger scales, the yield was significantly
higher and less catalyst was required. For example, the yield
increased from 61% on a 0.1 mmol scale to 80% on a 5 mmol
(gram) scale. In addition, on a 2 mmol scale, use of 2 mol %
Pd(OAc)2 provided a 74% yield of the product.
The kinetic isotope effect (KIE) was determined to briefly
explore the reaction mechanism. The KIE observed in the
intermolecular competition experiments (kH/D = 3.35) is
consistent with C−H cleavage being the rate-determining step
(eq 3).21
An unusual skeleton rearrangement was observed when
menthol-derived oxime 13 was used as the substrate (Scheme
3). A five-membered ring product (14) with 1,3-diol moieties
Scheme 3. An Unusual Oxidative Rearrangement
a
b
Isolated yields. Product consisted of a mixture of oxime E/Z
c
stereoisomers; see SI. Starting material and product consisted of a
mixture of oxime E/Z stereoisomers; see SI. 1 equiv of PhI(OAc)2
was used. 3 equiv of PhI(OAc)2 were used.
was isolated as a single diastereomer in good yield.22 Both the
starting material (13) and product (14) were unambiguously
identified by 1H and 13C NMR, IR, HRMS, and X-ray
crystallography. The mechanism for this rearrangement is
unclear23 but likely involves an intermediate such as 16 formed
via cis-palladation. Detailed mechanistic studies and extension of
this transformation to other substrates will be provided in a later
report.
Finally, we demonstrated the DG and Ac groups can be
removed in excellent yields using orthogonal chemical methods
(Scheme 4). Thus, the diols can be easily differentiated, which
shows great potential for use in syntheses of complex natural
products.
d
e
Furthermore, it is surprising to find that in the cases of
norbornyl-derived alcohols, both the endo- and exo-substrates
directed the C−H functionalization at the β-methine (CH)
group, instead of the less hindered β-methylene group (entries
11 and 12). The structures of the products (11b and 12b) were
determined by removal of both the DG and Ac groups to give the
known diols (vide infra, Scheme 4).18 To our knowledge, Pd-
catalyzed C−H functionalization of tertiary carbon centers is
extremely rare;19 thus this finding might shed light on the unique
reactivity of bridge-head C−H bonds in these reactions. Although
C
dx.doi.org/10.1021/ja3082186 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
(5) Simmons, E. M.; Hartwig, J. F. J. Am. Chem. Soc. 2010, 132, 17092.
(6) (a) Gormisky, P. E.; White, M. C. J. Am. Chem. Soc. 2011, 133,
12584. (b) Qi, X.; Rice, G. T.; Lall, M. S.; Plummer, M. S.; White, M. C.
Tetrahedron 2012, 66, 4816.
Scheme 4. Orthogonal Protecting Groups
(7) Use of amine-derived DGs for selective functionalization of
unactivated aliphatic C−H bonds has been recently developed by
Daugulis and Chen; for seminal works, see: (a) Zaitsev, V. G.; Daugulis,
O. J. Am. Chem. Soc. 2005, 127, 4156. (b) He, G.; Chen, G. Angew.
Chem., Int. Ed. 2011, 50, 5192.
(8) For a review, see: Hartung, J.; Gottwald, T.; Spehar, K. Synthesis
2002, 11, 1469.
(9) (a) Espino, C. G.; Du Bois, J. Angew. Chem., Int. Ed. 2001, 40, 598.
(b) Espino, C. G.; Wehn, P. M.; Chow, J.; Du Bois, J. J. Am. Chem. Soc.
2001, 123, 6935.
(10) (a) Chen, K.; Richter, J. M.; Baran, P. S. J. Am. Chem. Soc. 2008,
130, 7247. While this manuscript was in preparation, two additional
related works from Baran and co-workers were reported: (b) Michaudel,
Q.; Thevenet, D.; Baran, P. S. J. Am. Chem. Soc. 2012, 134, 2547.
(c) Voica, A. F.; Mendoza, A.; Gutekunst, W. R.; Fraga, J. O.; Baran, P. S.
Nat. Chem. 2012, 4, 629.
(11) (a) The directed activation of allylic C−H bonds to form 1,2-diols
was first developed by White; see ref 6a. (b) For a synthesis of catechols
from phenols via silanol-DGs, see: Huang, C.; Ghavtadze, N.;
Chattopadhyay, B.; Gevorgyan, V. J. Am. Chem. Soc. 2011, 133, 17630.
(12) (a) Desai, L. V.; Hull, K. L.; Sanford, M. S. J. Am. Chem. Soc. 2004,
126, 9542. (b) Neufeldt, S. R.; Sanford, M. S. Org. Lett. 2010, 12, 532.
(13) For a single example using an oxime as an exo-DG in
functionalization aromatic C−H bonds, see: Desai, L. V.; Stowers, K.
J.; Sanford, M. S. J. Am. Chem. Soc. 2008, 130, 13285.
(14) (a) Mawo, R. Y.; Mustakim, S.; Young, V. G., Jr.; Hoffmann, M.
R.; Smoliakova, I. P. Organometallics 2007, 26, 1801. (b) Keuseman, K.
J.; Smoliakova, I. P.; Dunina, V. V. Organometallics 2005, 24, 4159 and
references therein.
In summary, we have developed a new strategy for the Pd-
catalyzed site-selective functionalization of unactivated aliphatic
C−H bonds, which allows access to chemically differentiated 1,2-
diols from monoalcohol derivatives. The oxime was employed as
both a DG and an alcohol surrogate for this transformation. The
use of exo-DGs in C−H activation, as illustrated in this work,
would potentially open doors for the discovery of new
transformations and new cleavable DGs. On the basis of these
preliminary results, efforts toward enhancing the efficiency of the
DG for higher reactivity and easier installation, expanding the
substrate/reaction scope, and a detailed mechanistic study are
currently ongoing.
ASSOCIATED CONTENT
■
S
* Supporting Information
Experimental procedures, spectral data, and crystallographic data
(CIF). This material is available free of charge via the Internet at
http://pubs.acs.org.
(15) For an important discovery of the role of Ac2O in a Pd-catalyzed
C−H oxidation reaction, see: Giri, R.; Liang, J.; Lei, J.-G.; Li, J- J.; Wang,
D.-H.; Chen, X.; Naggar, I. C.; Guo, C.; Foxman, B. M.; Yu, J.-Q. Angew.
Chem., Int. Ed. 2005, 44, 7420.
(16) For a related imine-DG in a stoichiometric Pd-mediated C−H
activation, see: Dangel, B. D.; Godula, K.; Youn, S. W.; Sezen, B.; Sames,
D. J. Am. Chem. Soc. 2002, 124, 11856.
AUTHOR INFORMATION
Corresponding Author
gbdong@cm.utexas.edu
■
Author Contributions
†Z.R. and F.M. contributed equally.
(17) The reaction with a cyclohexyl-derived substrate only gave ∼20%
conversion. The reaction with a 2-pentanol-derived substrate has also
been attempted, and only a trace amount of C−H activation product was
detected. The difficulties with oxidation of acyclic secondary CH2
groups suggest the conformation of the substrate greatly affects the
reactivity.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank UT Austin and CPRIT for a start-up fund and thank
the Welch Foundation and Frasch Foundation for research
grants. G.D. thanks ORAU for a new faculty enhancement award.
We also thank faculty members from the organic division at UT
Austin for their generous support. Dr. Lynch is acknowledged for
X-ray crystallography. We thank Dr. Shoulders, Mrs. Spangen-
berg, and Mr. Sorey for their NMR advice.
(18) García Martínez, A.; Teso Vilar, E.; García Fraile, A.; De La Moya
Cerero, S.; Martínez Ruiz, P. Tetrahedron: Asymmetry 1998, 9, 1737.
(19) To the best of our knowledge, no such example has been reported.
(20) Olefins react under the reported conditions; see: Li, Y.; Song, D.;
Dong, V. M. J. Am. Chem. Soc. 2008, 130, 2962.
(21) For a similar KIE observation in sp3 C−H functionalization, see:
Wang, D.-H.; Wu, D.-F.; Yu, J.-Q. Org. Lett. 2006, 8, 3387.
(22) Compound 14 was isolated as a mixture of oxime E/Z isomers.
Cleavage of the DG revealed the corresponding alcohol as a single
diastereomer.
REFERENCES
■
(1) (a) Neufeldt, S. R.; Sanford, M. S. Acc. Chem. Res. 2012, 45, 936.
(b) Chen, K.; Baran, P. S. Nature 2009, 459, 824. (c) Simmons, E. M.;
Hartwig, J. F. Nature 2012, 483, 70.
(23) For a recent Fe-catalyzed C−H activation-initiated skeleton
rearrangement, see: Bigi, M. A.; Reed, S. A.; White, M. C. J. Am. Chem.
Soc. 2012, 134, 9721.
(2) (a) Davies, H. M. L.; Morton, D. Chem. Soc. Rev. 2011, 40, 1857.
(b) Newhouse, T.; Baran, P. S. Angew. Chem., Int. Ed. 2011, 50, 3362.
(c) Chen, M. S.; White, M. C. Science 2007, 318, 783.
(3) For a recent review on Pd-catalyzed directed C−H functionaliza-
tion, see: (a) Lyons, T. W.; Sanford, M. S. Chem. Rev. 2010, 110, 1147.
For recent reviews on sp3 C−H activation, see: (b) Li, H.; Li, B.-J.; Shi,
Z.-J. Catal. Sci. Technol. 2011, 1, 191. (c) Jazzar, R.; Hitce, J.; Renaudat,
A.; Sofack-Kreutzer, J.; Baudoin, O. Chem.Eur. J. 2010, 16, 2654.
(d) Davies, H. M. L. J. Mol. Catal. A: Chem. 2002, 189, 125.
(4) (a) Wang, X.; Lu, Y.; Dai, H.-D; Yu, J.-Q. J. Am. Chem. Soc. 2010,
132, 12203. (b) Lu, Y.; Wang, D.-H.; Engle, K. M.; Yu, J-.Q. J. Am. Chem.
Soc. 2010, 132, 5916.
D
dx.doi.org/10.1021/ja3082186 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX