Synthesis of Z- or E-Trisubstituted Allylic Alcohols and Ethers by Kinetically Controlled Cross-Metathesis with a Ru Catechothiolate Complex

Article abstract of DOI:10.1021/jacs.7b10010

The first examples of kinetically controlled cross-metathesis reactions that generate Z- or E-trisubstituted alkenes are disclosed. Transformations are catalyzed by ≤6.0 mol % of a Ru catechothiolate complex and afford trisubstituted allylic alcohols and

Full text of DOI:10.1021/jacs.7b10010

Communication
Cite This: J. Am. Chem. Soc. XXXX, XXX, XXX-XXX
pubs.acs.org/JACS
Synthesis of Z- or E‑Trisubstituted Allylic Alcohols and Ethers by
Kinetically Controlled Cross-Metathesis with a Ru Catechothiolate
Complex
Chaofan Xu, Zhenxing Liu, Sebastian Torker, Xiao Shen, Dongmin Xu, and Amir H. Hoveyda*
Department of Chemistry, Merkert Chemistry Center, Boston College, Chestnut Hill, Massachusetts 02467, United States
*
S Supporting Information
Scheme 1
ABSTRACT: The first examples of kinetically controlled
cross-metathesis reactions that generate Z- or E-trisub-
stituted alkenes are disclosed. Transformations are
catalyzed by ≤6.0 mol % of a Ru catechothiolate complex
and afford trisubstituted allylic alcohols and ethers in up to
8
1% yield and >98% stereoisomeric purity. The method
has considerable scope, as olefins containing an alcohol, an
aldehyde, an epoxide, a carboxylic acid, or an alkenyl group
may be used. Mechanistic models that account for the
observed levels and trends in efficiency and stereochemical
control are provided, based on DFT studies.
n 2013, we discovered that Ru catechothiolate complexes,
prepared from readily accessible starting materials, can
promote Z-selective ring-opening cross-metathesis (ROCM).
I
1
,2
We subsequently demonstrated that yields, diastereoselectivities,
and Z selectivities are higher when an allylic alcohol group is
3
4
involved in ROCM or cross-metathesis (CM), and mod-
ification of the bidentate and/or the N-heterocyclic carbene
(
NHC) ligand can be beneficial (e.g., Ru-1a−c, Scheme 1a).
Two features, the combination of which are unavailable in other
8a,c
4b,5
80% selectivity as the result of smaller energy gaps between
stereoselective Ru- or Mo- and W-based catalyst systems, are
notable: (1) The catalytic active Ru dithiolates are especially
robust, and reactions may be performed in the presence of a
Brønsted acid (e.g., a carboxylic acid), an electrophilic site (e.g.,
an aldehyde), a Lewis base (e.g., an amino acid), or a bulky allylic
substituent. (2) Acyclic 1,2-disubstituted olefins may be used
as substrates, a crucial attribute not shared by other Ru-based Z-
9
stereoisomers (vs 1,2-disubstituted alkenes). Herein, we
disclose the first examples of kinetically controlled CM processes
that furnish trisubstituted olefins efficiently and with high Z:E or
E:Z ratios.
4
a,6
We focused on synthesis of trisubstituted allylic alcohols
because these moieties are found in many biologically active
compounds and have considerable utility in chemical synthesis
4
a
4a
selective catalysts, providing the opportunity for the develop-
1
0
(
e.g., directed reactions ). There are a limited number of
approaches to stereoselective synthesis of trisubstituted allylic
12
ment of efficient stereoretentive transformations. We recently
6
demonstrated that with Z- or E-butene as capping agents, the
11
4a,7
alcohols, such as those involving α-alkoxy ketones or alkynes.
intermediacy of unstable methylidene complexes
can be
CM offers a distinct and important disconnection, with
considerable versatility owing to the relative stability of alkenes
in the presence of strongly nucleophilic or basic reagents.
We first examined representative CM with dichloro complex
Ru-2 (Scheme 2). Regardless of whether 1,1-disubstituted allylic
avoided and a considerable array of linear and macrocyclic Z- or
E-alkenes, including those containing the aforementioned polar
or hindered substituents, accessed efficiently and with high
stereoisomeric purity.
A compelling unaddressed question is whether, through6
13
alcohol 1a or Z- or E-1b was used, the thermodynamically
favored E-3a was formed predominantly (61−71% yield and 87−
bypassing unstable methylidene species by the capping strategy,
Ru catechothiolate catalysts can promote efficient stereo-
retentive CM to generate trisubstituted Z- or E-alkenes (Scheme
9
1% E selectivity).
We then probed the ability of Ru catechothiolate complexes to
1
b). Such transformations would be challenging for several
serve as catalysts (Table 1); all reactions were carried out in a
reasons, including the intermediacy of more congested metalla-
cyclobutane (mcb) intermediates. CM protocols designed for
synthesis of trisubstituted olefins are indeed scarce, and the few
8
Received: September 21, 2017
extant methods afford the lower energy E isomer only in up to
©
XXXX American Chemical Society
A
DOI: 10.1021/jacs.7b10010
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
a
a
Scheme 2. Reactions with a Dichloro-Ru Complex
Scheme 3. Scope I: Z-Trisubstituted Allylic Alcohols
a
1
Conversion (loss of 2a) determined by analysis of H NMR spectra
of unpurified mixtures. Yields correspond to purified products. See the
Supporting Information for details.
single vessel. There was 82% consumption of the terminal alkene
1
(
2a) with 1,1-disubstituted olefin 1a (entry 1). Analysis of the H
NMR spectrum of the product mixture indicated the major
component to be derived from homometathesis of Z-methyl-
6
substituted (capped) 2a (<5% 3a). Reaction of Z-1b with 2a in
the presence of Z-butene (10 equiv) and 6.0 mol % Ru-1a
afforded Z-3a in just 31% yield (81% conv) as a single
stereoisomer (>98% Z; entry 2); thus, homometathesis of the
capped terminal alkene derivative was again predominant. To
4c
increase CM efficiency, we turned to Ru-1b, a complex with a
smaller NHC ligand (entry 3). Accordingly, under otherwise
identical conditions, 3a was isolated in 76% yield and >98% Z:E
ratio. Further optimization (entries 4 and 5) revealed that with
5
.0 equiv Z-1b and Z-butene, 3a may be obtained in 74% yield, as
14
pure Z isomers. The data in entry 6 confirm the central role of
the capping agent. The transformation is scalable: 0.6 g 2a was
6
converted to 0.57g of Z-3a (77% yield, >98:2 Z:E).
Table 1. Initial Evaluation with Ru Catechothiolate
Complexes
a
a
Same conditions as Table 1, except 10 equiv of Z-butene for Z-3k.
Conversion (loss of Z-Me-substituted alkene from 2) determined by
analysis of H NMR spectra of unpurified mixtures. Yields correspond
1
to purified products. See the Supporting Information for details. Fc =
ferrocenyl.
Chemoselective synthesis of Z,E-diene Z-3m further underscores
utility.
The reaction leading to o-benzyloxy-substituted Z-3n (eq 1),
recently used to prepare and ascertain the structure of naturally
a
b
See the Supporting Information for details. Conversion (loss of Z-
1
Me-substituted alkene derived from 2a) determined by analysis of H
NMR spectra of unpurified mixtures. Yields correspond to purified
c
15
occurring antiproliferative agent xiamenmycin A, was challeng-
products. na = not applicable.
ing: use of 6.0 mol % Ru-1b in two equal portions was necessary.
The Z-trisubstituted allylic alcohol was synthesized in 62%
15
overall yield, compared to 67% reported before, from a
commercially available phenol (benzyl protection in 96% yield).
This is more step-economical than the more traditional sequence
(two vs four steps), as reduction of the enoate generated by a
Various Z-trisubstituted allylic alcohols were prepared in up to
1% yield and exceptional stereoisomeric purity (Z-3b-3m,
8
Scheme 3). This included compounds with a hydroxy (Z-3b), a
Lewis basic phthalimide (Z-3f), an epoxide (Z-3g), or an
aldehyde (Z-3h). Comparison of the yields for Z-3i (63%) and Z-
Horner−Wadsworth−Emmons reaction (LiAlH , thf, reflux)
4
15
was not necessary.
Another key attribute is showcased by reactions with E-1b.
13
3j (40%) demonstrates that CM with β-branched alkenes can be
more sluggish. Benzylic trisubstituted olefins Z-3k and Z-3l were
isolated in 55% yield with 98:2 Z:E ratio and 81% yield and >98:2
Z:E ratio, respectively. The transformation leading to Z-3k was
more efficient with 10 equiv Z-1b (vs 40% yield with 5.0 equiv).
The E-trisubstituted allylic alcohols were prepared with
efficiency and stereoretention (Scheme 4) similar to those of
the Z isomers (Scheme 3). Here too, the catalytic stereoretentive
approach was broadly applicable.
B
DOI: 10.1021/jacs.7b10010
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
a
Scheme 4. Scope II: E-Trisubstituted Allylic Alcohols
allylic ether (10 and 20 equiv, respectively) was needed (38% and
4
8%, respectively). The high conversion implies that homo-
metathesis of the Z-methyl-substituted olefin is facile but not the
formation of the trisubstituted alkenes. The lower efficiency for
Z-9 may be attributed to competitive decomposition of the active
Ru complex, a complication less problematic for reactions
4a
leading to 1,2-disubstituted alkenes. It should be emphasized
that no other stereoselective catalyst class, Ru-based or
otherwise, can be used for kinetically controlled synthesis of a
stereoisomerically pure trisubstituted alkene that contains a
carboxylic acid group.
Trisubstituted allylic ethers are formed more readily than their
1
,2-disubstituted variants likely because in the corresponding
mcb intermediates, the hydroxyl/alkoxy unit is attached to Cβ so
that steric pressure is minimized (II vs Cα in I, Figure 1). There is
accordingly lesser electron−electron repulsion between the
heteroatom and the apical sulfide, dispensing with the need for
3a
H-bonding to counter unfavorable interactions.
a
Same conditions as Table 1; 5.0 equiv Z-butene used except for 3m
(
2
10 equiv). Conversion (loss of Z-Me-substituted alkene derived from
1
) determined by analysis of H NMR spectra of unpurified mixtures.
b
Yields correspond to purified products. 10 equiv E-1b used. See the
Supporting Information for details.
In Z-selective and diastereoselective ROCM with Ru catecho-
thiolate catalysts, unlike allylic alcohols, there is typically <5%
3a
conversion to the desired products when allylic ethers are used.
This is however not the case here: CM under the same
conditions used to obtain trisubstituted allylic alcohols afforded
Z-4 in 72% yield (81% conv) and >98:2 Z:E ratio (Scheme 5).
Several other Z- or E-trisubstituted allylic ethers were accessed
likewise, although yields were somewhat lower (vs allylic
alcohols; Scheme 5). Carboxylic acid 9 and allylic acetate 10
were particularly difficult cases, and more of the trisubstituted
Figure 1. Unlike reactions leading to 1,2-disubstituted alkenes, H-
bonding and e−e repulsion play less of a role en route to trisubstituted
olefins.
CM between 2a and some other trisubstituted olefins (Scheme
6
) indicated that reactions involving the homoallylic alcohol
a
derivative of Z-1b (cf. 16) or those with an alkyl group give little
of the desired products (<5% yield; 70−83% homometathesis).
These findings, along with those in Table 1 and Schemes 3−5,
show that an allylic heteroatom is needed for efficient
transformation.
Scheme 5. Scope III: Z- and E-Trisubstituted Allylic Ethers
Scheme 6. Allylic Heteroatom Is Required for High
a
Efficiency
a
1
Same conditions as Table 1. Conversion determined by analysis of H
NMR spectra of unpurified mixtures. See the Supporting Information
for details.
To understand better why an allylic heteroatom is crucial to
16
efficiency, DFT calculations were carried out. We investigated
the model reactions represented as A−C (Figure 2a). Congruent
with the experimental results, the overall barriers for modes A
and C (16.3 and 15.8 kcal/mol) were found to be lower than for
B (17.9 kcal/mol).
Transition-state analysis points to a rationale regarding the
origin of the observed reactivity trends (Figure 2b). We propose
17
a
Same conditions as in Table 1; 10 and 20 equiv allylic ether used for
Z-9 and Z-10, respectively. Conversion (loss of Z-Me-substituted
alkene derived from α-olefin) determined by analysis of H NMR
1
spectra of unpurified mixtures. Yields correspond to purified products.
that ts1 is higher in energy due to A(1,2) involving (C3−C4 and
B
b
18
Same conditions as eq 1. See the Supporting Information for details.
C2′−C3′) and A(1,3) strain (involving C2−C3 and C2′−
C
DOI: 10.1021/jacs.7b10010
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
REFERENCES
■
(
2
1) (a) Khan, R. K. M.; Torker, S.; Hoveyda, A. H. J. Am. Chem. Soc.
013, 135, 10258. For investigations that shed light on the importance
of S-based (vs O-based) bidentate ligands, see: (b) Khan, R. K. M.;
Torker, S.; Hoveyda, A. H. J. Am. Chem. Soc. 2014, 136, 14337.
(2) For reviews on Z-selective olefin metathesis, see: (a) Hoveyda, A.
H. J. Org. Chem. 2014, 79, 4763. (b) Hoveyda, A. H.; Torker, S.; Khan, R.
M. K.; Malcolmson, S. J. In Handbook of Metathesis; Grubbs, R. H.,
O’Leary, D. J., Wenzel, A. G., Khosravi, E., Eds.; Wiley-VCH: Weinheim,
Germany, 2015; Vol. 2, pp 508−562.
(3) (a) Koh, M. J.; Khan, R. K. M.; Torker, S.; Hoveyda, A. H. Angew.
Chem., Int. Ed. 2014, 53, 1968. For applications of different types of Ru
catechothiolate complexes in stereoselective ROMP, see: (b) Mikus, M.
S.; Torker, S.; Hoveyda, A. H. Angew. Chem., Int. Ed. 2016, 55, 4997. For
the influence of bidentate anionic and chelating ligands on stereo-
selective ROMP with Ru catechothiolate catalysts, see: (c) Mikus, M. S.;
Torker, S.; Xu, C.; Li, B.; Hoveyda, A. H. Organometallics 2016, 35, 3878.
(4) (a) Koh, M. J.; Khan, R. K. M.; Torker, S.; Yu, M.; Mikus, M.;
Hoveyda, A. H. Nature 2015, 517, 181. For an application in total
synthesis, see: (b) Yu, M.; Schrock, R. R.; Hoveyda, A. H. Angew. Chem.,
Int. Ed. 2015, 54, 215. For related reports from others regarding E-
selective stereoretentive CM catalyzed by Ru catechothiolate species,
see: (c) Ahmed, T. S.; Grubbs, R. H. J. Am. Chem. Soc. 2017, 139, 1532.
(5) For example, see: (a) Nguyen, T. T.; Koh, M. J.; Shen, X.; Romiti,
F.; Schrock, R. R.; Hoveyda, A. H. Science 2016, 352, 569. (b) Koh, M. J.;
Nguyen, T. T.; Lam, J. K.; Torker, S.; Hyvl, J.; Schrock, R. R.; Hoveyda,
A. H. Nature 2017, 542, 80.
Figure 2. (a) Mechanistic analysis with model reactions that proceed via
A−C. (b) Selected transition states with free energy values (PBE0-
(6) Xu, C.; Shen, X.; Hoveyda, A. H. J. Am. Chem. Soc. 2017, 139,
D3BJ/Def2TZVPP
). See the Supporting Information for details.
thf(SMD)
10919.
Ar = 2-F,6-MeC H ; ts = transition state; SMD = solvation model based
6
3
(7) Hong, S. H.; Wenzel, A. G.; Salguero, T. T.; Day, M. W.; Grubbs, R.
H. J. Am. Chem. Soc. 2007, 129, 7961.
on density.
(8) (a) Chatterjee, A. K.; Grubbs, R. H. Org. Lett. 1999, 1, 1751.
(b) Chatterjee, A. K.; Sanders, D. P.; Grubbs, R. H. Org. Lett. 2002, 4,
1
2
939. (c) Morrill, C. M.; Funk, T. W.; Grubbs, R. H. Tetrahedron Lett.
004, 45, 7733. (d) Wang, Z. J.; Jackson, W. R.; Robinson, A. J. Org. Lett.
C3′); these unfavorable interactions are exacerbated as the mcb
2
3
2013, 15, 3006.
9) Cuvigny, T.; Herve du Penhoat, C.; Julia, M. Tetrahedron Lett.
1980, 21, 1331.
10) (a) Hoveyda, A. H.; Evans, D. A.; Fu, G. C. Chem. Rev. 1993, 93,
307. For recent examples, see: (b) Wu, R.; Beauchamps, M. G.;
Laquidara, J. M.; Sowa, J. R., Jr. Angew. Chem., Int. Ed. 2012, 51, 2106.
forms (namely, as sp -to-sp rehybridization occurs). With the
(
smaller methoxy group in C, steric pressure is diminished (Figure
2
3
1
b); this is reflected in the smaller C −C −O bond angles,
(
1
compared to those in ts1 and ts2 (107.1° and 110.5° vs 112.4°
and 112.5°, respectively), which alleviate most of the strain.
B
B
16,19
Development of additional Ru-based olefin metathesis
catalysts and studies of their applications in stereoselective
chemical synthesis are in progress.
(c) Li, H.; Mazet, C. J. Am. Chem. Soc. 2015, 137, 10720.
(11) Sreekumar, C.; Darst, K. P.; Still, W. C. J. Org. Chem. 1980, 45,
4260.
(12) For example, see: (a) Negishi, E.; Van Horn, D. E.; Yoshida, T. J.
ASSOCIATED CONTENT
Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/jacs.7b10010.
Am. Chem. Soc. 1985, 107, 6639. (b) Ma, S.; Negishi, E. J. Org. Chem.
■
1
997, 62, 784. (c) Chen, Y. K.; Walsh, P. J. J. Am. Chem. Soc. 2004, 126,
*
S
3
702.
(13) Stereoisomerically pure Z- and E-1b can be easily prepared by
reduction of commercially available carboxylic esters.
Experimental procedures, spectral and analytical data for
all products, and calculations (PDF)
(14) Reaction with the corresponding enoate (precursor to Z-1b) leads
to <2% conversion, probably due to low reactivity of the electron-
deficient alkene and perhaps internal chelation within the derived Ru
carbene. Studies to address this and related issues are in progress.
(15) Jiao, X.; Yao, Y.; Yang, B.; Liu, X.; Li, X.; Yang, H.; Li, L.; Xu, J.; Xu,
M.; Xie, P. Org. Biomol. Chem. 2016, 14, 1805.
AUTHOR INFORMATION
Corresponding Author
amir.hoveyda@bc.edu
■
*
(
16) For details of computational studies, see the Supporting
Information.
17) Yang, L.; Adam, C.; Nichol, G. S.; Cockroft, S. L. Nat. Chem. 2013,
, 1006.
ORCID
(
5
Amir H. Hoveyda: 0000-0002-1470-6456
(
18) Hoffmann, R. W. Chem. Rev. 1989, 89, 1841.
Notes
(19) The A value for a methyl or an ethyl group is ∼1.7−1.8, whereas it
The authors declare no competing financial interest.
is ∼0.6−0.7 for a hydroxy or methoxy group. The Charton values for the
latter two units are not known. See: (a) Eliel, E. L.; Wilen, S. H.
Stereochemistry of Organic Compounds; Wiley Interscience: New York,
1994; pp 696−697. (b) Charton, M. J. Am. Chem. Soc. 1975, 97, 1552.
ACKNOWLEDGMENTS
■
This work is dedicated to Prof. Richard R. Schrock. Financial
support was provided by the NSF (CHE-1362763). D.X. is a
Kozarich Undergraduate Fellow. We thank M. S. Mikus and F.
Romiti for insightful discussions.
D
DOI: 10.1021/jacs.7b10010
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX