High-value alcohols and higher-oxidation-state compounds by catalytic Z-selective cross-metathesis

Article abstract of DOI:10.1038/nature14061

Olefin metathesis catalysts provide access to molecules that are indispensable to physicians and researchers in the life sciences. A persisting problem, however, is the dearth of chemical transformations that directly generate acyclic Z allylic alcohols, including products that contain a hindered neighbouring substituent or reactive functional units such as a phenol, an aldehyde, or a carboxylic acid. Here we present an electronically modified ruthenium-disulfide catalyst that is effective in generating such high-value compounds by cross-metathesis. The ruthenium complex is prepared from a commercially available precursor and an easily generated air-stable zinc catechothiolate. Transformations typically proceed with 5.0 mole per cent of the complex and an inexpensive reaction partner in 4-8 hours under ambient conditions; products are obtained in up to 80 per cent yield and 98:2 Z:E diastereoselectivity. The use of this catalyst is demonstrated in the synthesis of the naturally occurring anti-tumour agent neopeltolide and in a single-step stereoselective gram-scale conversion of a renewable feedstock (oleic acid) to an anti-fungal agent. In this conversion, the new catalyst promotes cross-metathesis more efficiently than the commonly used dichloro-ruthenium complexes, indicating that its utility may extend beyond Z-selective processes.

Full text of DOI:10.1038/nature14061

LETTER
doi:10.1038/nature14061
High-value alcohols and higher-oxidation-state
compounds by catalytic Z-selective cross-metathesis
1
1
1
1
1
1
Ming Joo Koh , R. Kashif M. Khan , Sebastian Torker , Miao Yu , Malte S. Mikus & Amir H. Hoveyda
Olefin metathesis catalysts provide access to molecules that are in-
A readily accessible and easy-to-use Ru-based cross-metathesis cat-
15
1,2
dispensable to physicians and researchers in the life sciences . A alyst that generates linear molecules containing an allylic alcohol , a
persisting problem, however, is the dearth of chemical transforma- phenol, an aldehyde or a carboxylic acid (compare i–iii, Fig. 1b) would
tions that directly generate acyclic Z allylic alcohols, including pro- have a substantial impact on our ability to access a large assortment of
ducts that contain a hindered neighbouring substituent or reactive important molecules; these functionalunits commonly occur in nature
functional units such as a phenol, an aldehyde, or a carboxylic acid. and are widely employed in chemistry. It is, however, not a given that
Here we present an electronically modified ruthenium–disulfide the Z-selective carbene derivatives would exhibit the same degree of
catalyst that is effective in generating such high-value compounds robustness and chemoselectivity as the dichloro-Ru systems. This is
by cross-metathesis. The ruthenium complex is prepared from a because the development of Z-selective Ru catalysts has been achieved
commercially available precursor and an easily generated air-stable throughsubstitutionofthehalogenswithotheranionicligands(Fig.1a).
zinc catechothiolate. Transformationstypically proceed with5.0 mole In Ru-2, the chlorine atoms in a species such as Ru-1 are replaced by an
per cent of the complex and an inexpensive reaction partner in alkylandanitratogroup, andinRu-3awehavetwoRu–Sbondsinstead.
4
–8 hours under ambient conditions; products are obtained in up Anionic ligands in a Ru carbene catalyst certainly have a strong influ-
16
to 80 per cent yield and 98:2 Z:E diastereoselectivity. The use of this ence on its performance . For example, although allylic alcohols are
1
7
catalyst is demonstrated in the synthesis of the naturally occurring prone to competitive olefin isomerization , they can undergo unusu-
anti-tumour agent neopeltolide and in a single-step stereoselective ally efficient and selective olefin metathesis with Ru-dichloride com-
gram-scale conversion of a renewable feedstock (oleic acid) to an plexes (such as Ru-1) partly owing to H-bonding with the chloride
anti-fungal agent. In this conversion, the new catalyst promotes ligands ; such a feature could be lost as a result of alterations imple-
cross-metathesis more efficiently than the commonly used dichloro– mented in the design of Z-selective catalysts. Carbenes derived from
ruthenium complexes, indicating that its utility may extend beyond Ru-2 catalyse olefin metathesis in the presence of a remote alcohol but
1
8
7
,19
Z-selective processes.
yields are at times diminished , and the related Z-selective ring-
In 2009, it was demonstrated that a molybdenum (Mo) monopyr- opening/cross-metathesiswithallylicalcoholshasinvolvedonlyhighly
2
0
rolide aryloxide complex can promote ring-opening/cross-metathesis strained cyclobutene . Similarly, ring-opening/cross-metathesis with
3
with high Z selectivity . This initial discovery and the attendant stereo- dithiolate Ru-3a and an allylic alcohol is limited to strained cyclic
1
1
14
chemical model paved the way for development of other Mo and tung- olefins .Inthelattercases,aswiththeNHC-containing Mo alkylidenes ,
4
sten (W) alkylidenes that facilitate Z-selective cross-metathesis and rapid transformation with an exceedingly reactive cyclic alkene allows
macrocyclicring-closingmetathesis . Itwaslatershownthat olefinme- productive olefin metathesis to compete effectively with catalyst
5
tathesis can be promoted Z selectively by ruthenium (Ru) complexes decomposition.
containing a bidentate N-heterocyclic carbene (NHC) with a carbanio-
A pressing question then is whether any existing Ru catalyst can effi-
6–8
nic arm (such as Ru-2, Fig. 1a) . Monosulfide/monochloride Ru car- ciently promote cross-metathesis reactions that generate the types of Z
benes have also been used for Z-selective homocoupling of unhindered alkenes represented by i–iii in Fig. 1b. Z-selective cross-metathesis
terminal alkenes, but Z:E ratios (where E is the other isomer) generally transformations that directly afford allylic alcohols are yet to be dis-
plummet at later stages of the transformation owing to post-olefin closed. Catalytic Z-selective olefin metathesis involving an aldehyde-
9
metathesis isomerization . The most recently developed Ru complexes containing or a carboxylic-acid-containing substrate is unknown (see
for Z-selective olefin metathesis, represented by Ru-3a, feature a cate- ii–iii). Furthermore, unlike dichloro-Ru carbenes, complexes such as
1
0
chothiolate ligand ; the utility of these carbenes has been confined to Ru-2are generallyused with linear, unhindered aliphatic alkenes;with
10
ring-opening metathesis polymerization and ring-opening/cross- sterically hindered reaction partners, olefin metathesis is substantially
1
1
6
metathesis , processes that benefit from the driving force of strain less efficient (such as with a styrene; see i).
release.
We selected Z-2-butene-1,4-diol (1; Fig. 1c) as the cross partner for
That different families of complexes have been introduced for two reasons: stereoisomerically pure 1is commercially and inexpensively
Z-selective catalysts is not surprising, since the range of functional available, and use of 1,2-disubstituted olefin diminishes the amount of
group compatibility offered by Mo-, W- and Ru-based catalysts is a ethylene formation, discouraging the generation of unstable methyli-
1,12
4
key reason for the rapid rise of olefin metathesis . Ru dichloride spe- denes that areadept atpromotingpost-cross-metathesisisomerization .
cies (such as Ru-1) retain their structural integrity in the presence of a In the event, cross-metathesis of 1 and allylbenzene (2) with 3.0 mol%
hydroxyl, an aldehyde and a carboxylic acid group, but Mo and W Ru-2gave a Z:E 5 91:9 diastereoisomeric mixture forproduct 3 in 50%
alkylidenes arerendered inactiveunder such circumstances, necessita- yield (22 uC, 9 h); reaction with dithiolate Ru-3a was less efficient (42%
13
ting protection/deprotection schemes . Unlike high-oxidation-state yield) but more stereoselective (98:2 Z:E). Longer reaction times did
5
alkylidenes, Ru complexes are often deactivated in the presence of amines
not lead to further conversion, pointing to catalyst instability as a cul-
4
or enol ethers . An NHC–Mo complex with a triflate ligand has been prit.Wesurmisedthatadecompositionpathwaycouldinvolveasulfide
shown to tolerate alcohol, aldehyde and carboxylic acid functions, but shift (A R B, Fig. 1d), leading to generation of a catalytically inactive
only for extraordinarily facile ring-opening metathesis polymerization Ru-alkyl species. Catechothiolate complexes (such as Ru-3a) containa
14
and ring-closing metathesis reactions .
sulfideorientedoppositetotheNHC, adispositiondesignedtopromote
1
Department of Chemistry, Merkert Chemistry Center, Boston College, Chestnut Hill, Massachusetts 02467, USA.
8
J A N U A R Y 2 0 1 5 | V O L 5 1 7 | N A T U R E | 1 8 1
©
2015 Macmillan Publishers Limited. All rights reserved

RESEARCH LETTER
1
0
Z selectivity . It is therefore likely that the anionic ligand is especially
We pondered whether placement of electronegative atoms (X 5 Cl;
16
enriched in electron density (trans influence ) and prone to isomeriza- Fig. 1e) within the catechothiolate ligand (versus X 5 H) might dimin-
tion (compare A). The latter hypothesis gained credence when we iso- ish the rate of the proposed isomerization. We chose to assess the plau-
latedRu-4(Fig. 1d), whichcontainstheaforementionedC–SbondinB. sibility of such an approach by computationally (using densityfunctional
a Selected types of Ru complexes used in olefin metathesis
MesN
N
MesN
H
NMes
S
MesN
Cl
NMes
Widely used in all types
of olefin metathesis,
including those involving
allylic alcohols, styrenes,
carboxylic acids and
aldehydes
Used in Z-selective
ROMP, ROCM,
ethenolysis,
homocoupling,
cross-metathesis, and
macrocyclic RCM
–
+
N
O
Used in Z-selective
ROMP and ROCM
Ru
O
Ru
Ru
O
Cl
O
S
Oi-Pr
Ru-1
i-PrO
(non-stereoselective or
E-selective)
Ru-2
Ru-3a
Effective for stereoselective cross-metathesis reactions that generate i to iii?
b Issues unaddressed by state-of-the-art catalytic olefin metathesis regarding several critical functional units
O
O
OH
Efficient synthesis of Z-allylic
alcohols by cross-metathesis?
Cross-metathesis with hindered
alkenes and/or phenols?
HO
Efficient Z-selective
cross-metathesis
with aldehydes?
Efficient Z-selective
cross-metathesis
with carboxylic acids?
HO
HO
HO
i
ii
iii
e The energetics of sulfide shift versus olefin metathesis by DFT calculations
c Initial observations
HO OH
1
,2-shift
olefin metathesis
'
#$#%
1
MesN
X
NMes
Me
HO
Ph
3
.0 mol% Ru complex
THF, 22 °C, 9 h
Commercially available
MesN
NMes
X
MesN
NMes
H
S
+
3
X
S
Ru
H
S
Ph
Ru
Me
14e
S
Ru
Me
with Ru-2: 70% conv., 50% yield, 91:9 Z:E
with Ru-3a: 50% conv., 42% yield, 98:2 Z:E
+
H
X
2
"#$#%
&#$#%
S X
S
Me
shift tsPr
17.7
Me
Me
+
Me
Me
X
d Possible pathway for catalyst decomposition
tsBu
+
14.2
14.2
12.1
ArN
NAr
Me
Me
ArN
NAr
S
1,2-shift of
anti-to-NHC
sulfide
Me
1
2.1
R1
+
Probably
catalytically
inactive
13.4
R1
H
Ru
Ru
8.8
S
S
10.0
MesN
S
R
9.5
R
7.1
5
.8
.6
6
.9
NMes
Me
A
B
5
.3
X
4
MesN
X
NMes
S
MesN
NMes
Ru
0.0
Me
S
#$#%
Ru
H
H
H
S
IV
Ru
S
MesN
X
NMes
Me
Me Me
S
MesN
NMes
X
S
X
S
Me
S
S
mcbBu
S
X
Me
pcBu
Ru
H
H
Ru
Oi-Pr
Me
H
S
H
Me
O
S
S
X
parent complex (X = H)
dichloro complex (X = Cl)
dichloro complex (X = Cl)
!
&#$#%
X
H
pcPr
Me
shift prodPr
O
NMe2
Ru-4
X-ray of Ru-4
Reaction coordinate
f Effective protocol for synthesis of modified Ru catechothiolates
MesN
NMes
Cl
S
C(NHC)–Ru–S = 148.0°
(versus 143.4° for Ru-3a)
Cl
Cl
NH2
SH
SH
H N
S
S
Ru-1
2
H
Zn
Ru
Zn(OAc) •2H O, i-PrOH,
2
2
O
S
2
2 °C, 1 h
Cl
Cl
Cl
4
5
Commercially
available
95% yield
(stable white solid)
Ru-3b
85% yield
X-ray of Ru-3b
Figure 1
|
Unaddressed issues in Z-selective olefin metathesis. a, Ru
vB97XD/6-3111G(2df,2p)THF(SMD)//vB97XD/6-311G(d,p)gas-phase level of
theory. f, Ru-3b was prepared in 81% overall yield with air-stable zinc
dithiolate 5, dichloro-dithiocatechol 4, ethylene diamine, zinc acetate and
carbenes such as Ru-1 are widely used in olefin metathesis but products are
generated either non-stereoselectively or as thermodynamic mixtures.
Complexes Ru-2 and Ru-3a have recently been developed for Z-selective olefin Ru-1, all of which are commercially available. Conversions (conv.) and
1
metathesis, but neither has been used in reactions with an aldehyde, a phenol stereoselectivities were determined by analysis of H NMR (nuclear magnetic
or a carboxylic acid. b, Critical shortcomings remain in the state-of-the-art
resonance) spectra of the unpurified mixtures; yields correspond to purified
olefin metathesis. Among them is the lack of efficient catalytic cross-metathesis compounds; see the Supplementary Information for details. Ar, aryl; mcb,
reactions that afford the products represented by i–iii. c, Cross-metathesis with metallacyclobutane; ts, transition state; pc, p complex; 14e, 14-electron active
Ru-2 and Ru-3a are Z-selective but moderately efficient. d, X-ray structure
of byproduct Ru-4 indicates that the sulfide ligands might participate in a
complex; shift ts, transition state for 1,2,-shift process; shift prod, product
resulting from 1,2-shift; ROMP, ring-opening metathesis polymerization;
ROCM, ring-opening/cross-metathesis; RCM, ring-closing metathesis; Mes,
1
,2-shift (A R B). e, Density functional theory calculations indicate that the
dichlorocatecholate complex is less prone to decomposition by a 1,2-sulfide
3 6 2
2,4,6-(Me) C H ; DFT, density functional theory; THF, tetrahydrofuran.
shift. Energy values correspond to the free energy DG calculated with the
1
8 2 | N A T U R E | V O L 5 1 7 | 8 J A N U A R Y 2 0 1 5
©
2015 Macmillan Publishers Limited. All rights reserved

LETTER RESEARCH
Table 1
|
Cross-metathesis of 1 and 2 with modified complexes
MesN
H
NMes
MesN
H
NMes
Br
MesN
H
NMes
MesN
H
NMes
F
MesN
H
NMes
Cl
S
Br
S
Me
S
S
S
F
Ru
Ru
Ru
Ru
Ru
O
S
O
S
O
S
O
S
O
S
F
Cl
Br
Cl
F
Me
Ru-3b
Ru-3c
Ru-3d*
Ru-3e
Ru-3f
Entry
Ru complex
Reaction duration (h)
Conversion (%)
Yield of 3 (%)
Z:E
1
2
3
4
5
6
Ru-3b; 3.0 mol%
Ru-3c; 3.0 mol%
Ru-3d*; 3.0 mol%
Ru-3e; 3.0 mol%
Ru-3f; 3.0 mol%
Ru-3b; 5.0 mol%
9
9
9
9
9
4
77
75
76
75
56
84
61
60
57
58
47
71
96:4
94:6
95:5
96:4
96:4
96:4
*
Inseparable mixture of Ru complex isomers.
1
2
Reactions were carried out under a N atmosphere. The percentage conversion was determined by analysis of the 400 MHz H NMR spectra of the unpurified mixtures and the variance of the conversion values is
1
estimated to be less than 62%. The variance of the yield for the isolated and purified product 3 (Z/E mixture) is estimated to be less than 65%. Z/E ratios were determined by H NMR analysis of unpurified mixtures
and the variance of the ratios is estimated to be less than 62%. See Supplementary Information for all details.
theory) investigating the following processes (Fig. 1e): nonproductive dichlorocatechothiolate species (14e ) with Z-2-butene to generate
Cl
olefin metathesis involving Z-2-butene (Bu), and competitive decom- the metallacyclobutane structure (labelled as mcbBu,Cl; Fig. 1e, right)
2
1
position that entails the 1,2-shift of the more electron-rich trans-to- was calculated to be 2.1 kcal mol lower in energy compared to ts
Bu,H
21
21
NHC sulfide ligand to the electrophilicethylidene(compare Figure 1d; (14.2 kcal mol versus12.1 kcal mol ; red versusbluecurve);thispoints
fordetailsseeSupplementaryInformation). The turnover-limitingbar- toincreased RuLewisacidityin14e and alessendergonic olefin coor-
Cl
21
21
rier (transition state derived from Z-2-butene and the chloro complex dination (6.9 kcal mol versus 8.8 kcal mol for the pcomplexes pcBu,Cl
Ru-3btsBu,Cl)fortheolefinmetathesisreactionoftheactive14-electron versus pcBu,H). Additionally, the more stabilized metallacyclobutane
Figure 2
|
The range of Z allylic alcohols made
HO
OH
5.0 mol% Ru-3b
available through catalytic stereoselective
cross-metathesis with Ru-3b. An assortment of Z
allylic alcohols can be accessed through cross-
metathesis reactions performed with 5.0mol%
Ru-3b. Transformations afford the desired
products in 53%–80% yield and 87:13–98:2 Z:E.
Cross-metathesis can occur between 1 and
terminal alkenes that contain another hydroxyl
group (13), a ketone (14), an aldehyde (15), or a
carboxylic acid (16 and 17). Sterically hindered
alkenes (18–21), Z- or E-1,3-dienes including those
that bear an electron-releasing (23) or an electron-
withdrawing (24) substituent can be used as well.
For synthesis of Z-alkenes 20 and 21 the use of
complex Ru-3d led to higher yields (44% yield
for 20 and 55% yield for 21 with Ru-3b, and
identical Z:E ratios (93:7–94:6 Z:E) in all cases).
Conversions and stereoselectivities were
or Ru-3d
HO
G
1
(
2.0 equiv.)
THF, 22 ºC
G
OTBS
O
NO2
HO
C9H19
HO
HO
HO
On-Bu
6
7
8
9
with Ru-3b
1% conv. (4 h), 72% yield,
6:4 Z:E
with Ru-3b
with Ru-3b
82% conv. (4 h), 74% yield,
with Ru-3b
8
76% conv. (4 h), 65% yield,
65% conv. (12 h), 57% yield,
9
93:7 Z:E
96:4 Z:E
91:9 Z:E
CO2Bn
O
OH
HO
N
CO Bn
HO
2
HO
HO
O
Ph
NHFmoc
1
0
11
with Ru-3b
12
with Ru-3b
83% conv. (4 h), 73% yield,
98:2 Z:E
13
with Ru-3b
with Ru-3b
8% conv. (4 h), 80% yield,
8:2 Z:E
8
73% conv. (4 h), 64% yield,
72% conv. (4 h), 64% yield,
96:4 Z:E
9
98:2 Z:E
gram-scale)
(
HO
O
1
determined by analysis of H NMR spectra of
O
O
OH
the unpurified mixtures; yields correspond to
isolated and purified products; 5.0 equivalents
of 1 was used for 18–21. See the Supplementary
Information for reaction conditions and
other details. Bn, benzyl; Fmoc,
HO
HO
OH
HO
Me
HO
O
Me
1
4
Me
15
16
17
with Ru-3b
with Ru-3b
7% conv. (4 h), 68% yield,
8:2 Z:E
with Ru-3b
with Ru-3b
7
86% conv. (4 h), 80% yield,
85% conv. (4 h), 70% yield,
66% conv. (4 h), 61% yield,
9
94:6 Z:E
96:4 Z:E
98:2 Z:E
fluorenylmethoxycarbonyl; TBS, tert-
butyldimethylsilyl.
CF3
HO
HO
HO
HO
1
8
19
with Ru-3b
69% conv. (4 h), 63% yield,
98:2 Z:E
20
with Ru-3d
57% conv. (4 h), 53% yield,
94:6 Z:E
21
with Ru-3d
66% conv. (4 h), 60% yield,
93:7 Z:E
with Ru-3b
5% conv. (4 h), 59% yield,
8:2 Z:E
6
9
C6H13
OMe
CO2t-Bu
HO
HO
HO
HO
C5H11
2
2
23
with Ru-3b
72% conv. (8 h), 63% yield,
92:8 Z:E
24
with Ru-3b
62% conv. (8 h), 56% yield,
96:4 Z:E
25
with Ru-3b
61% conv. (8 h), 54% yield,
87:13 Z:E
with Ru-3b
8% conv. (8 h), 66% yield,
5:5 Z:E
8
9
8
J A N U A R Y 2 0 1 5 | V O L 5 1 7 | N A T U R E | 1 8 3
©
2015 Macmillan Publishers Limited. All rights reserved

RESEARCH LETTER
1
0
derived from the dichlorocatechothiolate species hints at a weakened
Efforts to prepare Ru-3b (Fig. 1f) via sodium dithiolate precursors
2
1
21
trans influence (10.0 kcal mol versus 12.1 kcal mol for mcbBu,Cl initially proved unsatisfactory, as the desired Ru catechothiolate was
versus mcbBu,H). Equally noteworthy is that, whereas generation of the inseparablefromunidentified sideproducts(.98%conv.;,10% yield).
propene-bound product resulting from a 1,2-shift within the modified We subsequently established that reaction of 3,6-dichloro-catechothiol
complex (labelled as shift prodPr,Cl) from the corresponding propene- 4 with ethylenediamine and Zn(OAc)
N2H
bound p complex (pcPr,Cl) is endergonic by 0.5 kcal mol , the same mercially available—at room temperature for one hour affords the air-
process in the parent system (that is, pc
R shift prodPr,H) is appre- stable white solid Zn-dithiolate 5 (95% yield). Treatment of 5 with
O—all of which are com-
2
2
21
Pr,H
21
ciably exergonic (–2.5 kcal mol ). We attribute the latter discrepancy purchasable Ru-1 gave Ru-3b in 85% yield with little or no byproduct
to the alleviation of a stronger electronic repulsion between the Lewis formation, rendering purification of the derived Ru complex straight-
basic NHC and non-chlorinated sulfide ligands upon cleavage of the forward. Carbenes Ru-3c to Ru-3f (Table 1) were synthesized similarly.
Ru–S bond; Ru complexes shift prodPr are then generated that can In support of the suggested effect of electronically altered catechotiolate
further react through as yet unidentified mechanisms. The appreciably ligands, the X-ray structure ofRu-3b(Fig. 1f) showsa wider C(NHC)–Ru–S
21
10
lower barrier (9.5 kcal mol ) for the 1,2-shift induced by the p-acidic angle of 148.0u (versus 143.4u for Ru-3a ); this indicates a lowering of
ethylene (structure of the complex not shown) suggests a detrimental trans influence between the NHC and the anionic ligand, presumably
effect in cross-metathesis between two terminal olefins versus when due to diminished electron density at the sulfide site.
diol 1 is used. Overall, comparison of the energy barriers for the model
Efficiency of the cross-metathesis to generate Z allylic alcohol 3 im-
olefinmetathesisprocessinvolvingRu-ethylideneandZ-2-butene(pro- proved noticeably when the dichloro complex Ru-3b was used (77%
ceeding through a transition state labelled as tsBu) versus the 1,2-shift versus 50% with Ru-3a; entry 1 of Table 1), and the desired product
induced by the propene (via a transition state labelled as shift tsPr) was obtained in 61% yield with minimal decrease in Z selectivity (96:4
implies a favourable partitioning between productive and decomposi- versus 98:2 Z:E with Ru-3a). Dibromo (entry 2 of Table 1), chloro-
2
1
21
tionpathwaysinthecaseofRu-3b(12.1 kcal mol and13.4kcal mol
for the transition states derived from olefin metathesis with Z-2-butene plexes Ru-3c to Ru-3e were equally effective and stereoselectivities
tsBu,Cl) versus the 1,2-shift in the presence of bound propene (shift werehighacrosstheboard(94:6–96:4Z:E). However, Ru-3b has aprac-
tsPr,Cl), respectively); for Ru-3a, the alternatives are equally preferred tical advantage: unlike the other halogenated catechothiols, dichloro-
bromo (entry 3 of Table 1) and tetrafluoro (entry 4 of Table 1) com-
(
2
1
(
14.2 kcal mol for tsBu,H and shift tsPr,H). The halogen substituents catechothiol 4 (Fig. 1f) is commercially available. The lower efficiency
thus lower for the p complex formed between Z-2-butene and the of 2,5-dimethylcatechothiolate Ru-3f (entry 5 of Table 1; 56% conver-
dichlorocatechothiolate system (pcBu,Cl) versus that which is derived sion, 47% yield) is consistent with the positive electronic effect of the
from the catechothiolate species (pcBu,H).
halogen substituents in Ru-3b to Ru-3e. The use of 5.0 mol% Ru-3b
a Comparison of the activity of Ru-2 and Ru-3b complexes in Z-selective cross-metathesis b Application to stereoselective synthesis of neopeltolide and leucascandrolide A
i-Pr
Me
OMe
O
Me
NMes
Cl
Cl
MesN
N
MesN
H
O
O
O
Me
O
OMe
Cl
S
N
–
+
N
O
N
MeO
HN
MeO
HN
O
O
O
O
Ru
Ru
W
t-Bu
O
O
O
O
S
i-PrO
O
Cl
O
O
N
O
O
O
N
Ru-2
Ru-3b
W-1
Neopeltolide
anti-tumour)
LeucascandrolideA
(cytotoxic)
HO
CF3
(
HO
C6H13
HO
HO
MeO
HN
MeO
O
HN
O
O
Me
1
4
21
withRu-2:
19% conv., yield ND,
Z:E ND
withRu-3b:
60% conv., 55% yield,
93:7 Z:E
22
withRu-2:
59% conv., 53% yield,
93:7 Z:E
withRu-3b:
88% conv., 66% yield,
95:5 Z:E
10 mol%
W or Ru complex
withRu-2:
8% conv., 50% yield,
2:18 Z:E
5
7
OH
O
N
O
8
B(pinacolato), oxidation
or
withRu-3b:
N
7% conv., 68% yield,
HO
OH
9
8:2 Z:E
26
27
O
1
O
Conditions A:
W-1, 5.0 equiv. (pinacolato)allylboron,
HO
HO
OH
1
00 torr, C6H6, 22 °C, 3 h;
NaBO3, THF, 3 h:
1
5
16
60% conv., 51% yield, 91:9 Z:E
withRu-2:
1%conv., 30% yield,
7:13 Z:E
withRu-3b:
6% conv., 80% yield,
withRu-2:
<2%conv.
withRu-3b:
Conditions B:
Ru-2, 3.0 equiv. 1,
THF, 22 °C, 12 h:
Conditions C:
Ru-3a, 3.0 equiv. 1,
THF, 22 °C, 12 h:
61% conv., 55% yield,
97:3 Z:E
Conditions D:
4
8
Ru-3b, 3.0 equiv. 1,
THF, 22 °C, 12 h:
85% conv., 70% yield,
98:2 Z:E
8
85% conv., 70% yield,
25% conv., yield ND,
96:4 Z:E
61:39 Z:E
9
4:6 Z:E
Figure 3
|
Assessing the performance of different Z-selective catalysts.
Z-selective cross-metathesis between the terminal alkene in 26 and an
a, A representative set of products that allow comparison of the two Ru-based appropriate cross partner. The optimal high-oxidation-state W-based
carbene systems. Reactions with Ru-2 can be more sensitive to the size of
the substituent on the reaction partner (see 14 and 21), and cross-metathesis
with dienes can be more efficient with Ru-3b (see 22). Unlike Ru-3b, when
Ru-2 is employed, the presence of an aldehyde (see 15) lowers reaction rates
and carboxylic acids (see 16) completely shut down reactivity. All processes
alkylidene (10 mol%) requires the two-step process involving cross-metathesis
with (pinacolato)allylboron followed by C–B bond oxidation to afford 27 in
51% yield and a 91:9 Z:E mixture of difficult-to-separate alkene isomers.
With Ru-2 and 3.0 equivalents of Z-2-butene-1,4-diol 1 there is only 25%
conversion to the desired product, which is formed in low stereoselectivity
were performed (in duplicate or triplicate) with 5.0 mol% of the Ru complex at (61:39 Z:E). Reaction with Ru-3a is highly Z-selective but proceeds to ,60%
2
2 uC for four hours; see the Supplementary Information for details. b, A key
conversion. In contrast, when Ru-3b was used, Z allylic alcohol 27 is obtained
in 70% yield and 98:2 Z:E ratio. ND, not determined. See the Supplementary
step in the stereoselective synthesis of the side chains of the anti-tumour and
cytotoxic natural products neopeltolide and leucascandrolide entails a site- and Information and ref. 25 for details of conditions A to D.
1
8 4 | N A T U R E | V O L 5 1 7 | 8 J A N U A R Y 2 0 1 5
©
2015 Macmillan Publishers Limited. All rights reserved

LETTER RESEARCH
led to 84% conversion within four hours, delivering 3 in 71% yield and to 21 was observed after four hours in the case of Ru-2 (55% yield and
9
6:4 Z:E selectivity (entry 6 of Table 1).
93:7 Z:E with Ru-3b). Synthesis of 1,3-diene 22 proved to be somewhat
The catalytic protocol exhibits considerable scope (Fig. 2). Reactions more efficient with Ru-3b (66% versus 53% yield with Ru-2). Catalytic
with various terminal alkenes, including those with heterocyclic and/ cross-metathesis affording aldehyde-containing allylic alcohol 15 pro-
or Lewis basic functional groups, afforded Z allylic alcohols 6–12 in ceeded in 30% yield and 87:13 Z:E selectivity with Ru-2 (versus 86%
5
7–80% yield and 91:9–98:2 Z:E selectivity. Z allylic alcohol 6 has been conversion, 80% yield and 94:6 Z:E with Ru-3b). Carboxylic acids are
used to prepare (1)-disparlure, a female sex pheromone of the gypsy detrimental to the catalytic activity of complexes represented by Ru-2
21
moth Porthertria dispar (L.) . To demonstrate practicality, the trans- (ref. 6): we were unable to observe any transformation in our attempts
formation leading to the formation of phthalamide 11 was successfully toprepare16or17(,2%conversionversus70%and61%yield, respec-
performed on the gram scale. The Z-selective cross-metathesis leading tively, and $96:4 Z:E with Ru-3b). It appears that the highly polarized
to diol 13 took place with useful efficiency and stereoselectivity levels and nucleophilic Ru–alkyl bond in Ru-2 gives rise to complications
(
9
(
64% yield, 96:4 Z:E). The transformation that delivered 14 (68% yield, with aldehyde- or carboxylic-acid-containing substrates, reminiscent
8:2 Z:E) involves a substrate with a particularly acidic phenolic unit of Mo- or W-based alkylidenes. All processes shown in Fig. 2 were
owing to a para ketone). Synthesis of aldehyde-substituted Z allylic performed under a N atmosphere in a typical fume hood at ambient
2
alcohol15wasefficient(80%yield) and similarlyselective(94:6Z:E), as temperature (22 uC).
were the processes that generated carboxylic acid 16 (70% yield, 96:4
An occasion to challenge the utility of the cross-metathesis strategy
Z:E) or 17 (61% yield, 98:2 Z:E). The method is applicable to reactions presented itself in the course of the total synthesis of the anti-tumour
2
3
with sizeable a-branched substrates (see 18–21); alkyl- and aryl- agent neopeltolide ; the same acyclic (Z,Z)-diene appendage is found
substituted olefins participated in cross-metathesis reactions that in the cytotoxic natural product leucascandrolide A (ref. 24; Fig. 3b).
affordedthe desiredproducts in 53–63%yield and 93:7–98:2Z:E selec- We required the conversion of diene 26 to Z-allylic alcohol 27; the
tivity. Prolonged reaction times (that is, exceeding four hours) did not optimal conditions involved a stereoselectivecross-metathesis with (pina-
improve conversion. In the case of 20 and 21 slightly higher yields were colato)allylboron in the presence of 10 mol% W-1 (ref. 22), followed by
obtained with Ru-3d. Stereochemically defined 1,3-Z,E-dienes 22–24 oxidation of the resulting Z-(pinacolato)allylboron product to deliver
were generated with high stereoselectivity (92:8–96:4 Z:E); formation 27 in 51% overall yield and 91:9 Z:E selectivity. We sought a more effi-
of the more sterically congested (Z,Z)-1,3-diene 25 proceededin 87% Z cientand selectiveprocess, particularlysincethe alkeneisomersproved
selectivity. The only other examplesof Z-selective cross-metathesis involv- to be inseparable. With 10 mol% Ru-2 and diol 1, there was 25% con-
ing 1,3-dienes correspond to processes catalysed by Mo- or W-based version to 27 (12 h), which was formed with 61% Z selectivity. When
2
2
alkylideneswithallyl-oralkenyl(pinacolato)boroncompounds , where catechothiolate Ru-3a was employed, 27 was isolated in 55% yield and
pinacolato is (CH ) C O . 97% Z selectivity. In contrast, with Ru-3b there was 85% conversion to
3
4 2 2
The above findings illustrate that replacing the chloride ligands of a thedesiredproduct, which was isolated in 70%yield and 98:2 Z:E selec-
Ru complex with an aryl disulfide unit does not alter the ability of the tivity. Oxidation of the alcohol delivered the neopeltolide side chain,
2
5
catalysttopromotecross-metathesis withsizeablesubstratesand/orthose allowing for the completion of the total synthesis .
with a carbonyl or a Brønsted acid group. For comparison, we per-
Another compelling area of application relates to transformations
formed a number of the reactions with Ru-2 (Fig. 3a). Allylic alcohol involving animal fats and vegetable oils, readily accessible and inex-
1
4 was obtained in 50% yield and 82:18 Z:E selectivity when Ru-2 was pensive compounds that represent viable substitutes for the dwindling
26
used(versus68%yieldand98:2Z:EwithRu-3b).Thereisalargergapin petrochemicals . Such starting materials, most of which consist of car-
efficiency with a sterically demanding aryl alkene, where 19% conversion boxylicacidsandalcohols(,80%), are beingemployed withincreasing
a
Different strategies for catalytic Z-selective cross-metathesis
Figure 4
|
Transformation of renewable raw
Desired
R G1
Undesired
materials to two different Z allylic alcohols by
stereoselective cross-metathesis. a, Catalytic
cross-metathesis of monosubstituted alkenes
generates ethylene as a byproduct, which can
lead to the formation of unstable methylidene
complexes (reduced catalyst lifetime) with the
ability to facilitate Z-to-E isomerization. In
contrast, a catalyst may facilitate cross-metathesis
between two Z-disubstituted olefins—one of which
is structurally symmetric and both are easily
accessible and inexpensive—to produce two new
Z alkenes of higher value without homocoupling or
detrimental effects of ethylene formation. b, Easily
accessible and highly abundant plant oils (oleyl
alcohol and oleic acid) can be converted to two
high-value and readily separable Z-alkene products
efficiently and stereoselectively. Transformations
proceed readily at ambient temperature and
pressure within six hours without the need for
purification of the substrates or protection/
deprotection of either the carboxylic acid or the
hydroxyl units. Reactions with Ru-3a are less
efficient and less stereoselective. G and R,
functional groups.
G1
G1
G1
Homocoupling byproducts
and ethylene byproduct:
carbon waste,
can lower catalyst lifetime,
promotes post-metathesis isomerization
catalyst
+
Involving terminal alkenes:
R
R
R
homocoupling
byproducts
G2
R
G1
R
R
G1
G2
G1
G2
G1
G2
1. Z-alkene substrates must be
readily accessible and inexpensive
2. Products must be easily separable
3. Reverse reaction must be less favoured
Involving Z-disubstituted
catalyst
+
alkenes:
R
4
. Self-metathesis must be minimal
self-metathesis
byproducts
b
Stereoselective cross-metathesis with two readily accessible Z-alkenes to access higher-value Z-alkene products
Readily accessible,
easy-to-use robust catalyst
Renewable raw
materials
Commercially
available,
inexpensive
cross partner
used as received)
High-value
Z-alkene
products
High-value
Z-alkene
products
+
as substrates
used as received)
(
Mild conditions, few hours,
no substrate purification,
no protection/deprotection
(
HO
HO
HO
C7H15
OH
HO
OH
5.0 mol% Ru-3b
C7H15
THF, 22 ºC, 6 h
1
28
29
Oleyl alcohol
(
2.0 equiv.)
62% yield, 96:4 Z:E
59% yield, 94:6 Z:E
(1.0 g)
(
0.43 g)
(0.37 g)
O
O
HO
HO
HO
HO
OH
5.0 mol% Ru-3b
C H
7
15
OH
C7H15
THF, 22 ºC, 6 h
1
Oleic acid
1.0 g)
30
29
(
2.0 equiv.)
(
65% yield, 94:6 Z:E
0.46 g)
60% yield, 94:6 Z:E
(
(0.36 g)
8
J A N U A R Y 2 0 1 5 | V O L 5 1 7 | N A T U R E | 1 8 5
©
2015 Macmillan Publishers Limited. All rights reserved

RESEARCH LETTER
2
7
9. Occhipinti, G., Hansen, F. R., T o¨ rnroos, K. W. & Jensen, V. R. Simple and highly
regularity . Olefin metathesis catalysts that directly convert renewable
natural resources to higher-value products efficiently and stereoselec-
Z-selective ruthenium-based olefin metathesis catalyst. J. Am. Chem. Soc. 135,
3
331–3334 (2013).
tivelyareexpectedtoplayacentralroleinthelong-termfutureofdiverse 10. Khan, R. K. M., Torker, S. & Hoveyda, A. H. Readily accessible and easily modifiable
2
7
industries . Additionally, the combined use of two different 1,2-
disubstituted alkene substrates offer definite advantages (over terminal
alkenes; Fig. 4a). Yet, previous attempts at catalysing cross-metathesis
with plentiful raw materials have led to inefficient and non-stereoselective
Ru-based catalysts for efficient and Z-selective ring-opening metathesis
polymerization and ring-opening/cross-metathesis. J. Am. Chem. Soc. 135,
1
0258–10261 (2013).
1
1. Koh, M. J., Khan, R. K. M., Torker, S. & Hoveyda, A. H. Broadly applicable Z- and
diastereoselective ring-opening/cross-metathesis catalyzed by a dithiolate Ru
complex. Angew. Chem. Int. Ed. 53, 1968–1972 (2014).
2
8
transformations . Despite the use of dichloro–Ru complexes, reactions
largely involvedprotectedoleicacidderivatives. Ina recenteffort, cross-
metathesis of oleic acid or methyl oleate with 1 or its bis-acetate deriv-
ativewithaRu-basedcomplexat50 uCaffordedproductsasequilibrium 13. Mann, T. J., Speed, A. W. H., Schrock, R. R. & Hoveyda, A. H. Catalytic Z-selective
mixtures of difficult-to-separate stereoisomers (70%–85% E) . More-
1
2. Hoveyda, A. H. Evolution of catalytic stereoselective olefin metathesis. From
ancillary transformation to purveyor of stereochemical identity. J. Org. Chem. 79,
4763–4792 (2014).
2
8
cross-metathesis with secondary silyl- and benzyl-protected allylic ethers:
mechanistic aspects and applications to natural product synthesis. Angew. Chem.
Int. Edn 52, 8395–8400 (2013).
over, to ensure high activity, 100 equivalents (versus the Ru complex)
of toxic PhSiCl had to bepresent, andself-metathesis byproducts were 14. Buchmeiser, M. R., Sen, S., Unold, J. & Frey, W. N-Heterocyclic carbene, high
3
2
8
oxidation state molybdenum alkylidene complexes: Functional-group-tolerant
cationic metathesis catalysts. Angew. Chem. Int. Edn 53, 9384–9388 (2014).
5. Lin, Y. A. & Davis, B. G. The allylic chalcogen effect in olefin metathesis. J. Org. Chem.
6, 1219–1228 (2010).
formed (,20%) . In contrast, treatment of one gram of oleyl alcohol
and 1 with 5.0 mol% Ru-3b for six hours at ambient temperature led to
theformationofdiol28(0.43g, 62%yield)andallylicalcohol29(0.37 g,
9% yield); these easily separable products were formed in 96:4 and 16. Torker, S., Khan, R. K. M. & Hoveyda, A. H. The influence of anionic ligands on
4:6 Z:E selectivity, respectively (Fig. 4b). In the same fashion, a gram
of oleic acid was converted to 0.46 g of acid-alcohol 30 (65% yield) with
4:6 Z:E selectivity and 0.36 g of 29 (94:6 Z:E). Carboxylic acid 30 is an
anti-fungal agent , and allylic alcohol 29 is a substrate for stereose-
lective organic synthesis (see the Supplementary Information for fur-
ther references).
SeveraladditionalpointsregardingthedatainFig. 4barenoteworthy. 19. Cannon, J. S. & Grubbs, R. H. Alkene chemoselectivity in ruthenium-catalyzed
1) Substrates were used as received from commercial vendors without
purification; there was no need for rigorous removal of water or oxy-
gen. (2) Cross-metathesis with the parent complex Ru-3a was mark-
edly less efficient and stereoselective; for example, in the reaction with
oleic acid, 30 and 29 were obtained in ,18–20% yield and ,86:14 Z:E
ratio.(3)Sideproductsfromself-metathesisofoleylalcoholoroleicacid
were not detected. It is possible that formation of the hydroxymethylene-
substituted carbene is strongly favoured (due in part to an internal
1
5
9
stereoisomerism of Ru carbenes and their importance to efficiency and selectivity
of catalytic olefin metathesis reactions. J. Am. Chem. Soc. 136, 3439–3455 (2014).
1
7. Werner, H., Gr u¨ nwald, C., St u¨ er, W. & Wolf, J. Deactivation of the Grubbs carbene
complex [RuCl
(5CHPh)(PCy ] by allylic alcohols. Organometallics 22,
1558–1560 (2003).
9
2
3 2
)
2
9
1
8. Hoveyda, A. H., Lombardi, P. J., O’Brien, R. V. & Zhugralin, A. R. H-bonding as a
control element in stereoselective Ru-catalyzed olefin metathesis. J. Am. Chem.
Soc. 131, 8378–8379 (2009).
Z-selective olefin metathesis. Angew. Chem. Int. Edn 52, 9001–9004 (2013).
(
2
0. Hartung, J. & Grubbs, R. H. Catalytic, enantioselective synthesis of 1,2-anti-diols by
asymmetricring-opening/cross-metathesis. Angew. Chem. Int.Ed. 53, 3885–3888
(2014).
2
1. Rossiter, B. E., Katsuki, T. & Sharpless, K. B. Asymmetric epoxidation provides
shortest routes to four chiral epoxy alcohols which are key intermediates in
syntheses of methymycin, erythromycin, leukotriene C-1, and disparlure. J. Am.
Chem. Soc. 103, 464–465 (1981).
2
2. Kiesewetter, E. T. et al. Synthesis of Z-(pinacolato)allylboron and
Z-(pinacolato)alkenylboron compounds through stereoselective catalytic cross-
metathesis. J. Am. Chem. Soc. 135, 6026–6029 (2013).
1
1
H-bond ); reaction of the latter species with 1 is degenerate and its
transformation with the oleyl alcohol and oleic acid would result in the
re-formation of the desired cross-metathesis products. However, such
processes are no longer degenerate if they lead to generation of the E
product isomer; the high Z selectivity even at advanced stages of the
processunderscores the fidelity ofstereoselective cross-metathesis with
Ru catechothiolate complexes.
2
3. Wright, A. E. et al. Neopeltolide, a macrolide from Lithistid sponge of the family
Neopeltidae. J. Nat. Prod. 70, 412–416 (2007).
4. D’Ambrosio, M., Guerriero, A., Debitus, C. & Pietra, F. 6. Leucascandrolide A, a new
type of macrolide: the first powerfully bioactive metabolite of calcareous sponges
2
(Leucascandra caveolata, a new genus from the coral sea). Helv. Chim. Acta 79,
51–60 (1996).
2
5. Miao. Yu. Schrock, R. R. & Hoveyda, A. H. Catalyst-controlled stereoselective olefin
metathesis as a principal strategy in multi-step synthesis design. A concise route
to (1)-neopeltolide. Angew. Chem. Int. Ed. http://dx.doi.org/doi: 10.1002/
anie.201409120 (2014).
Theaboveattributesdistinguishthepresentclassofolefinmetathesis
catalysts not only from Mo- and W-based alkylidenes but from other 26. Biermann, U., Bornscheuer, U., Meier, M. A. R., Metzger, J. & Sch a¨ fer, H. Oils and fats
as renewable raw materials in chemistry. Angew. Chem. Int. Ed. 50, 3854–3871
types of Ru complexes as well. That is, the advantage of the Ru cate-
(
2011).
chothiolate system extends beyond promoting olefin metathesis with
stereochemical control, as the previous studies show that dichloro–
Ru complexes (for example, Ru-1) cannot efficiently promote cross-
metathesis of common disubstituted alkene feedstocks such as oleic
2
2
2
7. Gunstone, F. D. in Oleochemical Manufacture and Applications (eds Gunstone, F. D.
& Hamilton, R. J.) Vol. 1 (Academic Press, 2001).
8. Behr, A. & Gomes, J. P. The cross-metathesis of methyl oleate with cis-2-butene-
1
,4-diylacetate and theinfluence of protecting groups. J. Org. Chem. 7, 1–8 (2011).
9. Suzuki, Y., Kurita, O., Kono, Y., Hyakutake, H. & Sakurai, A. Structure of a new
antifungalC11-hydroxyfatty acid isolated fromleaves ofwild rice(Oryza officinalis).
Biosci. Biotechnol. Biochem. 59, 2049–2051 (1995).
28,30
acid (considerable substrate self-metathesis is observed)
.
Received 15 September; accepted 4 November 2014.
30. Kajetanowicz, A., Sytniczuk, A. & Grela, K. Metathesis of renewable raw materials—
influence of ligands in the indenylidene type catalysts on self-metathesis of methyl
oleate and cross-metathesis of methyl oleate with (Z)-2-butene-1,4-diol diacetate.
Green Chem. 16, 1579–1585 (2014).
1.
2.
3.
Hoveyda, A. H. & Zhugralin, A. R. Theremarkable metal-catalyzed olefinmetathesis
reaction. Nature 450, 243–251 (2007).
F u¨ rstner, A. Teaching metathesis ‘‘simple’’ stereochemistry. Science 341,
Supplementary Information is available in the online version of the paper.
1357–1364 (2013).
Ibrahem, I., Yu, M., Schrock, R. R. & Hoveyda, A. H. Highly Z- and enantioselective
ring-opening/cross-metathesis reactions catalyzed by stereogenic-at-Mo
adamantylimido complexes. J. Am. Chem. Soc. 131, 3844–3845 (2009).
Meek, S. J., O’Brien, R. V., Llaveria, J., Schrock, R. R. & Hoveyda, A. H. Catalytic
Z-selective olefin cross-metathesis for natural product synthesis. Nature 471,
Acknowledgements This research was supported by a grantfrom the National Science
Foundation (CHE-1362763). R.K.M.K. and M.Y. were partially supported as
AstraZeneca Graduate Fellows. We thank Boston College for access to computational
facilities.
4.
4
61–466 (2011).
Author Contributions M.J.K. and R.K.M.K. carried out the catalyst synthesis, method
development studies and applications related to renewable feedstock, S.T. performed
the computational investigations, M.Y. carried out the experiments in connection with
neopeltolide, and M.S.M. studied modes of catalyst decomposition. A.H.H. conceived
and directed the investigations and composed the manuscript with revisions provided
by the other authors.
5
.
.
Yu, M. et al. Synthesis of macrocyclic natural products by catalyst-controlled
stereoselective ring-closing metathesis. Nature 479, 88–93 (2011).
Keitz, B. K., Endo, K., Herbert, M. B. & Grubbs, R. H. Z-selective homodimerization of
terminal olefins with a ruthenium metathesis catalyst. J. Am. Chem. Soc. 133,
6
9686–9688 (2011).
7
8
.
.
Keitz, B. K., Endo, K., Patel, P. R., Herbert, M. B. & Grubbs, R. H. Improved ruthenium
catalysts for Z-selective olefin metathesis. J. Am. Chem. Soc. 134, 693–699 (2012).
Rosebrugh, L. E., Herbert, M. B., Marx, V. M., Keitz, B. K. & Grubbs, R. H. Highly active
ruthenium metathesis catalysts exhibiting unprecedented activity and Z
selectivity. J. Am. Chem. Soc. 135, 1276–1279 (2013).
Author Information Reprints and permissions information is available at
www.nature.com/reprints. The authors declare no competing financial interests.
Readers are welcome to comment on the online version of the paper. Correspondence
and requests for materials should be addressed to A.H.H. (amir.hoveyda@bc.edu).
1
8 6 | N A T U R E | V O L 5 1 7 | 8 J A N U A R Y 2 0 1 5
©
2015 Macmillan Publishers Limited. All rights reserved