Catalytic Z-selective olefin cross-metathesis for natural product synthesis

Article abstract of DOI:10.1038/nature09957

Alkenes are found in many biologically active molecules, and there are a large number of chemical transformations in which alkenes act as the reactants or products (or both) of the reaction. Many alkenes exist as either the E or the higher-energy Z stereoisomer. Catalytic procedures for the stereoselective formation of alkenes are valuable, yet methods enabling the synthesis of 1,2-disubstituted Z alkenes are scarce. Here we report catalytic Z-selective cross-metathesis reactions of terminal enol ethers, which have not been reported previously, and of allylic amides, used until now only in E-selective processes. The corresponding disubstituted alkenes are formed in up to >98% Z selectivity and 97% yield. These transformations, promoted by catalysts that contain the highly abundant and inexpensive metal molybdenum, are amenable to gram-scale operations. Use of reduced pressure is introduced as a simple and effective strategy for achieving high stereoselectivity. The utility of this method is demonstrated by its use in syntheses of an anti-oxidant plasmalogen phospholipid, found in electrically active tissues and implicated in Alzheimers disease, and the potent immunostimulant KRN7000.

Full text of DOI:10.1038/nature09957

ARTICLE
doi:10.1038/nature09957
Catalytic Z-selective olefin
cross-metathesis for natural product
synthesis
Simon J. Meek1, Robert V. O’Brien1, Josep Llaveria1, Richard R. Schrock2 & Amir H. Hoveyda1
Alkenes are found in many biologically active molecules, and there are a large number of chemical transformations in
which alkenes act as the reactants or products (or both) of the reaction. Many alkenes exist as either the E or the
higher-energy Z stereoisomer. Catalytic procedures for the stereoselective formation of alkenes are valuable, yet
methods enabling the synthesis of 1,2-disubstituted Z alkenes are scarce. Here we report catalytic Z-selective
cross-metathesis reactions of terminal enol ethers, which have not been reported previously, and of allylic amides,
used until now only in E-selective processes. The corresponding disubstituted alkenes are formed in up to .98% Z
selectivity and 97% yield. These transformations, promoted by catalysts that contain the highly abundant and
inexpensive metal molybdenum, are amenable to gram-scale operations. Use of reduced pressure is introduced as a
simple and effective strategy for achieving high stereoselectivity. The utility of this method is demonstrated by its use in
syntheses of an anti-oxidant plasmalogen phospholipid, found in electrically active tissues and implicated in
Alzheimer’s disease, and the potent immunostimulant KRN7000.
Carbon-carbon double bonds reside within a large variety of molecules catalytic cycle. As a result, the stereogenic-at-Mo complexes are
that possess desirable properties1, and catalytic cross-metathesis2 (CM; generally more effective olefin metathesis catalysts than other Mo-
Fig. 1) represents one of the most attractive approaches to stereoselec- based complexes 39 and 410 or Ru carbene 511. We thus established
tive preparation of these versatile functional groups. Through fusion of that alkylidene 2 readily catalyses Z-selective alkene formation
two terminal alkenes, available in ample quantities as by-products of through ring-opening/cross-metathesis (ROCM)12 with strained
petroleum purification or readily accessed by a variety of methods, 1,2- oxabicyclic alkenes and styrenes. Homocoupling of terminal alkenes
disubstituted alkenes can be obtained; the other product generated is was subsequently shown to proceed with high efficiency and Z selec-
gaseous ethylene. However, the only reported instances of Z-selective tivity in the presence of members of the same catalyst class13. The
CM (65–90% Z) involve substrates with an sp-hybridized substituent general mechanistic features that engender Z selectivity in the above
(acrylonitrile or enynes3,4). In an efficient Z-selective CM, it is not only reactions, and would be expected to do so in a CM process, are
required that reaction between the two substrates proceed selectively depicted in Fig. 1. The preference for Z alkene formation can be
(versus homocoupling), it must exhibit a preference for the thermo- attributed to the ability of the large monodentate aryloxide to rotate
dynamically less favoured stereoisomer (Fig. 1). The inherent reversi- freely (compare I in Fig. 1), causing the incoming alkene to be
bility of olefin metathesis (products can re-enter the catalytic cycle) oriented such that its substituent (R2) is situated syn to that of the
and thehigherreactivity of Z alkenes(versus E isomers) further exacer- alkylidene (R1).
bate the problem. Through careful consideration of various mech-
anistic aspects of the process, conditions must be identified where
the catalyst promotes CM but fails to react with the product Z alkene
to effect equilibration, favouring the lower energy E isomer.
Designing a Z-selectiveCMissubstantiallymore difficult:ina homo-
coupling, only one type of alkene is involved and no more than two
stereoisomeric alkenes can be formed; in contrast, there are two sub-
strates in CM, which can generate up to six different products. In the
case of a catalytic ROCM14, a strained cyclic alkene and a terminal
alkene that are reluctant to undergo homocoupling (for example, a
Challenges of catalytic Z-selective cross-metathesis
As the preliminary steps towards the eventual development of an effi- styrene) must be selected as substrates for the catalytic process to be
cient class of Z-selective CM reactions, we investigated two related— efficient15.Transformationsarecarefullycraftedsuchthatthealkylidene
but much simpler—versions of the process. Alkylidenes and carbenes derived from the terminal alkene favours association with the cyclic
1–5 (Fig. 1) represent the catalyst classes used in our studies. alkene (versus another of the same type) in the ring-opening stage,
Stereogenic-at-Mo 1a, 1b and 25,6 were recently designed in these generating a new Mo complex that strongly prefers to react with a
laboratories to promote enantioselective ring-closing metathesis; sterically less demanding terminal alkene (CM stage). The possibility
theoretical7 and experimental explorations8 suggest that these com- of a transformation between the alkylidene generated through ring-
plexesexhibit high activitypartly asthe resultofstereoelectroniceffects opening and another strained—but more hindered—cyclic alkene is
induced by the electron donor pyrrolide and acceptor monoaryloxide thus discouraged (that is, minimal homocoupling or oligomerization).
ligands. The fluxional nature of complexes such as 1a, 1b and 2, Such deliberate orchestration is not feasible with catalytic CM, where
facilitated by the absence of rigid bidentate ligands, allows the metal both alkenes are mono-substituted and, in contrast to ROCM, there is
alkylidenes to adapt to the structural strains imposed during the no relief of ring strain to be manipulated.
1Department of Chemistry, Merkert Chemistry Center, Boston College, Chestnut Hill, Massachusetts 02467, USA. 2Department of Chemistry, Massachusetts Institute of Technology, Cambridge,
Massachusetts 02139, USA.
2 4 M A R C H 2 0 1 1
| V O L 4 7 1 | N A T U R E | 4 6 1
Macmillan Publishers Limited. All rights reserved
©2011

RESEARCH ARTICLE
Z-Selective cross-metathesis of terminal alkenes
Cross-metathesis
products
Homocoupling
products
G1
G2
G1
G1
G1 G2
G2
G1
Z alkenes
higher in energy
than corresponding
E alkenes
Catalyst-induced kinetic control
is required for
stereoselective Z-alkene synthesis
(vs substrate-induced thermodynamic control)
Catalyst
G1
G2
G2
G2
G1
G2
R
N
R
i-Pr
i-Pr
i-Pr
i-Pr
MesN
Cl
NMes
Cl
N
N
N
N
N
Mo
Ph
Ph
Mo
Mo
Ru
Mo
Ph
Ph
O
O
O
F3C
F3C
O
O
O
Br
Br
Br
Br
TBSO
Oi-Pr
TBSO
F3C CF3
1a R = Me
1b R = i-Pr
2
3
4
5
Rotation
around
Mo–O bond
N
N
N
N
N
N
N
N
R2
R1
Mo
R1
Br
Mo
R1
R2
Mo
Mo
Br
O
O
O
R1
R2
R2
Br
Br
O
Br
TBSO
Br
Br
TBSO
TBSO
Br
OTBS
Rotating and large
monodentate aryloxide
ligand generates a
I
II
III
IV
significant steric presence
R1
R2
Figure 1
|
A catalytic CM reaction can afford as many as six alkenes, so the (3–5). The structural flexibility of the stereogenic-at-metal complexes 1 and
challenge is designing an efficient process that favours formation of the
2 can give rise to exceptional reactivity, and free rotation around the Mo–O
cross products. Particularly difficult is the development of a process that
bond of these alkylidenes might serve as the basis for development of highly
affords the higher-energy Z alkene predominantly. To accomplish a Z-selective Z-selective olefin metathesis reactions of terminal alkenes. The sphere
CM, a variety of catalysts were considered, such as stereogenic-at-Mo
represents an appropriate size imido substituent.
complexes (1, 2) or other previously reported Mo- and Ru-based complexes
Z-Selective cross-metathesis of enol ethers
.98% Z-8a is generated; further transformation is not observed after
six hours. Mo-based diolate 3 and Ru carbene 5 do not promote CM,
and achiral Mo complex 4 catalyses a non-selective transformation
(47.5% Z). Thus, stereogenic-at-Mo complexes prove to be effective in
promoting enol ether CM, and although 1b or the less hindered 2 also
afford exceptional stereoselectivity, neither delivers the efficiency of
1a. The 2,6-dimethylphenylimido 1a therefore offers the best balance
between activity and stereoselectivity. Such performance variations
may be observed because catalyst turnover is slower with the more
sizeable 1b whereas the methylidene of the relatively unhindered 2
(compare IV, Fig. 1) might suffer from a shorter life span. Consistent
with the above scheme, 82% 8a is formed when CM with 1b is allowed
to continue for 16 hours; in contrast, conversion with 2 after 10minutes
or two hours is nearly identical (,38%).
Thereareseveral, mechanisticallyrevealing, reasonsforuseofexcess
enol ether. CM generates a Mo-methylidene; this unhindered alkyli-
dene can readily react with the Z-alkene product, reverse CM, cause
equilibration and lower stereoselectivity. An enol ether reacts with a
methylidene complex, circumventing diminution in Z selectivity. The
more stablealkoxy-substitutedalkylidene, generatedfromreactionofa
methylidene complex and an enol ether (I in Fig. 1 with R1 5 On-Bu),
can undergo productive CM, giving rise to longer catalyst lifetime and
improved turnover numbers. Furthermore, generation of the afore-
mentioned alkoxy- or aryloxy-containing alkylidene means less of
the alkyl-substituted derivative is formed and homocoupling of the
aliphatic alkene is minimized. Owing to electronic factors, productive
reaction between an enol ether-derived alkylidene and another
O-substituted alkene is disfavored2. However, as use of excess enol
ether is wasteful, we decided to examine the efficiency of the CM with
We began by evaluating the ability of stereogenic-at-Mo complexes to
promote transformations of enol ethers, a class of substrates for which
a CM reaction has not been previously reported (E-or Z-selective); the
resulting products have proven to be of utility in chemical synthesis
and can be found in biologically active molecules (see below). In the
presence of 2.5 mol% 1a, CM between 6 and 7 (entry 1, Table 1)
proceeds to 85% conversion to afford disubstituted enol ether 8a in
98% Z selectivity and 73% yield. With 1b, which bears a more sizeable
2,6-di-i-propyl-arylimido unit, the reaction is completely Z-selective
(.98% Z) but 47% conversion is achieved within the same time span.
When alkylidene 2 is used, CM proceeds to 37% conversion and
Table 1
|
Examination of various catalysts for CM with an enol ether
2.5 mol%
Mo or Ru complex
Bn
On-Bu
On-Bu
Ph
C6H6, 22 °C
6
7
8a
Entry no.
Complex
Time
Conv. (%)*
Yield (%){
Z:E*
1
2
3
4
5
6
1a
1b
2
3
4
2 h
2 h
2 h
2 h
10 min
24 h
85
47
37
,2
80
,2
73
98:2
ND
ND
NA
ND
NA
.98:2
.98:2
NA
47.5:52.5
NA
5
The reactions were carried out in purified benzene under an atmosphere of nitrogen gas; 10 equiv. of 6
was used (see Supplementary Information for details). NA, not available; ND, not determined.
* Conversion (conv.) and Z:E ratios were measured by analysis of 400 MHz 1H NMR spectra of
unpurified mixtures; the variance of values is estimated to be ,62%.
{ Yield of isolated product after purification; the variance of values is estimated to be ,65%.
4 6 2
| N A T U R E | V O L 4 7 1 | 2 4 M A R C H 2 0 1 1
Macmillan Publishers Limited. All rights reserved
©2011

ARTICLE RESEARCH
varying amounts of 6 (see Supplementary Information for details). Synthesis of natural product C18 (plasm)-16:0 (PC)
The latter studies established that, although fewer equivalents of 6
lead to reduced Z selectivity and competitive homocoupling, with
5 equiv. of the inexpensive and commercially available enol ether,
8a can be obtained in 93:7 Z:E selectivity and 71% yield (7% homo-
coupled product). Excess enol ether 6 does not complicate product
isolation, as this inexpensive reagent is volatile and can be easily
removed in vacuo.
Z-Disubstituted enol ethers are obtained in 57–77% yield through
exceptionally stereoselective (94% to .98% Z) CM with Mo alkylidene
1a (Fig. 2). Alkyl- (8) or aryl-substituted (10) Z enol ethers as well as
those that bear a carboxylic ester (8c), a secondary amine (8e), a brom-
ide (10b) or an alkyne (10c) are readily accessed. Reactions with the
more electron-deficient enol ether 9 and the relatively electron-rich
alkenes proceed with 2.0 equiv. of the aryl-substituted enol ether; in
contrast, 10 equiv. of alkyl-substituted and easily removable 6 are
required for similar efficiency. Such variations probably occur because
when 9 is used there is a better electronic match3 between the Mo-
alkylidenes derived from the cross partners and either of the two
alkenes, favouring CM versus homocoupling. Only 1.2 mol% 1a and
2.0 equiv. of the p-methoxyphenylenol ether(for example, 10a, 10b and
Next, weset outtodemonstratetheutilityofthecatalyticCMprocessby
a diastereo- and enantioselective synthesis of an anti-oxidant plasmalo-
gen phospholipid, C18 (plasm)-16:0 (PC) (Fig. 2)16,17, the correspond-
ing E isomer of which has been shown to be less active17. This initiative
required addressing a challenge that is of general concern in catalytic
CM: the inefficiency associated with the use of excess of one cross
partner. The enol ether to be used (11) in the CM step is more valuable
than the commercially available and inexpensive 1-octadecene (12),
rendering utilization of excess amounts of the former unfavourable.
Reducing the enol ether concentration diminishes efficiency and
Z-selectivity, as detailed above and substantiated by the data in
Table 2 (85% and 47% conversion with 5:1 and 1:1 11:12; entries 1
versus 2). Larger quantities of the less valuable 12 could improve yield
and selectivity, as Mo-methylidene concentration is probably lowered
through its reaction with excess alkene. However, increased amounts of
an aliphatic alkene, unlike an enol ether, give rise to homocoupling and
ethylene generation. Ethylene, in addition to being detrimental to the
rateofCM(becauseitcompeteswiththesubstratesfor reaction withthe
available alkylidene), causes diminished stereoselectivity by increasing
methylidene concentration, which promotes Z alkene isomeriza-
10d, Fig. 3c) are sufficient for an effective and exceptionally Z-selective tion (see above). We thus surmised that, if the negative effects of the
CM to take place. generated ethylene were to be attenuated by performing the reaction
O
On-Bu
G
O
OMe
G
On-Bu
1.2–5.0 mol% 1a
G
or
or
6
9
C6H6, 22 °C, 2 h
OMe
8
10
(10 equiv.)
(2.0 or 10 equiv.)
H17C8
On-Bu
On-Bu
PhHN
On-Bu
(i-Pr)3SiO
On-Bu
PhO
O
8b
8c
8d
8e
76% conv., 68% yield,
76% conv., 73% yield,
86% conv., 77% yield,
57% conv., 51% yield,
98% Z
98% Z
94% Z
>98% Z
( )6
Bn
OPMP
Br
OPMP
( )6
Me3Si
OPMP
Cy
OPMP
10a
10b
10c
10d
71% conv., 57% yield,
74% conv., 70% yield,
66% conv., 59% yield,
81% conv., 75% yield,
>98% Z
>98% Z
>98% Z
>98% Z
Si(i-Pr)3
O
H33C16
14
O
a
b
11 (1.0 equiv.)
Ph
Ph
O
85% overall yield,
H33C16
i-Pr
>98% Z
S
O
N
N
O
12 (2.0 equiv.)
i-Pr
i-Pr
15
NMe3
O
P
Four steps
O
OH
O
O
O
O
O
H33C16
OH
H33C16
O
86% overall yield
(see ref. 12)
16
C
15H31
64% yield,
98:2 e.r., >98% Z
C18 (plasm)-16:0 (PC)
Figure 2
|
Z-selective CM reactions of enol ethers with terminal alkenes and and used in situ. Conversions and Z selectivities were determined by analysis of
application to stereoselective synthesis of C18 (plasm)-16:0 (PC). Various Z 400MHz 1H NMR spectra of unpurified mixtures; the variance of selectivity
enol ethers are synthesized with 1.2–5.0 mol% of Mo complex 1a, and typically values is estimated to be ,62%. Yields of isolated products are shown (65%).
require 2.0 equiv. (in the case of p-methoxyphenylvinyl ether) or 10.0 equiv.
(with butylvinyl ether) of the terminal enol ether; excess butyl vinyl ether (6) is used 1.2 mol% 1a and 2.0equiv. 9; for 10c, 10d, we used 5.0 mol% 1a and 10
easily removed in vacuo. The desired Z alkenes are obtained in 51–77% yield equiv. (10c) or 2.0 equiv. 9 (10d). See Supplementary Information for
Reactions: for 8b–8e, we used 2.5mol% 1a and 10 equiv. 6; for 10a, 10b, we
and in 94% to .98% Z selectivity. Application to synthesis of C18 (plasm)-16:0 experimental details. Conditions for synthesis of 16. Route a, step 1; 2.5 mol%
(PC) demonstratestheutility of theZ-selectiveMo-catalysedCM, whichis used 1a, C6H6, 22 uC, 2.0 h, decalin, 1.0torr: step 2; 5.0equiv. (n-Bu)4NF, THF,
in conjunction with a site- and enantioselective Cu-catalysed dihydroboration 22 uC, 2 h. Route b; 2.5 mol% 15, 2.5 mol% CuCl, 20 mol% NaOt-Bu, 2.1 equiv.
of the terminal alkyne in 14 (see Supplementary Information for details). All
bis(pinacolato)diboron, 3.0equiv. MeOH, THF, 0 uC, 24 h; 30% H2O2, NaOH
reactions shown were performed under N2 atmosphere;catalysts were prepared in aqueous THF, 1.0 h.
2 4 M A R C H 2 0 1 1
| V O L 4 7 1 | N A T U R E | 4 6 3
Macmillan Publishers Limited. All rights reserved
©2011

RESEARCH ARTICLE
Br
PMBO
PhO
( )6
( )6
N(phth)
OTBS
N(phth)
OTBS
N(phth)
OTBS
O
19b
19c
19d
96% conv., 93% yield
98% conv., 97% yield
65% conv., 63% yield
96% Z
93% Z
96% Z
Cy N(phth)
OTBS
H29C14
H29C14
N(H)Boc
19f
N(phth)
19e
65% conv., 65% yield
97% Z
19g
80% conv., 75% yield
81% Z
87% conv., 87% yield
85% Z
OBn
OBn
OBn
OBn
OBn
HO
H29C14
N(Boc)2
O
N(Boc)2
O
H29C14 N(Boc)2
O
O
O
O
OBn
OBn
a
b
H29C14
OBn
OBn
OH
BnO
BnO
BnO
20
18
(5.0 equiv.)
21
22
(98:2 e.r.; 1.0 equiv.)
94% conv., 85% yield,
89% yield, 92:8 d.r.
96% Z
c–d
C25H51
C
25H51
OBn
OBn
OH
OH
O
HO HN
OH
O
O
HO HN
OH
O
O
C25H51
O
O
N
O
O
H29C14
H
29C14
(95% yield;
see ref. 23)
O
OBn
OH
BnO
HO
23
24
87% overall yield
KRN7000
Figure 3
|
Z-selective CM reactions of allylic amides with terminal alkenes
N-containing alkenes (19b, 19c) or 5.0 mol% 2 and 10.0 equiv. of cross partner,
and application to stereoselective synthesis of KRN7000. A range of Z-1,2- 7.0torr, 5.0 hours, 22 uC; catalysts were prepared and used in situ. Conversions,
disubstituted allylic amides can be synthesized; in most cases, use of reduced Z selectivities, yields and Z:E ratios determined as in Fig. 1. For 19e, reduced
pressure leads to substantially improved yield and stereoselectivity. Application pressure was not used; reaction performed at 50 uC for 12 h. For
to the stereoselective synthesis of KRN7000, involving catalytic
diastereoselective dihydroxylation of the Z alkene obtained by Mo-catalysed
19f, 19g, reaction time was one hour. See Supplementary Information for
experimental details. Conditions for synthesis of 24: route a; 8.0 mol% 2 (in
CM, leads to an expeditious route for preparation of this biologically significant situ-generated), C6H6, 22 uC, 5.0h, 1.0torr. Route b; 5 mol% OsO4, 2.5equiv.
molecule (see Supplementary Information for details). All reactions shown N-Me-morpholine oxide, CH2Cl2, 0 uC, 24 h. Route c; 10% trifluoroacetic acid,
were performed under N2 atmosphere with 3.0mol% 2, 3.0equiv. of the non- CH2Cl2, 22 uC, 30 min. Route d; 1.2equiv. 23, Et3N, THF, 50 uC, 12 h.
undervacuum, anefficientCM might beinducedtoproceed withonlya C18 (plasm)-16:0 (PC) in four steps and 86% overall yield19. Two addi-
relatively slight excess of the aliphatic alkene (12). Indeed, when cata- tional points merit mention: (1) catalytic CM between 11 and 12 has
been performed on the gram-scale with 1.0 mol% of in situ-generated 1a
and 2 equiv. of 12, affording Z-13 with .98% stereoselectivity and in
71% yield after purification (3 h, 1.0 torr, 79% conversion; see
Supplementary Information for details). (2) The only previous synthesis
of 16 involves nine transformations starting with (S)-isopropylidene
glycerol (versus five reactions from (i-Pr)3Si-acetylene, Fig. 2) by a
sequence that includes the use of highly toxic hexamethylphosphor-
amide and a catalytic hydrogenation with lead-containing salts19.
lyticCMis performed withan equal amountof11 and 12under 1.0 torr
(entry 3, Table 2), efficiency (78% versus 47% conversion in entry 2) as
well as stereoselectivity is substantially improved (97% versus 91.5% Z).
With reduced pressure, 2 equiv. of 12 (versus 11) and decalin as solvent
(to prevent precipitation of the homocoupled by-product causing cata-
lyst sequestration), 89%conversionisobservedintwohoursandZ-13is
obtained with 97% selectivity (entry 4).
Removal of the silyl group delivers stereoisomerically pure Z-14 in
85% overallyield (Fig.2); the desired product cannotbe accessed through
catalytic hydrogenation of the corresponding alkyne (see also 10c). Cu-
catalysed site- and enantioselective dihydroboration18 furnishes 16 (in
98:2 enantiomeric ratio, e.r.), which has been previously converted to
Z-Selective cross-metathesis of allylic amides
Another class of reactions that we examined involves allylic amides as
substrates. Such catalytic CM reactions are of considerable value, as a
large number of biologically active molecules are nitrogen-containing,
and 1,2-disubstituted alkenes bearing a C–N bond at the allylic posi-
tion can be functionalized in a variety of ways. Furthermore, in con-
trast to transformations with enol ethers, CM with allylic amides
poses the added complication that both substrates can undergo
homocoupling. Preliminary investigations with enantiomerically
pure allylic amide 17 (from commercially available alcohol) and
1-hexadecene 18 indicated that the optimal catalyst for this class of
processes is derived from adamantylimido complex 2, affording the
desired Z alkene in 88% yield and with 97% stereoselectivity (entry 3,
Table 3). Although arylimido derivatives 1a and 1b generate 19a with
similar selectivity (entries 1 and 2, Table 3), reactions are inefficient
(26–44% versus 88% conversion), perhaps because an alkylidene
derived from 2 is less congested and can more readily promote CM
of the relatively hindered 17. The higher efficiency of CM with 2, in
Table 2
|
Effect of reduced pressure on efficiency and Z selectivity
Si(i-Pr)3
Si(i-Pr)3
O
11
2.5 mol% 1a
22 °C
H33C16
O
H33C16
12
13
Entry no. 11:12
Time (h)
Conv. (%)*
Solvent
Pressure
Z:E*
1
2
3
4
5:1
1:1
1:1
1:2
2
2
2
2
85
47
78
88
Benzene Ambient
Benzene Ambient
Benzene 1.0 torr
.98:2
91.5:8.5
97:3
Decalin
1.0 torr
97:3
The reactions were carried out in purified benzene or decalin under an atmosphere of nitrogen gas (see
Supplementary Information for details).
* Conversion, Z:E ratios and the amount of the homocoupled product were measured by analysis of
400 MHz 1H NMR spectra of unpurified mixtures; the variance of values is estimated to be ,62%.
4 6 4
| N A T U R E | V O L 4 7 1 | 2 4 M A R C H 2 0 1 1
Macmillan Publishers Limited. All rights reserved
©2011

ARTICLE RESEARCH
Table 3
amide
|
Examination of various catalysts for CM with an allylic
and adventitious homocoupling. Transformations performed under
ambient conditions (no vacuum) may not proceed beyond 80% con-
version, probably because the ethylene by-product competes with the
remainingcrosspartnermolecules. Highethyleneconcentration might
also diminish CM rate through formation of relatively stable unsub-
stitutedmetallacyclobutanes7. Asmentionedbefore, methylidenecom-
plexes can engender reduction of Z selectivity; time-dependent studies
indicate that stereoselectivities suffer with prolonged reaction times.
With non-volatile substrates, if reactions are carried out under
vacuum, complete consumption of the limiting alkene is observed only
when excess amounts (,10 equiv.) of one cross partner are present.
Under such regimes, however, difficulties associated with removal of
the excess substrate and the homocoupled product might arise, ren-
dering the use of lesser alkene amounts preferable. With lower sub-
strate ratios, .98% consumption of the limiting substrate is difficult to
achieve, as terminal alkene concentration is diminished as a result of
partial homocoupling.
N(phth)
OTBS
H29C14 N(phth)
OTBS
3.0 mol% complex
C14H29
C6H6, 22 °C, 5 h,
7.0 torr
19a
17
18
(1.0 equiv.)
(3.0 equiv.)
Entry no.
Complex
Conv. (%)*
Yield (%){
Z:E*
1
2
3
4
5
6
1a
1b
2
3
4
44
26
93
9
71
73
35
21
88
6
68
64
96:4
97:3
97:3
21:79
12:88
11:89
5
The reactions were carried out in purified benzene under an atmosphere of nitrogen gas (see
Supplementary Information for details). N(phth) 5 N-phthalamide.
* Conversion and Z:E ratios were measured by analysis of 400 MHz 1H NMR spectra of unpurified
mixtures; the variance of values is estimated to be ,62%.
{ Yield of isolated product after purification; the variance of values is estimated to be ,65%.
contrast to those involving enol ethers (Fig. 2), might be the result of
CM with 17 being performed under vacuum, allowing minimal
amounts of the relatively unstable methylidene to be formed. Chiral
complex 3 is ineffective, and achiral Mo alkylidene 4 and Ru carbene 5
furnish the E isomer predominantly (79–89%). A weaker vacuum
(7.0 torr versus 1.0 torr CM with enol ethers) is sufficient, indicating
that such conditions can be applied to cases that involve relatively
volatile substrates.
An assortment of allylic amides and terminal alkenes, including
those that contain a halide (19b), a Lewis basic group (19c, 19d) or a
sterically demanding substituent (19e), can be used (Fig. 3). Stereo-
selective formation of 19f and 19g is noteworthy as the relatively less
hindered unsaturated amides are more prone to homocoupling and
the Z alkene products undergo equilibration to the E isomer readily, as
manifested by the lower Z:E ratios. Although in certain cases 10 equiv.
of a cross partner is used for maximum efficiency, lower amounts of
alkene substrates lead to reasonably efficient processes. For example,
with 3.0 mol% 2 and 3 equiv. of the aliphatic alkene (versus 5 mol%
and 10 equiv.), 19g is isolated in 62% yield and 90% Z selectivity (80%
conversion, 5 min, 22 uC). It should be noted that in all the above trans-
formations, use of catalysts bearing a racemic binol ligand furnishes
similar levels of reactivity and stereoselectivity (see Supplementary
Information).
Conclusions and discussions
In addition to catalytic olefin CM, Wittig reactions25, catalytic alkyne
hydrogenation and cross-coupling1,26 are notable approaches for syn-
thesis of Z-disubstituted alkenes. The above four types of transforma-
tions are distinct—each delivers the desired product through a
different bond disconnection. Similarly to CM, in cross-coupling
alkenes serve as starting materials; in contrast to CM, however, it is
through the synthesis of the substrate (for example, a Z vinyl halide)—
and not in the cross-coupling step—that the stereochemical identity
of the product is determined. Wittig-type processes are typically not
catalytic and involve reaction of aldehydes (versus the more stable
alkenes) and triphenylphosphonium ylides. Catalytic alkyne hydro-
genation requires substrates derived from functionalization of a ter-
minal alkyne; currently, relative to alkenes, methods for preparation
of alkynes are less common and related synthesis routes are often
lengthier. Moreover, partial hydrogenation of alkynes involves metal
catalysts that contain poisonous lead salts and must be controlled to
avoid over-reduction and generation of alkane by-products that can
be difficult to separate from the desired Z alkene. Catalytic CM thus
offers a desirable alternative to synthesis of Z alkenes, particularly as it
requires as starting material a functional group that is stable, easily
accessible and distinct from the other commonly used protocols men-
tioned above.
The strategies outlined here—including the use of reduced pressure
to enhance stereoselectivity in catalytic CM, and the Z-selective Mo-
catalysed transformations—offer a unique solution to a long-standing
problem in organic chemistry27. Our findings offer additional
evidence regarding the unique ability of stereogenic-at-Mo mono-
aryloxypyrrolides to effect olefin metathesis reactions that extend
beyond enantioselective processes28, with efficiency and selectivity
levels that are not achievable with other catalyst classes. The catalytic
processes described here are expected to affect significantly activities
Synthesis of immunostimulant KRN7000
Stereoselective synthesis of anti-tumour agent KRN700020,21 under-
lines the utility of the method (Fig. 3). Catalytic CM of carbohydrate-
containing allylic amide 20, prepared in four steps from commercially
available agents, affords 21 in 85% yield and with 96% Z-selectivity.
Diastereoselective dihydroxylation (89% yield, 92:8 diastereomeric
ratio (d.r.)) of the Z alkene delivers 22; it should be noted that similar
functionalization of the corresponding E alkene isomer would afford
an undesired diol diastereomer22. Dihydroxylamide 24 is secured in
two steps and the target is obtained after carbohydrate deprotection23.
Z-Selective CM thus provides access to a route that is significantly
more concise than the 14-step sequence (compared to nine steps in
Fig. 3) reported thus far as the shortest synthesis of KRN700024. It is
noteworthy that the convergent nature of a synthesis approach invol-
ving catalytic CM, such as the two examples provided here, can easily
translate to preparation of a variety of related analogues; for example,
in connection with preparation of 21 (Fig. 3), a wide range of other
terminal alkenes may be used.
that require the stereoselective synthesis of organic molecules29,30
.
METHODS SUMMARY
General procedure for catalytic Z-selective cross-metathesis. In an N2-filled dry
box, an oven-dried (135 uC) 20-ml vial equipped with a magnetic stir bar was
charged with vinyl ether 11 (1.00 g, 4.19 mmol) and 1.0 mol% of in situ-generated
complex 1a (419 ml, 0.100 M, 41.9 mmol; final substrate concentration 5 1.70 M).
A separate 2.0-ml vial was charged with 1-octadecene (12, 2.12 g, 8.39 mmol) and
decalin (2.10 ml). The resulting solution was transferred to the mixture of 11 and
1a by syringe; a septum, fitted with an outlet needle, was attached to the vial and
an adapter was attached to the top of the septum and vacuum (,1.0 torr) applied.
The resulting solution was stirred for 3 h. The vessel was removed from the dry
box and the reaction was quenched by the addition of wet Et2O (,1.0 ml). The
unpurified product is .98% Z (as determined by 400 MHz 1H NMR analysis).
The residue was dissolved in Et2O and passed through a 2.5-cm plug of neutral
alumina to remove inorganic salts, and the solution was concentrated. In a 25-ml
The balance between conversion and Z selectivity
The relationship between efficiency and stereoselectivity is critical and
merits a brief discussion. The conversion values, at times less than
complete, represent a balance struck between achieving the highest
yield and maximal Z selectivity with minimal substrate equivalents round-bottom flask equipped with a stir bar, the resulting residue was treated
2 4 M A R C H 2 0 1 1
| V O L 4 7 1 | N A T U R E | 4 6 5
Macmillan Publishers Limited. All rights reserved
©2011

RESEARCH ARTICLE
17. Lankalapalli, R. S. et al. Synthesis and antioxidant properties of an unnatural
plasmalogen analogue bearing a trans O-vinyl ether linkage. Org. Lett. 11,
2784–2787 (2009).
18. Lee, Y., Jang, H. & Hoveyda, A. H. Vicinal diboronates in high enantiomeric purity
through tandem site-selective NHC–Cu-catalyzed boron–copper additions to
terminal alkynes. J. Am. Chem. Soc. 131, 18234–18235 (2009).
19. Qin, D., Byun, H.-S. & Bittman, R. Synthesis of plasmalogen via 2,3-bis-O-(4’-
methoxybenzyl)-sn-glycerol. J. Am. Chem. Soc. 121, 662–668 (1999).
20. Borg, N. A. et al. CD1d-lipid-antigen recognition by the semi-invariant NKT T-cell
receptor. Nature 448, 44–49 (2007).
with (n-Bu)4NF (1.0 M in THF, 21.0 ml, 21.0 mmol), and stirred for 2 h. The
mixture was diluted with Et2O (200 ml), passed through a 5-cm plug of neutral
alumina, and concentrated. The resulting white solid was purified by chromato-
graphy on neutral alumina (100% hexanes) to afford 15 as a white solid (m.p. 30–
31 uC, 0.914 g, 2.98 mmol, 71.0% yield; .98% Z isomer).
Received 25 January; accepted 22 February 2011.
1. Negishi, E.-i. et al. Recent advances in efficient and selective synthesis of di-, tri, and
tetrasubstituted alkenes via Pd-catalyzed alkenylation–carbonyl olefination
strategy. Acc. Chem. Res. 41, 1474–1485 (2008).
2. Chatterjee, A. K. in Handbook of Metathesis Vol. 1 (ed. Grubbs, R. H.) 246–295
(Wiley-VCH, 2003).
3. Crowe, W. E. & Goldberg, D. R. Acrylonitrile cross-metathesis: coaxing olefin
metathesis reactivity from a reluctant substrate. J. Am. Chem. Soc. 117,
5162–5163 (1995).
4. Hansen, E. C. & Lee, D. Efficient and Z-selective cross-metathesis of conjugated
enynes. Org. Lett. 6, 2035–2038 (2004).
5. Malcolmson, S. J., Meek, S. J., Sattely, E. S., Schrock, R. R. & Hoveyda, A. H. Highly
efficient molybdenum-based catalysts for alkene metathesis. Nature 456,
933–937 (2008).
6. Lee, Y.-J., Schrock, R. R. & Hoveyda, A. H. Endo-selective enyne ring-closing
metathesis promoted by stereogenic-at-Mo monoalkoxide and monoaryloxide
complexes. Efficient synthesis of cyclic dienes not accessible through reactions
21. Trappeniers, M. et al. Synthesis and in vitro evaluation of a-GalCer epimers.
ChemMedChem 3, 1061–1070 (2008).
´
´
22. Llaveria, J., Dıaz, Y., Matheu, I. & Castillion, S. An efficient and general
enantioselective synthesis of sphingosine, phytosphingosine, and 4-substituted
derivatives. Org. Lett. 11, 205–208 (2009).
23. Kim, S., Song, S., Lee, T., Jung, S. & Kim, D. Practical synthesis of KRN7000 from
phytosphingosine. Synthesis 847–850 (2004).
24. Michieletti, M. et al. Synthesis of a-galactosyl ceramide (KRN7000) and analogues
thereof via a common precursor and their preliminary biological assessment. J.
Org. Chem. 73, 9192–9195 (2008).
25. Nicolaou, K. C., Ha¨rter, M. W., Gunzner, J. L. & Nadin, A. The Wittig and related
reactionsinnaturalproductsynthesis.JustusLiebigsAnn. Chem. 1283–1301(1997).
26. Miyaura, N. & Suzuki, A. Palladium-catalyzed cross-coupling reactions of
organoboron compounds. Chem. Rev. 95, 2457–2483 (1995).
27. Hoveyda, A. H. & Zhugralin, A. R. The remarkable metal-catalysed olefinmetathesis
reaction. Nature 450, 243–251 (2007).
28. Hoveyda, A. H., Malcomson, S. J., Meek, S. J. & Zhugralin, A. R. Catalytic
enantioselective olefin metathesis in natural product synthesis. Chiral metal-
based complexes that deliver high enantioselectivity and more. Angew. Chem. Int.
Edn 49, 34–44 (2010).
with Ru carbenes. J. Am. Chem. Soc. 131, 10652–10661 (2009).
0
´
7. Solans-Monfort, X., Clot, E., Coperet, C. & Eisenstein, O. d -Re-based olefin
metathesis catalysts, Re(;CR)(5CHR)(X)(Y): the key role of X and Y ligands for
efficient active sites. J. Am. Chem. Soc. 127, 14015–14025 (2005).
8. Meek, S. J., Malcolmson, S. J., Li, B., Schrock, R. R. & Hoveyda, A. H. The significance
of degenerate processes to enantioselective olefin metathesis reactions promoted
by stereogenic-at-Mo complexes. J. Am. Chem. Soc. 131, 16407–16409 (2009).
9. Alexander, J. B., La, D. S., Cefalo, D. R., Hoveyda, A. H. & Schrock, R. R. Catalytic
enantioselective ring-closing metathesis by a chiral biphen–Mo complex. J. Am.
Chem. Soc. 120, 4041–4042 (1998).
10. Schrock, R. R. & Hoveyda, A. H. Molybdenum and tungsten imido alkylidene
complexes as efficient olefin- metathesis catalysts. Angew. Chem. Int. Edn 42,
4592–4633 (2003).
11. Garber, S. B., Kingsbury, J. S., Gray, B. L. & Hoveyda, A. H. Efficient and recyclable
monomeric and dendritic Ru-based metathesis catalysts. J. Am. Chem. Soc. 122,
8168–8179 (2000).
12. 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).
13. Jiang, A. J., Zhao, Y., Schrock, R. R. & Hoveyda, A. H. Highly Z-selective metathesis
homocoupling of terminal olefins. J. Am. Chem. Soc. 131, 16630–16631 (2009).
14. Schrader, T. O. & Snapper, M. L. in Handbook of Metathesis (ed. Grubbs, R. H.)
205–237 (Wiley-VCH, 2003).
29. Nicolaou, K. C., Bulger, P. G. & Sarlah, D. Metathesis reactions in total synthesis.
Angew. Chem. Int. Edn 44, 4490–4527 (2005).
30. Prunet, J. & Grimaud, L. in Metathesis in Natural Product Synthesis (eds Cossy, J.,
Arsenyadis, S. & Meyer, C.) 287–312 (Wiley-VCH, 2010).
Supplementary Information is linked to the online version of the paper at
www.nature.com/nature.
Acknowledgements This research was supported by the US National Institutes of
Health, Institute of General Medical Sciences (GM-59426 to A.H.H. and R.R.S.) and the
National Science Foundation (CHE-0715138 to A.H.H.). R.V.O. and J.L. were LaMattina
and Spanish government Visiting Scholar Fellows, respectively. We thank S. Castillon,
S. J. Malcolmson and M. Yu for discussions. Mass spectrometry facilities at Boston
College are supported by the US National Science Foundation (DBI-0619576).
´
Author Contributions S.J.M., R.V.O. and J.L. were involved in the discovery, design and
development of the new Z-selective cross-metathesis strategies and applications to the
natural product syntheses. A.H.H. and R.R.S. conceived the research programme.
A.H.H. directed the investigations and composed the manuscript, with revisions
provided by S.J.M. and R.V.O.
15. La, D. S., Sattely, E. S., Ford, J. G., Schrock, R. R. & Hoveyda, A. H. Catalytic
asymmetric ring-opening metathesis/cross metathesis (AROM/CM) reactions.
Mechanism and application to enantioselective synthesis of functionalized
cyclopentanes. J. Am. Chem. Soc. 123, 7767–7778 (2001).
16. Nagan, N. & Zoeller, R. A. Plasmalogens: biosynthesis and functions. Prog. Lipid
Res. 40, 199–229 (2001).
Author Information Reprints and permissions information is available at
www.nature.com/reprints. The authors declare competing financial interests; details
accompany the full-text HTML version of the paper at www.nature.com/nature.
Readers are welcome to comment on the online version of this article at
www.nature.com/nature. Correspondence and requests for materials should be
addressed to A.H.H. (amir.hoveyda@bc.edu).
4 6 6
| N A T U R E | V O L 4 7 1 | 2 4 M A R C H 2 0 1 1
Macmillan Publishers Limited. All rights reserved
©2011