A Two-Component Alkyne Metathesis Catalyst System with an Improved Substrate Scope and Functional Group Tolerance: Development and Applications to Natural Product Synthesis

Article abstract of DOI:10.1002/chem.201601163

Although molybdenum alkylidyne complexes such as 1 endowed with triarylsilanolate ligands are excellent catalysts for alkyne metathesis, they can encounter limitations when (multiple) protic sites are present in a given substrate and/or when forcing conditions are necessary. In such cases, a catalyst formed in situ upon mixing of the trisamidomolybenum alkylidyne complex 3 and the readily available trisilanol derivatives 8 or 11 shows significantly better performance. This two-component system worked well for a series of model compounds comprising primary, secondary or phenolic -OH groups, as well as for a set of challenging (bis)propargylic substrates. Its remarkable efficiency is also evident from applications to the total syntheses of manshurolide, a highly strained sesquiterpene lactone with kinase inhibitory activity, and the structurally demanding immunosuppressive cyclodiyne ivorenolide A; in either case, the standard catalyst 1 largely failed to effect the critical macrocyclization, whereas the two-component system was fully operative. A study directed toward the quinolizidine alkaloid lythrancepine I features yet another instructive example, in that a triyne substrate was metathesized with the help of 3/11 such that two of the triple bonds participated in ring closure, while the third one passed uncompromised. As a spin-off of this project, a much improved ruthenium catalyst for the redox isomerization of propargyl alcohols to the corresponding enones was developed.

Full text of DOI:10.1002/chem.201601163

DOI: 10.1002/chem.201601163
Full Paper
&
Natural Products |Very Important Paper|
A Two-Component Alkyne Metathesis Catalyst System with an
Improved Substrate Scope and Functional Group Tolerance:
Development and Applications to Natural Product Synthesis
Sebastian Schaubach, Konrad Gebauer, Felix Ungeheuer, Laura Hoffmeister, Marina K. Ilg,
Conny Wirtz, and Alois Fürstner*[a]
Chem. Eur. J. 2016, 22, 8494 – 8507
8494
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Abstract: Although molybdenum alkylidyne complexes such
as 1 endowed with triarylsilanolate ligands are excellent cat-
alysts for alkyne metathesis, they can encounter limitations
when (multiple) protic sites are present in a given substrate
and/or when forcing conditions are necessary. In such cases,
a catalyst formed in situ upon mixing of the trisamidomoly-
benum alkylidyne complex 3 and the readily available trisila-
nol derivatives 8 or 11 shows significantly better per-
formance. This two-component system worked well for
a series of model compounds comprising primary, secondary
or phenolic -OH groups, as well as for a set of challenging
(bis)propargylic substrates. Its remarkable efficiency is also
evident from applications to the total syntheses of manshur-
olide, a highly strained sesquiterpene lactone with kinase in-
hibitory activity, and the structurally demanding immuno-
suppressive cyclodiyne ivorenolide A; in either case, the
standard catalyst 1 largely failed to effect the critical macro-
cyclization, whereas the two-component system was fully
operative. A study directed toward the quinolizidine alkaloid
lythrancepine I features yet another instructive example, in
that a triyne substrate was metathesized with the help of 3/
11 such that two of the triple bonds participated in ring clo-
sure, while the third one passed uncompromised. As a spin-
off of this project, a much improved ruthenium catalyst for
the redox isomerization of propargyl alcohols to the corre-
sponding enones was developed.
Introduction
roborated by the results shown in Table 1, which surveys sub-
stituents that are known to be problematic for many transition
metal catalysts. As can be seen, 1 is fully operational in the
presence of primary alkyl iodides or bromides, an azide, a thio-
ether, a sulfoxide, a sulfinamide, and even secondary or tertiary
amines.[13–15] Suffice it to say that any of these groups is chal-
lenging—if not detrimental—for the otherwise so highly ac-
complished ruthenium-based olefin metathesis catalysts;[16] the
latter, however, excel whenever it comes to protic substituents,
where they encounter hardly any serious limitation. Although
1 and congeners also allow certain protic groups to be han-
dled, the tolerance is not nearly as general. For example, sec-
ondary alcohol derivatives can often be transformed in accept-
able yields under somewhat forcing conditions (Table 1, en-
tries 12/13), whereas the attempted metathesis of a substrate
containing a primary alcohol basically met with failure (en-
tries 14/15).
Alkyne metathesis has greatly benefitted from the introduction
of molybdenum alkylidyne complexes endowed with silyloxy
groups as ancillary ligands such as 1 and 2,[1–5] which are dis-
tinguished by high activity as well as by a remarkable function-
al group tolerance.[6,7]
This outcome is perhaps not overly surprising for a catalyst
that owes its excellent performance to an intact siloxide ligand
sphere.[1,2] It seems likely that the Ph3SiO groups in 1 might be
prone to partial or even complete exchange on exposure to
a protic substrate; the more acidic and more readily accessible
the protic site, the higher the likelihood for damaging the cata-
lyst. If this analysis is true, the use of multivalent ancillary li-
gands might be one way to counteract this fatal exchange pro-
cess. Therefore we investigated whether or not chelating silox-
ides improve the functional group tolerance of molybdenum
alkylidyne complexes, in particular vis-à-vis protic groups.
In this context, recent contributions by Zhang and co-work-
ers deserve mentioning, who introduced podand ligands of
the general type 4 that afford chelate complexes of type 5 and
6 on reaction with 3[17] as the precatalyst; 5 and 6 were ap-
plied with considerable success to material science, although
their functional group tolerance has not been comprehensively
tested.[18–22] Yet, the amine tether in 5 is positioned such that it
necessarily coordinates the metal center, which in turn reduces
the catalytic activity; although N-methylation fixes the prob-
lem, the synthesis of the required charged ligand 4b (X=
NMe+) proved low-yielding and the solubility of the resulting
cationic complex is poor.[18] Therefore, a second ligand genera-
A rapidly growing number of applications to exceedingly
challenging target molecules bears witness of their excellent
profile; a few selected examples of key intermediates of previ-
ous total syntheses formed by ring closing alkyne metathesis
(RCAM) catalyzed by complexes 1, 2 or congeners[8,9] are
shown below.[8–11]
While occasional limitations were encountered with sterically
hindered or very electron-deficient substrates, surprisingly few
functionalities seem to be incompatible.[1–6] This notion is cor-
[a] S. Schaubach, Dr. K. Gebauer, F. Ungeheuer, Dr. L. Hoffmeister, M. K. Ilg,
C. Wirtz, Prof. A. Fürstner
Max-Planck-Institut für Kohlenforschung
45470 Mülheim/Ruhr (Germany)
E-mail: fuerstner@kofo.mpg.de
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201601163.
Chem. Eur. J. 2016, 22, 8494 – 8507
www.chemeurj.org
8495
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Results and Discussion
Table 1. Brief survey of the compatibility of 1 with challenging functional groups.
Ligand Synthesis and Evaluation
Possible chelate trisilanol ligands 8 were obtained in
only one operation by hydrosilylation of 7 with di-
phenylchlorosilane followed by in situ hydrolysis of
the SiÀCl unit (Scheme 1). As the substrates are
cheap and commercial, such compounds are avail-
able in quantity. The preparation of 11 a (R=H) com-
prising a tethering phenyl group is only one step
longer but equally scalable. With the idea in mind
that a hexasubstituted arene unit might preorganize
the ligand in a favorable alternating up/down confor-
mation,[23] compound 11 b (R=Et) was prepared anal-
ogously.
Entry Product
1
mol% t [h] T [8C] Yield
[%]
1
1.5
20
91 (X=Br)
2
3
4
5
6
7
8
1
4
3
2
15
1
16
2
16
3
20
20
20
20
80
20
20
96 (X=N3)
83 (X=piperidin-1-yl)
77 (X=NHBu)
85 (X=S(CH2)11Me)
75 (X=S(O)(CH2)11Me)
80 (X=NHBoc)
5
2
5[a]
22
67 (X=NS(O)p-Tol)
9
5
4
1.5
14
20
80
78 (X=I)
For the preparation of complex 1, Ph3SiOH is first
deprotonated and the resulting sodium (or potassi-
um) salt is then slowly added to a solution of the
readily accessible tribromoalkylidyne complex [ArCꢀ
MoBr3]·dme (Ar=p-MeOC6H4-).[1,2] The very poor solu-
bility of the triple salts derived from 8 or 11, howev-
er, precluded this procedure from being used. There-
fore the protic ligands themselves were added to
a solution of the trisamidomolybdenum alkylidyne
complex 3[17] in toluene; NMR inspection showed
that the ligand exchange is rapid but a mixture of
products is formed, independent of the addition
time and the structure of the ligand precursor.[24] In contrast
to the triphenolamine ligand 4a (X=N), which caps a single al-
kylidyne unit with formation of the chelate complex 5,[12] the
trisilanols of type 8 and 11 act primarily as “cross-linking”
agents.
10
52
11
4
14
80
87
12
13
10
7
14
24
20
80
34
66
14
15
10
10
14
14
20
0
7
110
[a] A solution of the catalyst in toluene was added over the course of 1 h.
tion comprising a benzylsilane tether was introduced (4c, X=
SiMe). Since the preparation requires several steps and the de-
rived catalysts of type 6 work best in CCl4 as the solvent
(which is particularly problematic in conceived applications to
macrocyclization reactions that usually require high dilution),[18]
we set out to find practical alternatives that should be easy to
make on scale, retain the many proven advantages of silanol li-
gands,[1–4] and allow alkyne metathesis to be applied to other-
wise problematic substrates.
In this context, the X-ray structure of compound 11 b is in-
formative (Figure 1), which shows the expected alternating up/
down pattern of the six alkyl substituents on the central ben-
zene ring. Interestingly, all three silanol groups are oriented to
the outside, away from each other. Although it is not clear if
this particular conformation is retained in solution, it might
provide an explanation why this compound failed to afford
Chem. Eur. J. 2016, 22, 8494 – 8507
www.chemeurj.org
8496
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 2. Comparison of the performance of 1 with that of the new two-
component system in homo-metathesis reactions.[a]
Entry
Catalyst Product
t [h] Yield [%]
1
2
3
1
14
14
14
84
80
80
3/8a
3/11 b
4
5
6
1
14
1
4
69
81
84
3/8a
3/11 b
7
8
1
3/8a
16
0.5
47
86
Scheme 1. a) Ph2Si(H)Cl, Karstedt’s catalyst (0.5 mol%), toluene, 608C;
b) H2O, Et3N, RT, 56% (8a), 58% (8b), 47% (11 a), 54% (11 b); c) allyltributyl-
stannane, Pd(PPh3)4 (16 mol%), CsF, DMF, 120–1358C, 76% (10a), 60%
(10b); Karstedt’s catalyst=platinum(0)-1,3-divinyl-1,1,3,3-tetramethyldisilox-
ane.
9
10
11
1
14
0.5
4
34
81
76
3/8a
3/11 b
12
13
14
1
14
14
4
0
3/8a
3/11 b
69
72
[a] For the sake of comparison, all reactions were performed at ambient
temperature with 10 mol% of catalyst in toluene in the presence of MS
5 Š.
Propargylic and Bis-Propargylic Alcohol Derivatives
Until recently, it had been basically impossible to subject prop-
argyl alcohol derivatives to alkyne metathesis mainly for two
reasons: all well-defined alkyne metathesis catalysts are
Schrock alkylidynes, and as such comprise an early transition
metal in its highest possible oxidation state.[25,26] Unless appro-
priately tempered by the ligand set, the Lewis acidity inherent
to such complexes endangers substituents at any activated po-
sition; propargylic alcohol derivatives fall into this category
(see the generic structure A) for the resonance stabilization of
the resulting carbocations. Even if this serious pitfall is over-
come and the chosen catalyst engages productively with the
triple bond, the ensuing alkylidyne of type B might decom-
pose by extrusion of the potential leaving group next to the
nucleophilic site.
Figure 1. Structure of compound 11 b in the solid state; for details, see the
Supporting Information.
a defined monomeric chelate complex on reaction with alkyli-
dyne 3.
The oligomeric character of the active species notwithstand-
ing, the resulting mixtures proved effective in a number of
challenging cases (Table 2). While a meta-propynyl phenol de-
rivative was metathesized by 1 and 3/8a or 3/11 b with similar
ease (entries 1–3), the new catalyst system worked better for
the para-analogue (entries 4–6). The difference is more clear-
cut with the Weinreb amide derivative shown in entries 7/8,
which reacted poorly with 1 but was transformed with remark-
able ease with the help of 3/8a; the very same is true for a sec-
ondary alcohol (entries 9–11). In the case of a substrate bearing
a primary alcohol, the standard catalyst 1 failed, whereas the
new two-component system furnished the desired product in
respectable yields (entries 12–14).
Somewhat unexpectedly, the different trisilanol ligands 8
and 11 led to catalysts of similar activity and selectivity. Dis-
tinct differences became evident only with more elaborate
substrates; the study directed toward the quinolizidine alkaloid
lythrancepine I is representative, where the combination 3/
11 b gave the best result (vide infra). In any case, the target ori-
ented synthesis projects outlined below provide additional evi-
dence for the superior performance of the new two-compo-
nent catalyst when it comes to challenging applications.
Only after the advent of complex 1 and congeners has this
important gap in substrate scope been closed, at least in
part.[3] A couple of exigent applications to natural product
total synthesis are encouraging (a representative example is
shown in the Introduction), but the number of successful cases
is still small and certain limitations persist.[27–31] Most notably,
only substrates comprised of only one propargylic site but one
regular alkyne have been successfully metathesized. In con-
Chem. Eur. J. 2016, 22, 8494 – 8507
www.chemeurj.org
8497
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
11 b afforded the product in an isolated yield of 64%.
On the one hand, this control experiment suggests that
the difficulty in converting bis-propargylic substrates
arises—to a large part—from the initiation step, which
has not been recognized as a possible limiting factor in
previous discussions of alkyne metathesis.[6,7] Precata-
lysts with an unbranched alkyl substituent on their alky-
lidyne unit seem to have a distinct advantage in this
regard. This aspect provides valuable guidance for fur-
ther catalyst optimization, although it needs to be seen
if and how it can be reconciled with stability considera-
tions. On the other hand, the lower net efficiency of 3/
Ph3SiOH (ratio 1:3) compared to 3/11 b indicates that
a chelating or cross-linking ligand set provides addition-
al advantages, be it for a higher resistance toward
protic groups and/or for higher thermal stability since
all tested bis-propargylic substrates required rather
forcing conditions.
Table 3. Ring closing metathesis of bis-propargylic and related substrates.[a]
Entry Substrate
Product
Yield
[%][b]
1
3/
11 b
1
2
R1 =R2 =MOM 54 70
R1 =MOM,
R2 =OH
0
67
3
4
R1 =R2 =TES
R1 =TES,
0
0
71
64
R2 =OH
5
6
R1 =R2 =OH
R1 =R2 =Ac
0
0
53
0
7
8
0
0
90
76
Total Synthesis of Manshurolide
The metathesis of propargylic alcohol derivatives has
the charm of providing access to many different struc-
tural motives when combined with appropriate post-
metathetic transformations of this functional group.[33]
Previous work from this laboratory showcased this
aspect by engaging such products, inter alia, into
a ruthenium-catalyzed redox isomerization/transannular
Michael addition process en route to the biphenyl alkaloid ly-
thranidine,[27] into a hydroxyl-directed trans-hydrostannation/
methyl-Stille coupling sequence for the total synthesis of the
antibiotic 5,6-dihydrocineromycin B,[29] or into an intricate
gold-catalyzed [3,3]-sigmatropic rearrangement/transannular
hydroalkoxylation cascade as the key step of a recent total syn-
thesis of the cytotoxic macrolide enigmazole A.[31] We conjec-
tured that butenolides are another important structural motif
commonly found in nature that should be within reach.
The sesquiterpene lactone manshurolide (12), isolated from
the stems of the climbing plant Aristolochia manushuriensis
Kom., was deemed an interesting and relevant target
(Scheme 2).[35] Compound 12 has been shown to inhibit mito-
gen-activated protein (MAP) kinases signaling pathways by se-
lective interference with the extracellular kinase ERK1/2.[36]
MAP kinases are implicated in a number of critical physiologi-
cal processes, including mitosis, gene expression, cell motility,
and programmed cell death, and therefore represent potential
biological targets for drug development. From the purely
chemical viewpoint, the three E-configured alkenes within the
12-membered ring of 12 impose substantial strain; this issue
will arguably be potentiated at the stage of the envisaged
propargyl alcohol precursor of type C, which is therefore
deemed a serious testing ground for the alkyne metathesis
catalysts at hand.
[a] All reactions were performed with a catalyst loading of 20 mol% in reflux-
ing toluene (1 mm) in the presence of MS 5 Š. [b] Isolated yields.
trast, compounds, in which both triple bonds bear propargylic
leaving groups, largely fail to react when exposed to 1 even
under forcing conditions (see Table 3); only a fully MOM-pro-
tected derivative was a positive outlier (entry 1). Despite the
modest yield, this result does indicate that there is nothing in-
herently wrong with the projected ring closure of bis-propar-
gylic substrates; rather, the problem must be due to an incom-
petence of the standard catalyst.
Therefore it seemed worthwhile to test the new two-compo-
nent system. Gratifyingly, the use of 3/11 b allowed all sub-
strates to be converted into the corresponding products,
except for the ester derivative shown in entry 6, which re-
mained unchanged for reasons that are not entirely clear. The
additional substrates shown in entries 7 and 8 corroborate the
trend, in that 1 failed completely whereas 3/11 b gave the cor-
responding strained cyclic enynes with flanking -OR groups in
good to excellent yield.
Although these results represent a step forward in prepara-
tive terms, they are mechanistically perplexing. The triphenylsi-
lanolate ligands in 1 and the new cross-linking silanolates 8
and 11 are arguably too similar in electronic terms to explain
the reactivity difference. Therefore we wondered if the sub-
stituents on the alkylidyne unit of the precatalysts (aryl in
1 versus ethyl in 3) also play a role. To test this hypothesis,
Ph3SiOH was added to a solution of 3 in toluene;[32] indeed,
the resulting mixture was able to convert the mono-TES pro-
tected substrate shown in entry 4; the conversion, however, re-
mained incomplete (ca. 60%) and a mixture of the desired
cyclic monomer and the cyclic dimer was formed, whereas 3/
Commercial 13 served as convenient starting point, which
was subjected to a zirconium-assisted carboalumination/iodi-
nation to give product 14 (Scheme 3).[37,38] One part of the
sample was then subjected to oxidation and the resulting alde-
Chem. Eur. J. 2016, 22, 8494 – 8507
www.chemeurj.org
8498
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Scheme 2. Retrosynthetic analysis of manshurolide (12); the configuration is
drawn as shown in the isolation paper.[34]
hyde reacted with propynylmagnesium bromide to give the re-
quired building block 15. The second half of the material was
transformed into the corresponding iodide 17;[37] as expected,
this compound could be selectively cross coupled at the alken-
yl halide site under standard palladium catalysis when treated
with the organozinc reagent derived from 18.[39] Unfortunately,
attempted coupling of 19 and unprotected 15 failed, whereas
the derived TBS-ether 16 was well behaved under the condi-
tions of the 9-MeO-9-BBN variant of the Suzuki coupling previ-
ously developed in our laboratory.[40]
Scheme 3. a) (i) Cp2ZrCl2(22 mol%), AlMe3, CH2Cl2, À258C!RT; (ii) I2, Et2O,
À258C!RT, 65%; b) DMP, CH2Cl2, 08C!RT; c) propynylmagnesium bromide,
THF, 08C!RT, 82% (over both steps); d) TBSCl, imidazole, DMF, 08C, 87%;
e) I2, PPh3, imidazole, Et2O/MeCN, 72%; f) (i) 18, tBuLi, Et2O, À788C, ZnBr2,
then 17; (ii) (dppf)PdCl2(2 mol%), THF, 80%; g) (i) 19, tBuLi, Et2O, À788C, then
9-MeO-9-BBN, À788C!RT; (ii) (dppf)PdCl2(10 mol%), K3PO4, DMF, 508C, 60%;
h) CSA cat., MeOH, 08C, 82%; i) 3 (10 mol%), 8b (10 mol%), MS 5 Š, toluene,
reflux, 82%; j) Pd2(dba)3·CHCl3 (4 mol%), dppb (8 mol%), CH2Cl2, CO (41 bar),
H2 (14 bar), 958C, ca. 17%; k) (i) Red-Al, Et2O, 08C!RT; (ii) I2, Et2O, À258C,
97%; l) Pd(tfa)2 (15 mol%), DPE-Phos (15 mol%), CO (1 bar), Et2N(iPr), MeCN/
MeOH (4:1), 95%; BBN=9-borabicyclononane; Cp=cyclopentadienyl;
CSA=camphorsulfonic acid; dba=dibenzylideneacetone; DMP=Dess–
Martin periodinane; DPE-Phos=bis[(2-diphenylphosphino)phenyl]ether;
dppb=1,4-bis(diphenylphosphino)butane; dppf=1,1’-bis(diphenylphosphi-
no)ferrocene; MS=molecular sieves; TBS=tert-butyldimethylsilyl; tfa=tri-
fluoroacetate.
Prior to the crucial macrocyclization, the bulky silyl group
had to be removed;[41] the resulting propargyl alcohol 20 was
then subjected to RCAM using either 1 or 3/8b as the catalyst,
the efficiency of which proved strikingly different: while
1 failed completely and 20 was largely recovered, the new
two-component system 3/8b afforded cycloalkyne 21 in excel-
lent yield, even though unusually high dilution (0.5 mm) was
mandatory to counteract competing oligomerization. This
technical issue notwithstanding, this example provides a nice
illustration of the superior performance of 3/8b in exigent
cases, which is tentatively ascribed to the stability of the cata-
lyst toward the protic substrate even in refluxing toluene.
We had envisaged to install the yet missing butenolide ring
in a single step by direct carbonylation of the propargyl alco-
hol subunit of 21 according to a literature procedure.[42] De-
spite many attempts, these efforts were to no avail; only the
corresponding b-lactone 22 could be isolated from the crude
material, albeit in low yield. In any case, the problem was fixed
The spectral data of the synthetic sample matched those of
the natural product in all respects. Surprisingly, however, the
rotatory power of the enantiomers (obtained by separation on
a chiral stationary phase) was found to be almost an order of
magnitude larger than the [a]D reported in the literature.[47] For
this massive discrepancy and the uncertainty of the original
stereochemical assignment,[34] we refrain from drawing a final
conclusion as to the absolute configuration of natural man-
shurolide.
by
a classical hydroxyl-directed hydroalumination/iodina-
tion,[43,44] followed by palladium-catalyzed carbonylation of the
resulting product 23. This step was best achieved by following
the conditions previously optimized by our group in a different
synthetic context.[45] Specifically, the use of Pd(tfa)2 in combina-
tion with DPE-Phos as the ligand in a mixed solvent system
proved highly effective, furnishing the desired butenolide
almost quantitatively. Overall, this route provided manshuro-
lide (12) in only nine steps with >17% overall yield for the
longest linear sequence. Moreover, the tactics manifest in this
approach should have a good chance to be applicable to
other targets featuring a butenolide ring-fused to a macrocyclic
skeleton too, which is a common structural motif in nature.[46]
Total Synthesis of Ivorenolide A
An arguably even more challenging target is ivorenolide A
(24), a potential lead in the quest for novel immunosuppres-
sive agents for medical use.[48,49] Compound 24 was isolated
together with its one-carbon shorter homologue ivorenolide B
(25)[50] from the stem bark of the mahogany Khayaivorensis A.
Chev. (Meliaceae); it was found to inhibit concanavalin A-in-
duced T-cell proliferation as well as lipopolysaccharide-induced
B-cell proliferation with selectivity indices comparable to or
better than those of the clinically approved drugs cyclospori-
ne A and periplocoside A.
Chem. Eur. J. 2016, 22, 8494 – 8507
www.chemeurj.org
8499
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
The streamlined preparation of the cyclization precursor is
shown in Scheme 4. It commences with an asymmetric addi-
tion of 1,3-pentadiyne (29)[53] to 3-iodoacrolein (28), which
worked best under the conditions described by Trost and co-
workers using (R,R)-ProPhenol (41) as ligand;[54] the modest ee
of 70% is not unexpected for a nucleophilic partner as slim as
29, but must be seen in the light of the fact that the available
alternative protocols gave inferior results[55] or failed complete-
ly.[56] Product 30 is unstable and had to be processed without
delay into the carboxylic acid building block by a sequence of
MOM-protection, Negishi cross-coupling with the organozinc
reagent 33,[39,57] and saponification of the methyl ester termi-
nus in 34.
The second fragment was also secured by a Trost alkynyla-
tion[54] of the lactic-acid derived aldehyde 38, which furnished
compound 39 in enantio- and diastereomerically pure form
after flash chromatography. Following a routine protecting
group maneuver, the derived alcohol 40 was esterified under
Yamaguchi conditions to give the required cyclization precur-
sor 35 in appreciable yield.[58] This protocol proved more con-
venient than the EDC-mediated esterification previously used
in our route to ivorenolide B (25).[30]
While the known routes to these structurally unique targets
used conventional macrolactonization,[48–50] we saw an oppor-
tunity to employ RCAM as the key strategic element. Such
a strategy, however, bears considerable risk as the conspicuous
1,3-diyne substructure might be prone to ring contraction
and/or competing oligomerization if the catalyst is kinetically
competent to attack the strained product once formed.[51]
These issues notwithstanding, we have recently been able to
communicate a concise entry into ivorenolide B (25) as well as
the non-natural analogue 26 for biological testing.[30,52]
In view of this success, we were stunned that our initial at-
tempts to extend the program to the parent compound ivore-
nolide A (24) faced major problems. This particular compound
features exocyclic propargylic -OH groups on either side of the
triple bonds and hence shows the exact substitution pattern
recognized in the model study as rendering the initiation step
difficult. Since this hurdle can be overcome in many cases with
the help of the two-component catalyst system 3/11 b in lieu
of 1, it seemed sensible to reinvestigate the total synthesis of
this unusual lead compound.
In concord with our model study on bis-propargylic sub-
strates (vide supra), 35 underwent a clean and productive cyc-
lization when exposed to catalytic amounts of 3/11 b in tolu-
ene in the presence of molecular sieves to sequester the re-
leased hexadiyne byproduct (Table 4). In contrast, this critical
step was problematic when the standard catalyst 1 was used,
which afforded only modest yields despite the higher loading.
Together with the manshurolide example outlined in the previ-
ous section, this particular transformation therefore illustrates
the power of the novel two-component catalyst system.
Scheme 4. a) Dibal-H, CH2Cl2, À788C, 86%; b) 29, Me2Zn, (R,R)-41 (10 mol%), Ph3P=O (20 mol%), toluene, 48C, 86% (70% ee); c) MOMCl, iPrNEt2, CH2Cl2, 86%;
d) Zn, THF, 408C; e) (Ph3P)2PdCl2 (5 mol%), tmeda, THF, 508C, 75%; f) LiOH, THF/MeOH (2:1), 95%; g) 2,4,6-trichlorobenzoic acid chloride, Et3N, toluene, 08C,
then 40, DMAP, RT, 79%; h) 3 (20 mol%), 11 b (20 mol%), toluene, MS 5 Š, 608C, 78%; i) aq. HCl, EtOH, 708C, 64%; j) mCPBA, CH2Cl2, 74%; k) 29, Me2Zn, (R,R)-
41 (10 mol%), Ph3P=O (20 mol%), toluene, 48C, 61%; l) MOMCl, iPr2NEt, CH2Cl2, 85%; m) DDQ, CH2Cl2, pH 7 buffer, 08C, 72%; DDQ=2,3-dichloro-5,6-dicyano-
p-benzoquinone; Dibal-H=diisobutylaluminum hydride; DMAP=4-dimethylaminopyridine; mCPBA=meta-chloroperbenzoic acid; MOM=methoxymethyl;
TBAF=tetra-n-butylammonium fluoride; tmeda=N,N,N’,N’-tetramethylethylendiamine.
Chem. Eur. J. 2016, 22, 8494 – 8507
www.chemeurj.org
8500
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Studies toward (+)-Lythrancepine I
Table 4. Comparison of the performance of different catalysts in the
RCAM reaction of substrate 35 with formation of cyclodiyne 36.[a]
The first asymmetric synthesis of the biphenyl alkaloid (À)-ly-
thranidine previously reported by our group had been based
on an unprecedented sequence comprised of a RCAM reaction
to form a macrocyclic propargyl alcohol skeleton followed by
a ruthenium-catalyzed redox isomerization and a transannular
aza-Michael addition to build the signature trans-substituted
piperidine ring (Scheme 5).[27] It seemed reasonable to assume
that a related—but clearly more involved—strategy might
bring the quinolizidine alkaloid (+)-lythrancepine I (42) into
reach, which is representative for a large family of natural
products isolated from plants of the Lythraceae genus
(Scheme 5).[59–62] Ideally, an intricate cascade comprised of two
consecutive transannular aza-Michael reactions should trans-
form a bis-enone of type I into the target after reduction of
the carbonyls; I in turn could be accessed by twofold redox
isomerization of a substrate such as J, which might be closed
by RCAM from an acyclic triyne K comprising two propargylic
entities and one regular alkyne.
Entry
Catalyst
Loading [mol%]
T [8C]
t [h]
Yield [%]
1
2
3
4
1
1
35
35
20
20
60
120
60
4
4
1
<1
38
33
78
75
3/11 b
3/11 b
120
[a] All reactions were performed in toluene (0.008m) in the presence of
MS 5 Š.
Compound 36 intercepts a previous total synthesis of 24;[48]
hence the remaining two steps were carried out in analogy to
this literature precedent. To this end, 36 was deprotected
under acidic conditions and the resulting diol 37 subjected to
a substrate controlled stereoselective epoxidation, which oc-
curred selectively at the double bond and left the propargylic
alkynes untouched. Synthetic ivorenolide A (24) was identical
with the natural product in all spectroscopic and analytical re-
gards.[48] With only nine steps and an overall yield of ꢁ13%
(longest linear sequence from commercial starting materials),
the new route described herein is not only competitive (com-
pare ref. [48]: 15 steps, 22%; ref. [49]: 17 steps, 13.4%) but also
chemically instructive with regard to the remaining challenges
of alkyne metathesis chemistry and how they can possibly be
solved.
This plan bears considerable risks, with the most obvious
possible pitfalls being as follows:
i) The stereochemical course of the aza-Michael tandem is
difficult to predict, even though a sufficiently large N-sub-
stituent Z is almost certainly necessary to reach the critical
trans-relationship between H5 and H9 in the first place
(Scheme 6):[63] in its absence, a transition state of type O
with a diequatorial orientation of the incipient 2,6-substitu-
Scheme 5. Key strategic elements of our previous total synthesis of lythranidine (ref. [27]) and their possible implementation into the synthesis of lythrance-
pine I (42); the helicity of the biaryl axis in 42 is known to be induced by the preexisting chiral centers, see ref. [62]; R, Z=protecting groups.
Chem. Eur. J. 2016, 22, 8494 – 8507
www.chemeurj.org
8501
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
ents is favored that converts the substrate N into a cis-pi-
peridine P. A transition state of type T leading to the trans-
piperidine U can only be envisaged if Z ¼ H;[64] in this
case, the actual cis/trans ratio will depend on the differen-
tial between the allylic strain in T and the transannular
strain of the 1,3-diaxial substituents in the diastereomeric
transition state R. Once a first trans-piperidine is in place,
the second cyclization meant to yield the desired quinolizi-
dine core of the target compound 42 might be under ther-
modynamic control.[65]
Pd(OAc)2. The required organozinc reagent 47 was derived
from iodide 46 bearing a TMS group as a masked site for
downstream fragment coupling. To this end, the biaryl 48 was
treated with NIS in MeCN,[72] the aldehyde group was unveiled
and reacted with propynyllithium to install the first propargyl
alcohol entity; this latter step was much more effective when
stoichiometric amounts of LaCl3·2LiCl were added.[73] Overall,
this sequence proved productive and scalable, providing more
than five grams of 50 for further elaboration.
Scheme 6. Rationale for the formation of 2,6-disubstituted piperidines by
aza-Michael addition; Z=generic N-protecting group of sufficient size to
induce 1,3-allylic strain; lythrancepine numbering.
ii) It is difficult to predict if the transannular context of the
envisaged Michael addition cascade imposes additional
constraints.[66]
Scheme 7. a) Allyl alcohol, Pd(OAc)2 (1 mol%), NaHCO3, TBAC, DMF, 508C,
81%; b) Ag2SO4, I2, MeOH, 87%; c) nBuLi, TMSCl, THF, À788C!RT, 99%;
d) nBuLi, tmeda, Et2O, then I2, THF, À788C!RT, 82%; e) Zn, LiCl, TMSCl, 1,2-
dibromoethane, THF, 808C; f) Pd(OAc)2 (2 mol%), SPhos (4 mol%), THF, 93%;
g) NIS, MeCN, 73%; h) PPTS, acetone/H2O, 508C, quant.; i) propynyllithium,
LaCl3·2LiCl, THF, 08C, 87%; NIS=N-iodosuccinimide; PPTS=pyridinium p-tol-
uenesulfonate; TBAC=tetra-n-butyl-ammonium chloride; tmeda=N,N,N’,N’-
tetramethylethylendiamine.
iii) We are unaware of any precedent in which two redox iso-
merizations were carried out concomitantly (see J!I);[67,68]
with several heteroelements and nucleophilic C-atoms
present within the congested framework of a substrate of
type J, many side tracks are conceivable.
iv) For the macrocyclic frame in J to be forged, two out of
three triple bonds in the cyclization precursor K have to
participate in metathesis while the third one has to pass
untouched; although this type of selectivity is unprece-
dented, we presumed that a bulky protecting group R on
As model studies had suggested that the efficiency of
a redox isomerization of an acyclic b-aminoalcohol derivative
depends on the stereostructure, with a 1,3-syn relationship
the propargyl alcohol at C3 might turn the adjacent C1À being beneficial,[74] the second required building block was
C2 triple bond into an innocent by-stander during RCAM.
formed in this particular stereochemical format (Scheme 8). A
Stahl oxidation[75] proved convenient for the preparation of al-
dehyde 52, which was converted into the homoallylic amine
53 on treatment with 58[76] in the presence of catalytic
amounts of the (S)-VANOL B3 catalyst (generated in situ) and
co-catalytic benzoic acid;[77] not only did this asymmetric aza-
Cope reaction scale well, but it afforded the desired product
53 with impeccable enantiomeric excess (98% ee as deter-
mined after N-Boc protection).
Based on this analysis and encouraged by the successful
total synthesis of the sister compound lythranidine,[27] the syn-
thesis of 42 was tackled as shown in Scheme 7. A Heck cou-
pling of 43 (X=I) with allyl alcohol[69] furnished multigram
amounts of the corresponding aldehyde, which was selectively
iodinated on treatment with Ag2SO4 and iodine in MeOH as
the preferred solvent;[70} under these conditions, concomitant
acetal formation was observed. The resulting compound 44
was subjected to a Negishi reaction,[37] which was rendered re-
markably effective by the SPhos ligand[71] in combination with
After routine N-protection of 53, the seemingly simple ozo-
nolysis of the double bond proved erratic in that variable but
substantial amounts of the secondary ozonide were present in
Chem. Eur. J. 2016, 22, 8494 – 8507
www.chemeurj.org
8502
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Gratifyingly, this expectation proved correct: thus, a Sonoga-
shira coupling of 50 with 57 afforded triyne 59 (Scheme 9)
which was cyclized within 15 min on treatment with the two-
component system 3/11 b in refluxing toluene to give the de-
sired product 60 in 66% yield (2.23 g scale). As evident from
Table 5, the standard catalyst 1 was largely incompetent, in
that it led to low conversion as well as to an unfavorable prod-
uct distribution. Moreover, the data reveal subtle differences
amongst the novel trisilanol ligands, which were hardly notice-
able in the model studies outlined above. The trend that the
phenyl-tethered ligands 11 give better results than their more
flexible relatives 8 is consistent with the excellent performance
of 11 b in the ivorenolide project.
Table 5. Catalyst screening for the formation of product 60.[a]
Entry Catalyst T [8C] HPLC [rel.%][b]
Conversion 60 S of Dimers[c]
Yield [%][d]
1
2
3
4
5
6
1
80
80
80
80
80
45
75
79
92
91
90
16 20
nd
41
52
58
60
66
3/8a
3/8b
3/11 a
3/11 b
66
70
84
83
84
8
9
8
8
6
3/11 b[e] 110
[a] For the sake of comparison, all reactions were performed with
20 mol% of catalyst in toluene (0.002m) in the presence of MS 5 Š.
[b] Data of the HPLC analysis of the crude reaction mixtures. [c] The sum
of the cyclic dimer and an acyclic dimer is given. [d] Isolated yield of 60.
[e] With 18 mol% catalyst loading.
Scheme 8. a) TEMPO (5 mol%), [Cu(MeCN)4]BF4 (5 mol%), bipy (5 mol%), air,
NMI (10 mol%), MeCN, 91%; b) 58, (S)-VANOL B3 (5 mol%), benzoic acid
(5 mol%), MS 5 Š, m-xylene, 608C, then aq. HCl, THF, 91%; c) Boc2O, Et3N,
CH2Cl2, 08C!RT, 89% (98% ee); d) O3, N-methylmorpholine-N-oxide, CH2Cl2,
08C, 76%; e) ethynylmagnesium bromide, Et2O, À788C!RT, 56 (39%) + 55
(33%); f) DMP, CH2Cl2, 08C!RT, 87%; g) LiAl(OtBu)3H, LiCl, Et2O, 08C, 92%;
h) TIPSCl, imidazole, DMAP, CH2Cl2, 08C!RT, 84%; bipy=2,2’-bipyridine;
Boc=tert-butyloxycarbonyl; DMAP=4-dimethylaminopyridine; NMI=N-
methyl-imidazole; TIPS=tri(isopropyl)silyl.
A particular challenge of this project deserves brief mention-
ing at this point: Once the macrocycle is closed all NMR spec-
tra are difficult to interpret by virtue of massive line broaden-
ing and/or multiple signal sets; in compounds 60 and 63, the
complexity is further increased by the presence of diastereo-
mers. Yet, the structural assignments are unambiguous; in all
cases could the data sets be deconvoluted and the signals be
attributed to individual interconverting conformers by exten-
sive 1D and 2D experimentation at variable temperatures (for
details, see the Supporting Information).
the mixture even after extended exposure of the crude materi-
al to excess PPh3. This problem was circumvented by running
the ozonolysis in the presence of N-methylmorpholine-N-oxide
which prevents secondary ozonides from being formed (pre-
sumably by trapping the preceding carbonyl oxide derived
from the primary ozonide);[78] close monitoring of the reaction
allowed any significant degradation of the triple bond to be
avoided. This method proved safe and scalable, affording alde-
hyde 54 in well reproducible yield. Likewise, the subsequent
formation of the propargyl alcohol moiety required some opti-
mization. As none of the tested asymmetric protocols allowed
the ethynyl residue (or a silyl protected equivalent) to be intro-
duced satisfactorily,[79] we resorted to the substrate controlled
addition of ethynylmagnesium bromide, separation of the dia-
stereomers, and subsequent conversion of the anti-isomer 55
into 56 by oxidation/syn-reduction. This sequence was well
suited for material throughput and further optimization of the
addition step therefore postponed to a later stage. Final TIPS-
protection gave more than three grams of the required build-
ing block 57; this bulky silyl ether was supposed to shield the
flanking triple bond during the projected regioselective alkyne
metathesis reaction.
With ample material at hand, the stage was set for the two
redox isomerizations meant to lead to a bis-enone of type I. By
virtue of the protecting group pattern in 60, it was possible to
perform this transformation either in two consecutive steps or,
after cleavage of the TIPS-ether, in one pot. The more conser-
vative stepwise approach taught us that the conversion of 60
into enone 61 was easier (47%, 10 mol% catalyst loading)
than the subsequent redox isomerization of 62 into 64 (<
40%, 40 mol% catalyst loading), but neither step was satisfac-
tory under the standard conditions using catalyst 66/
In(OTf)3.[67,68] An extensive optimization exercise was therefore
inevitable. In essence, it was found that replacement of 66 by
cationic mono-phosphine complexes of type 67 (conveniently
prepared in situ)[80] and substitution of camphor-10-sulfonic
acid (CSA) co-catalyst by NH4PF6 greatly improved the out-
come. Best results in the first step were obtained with 67a
bearing commercial PCy3 as the ligand, whereas 67b carrying
Chem. Eur. J. 2016, 22, 8494 – 8507
www.chemeurj.org
8503
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
the newly optimized Cy2P(3,5-dimethylphenyl) ligand worked
decisively better for the second isomerization;[81] this modifica-
tion allowed the yield to be improved to respectable 55%.
Complex 67b was also the catalyst of choice for the one-pot
double redox isomerization of diol 63 into 64 (43–47%, corre-
sponding to ~69% per step; 400–500 mg scale). To fully appre-
ciate this outcome one has to keep in mind that the literature
conditions (66/CSA cat.) failed completely when applied to this
particular substrate. In any case, complexes of type 67 seem to
be powerful catalysts that merit attention in the context of
redox isomerization chemistry in general; the good per-
formance of the new ancillary ligand in 67b and the inherent
modularity of the catalyst preparation promise a potentially
broad scope of this new method.[82]
At this stage, however, the projected total synthesis of
(+)-lythrancepine I (42) came to an end, since 64 resisted all
attempts to engage it in transannular aza-Michael reactions
under a large variety of conditions as long as the N-Boc group
was in place. This failure was unanticipated because a related
transformation had served our previous total synthesis of the
sister compound (À)-lythranidine very well (see Scheme 5).[27]
Steric factors are more likely the malefactor than the stiffness
of the di-unsaturated macrocyclic frame, since the deprotected
amine derived from 64 readily cyclized with formation of pi-
peridine 65 as the major product; as expected, however, the
cis- rather than trans-piperidine was favored in this case (cis/
trans 7.9:1). Changing the temperature, solvent and/or the
base did not allow this stereochemical outcome to be signifi-
cantly improved or even inverted.[83] This result corroborates
that a bulky N-substituent is essential to enforce formation of
an 2,6-trans-disubstituted piperidine, as contemplated during
the retrosynthetic planning (Scheme 6). Preliminary attempts
at forging the quinolizidine core via p-acid catalysis or under
free radical conditions were also unfruitful.[74]
The envisaged route to (+)-lythrancepine I (42) was there-
fore abandoned at this point and the project is back at the
strategic planning stage. Yet, the selective metathesis of two
out of three triple bonds present in compound 59 is without
precedent; this possibility illustrates the power of the method-
ology and might provide inspiration for other projects. More-
over, we are convinced of the utility of the much improved
conditions for the redox isomerization of propargylic alcohols.
Scheme 9. a) (Ph3P)2PdCl2 (5 mol%), CuI (10 mol%), Et3N, DMF, 69%; b) 3
(18 mol%), 11 b (18 mol%), MS 5 Š, toluene, reflux, 66%; c) 67a (10 mol%),
CSA (10 mol%), THF, 808C, 74%; d) TBAF, HOAc, THF, 08C!RT, quant.;
e) 67b (10 mol%), NH4PF6 (10 mol%), THF, 808C, 55%; f) TBAF, HOAc, THF,
08C!RT, 99%; g) 67b (20 mol%), NH4PF6 (20 mol%), THF, 808C, 47%; h) tri-
fluoroacetic acid, CH2Cl2, 08C; i) DBU, CH2Cl2, À908C, 88% (dr=7.9:1);
CSA=camphor-10-sulfonic acid.
Conclusion
Although the structurally well-defined and commercially avail-
able trisilanolate complex 1 will continue to serve alkyne meta-
thesis well,[1,2] it finds limitations when it comes to converting
substrates comprising protic functionality and/or when forcing
conditions are needed. In such cases, a catalyst mixture
formed in situ from the trisamido molybdenum alkylidyne
complex 3[17] and the easily made trisilanol ligands 8 or 11 can
help out. The fact that a multivalent ligand set together with
an alkyl (rather than aryl) substituent on the alkylidyne residue
of the precatalyst leads to a superior performance provides im-
portant guidance for future catalyst design.
The new two-component system excels in reactions of al-
kynes bearing primary, secondary or phenolic -OH groups, as
well as in reactions of (bis)propargylic alcohol derivatives. This
aspect is nicely illustrated by the total syntheses of the MAPK-
inhibitory sesquiterpene manshurolide (12), the immunosup-
pressive macrolide ivorenolide A (24), as well as by a study
toward the structurally unusual cyclophanic quinolizidine alka-
loid lythrancepine I (42). This latter study also features the first
example of an alkyne metathesis reaction of a substrate com-
Chem. Eur. J. 2016, 22, 8494 – 8507
www.chemeurj.org
8504
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Valot, D. Maihol, C. S. Regens, D. P. O’Malley, E. Godineau, H. Takikawa, P.
Philipps, A. Fürstner, Chem. Eur. J. 2015, 21, 2398–2408; e) J. Willwacher,
B. Heggen, C. Wirtz, W. Thiel, A. Fürstner, Chem. Eur. J. 2015, 21, 10416–
10430; f) L. Hoffmeister, P. Persich, A. Fürstner, Chem. Eur. J. 2014, 20,
4396–4402.
prising three different triple bonds, two of which were selec-
tively addressed while the third one remained untouched. As
a valuable spin-off of the lythrancepine project, a powerful
new ruthenium catalyst system for redox isomerization of
propargyl alcohols to the corresponding enones was estab-
lished that should be of interest well beyond the present con-
text.
[10] See also: a) M. G. Nilson, R. L. Funk, Org. Lett. 2010, 12, 4912–4915;
b) M. FouchØ, L. Rooney, A. G. M. Barrett, J. Org. Chem. 2012, 77, 3060–
3070.
[11] a) K. Micoine, A. Fürstner, J. Am. Chem. Soc. 2010, 132, 14064–14066;
b) K. Micoine, P. Persich, J. Llaveria, M.-H. Lam, A. Maderna, F. Loganzo,
A. Fürstner, Chem. Eur. J. 2013, 19, 7370–7383; c) W. Chaładaj, M.
Corbet, A. Fürstner, Angew. Chem. Int. Ed. 2012, 51, 6929–6933; Angew.
Chem. 2012, 124, 7035–7039; d) L. Hoffmeister, T. Fukuda, G. Pototsch-
nig, A. Fürstner, Chem. Eur. J. 2015, 21, 4529–4533; e) L. Brewitz, J. Lla-
veria, A. Yada, A. Fürstner, Chem. Eur. J. 2013, 19, 4532–4537.
[12] Actually, complex 5 is dimeric in the solid state by virtue of one phen-
oxide O-atom serving as a bridging ligand, see ref. [18a].
Experimental Section
All experimental details can be found in the Supporting Informa-
tion. The material includes compound characterization and copies
of spectra of new compounds.
[13] These results stand in line with our previous findings that 1, 2 and con-
geners tolerate various kinds of nitrogen- and/or sulfur containing aro-
matic heterocycles, including pyridines, thiophenes, thiazoles, oxazoles,
and carbazoles, see ref. [1–11]].
[14] For an impressive illustration of the compatibility with tertiary amines,
see the total synthesis of haliclonacyclamine C: B. J. Smith, G. A. Suli-
kowski, Angew. Chem. Int. Ed. 2010, 49, 1599–1602; Angew. Chem.
2010, 122, 1643–1646.
Acknowledgements
Generous financial support by the MPG and by the Fonds der
Chemischen Industrie (KekulØ stipend to S.S.) is gratefully ac-
knowledged. We thank Mrs. J. Lenartowicz for technical assis-
tance and Dr. R. Goddard for solving the X-ray structure.
[15] It needs to be emphasized, however, that attempted cross metathesis
reactions of azide- or tert-amine containing acetylene derivatives with
terminal alkynes failed for unknown reasons, cf: R. Lhermet, A. Fürstner,
Chem. Eur. J. 2014, 20, 13188–13193.
[16] a) T. M. Trnka, R. H. Grubbs, Acc. Chem. Res. 2001, 34, 18–29; b) A. Fürst-
ner, Angew. Chem. Int. Ed. 2000, 39, 3012–3043; Angew. Chem. 2000,
112, 3140–3172; c) K. C. Nicolaou, P. G. Bulger, D. Sarlah, Angew. Chem.
Int. Ed. 2005, 44, 4490–4527; Angew. Chem. 2005, 117, 4564–4601;
d) A. H. Hoveyda, A. R. Zhugralin, Nature 2007, 450, 243–251; e) A.
Fürstner, Chem. Commun. 2011, 47, 6505–6511.
Keywords: alkyne metathesis
products · propargyl alcohols · redox isomerization
·
molybdenum
·
natural
[1] J. Heppekausen, R. Stade, R. Goddard, A. Fürstner, J. Am. Chem. Soc.
2010, 132, 11045–11057.
[2] J. Heppekausen, R. Stade, A. Kondoh, G. Seidel, R. Goddard, A. Fürstner,
Chem. Eur. J. 2012, 18, 10281–10299.
[3] P. Persich, J. Llaveria, R. Lhermet, T. de Haro, R. Stade, A. Kondoh, A.
Fürstner, Chem. Eur. J. 2013, 19, 13047–13058.
[4] See also: A. D. Lackner, A. Fürstner, Angew. Chem. Int. Ed. 2015, 54,
12814–12818; Angew. Chem. 2015, 127, 13005–13009.
[17] W. Zhang, Y. Lu, J. S. Moore, Org. Synth. 2007, 84, 163–176.
[18] a) K. Jyothish, W. Zhang, Angew. Chem. Int. Ed. 2011, 50, 3435–3438;
Angew. Chem. 2011, 123, 3497–3500; b) H. Yang, Z. Liu, W. Zhang, Adv.
Synth. Catal. 2013, 355, 885–890.
[5] For the use of molybdenum amido complexes as functional group tol-
erant alkyne metathesis catalysts, see: a) A. Fürstner, C. Mathes, C. W.
Lehmann, J. Am. Chem. Soc. 1999, 121, 9453–9454; b) A. Fürstner, C.
Mathes, C. W. Lehmann, Chem. Eur. J. 2001, 7, 5299–5317; c) A. Fürstner,
K. Grela, C. Mathes, C. W. Lehmann, J. Am. Chem. Soc. 2000, 122, 11799–
11805; d) A. Fürstner, K. Radkowski, J. Grabowski, C. Wirtz, R. Mynott, J.
Org. Chem. 2000, 65, 8758–8762; e) W. Zhang, S. Kraft, J. S. Moore, J.
Am. Chem. Soc. 2004, 126, 329–335.
[6] a) A. Fürstner, Angew. Chem. Int. Ed. 2013, 52, 2794–2819; Angew.
Chem. 2013, 125, 2860–2887; b) A. Fürstner, Science 2013, 341,
1229713.
[7] a) W. Zhang, J. S. Moore, Adv. Synth. Catal. 2007, 349, 93–120; b) R. R.
Schrock, C. Czekelius, Adv. Synth. Catal. 2007, 349, 55–77; c) A. Mor-
treux, O. Coutelier, J. Mol. Catal. A 2006, 254, 96–104; d) A. Fürstner,
P. W. Davies, Chem. Commun. 2005, 2307–2320.
[8] a) K. J. Ralston, H. C. Ramstadius, R. C. Brewster, H. S. Niblock, A. N.
Hulme, Angew. Chem. Int. Ed. 2015, 54, 7086–7090; Angew. Chem. 2015,
127, 7192–7196; b) C. M. Neuhaus, M. Liniger, M. Stieger, K.-H. Altmann,
Angew. Chem. Int. Ed. 2013, 52, 5866–5870; Angew. Chem. 2013, 125,
5978–5983; c) J. Burnley, W. R. Jackson, A. J. Robinson, J. Org. Chem.
2015, 80, 9057–9063; d) L.-D. Guo, X.-Z. Huang, S.-P. Luo, W.-S. Cao, Y.-P.
Ruan, J.-L. Ye, P.-Q. Huang, Angew. Chem. Int. Ed. 2016, 55, 4064–4068;
Angew. Chem. 2016, 128, 4132–4136; e) A. Guy, C. Oger, J. Heppekau-
sen, C. Signorini, C. De Felice, A. Fürstner, T. Durand, J.-M. Galano, Chem.
Eur. J. 2014, 20, 6374–6380; f) G. Herstad, P. P. Molesworth, C. M. Miller,
T. Benneche, M. A. Tius, Tetrahedron 2016, 72, 2084–2093.
[9] a) V. Hickmann, A. Kondoh, B. Gabor, M. Alcarazo, A. Fürstner, J. Am.
Chem. Soc. 2011, 133, 13471–13480; b) S. Benson, M.-P. Collin, A. Arlt, B.
Gabor, R. Goddard, A. Fürstner, Angew. Chem. Int. Ed. 2011, 50, 8739–
8744; Angew. Chem. 2011, 123, 8898–8903; c) G. Valot, C. S. Regens,
D. P. O’Malley, E. Godineau, H. Takikawa, A. Fürstner, Angew. Chem. Int.
Ed. 2013, 52, 9534–9538; Angew. Chem. 2013, 125, 9713–9717; d) G.
[19] K. Jyothish, Q. Wang, W. Zhang, Adv. Synth. Catal. 2012, 354, 2073–
2078.
[20] For applications, see: a) D. W. Paley, D. F. Sedbrook, J. Decatur, F. R.
Fischer, M. L. Steigerwald, C. Nuckolls, Angew. Chem. Int. Ed. 2013, 52,
4591–4594; Angew. Chem. 2013, 125, 4689–4692; b) K. Hu, H. Yang, W.
Zhang, Y. Qin, Chem. Sci. 2013, 4, 3649–3653; c) Y. Zhu, H. Yang, Y. Jin,
W. Zhang, Chem. Mater. 2013, 25, 3718–3723; d) Q. Wang, C. Zhang,
B. C. Noll, H. Long, Y. Jin, W. Zhang, Angew. Chem. Int. Ed. 2014, 53,
10663–10667; Angew. Chem. 2014, 126, 10839–10843; e) Z. Chen, D. W.
Paley, L. Wei, A. L. Weisman, R. A. Friesner, C. Nuckolls, W. Min, J. Am.
Chem. Soc. 2014, 136, 8027–8033; f) Q. Wang, C. Yu, H. Long, Y. Du, Y.
Jin, W. Zhang, Angew. Chem. Int. Ed. 2015, 54, 7550–7554; Angew.
Chem. 2015, 127, 7660–7664.
[21] For the use of an ONO-pincer ligand, which leads to a less active but
more selective catalyst for living metathesis polymerizations of strained
alkynes, see: D. E. Bellone, J. Bours, E. H. Menke, F. R. Fischer, J. Am.
Chem. Soc. 2015, 137, 850–856.
[22] For metal alkylidyne complexes with pincer ligands, the catalytic activi-
ty and functional group tolerance of which has not been thoroughly in-
vestigated, see: M. E. O’Reilly, I. Ghiviriga, K. A. Abboud, A. S. Veige, J.
Am. Chem. Soc. 2012, 134, 11185–11195.
[23] Compare: M. Fukawa, T. Sato, Y. Kabe, Chem. Commun. 2015, 51,
14746–14749.
[24] This situation did not change when yet other trisilanols with shorter or
longer tethers between the silanol groups were used; S. Schaubach,
projected doctoral thesis, TU Dortmund.
[25] R. R. Schrock, Chem. Rev. 2002, 102, 145–179.
[26] R. R. Schrock, Angew. Chem. Int. Ed. 2006, 45, 3748–3759; Angew. Chem.
2006, 118, 3832–3844.
[27] K. Gebauer, A. Fürstner, Angew. Chem. Int. Ed. 2014, 53, 6393–6396;
Angew. Chem. 2014, 126, 6511–6514.
Chem. Eur. J. 2016, 22, 8494 – 8507
www.chemeurj.org
8505
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[28] a) J. Willwacher, N. Kausch-Busies, A. Fürstner, Angew. Chem. Int. Ed.
2012, 51, 12041–12046; Angew. Chem. 2012, 124, 12207–12212; b) D.
Mailhol, J. Willwacher, N. Kausch-Busies, E. E. Rubitski, Z. Sobol, M. Schu-
ler, M.-H. Lam, S. Musto, F. Loganzo, A. Maderna, A. Fürstner, J. Am.
Chem. Soc. 2014, 136, 15719–15729.
[29] S. M. Rummelt, J. Preindl, H. Sommer, A. Fürstner, Angew. Chem. Int. Ed.
2015, 54, 6241–6245; Angew. Chem. 2015, 127, 6339–6343.
[30] F. Ungeheuer, A. Fürstner, Chem. Eur. J. 2015, 21, 11387–11392.
[31] A. Ahlers, T. de Haro, B. Gabor, A. Fürstner, Angew. Chem. Int. Ed. 2016,
55, 1406–1411; Angew. Chem. 2016, 128, 1428–1433.
[32] Such mixtures were previously used in material science, see: a) S. W.
Sisco, J. S. Moore, J. Am. Chem. Soc. 2012, 134, 9114–9117; b) A. D.
Finke, D. E. Gross, A. Han, J. S. Moore, J. Am. Chem. Soc. 2011, 133,
14063–14070; c) S. W. Sisco, J. S. Moore, Chem. Sci. 2014, 5, 81–85;
d) S. W. Sisco, B. M. Larson, J. S. Moore, Macromolecules 2014, 47, 3829–
3836.
bromoethane in THF according to a method described by Knochel is
more convenient and productive, cf: A. Krasovskiy, V. Malakhov, A. Gav-
ryushin, P. Knochel, Angew. Chem. Int. Ed. 2006, 45, 6040–6044; Angew.
Chem. 2006, 118, 6186–6190.
[58] a) J. Inanaga, K. Hirata, H. Saeki, T. Katsuki, M. Yamaguchi, Bull. Chem.
Soc. Jpn. 1979, 52, 1989–1993; b) A. Parenty, X. Moreau, J.-M. Cam-
pagne, Chem. Rev. 2006, 106, 911 –939.
[59] E. Fujita, K. Bessho, Y. Saeki, M. Ochiai, K. Fuji, Lloydia 1971, 34, 306–
309.
[60] a) E. Fujita, Y. Saeki, J. Chem. Soc. D 1971, 368–369; b) E. Fujita, Y. Saeki,
J. Chem. Soc. Perkin Trans. 1 1972, 2141–2146; c) E. Fujita, Y. Saeki, J.
Chem. Soc. Perkin Trans. 1 1973, 297–301; d) E. Fujita, Y. Saeki, Chem.
Pharm. Bull. 1971, 19, 1515–1516.
[61] M. J. Barrow, P. D. Cradwick, G. A. Sim, J. Chem. Soc. Perkin Trans. 2 1974,
1812–1818.
[62] For the only total synthesis of a member of this series, see: a) D. J. Hart,
W. P. Hong, J. Org. Chem. 1985, 50, 3670–3672; b) D. J. Hart, W. P. Hong,
L. Y. Hsu, J. Org. Chem. 1987, 52, 4665–4673.
[33] E. B. Bauer, Synthesis 2012, 1131–1151.
[34] The published X-ray data do not allow for an unambiguous assignment
of the absolute configuration; we thank Prof. C. W. Lehmann, this Insti-
tute, for this assessment.
[35] G. Rücker, C. W. Ming, R. Mayer, G. Will, A. Güllmann, Phytochemistry
1990, 29, 983–985.
[36] Z.-B. Cheng, W.-W. Shao, Y.-N. Liu, Q. Liao, T.-T. Lin, X.-Y. Shen, S. Yin, J.
Nat. Prod. 2013, 76, 664–671.
[37] E. Negishi, S.-Y. Liou, C. Xu, S. Huo, Org. Lett. 2002, 4, 261–264.
[38] P. Wipf, S. Lim, Angew. Chem. Int. Ed. Engl. 1993, 32, 1068–1071; Angew.
Chem. 1993, 105, 1095–1097.
[63] For the formation of disubstituted piperidines by 5-exo cyclization reac-
tions, see: a) C. Zhong, Y. Wang, A. W. Hung, S. L. Schreiber, D. W.
Young, Org. Lett. 2011, 13, 5556–5559; b) I. J. McAlpine, R. W. Arm-
strong, Tetrahedron Lett. 2000, 41, 1849–1853; c) S. Chandrasekhar,
G. S. K. Babu, C. R. Reddy, Tetrahedron: Asymmetry 2009, 20, 2216–2219;
d) S. Fustero, S. Monteagudo, M. Sµnchez-Roselló, S. Flores, P. Barrio, C.
del Pozo, Chem. Eur. J. 2010, 16, 9835–9845; e) P. Radha Krischna, B. K.
Reddy, Tetrahedron: Asymmetry 2013, 24, 758–763.
[64] P. Beak, W. K. Lee, J. Org. Chem. 1993, 58, 1109–1117.
[65] A. O. Maldaner, R. A. Pilli, Synlett 2004, 1343–1346.
[66] Reviews: a) P. A. Clarke, A. T. Reeder, J. Winn, Synthesis 2009, 691–709;
b) E. Reyes, U. Uria, L. Carrillo, J. L. Vicario, Tetrahedron 2014, 70, 9461–
9484.
[67] a) B. M. Trost, R. C. Livingston, J. Am. Chem. Soc. 1995, 117, 9586–9587;
b) B. M. Trost, R. C. Livingston, J. Am. Chem. Soc. 2008, 130, 11970–
11978.
[68] a) B. M. Trost, N. Maulide, R. C. Livingston, J. Am. Chem. Soc. 2008, 130,
16502–16503; b) B. M. Trost, A. C. Gutierrez, R. C. Livingston, Org. Lett.
2009, 11, 2539–2542; c) B. M. Trost, A. Breder, B. M. O’Keefe, M. Rao,
A. W. Franz, J. Am. Chem. Soc. 2011, 133, 4766–4769; d) B. M. Trost, A.
Breder, Org. Lett. 2011, 13, 398–401; e) B. M. Trost, A. Breder, B. Kai, Org.
Lett. 2012, 14, 1708–1711.
[39] E. Negishi, Bull. Chem. Soc. Jpn. 2007, 80, 233–257.
[40] a) G. Seidel, A. Fürstner, Chem. Commun. 2012, 48, 2055–2070; b) A.
Fürstner, G. Seidel, Tetrahedron 1995, 51, 11165–11176.
[41] Macrocyclization can also be enforced with the TBS-group in place
using complex 1 in refluxing toluene; the yield was moderate (29–55%)
and significant amounts of dimeric product as well as anisole transfer
from the alkylidyne group rendered product isolation difficult.
[42] W.-Y. Yu, H. Alper, J. Org. Chem. 1997, 62, 5684–5687.
[43] S. E. Denmark, T. K. Jones, J. Org. Chem. 1982, 47, 4595–4597.
[44] a) K. Lehr, R. Mariz, L. Leseurre, B. Gabor, A. Fürstner, Angew. Chem. Int.
Ed. 2011, 50, 11373–11377; Angew. Chem. 2011, 123, 11575–11579;
b) K. Lehr, S. Schulthoff, Y. Ueda, R. Mariz, L. Leseurre, B. Gabor, A. Fürst-
ner, Chem. Eur. J. 2015, 21, 219–227.
[45] A. Kondoh, A. Arlt, B. Gabor, A. Fürstner, Chem. Eur. J. 2013, 19, 7731–
7738.
[46] For representative examples, see the following and literature cited
therein: J.-C. Han, C.-C. Li, Synlett 2015, 1289–1304.
[47] The isolation paper (ref. [35]) reports an [a]2D4 = +568 (EtOH, c=0.79),
whereas the two enantiomers obtained by separation of the racemic
synthetic sample showed the following rotatory power (in the order, in
which the enantiomers were eluting from the column, cf. the Support-
ing Information): [a]2D0 = +519.88 (c=0.5, EtOH); [a]2D0 =À508.18 (c=
0.53, EtOH).
[69] L. F. Tietze, K. Kahle, T. Raschke, Chem. Eur. J. 2002, 8, 401–407.
[70] In our hands, much better results were obtained on larger scale with
MeOH than EtOH as the solvent which had previously been used for
similar transformations, see: S. Wing-Wah, Tetrahedron Lett. 1993, 34,
6223–6224.
[71] D. S. Surry, S. L. Buchwald, Chem. Sci. 2011, 2, 27–50.
[72] A. Kuboki, C. Maeda, T. Arishige, K. Kuyama, M. Hamabata, S. Ohira, Tet-
rahedron Lett. 2008, 49, 4516–4518.
[73] a) A. Krasovskiy, F. Kopp, P. Knochel, Angew. Chem. Int. Ed. 2006, 45,
497–500; Angew. Chem. 2006, 118, 511 –515; b) A. Metzger, A. Gav-
ryushin, P. Knochel, Synlett 2009, 1433–1436.
[48] B. Zhang, Y. Wang, S.-P. Yang, Y. Zhou, W.-B. Wu, W. Tang, J.-P. Zuo, Y. Li,
J.-M. Yue, J. Am. Chem. Soc. 2012, 134, 20605–20608.
[49] D. K. Mohapatra, G. Umamaheshwar, R. N. Rao, T. S. Rao, S. Kumar, J. S.
Yadav, Org. Lett. 2015, 17, 979–981.
[50] Y. Wang, Q.-F. Liu, J.-J. Xue, Y. Zhou, H.-C. Yu, S.-P. Yang, B. Zhang, J.-P.
Zuo, Y. Li, J.-M. Yue, Org. Lett. 2014, 16, 2062–2065.
[74] K. Gebauer, doctoral thesis, TU Dortmund, 2016.
[75] a) J. M. Hoover, S. S. Stahl, J. Am. Chem. Soc. 2011, 133, 16901–16910;
b) B. L. Ryland, S. S. Stahl, Angew. Chem. Int. Ed. 2014, 53, 8824–8838;
Angew. Chem. 2014, 126, 8968–8983.
[76] H. Ren, W. D. Wulff, J. Am. Chem. Soc. 2011, 133, 5656–5659.
[77] G. Hu, A. K. Gupta, R. H. Huang, M. Mukherjee, W. D. Wulff, J. Am. Chem.
Soc. 2010, 132, 14669–14675.
[78] C. Schwartz, J. Raible, K. Mott, P. H. Dussault, Org. Lett. 2006, 8, 3199–
3201.
[51] S. Lysenko, J. Volbeda, P. G. Jones, M. Tamm, Angew. Chem. Int. Ed. 2012,
51, 6757–6761; Angew. Chem. 2012, 124, 6861–6865.
[52] In our publication (ref. [30]), ivorenolide A was erroneously depicted as
a 17-membered rather than 18-membered macrolide; we apologize for
this mistake in the drawing of the structure.
[79] Note that substrate and product are aminoalcohol derivatives, which
themselves might ligate acetylide reagents and hence intervene in an
attempted asymmetric addition process.
[80] These catalysts were conveniently prepared in situ from [CpRu-
(MeCN)3]PF6 and the corresponding phosphine according to: E. Rüba,
W. Simanko, K. Mauthner, K. M. Soldouzi, C. Slugovc, K. Mereiter, R.
Schmid, K. Kirchner, Organometallics 1999, 18, 3843–3850; control ex-
periments showed that isolated 67a or 67b lead to the same outcome.
[81] The Supporting Information contains the procedure for the preparation
of Cy2P(C6H3Me2).
[53] L. Brandsma, Synthesis of Acetylenes, Allenes and Cumulenes. Methods
and Techniques, Elsevier, Oxford, 2004.
[54] B. M. Trost, V. S. Chan, D. Yamamoto, J. Am. Chem. Soc. 2010, 132,
5186–5192.
[55] M. Turlington, Y. Du, S. G. Ostrum, V. Santosh, K. Wren, T. Lin, M. Sabat,
L. Pu, J. Am. Chem. Soc. 2011, 133, 11780–11794.
[56] S. Reber, T. F. Knçpfel, E. M. Carreira, Tetrahedron 2003, 59, 6813–6817.
[57] In our previous synthesis of ivorenolide B (ref. [30]), the same organo-
zinc reagent was prepared on treatment of iodide 32 with zinc dust in
DMF; we found that the use of zinc dust activated with TMSCl/1,2-di-
Chem. Eur. J. 2016, 22, 8494 – 8507
www.chemeurj.org
8506
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[82] A variety of aryl (aryl)dicyclohexylphosphines with different substitution
patterns on the aromatic ring were prepared according to: M. Murata,
S. L. Buchwald, Tetrahedron 2004, 60, 7397–7403. It was observed that
minor changes in ligand architecture had strong impact on the catalytic
activity of the resulting ruthenium complexes in the given transforma-
tion, see ref. [74].
[83] Moreover, the minor trans-isomer is configurationally unstable and rap-
idly converts to the cis-configured product 65 under basic or acidic
conditions.
Received: March 11, 2016
Published online on May 20, 2016
Chem. Eur. J. 2016, 22, 8494 – 8507
www.chemeurj.org
8507
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim