Optimized synthesis, structural investigations, ligand tuning and synthetic evaluation of silyloxy-based alkyne metathesis catalysts

FULL PAPER

DOI: 10.1002/chem.201200621

Optimized Synthesis, Structural Investigations, Ligand Tuning and Synthetic

Evaluation of Silyloxy-Based Alkyne Metathesis Catalysts**

Johannes Heppekausen, Robert Stade, Azusa Kondoh, Gꢀnter Seidel,

Richard Goddard, and Alois Fꢀrstner*^[a]

Abstract: Nitride- and alkylidyne com-

plexes of molybdenum endowed with

triarylsilanolate ligands are excellent

(pre)catalysts for alkyne-metathesis re-

actions of all sorts, since they combine

high activity with an outstanding toler-

ance toward polar and/or sensitive

functional groups. Structural and reac-

tivity data suggest that this promising

application profile results from a favor-

able match between the characteristics

of the high-valent molybdenum center

and the electronic and steric features

of the chosen Ar₃SiO groups. This in-

terplay ensures a well-balanced level of

Lewis acidity at the central atom,

which is critical for high activity. More-

over, the bulky silanolates, while disfa-

voring bimolecular decomposition of

the operative alkylidyne unit, do not

obstruct substrate binding. In addition,

Ar₃SiO groups have the advantage that

they are more stable within the coordi-

nation sphere of a high-valent molyb-

denum center than tert-alkoxides,

which commonly served as ancillary li-

gands in previous generations of

alkyne metathesis catalysts. From

a practical point of view it is important

to note that complexes of the general

phenanthroline, 2,2’-bipyridine or de-

rivatives thereof. Although the result-

ing adducts are themselves catalytically

inert, treatment with Lewis acidic addi-

tives such as ZnCl₂or MnCl₂removes

the stabilizing N-donor ligand and

gently releases the catalytically active

template into the solution. This proce-

dure gives excellent results in alkyne

metathesis starting from air-stable and

hence user-friendly precursor com-

plexes. The thermal and hydrolytic sta-

bility of representative molybdenum al-

kylidyne and -nitride complexes of this

series was investigated and the struc-

ture of several decomposition products

elucidated.

ꢀ

type [(Ar₃SiO)₃Mo X] (X = N, CR ; R

= aryl, alkyl, Ar = aryl) can be ren-

dered air-stable with the aid of 1,10-

Keywords: alkynes · metathesis ·

molybdenum · Schrock alkylidynes ·

silanol

Introduction

strated that such molybdenum silanolate complexes can be

rendered air-stable upon coordination to 1,10-phenanthro-

line.^[1]Although the resulting adducts themselves are cata-

lytically inactive, the operative metal fragment of the gener-

al type 1 can be released with the aid of MnCl₂or other

salts that bind the 1,10-phenanthroline ligand more tightly

than the Mo^VIcenter does. As a result, adducts such as 3

and 8 combine the convenience of air-stable precatalysts

with a remarkable application profile.^[1,2]The new com-

plexes have already enabled the total synthesis of structural-

ly challenging and bioactive natural products, including alka-

loids,^[5,6]oxylipins^[7]and macrolides,^[8–11]which clearly attests

to their utility.

In two previous publications we reported that molybdenum

alkylidyne and -nitride complexes endowed with triphenylsi-

lanolate ligands excel in alkyne metathesis, as they combine

outstanding activity with an unrivaled functional group tol-

erance (Scheme 1).^[1–4]Most transformations proceed with

remarkable rates at ambient temperature and require only

low catalyst loadings, provided that the reactions are per-

formed in the presence of molecular sieves, preferably MS

5 ꢀ.^[1]This additive prevents 2-butyne from accumulating in

the mixture, which is formed as a generic by-product in all

reactions of alkyne substrates capped by a methyl group;

this is exemplified in Scheme 2 for the conversion of

1-phenyl-1-propyne into tolane. Moreover, it was demon-

The excellent performance of molybdenum alkylidynes of

ꢀ

the general type [(Ph₃SiO)₃Mo CR] (1) (R = aryl, alkyl), in

particular in challenging ring-closing alkyne metathesis reac-

tions (RCAM),^[13]is thought to derive from a combination

of favorable properties of the triarylsilanolate groups as

compared to (fluorinated) alkoxides, phenolates or amides,

which were previously used as the ancillary ligands

(Scheme 3).^[14–25]As fairly poor net donors, silanolates leave

the Mo^VIcenter in 1 sufficiently Lewis acidic, which favors

substrate binding and ensures low barriers to metallacycle

formation.^[26–28]Their bulkiness prevents more than one

alkyne from coordinating to the metal and hence polymeri-

[a] M. Sc. J. Heppekausen, Dipl.-Chem. R. Stade, Dr. A. Kondoh,

Dipl.-Ing. G. Seidel, Dr. R. Goddard, Prof. A. Fꢁrstner

Max-Planck-Institut fꢁr Kohlenforschung

45470 Mꢁlheim/Ruhr (Germany)

Fax : (+49) 208-306-2994

E-mail: fuerstner@kofo.mpg.de

[**] Silyloxy-Based Alkyne Metathesis Catalysts, Part 3, for Part

see ref. [1].

2

Supporting information for this article is available on the WWW

under http://dx.doi.org/10.1002/chem.201200621.

Chem. Eur. J. 2012, 00, 0 – 0

ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

1

&

ÞÞ

These are not the final page numbers!

present the results of a first

round of investigations, in

which the alkylidyne unit, the

silanolate belt and the stabiliz-

ing N-donor ligand were sys-

tematically varied; related stud-

ies in which molybdenum as the

central atom was replaced by

tungsten will be reported in

a separate publication in due

course. As our primary objec-

tives, we intended to develop

a scalable synthesis route, re-

solve the remaining structural

issues, study pertinent struc-

ture–activity relationships, gain

insights into the deactivation

pathways, and map the perfor-

mance of these catalysts in

detail by applications to proto-

type alkyne metathesis reac-

tions.

Scheme 1. Previously reported molybdenum alkylidyne (red) and -nitride (blue) complexes endowed with tri-

phenylsilanolate ligands, which combine good to excellent catalytic activity with outstanding functional group

tolerance.^[1,2]The different precatalysts are thought to afford the same active species 1 in solution.

zation from occurring; likewise bimolecular decomposition

pathways of the catalyst are disfavored on steric grounds.

Results and Discussion

ꢁ

Since R₃Si O bonds, when bound to a Lewis acidic metal

Optimized syntheses: The preparation of the molybdenum

nitride phenanthroline adduct 3 outlined in our earlier pub-

lication starts from cheap substrates, takes only three opera-

tions (complex 20 primarily formed upon addition of

Ph₃SiOH to 19 needs not be isolated), has already been

used to prepare 20 gram lots of the fully air-stable adduct 3,

ꢁ

center, are arguably more robust than the R₃C O bond of

tertiary alkoxides, catalysts ligated with triarylsilanolates

may also have longer lifetimes.

From the coordination chemistry perspective, however,

the previously reported complexes shown in Scheme 1 invite

a more comprehensive molecular editing exercise. We now

Scheme 2. Generally accepted mechanism of alkyne metathesis illustrated

for the conversion of 1-phenyl-1-propyne into tolane;^[12,16a]for the remov-

al of the 2-butyne, formed as the generic by-product in all metathesis re-

actions employing methyl-capped alkyne substrates, by molecular sieves

(MS) 5 ꢀ, see ref. [1].

Scheme 3. Assortment of established alkyne metathesis catalysts bearing

tert-alkoxides, fluorinated alkoxides, amides, phenolates or neopentyl

groups as ancillary ligands; in case of complex 14, the alkylidyne unit is

generated in situ on treatment with CH₂Cl₂and exposure to the alkyne

substrate, see ref. [21].

&

2

&

www.chemeurj.org

ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 0000, 00, 0 – 0

ÝÝ

These are not the final page numbers!

Silyloxy-Based Alkyne Metathesis Catalysts

FULL PAPER

and is expected to scale to even larger batches without prob-

lems (Scheme 4);^[1,2,29]as will be discussed below in more

detail, bipyridine and derivatives thereof can also be used to

stabilize the active core. Such nitride complexes slowly con-

vert into the corresponding alkylidynes upon reaction with

an alkyne substrate (Scheme 1).^[30]Although this critical

transformation can take place—to a limited extent—even at

ambient temperature, it is beneficial to heat the reaction

mixture to ensure reasonable rates. Complexes with a pre-

formed alkylidyne unit clearly do not require such an initia-

tion step. Metal alkylidynes carrying silanolates as ancillary

ligands arguably represent the most effective and selective

alkyne metathesis catalysts known to date.^[1]Thus, the cur-

rent study was largely focused on this particular class of

compounds.

temperature control and the rapid removal of the Me₄NBr

at ꢁ788C prior to bromine oxidation allowed multigram

amounts of 25a to be prepared in analytically pure form.

The use of lithium reagents other than PhLi in the first

step allows the alkylidyne unit to be readily modified

(Scheme 5). In this context, the p-methoxybenzylidyne var-

iant 25b (Ar = MeOC₆H₄-) is of particular note, as this and

the derived silanolate complexes are more readily isolated

than the unsubstituted benzylidyne analogues. As a result,

the synthesis of the derived precatalysts 28b and 8b is high

yielding and scales particularly well, as described in detail in

1

the Experimental Section. Moreover, the H NMR signals

of the MeO substituent and of the para-disubstituted aryl

group are not obscured by other peaks and may hence serve

as mechanistic probes. Similar reasons let us target complex

8c with a -CF₃substituent on the benzylidyne unit, which

can be monitored ¹⁹F NMR spectroscopy. Complexes featur-

ing a sterically more demanding 2,6-dimethylphenyl moiety

(d series) were prepared to study whether the encumbered

alkylidyne affects the coordination behavior at the Mo^VI

center or not (see below).

Scheme 4. Optimized synthesis of bench-stable molybdenum-nitride com-

plexes. a) TMSCl, 1,2-dimethoxyethane (DME), reflux; b) LiHMDS,

hexane, 64% (over both steps); c) Ph₃SiOH (3 equiv), toluene, 808C;

d) 1,10-phenanthroline, toluene, 82%; e) toluene, 2,2’-bipyridine (79%,

21a) or 4,4’-di-tert-butyl-2,2’-bipyridine (68%, 21b).

The preferred access route to the alkylidyne series adopts

a literature procedure originally developed by Mayr and co-

workers (Scheme 5).^[31]Reaction of [Mo(CO)₆] with PhLi

followed by cation exchange provides the Fischer carbene

complex 23a (Ar = Ph) in good yield.^[32]Treatment of this

compound with oxalyl bromide leads to alkylidyne 24a,

which is subjected to oxidation with Br₂to give the tribromo

alkylidyne complex 25a, isolated in the form of its 1,2-dime-

thoxyethane (dme) adduct.^[31]Although this transformation

worked well on a gram scale, it was somewhat erratic with

larger batches. A careful optimization showed that these

problems are mainly caused by the limited stability of 24a

under the reaction conditions. Gratifyingly, however, strict

Scheme 5. Preparation of molybdenum alkylidyne complexes: a) ArLi,

Et₂O; b) Me₄NBr, H₂O, 64 % (23a), 84% (23b), 65% (23c), 63% (23d);

c) oxalyl bromide, CH₂Cl₂, ꢁ78 ! ꢁ308C; d) 1,2-dimethoxyethane, Br₂,

CH₂Cl₂, ꢁ788C ! RT, 50 % (25a), 80% (25b), 71% (25c), 85% (25d);

e) Ph₃SiOK (4 equiv), toluene; f) Ph₃SiOK (4 equiv), toluene, then Et₂O,

71% (27a); g) recrystallization from CH₂Cl₂/pentane; h) Ph₃SiOK

(5 equiv), toluene; recrystallization from CH₂Cl₂/pentane, 62% (28a),

68% (28b); i) 1,10-phenanthroline, toluene, 90% (8a), 84% (8b), 71%

(8c, over two steps), 73% (8d, over two steps).

Chem. Eur. J. 2012, 00, 0 – 0

ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

&

3

&

ÞÞ

These are not the final page numbers!

A. Fꢁrstner et al.

Silanolate alkylidyne complexes: Reactions of the tribro-

mide 25a (Ar = Ph) with three equivalents of Ph₃SiOK

gave product mixtures which are difficult to purify. Inspec-

at molybdenum permanently. This notion explains why

Lewis-basic functional groups do not interfere with produc-

tive alkyne metathesis catalyzed by complexes of type 1,

and why such an impressive assortment of polar substituents

survive undamaged and uncompromised (see Tables 2–4, see

below and ref. [1,7–10]). Moreover, the integrity of 29 sug-

gests that the operative alkylidyne unit as well as the ancil-

lary ligand sphere even resist mildly acidic conditions, since

ligation to the high-valent Mo center will enhance the acidi-

ty of Ph₃SiOH (pK_aꢂ16.6 in DMSO).^[34]

1

tion of the crude material by H NMR showed that the ate-

complex 26a comprising four silanolates is formed as

a major constituent, which will be described below in more

detail. Although the other species present in the crude ma-

terial could not be identified by NMR, X-ray quality crystals

were obtained from one such reaction mixture, which had

been kept for days at low temperature. The harvested crys-

tals consisted of the new alkylidyne complex 29, which is

mechanistically relevant (Scheme 6).

Somewhat surprisingly, complex 29 displays the heteroele-

ment-donor trans to the alkylidyne unit as the ligand exert-

ing the strongest trans-influence, in an arrangement that can

be described as distorted trigonal bipyramidal. Should the

primary site of alkyne binding in catalysts of the general

ꢀ

type [(Ar₃SiO)₃Mo CR] (1) also be trans to the alkylidyne,

the resulting p-complex 30 must rearrange to 30’ before

metallacycle formation and productive metathesis can ensue

(Scheme 7). In this context it is noteworthy that the previ-

ously described acetonitrile adduct 5 carries the neutral

donor MeCN cis to the alkylidyne in a distorted square pyr-

amidal array.^[1]Adducts 5 and 29 hence show that the Mo

center can entertain different coordination geometries, the

interconversion of which by a Berry-type pseudorotation is

supposedly facile. The question as to whether alkyne bind-

ing occurs trans or cis to the alkylidyne and, as a conse-

quence, whether a geometrical reorganization has to pre-

cede productive metathesis must be left open to future stud-

ies.

Scheme 6. Adducts that could be crystallized out of a reaction mixture in

which the siloxide-for-bromide exchange was performed with only three

equivalents of Ph₃SiOK in toluene at ambient temperature; the crude

material must contain additional species, which have not yet been identi-

fied.

Complex 29 is remarkable, because free Ph₃SiOH^[33]has

entered the coordination sphere of the metal alkylidyne

(Figure 1). The molybdenum atom of the neutral entity

ꢀ

[(Ph₃SiO)₃Mo CPh] (1a) is obviously Lewis-acidic enough

to pick up even a weak donor such as Ph₃SiOH from the so-

lution. Yet, the large Mo···O4 distance (average, 2.68(1) ꢀ)

suggests that the interaction with the heteroelement is

feeble and therefore unlikely to block the coordination site

ꢀ

Scheme 7. Neutral complexes of the type [(Ph₃SiO)₃Mo CR] (1) are ca-

pable of binding donors either trans- or cis to the alkylidyne unit, as evi-

denced by 29 and 5. This leaves the question open as to the primary bind-

ing site of an alkyne substrate (30 versus 30’).

Because the silanolate-for-bromide exchange leads to

product mixtures when only three equivalents of Ph₃SiOK

are added to tribromide 25a (exceptions are described

below), excess reagent was employed for preparative pur-

poses (Scheme 5). With ꢃ4 equivalents, the corresponding

ate complex could be obtained in form of the dme adduct

26a (Ar = Ph, Figure 2).^[35,36]This species catalyzes the

metathesis of 1-phenyl-1-propyne in the presence of MS 5 ꢀ

Figure 1. Structure of complex 29 in the solid state. Selected bond lengths

ꢁ

[ꢀ] and angles [8]: Mo1 C1 1.753(2), Mo1 O4 2.669(2), Mo1 O1

ꢁ

1.903(2), Mo1 O2 1.894(2), Mo1 O3 1.890(2), Mo1-C1-C2 177.2(2),

Mo1-O1-Si1 134.7(1), Mo1-O2-Si2 164.0(1), Mo1-O3-Si3 154.1(1).

&

4

&

www.chemeurj.org

ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 0000, 00, 0 – 0

ÝÝ

These are not the final page numbers!

Silyloxy-Based Alkyne Metathesis Catalysts

FULL PAPER

Figure 2. Structure of the ate-complex 26a in the solid state.^[35]Selected

Figure 3. Structure of complex 28a in the solid state.^[38]Selected bond

ꢁ

bond lengths [ꢀ] and angles [8]: Mo1 C73 1.743(2), Mo1 O1 2.022(2),

lengths [ꢀ] and angles [8]: Mo1 C1 1.729(3), Mo1 C1 C2 174.1(3), O3

ꢁ

Mo1 O2 2.009(2), Mo1 O3 1.933(2), Mo1 O4 1.962(2), O1 K1 2.624(2),

Mo1-O1-Si1 131.0(1), Mo1-O2-Si2 138.8(1), Mo1-O3-Si3 167.4(1), Mo1-

O4-Si4 148.4(1), Mo1-C73-C74 171.5(2).

Mo1 2.041(2), O1 Mo1 1.971(2), O2 Mo1 1.984(2), O4 Mo1 1.968(2),

Mo1-O1-Si1 136.41(14), Mo1-O2-Si2 168.68(15), Mo1-O3-Si3 148.86(14),

Mo1-O4-Si4 132.33(13), Mo1-C1-C2 174.1(3), O3····K1 2.691(2) ꢀ.

at ambient temperature with remarkable ease.^[1]This finding

suggests that such ate complexes readily loose one silanolate

unit in situ with formation of the neutral alkylidyne

terion. To this end, the tribromoalkylidyne complex 25a was

treated with five equivalents of Ph₃SiOK in toluene and the

product recrystallized from CH₂Cl₂/pentane. Under these

conditions, the resulting bulk material was shown to be the

ether-free ate-complex 28a in analytically pure form; its

constitution was confirmed by X-ray data (Figure 3). This

modified procedure, as described in detail in the Experimen-

tal Section for the larger scale preparation of the p-methox-

ybenzylidyne variant 28b (Ar = p-MeOC₆H₄-), is recom-

mended.

ꢀ

[(Ph₃SiO)₃Mo CPh] (1a), which is thought to represent the

actual catalyst in solution.

The dme ligated to the potassium counterion is removed

on treatment of 26a with Et₂O, giving rise to an adduct

ꢀ

which we tentatively formulated as [(Ph₃SiO)₃Mo

CPh]·Et₂O (7) (Scheme 5).^[1]This assignment was made in

analogy to the acetonitrile complex 5, which had been fully

characterized by X-ray diffraction.^[1]X-ray quality crystals

of this ether adduct itself could not be grown, despite con-

siderable experimentation. During the current up-scaling

campaign, however, we noticed that the composition of this

material is more consistent with the ether containing ate-

ꢀ

As can be seen from Figure 3,^[38]the potassium cation of

28a is strongly associated with only one of the four silano-

late O atoms (O3···K1 2.691(2) ꢀ), whereas the contacts to

the other oxygen atoms are much weaker (e.g., O2···K1

3.126(3) ꢀ). It is worth noting here that the potassium ion

also makes short contacts in the range of 3.0–3.3 ꢀ

(K1···C_Ph) to three of the phenyl groups, which surround it.

Since one of the lone pairs of O3 is engaged with the escort-

complex [(Ph₃SiO)₄Mo CPh][KACTHNUGTRNEUNG(Et₂O)_n] (n = 1, 2) (27a), al-

though neither the error bars of the integrals in the ¹H

NMR spectrum recorded at 600 MHz^[37]nor the combustion

data did allow for a definite assignment.

ꢁ

ing potassium ion, the O3 Mo1 bond is distinctly longer

ꢁ

To firmly resolve this issue, samples of this ether adduct

were recrystallized from CH₂Cl₂/pentane, which afforded

than the other O Mo bonds of this complex. The Si-O-Mo

bond angles V greatly differ from each other, with one

being fairly linear, while the three others are noticeably

bent (168.68(15), 148.86(14), 136.41(14), 132.33(13)8) (see

below). The ¹H NMR spectrum of 28a in CD₂Cl₂shows

marked temperature dependence, which is subject to further

investigations. In any case, the solvent-free ate-complex 28a

serves as an excellent metathesis catalyst in its own right

and as a convenient starting point for the preparation of var-

ious stabilized 1,10-phenanthroline- or 2,2’-bipyridine ad-

ducts (see below). Its substituted cousin 28b was prepared

and used analogously (see the Experimental Section).

ꢀ

crops of an ether-free ate-complex [(Ph₃SiO)₄Mo CPh]K

(28a) complex. This species was unambiguously character-

ized by X-ray diffraction (Figure 3). Since 28a derives from

the ether containing precursor complex, we need to reassign

the composition of the material previously thought to be

ꢀ

[(Ph₃SiO)₃Mo CPh]·Et₂O

(7)^[1]

as

consisting

of

ꢀ

[(Ph₃SiO)₄Mo CPh][K

N

ly, however, this reassignment has no bearing on the catalyt-

ic ability of this very active and selective alkyne metathesis

catalyst.^[1]

With this additional information in hand, we slightly

modified the preparative procedure such that no ambiguities

remain as to the ether ligands escorting the potassium coun-

Next, attempts were made to obtain the neutral core

ꢀ

[(Ph₃SiO)₃Mo CR] (1) in base-free form, which supposedly

constitutes the operating unit in alkyne metathesis reactions.

Chem. Eur. J. 2012, 00, 0 – 0

ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

&

5

&

ÞÞ

These are not the final page numbers!

A. Fꢁrstner et al.

Our efforts were rewarded in two cases: The first successful

foray started with the sterically more encumbered precursor

25d carrying a 2,6-dimethylphenyl group on the alkylidyne

(Scheme 8). Siloxide-for-bromine exchange followed by re-

Scheme 8. Donor-free neutral molybdenum alkylidynes endowed with tri-

phenylsilanolate groups. a) Ph₃SiOK (3 equiv, slow addition), toluene,

62% (31), 66% (32), see text.

Figure 4. Structure of the donor-free alkylidyne complex 31 in the solid

ꢁ

state. Selected bond lengths [ꢀ] and angles [8]: C1 Mo1 1.748(2), C2

ꢁ

C1 Mo1 171.37(13), Si1-O1-Mo1 168.8(1), Si2-O2-Mo1 146.1(1), Si3-O3-

ꢁ

Mo1 154.0(1), Mo1····Cl2 3.417, C1 Mo1····Cl2 173.9(1).

crystallization gave product 31, free of any neutral donor. It

is believed that the two methyl substituents on the benzyli-

dyne pointing toward the siloxide region disfavor the uptake

of an additional silanolate ligand. As can be seen from

Figure 4, this complex features the expected tetrahedral co-

ordination geometry. Its alkylidyne group deviates from lin-

earity (C2-C1-Mo1 171.37(13)8), but the alkylidyne bond

only differ in conformational detail, it suffices to address

here the metric parameters of only one of them. Interesting-

ꢀ

ly, the alkylidyne bond (Mo1 C1 1.745(1) ꢀ) is slightly

ꢀ

longer than that of the ether-free ate-complex 28a (Mo1 C1

ꢀ

length C1 Mo1 falls into the normal range (1.748(2) ꢀ).

1.729(3) ꢀ). All Mo-O-Si bond angles V deviate from line-

arity (141.3(1), 147.7(1), 159.5(1)8 in molecule 1), which may

impact on the Lewis acidity of the Mo^VIcenter and, in turn,

on the outstanding catalytic activity of this particular com-

plex (see below).

The Si-O-Mo bond angles are again rather variable

(168.8(1), 154.0(1), 146.1(1)8).

A second base-free, neutral molybdenum alkylidyne en-

dowed with silanolate ligands was reached when a solution

of

three

equivalents

of

Ph₃SiOK in toluene was slowly

added to a dilute solution of

25b bearing a p-methoxyben-

zylidyne entity; complex 32

can thus be obtained in re-

spectable yield (Scheme 8). It

should be pointed out, howev-

er, that the p-methoxybenzyli-

dyne group is essential for the

success of the reaction, be-

cause similarly slow addition

of Ph₃SiOK to the parent com-

plex 25a with an unsubstituted

benzylidyne afforded a product

mixture, of which the ate-com-

plex 28a mentioned above is

a significant constituent (see

above).

Figure 5 shows the structure

of complex 32 in the solid

state, which unequivocally con-

firms the proposed donor-free

constitution. Since the two in-

dependent molecules in the

unit cell are very similar and

Figure 5. Structure of the two independent molecules of the donor-free alkylidyne complex 32 in the solid

ꢁ

state. Selected bond lengths [ꢀ] and angles [8]: Mo1 C1 1.745(1), Mo2 C71 1.747(1), C2-C1-Mo1 175.8(1),

C72-C71-Mo2 175.3(1), Si1-O1-Mo1 141.3(1), Si2-O2-Mo1 147.7(1), Si3-O3-Mo1 159.5(1), Si5-O5-Mo2

149.9(1), Si6-O6-Mo2 162.9(1), Si7-O7-Mo2 142.5(1).

&

6

&

www.chemeurj.org

ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 0000, 00, 0 – 0

ÝÝ

These are not the final page numbers!

Silyloxy-Based Alkyne Metathesis Catalysts

FULL PAPER

Although the ate complexes described above perform

very well in alkyne metathesis and were routinely used in

this and in our previous studies,^[1,8–10]it is reasonable to

expect their donor-free cousins to show even higher reaction

rates. The direct comparison of complexes 28b and 32 nicely

confirms this notion (Figure 6). Methyl 2-(prop-1-ynyl)ben-

zoate was chosen for this comparative study, as this elec-

disfavor collisions of two alkylidynes and therefore retards

bimolecular decomposition pathways (see below).

In electronic terms, siloxides are weaker net donors than

alkoxides because p_p!d donation has to compete with

ꢁ

backbonding from oxygen into the low lying Si C s* orbi-

tals.^[39]Furthermore, their donor capacity is angle-depen-

dent: electron donation into the vacant metal d orbitals is

maximized when the Mo-O-Si bond angle V is close to

1808. This geometry enables both orthogonal electron pairs

of oxygen to engage in O!Mo bonding (Scheme 9) and

ꢁ

should hence go hand in hand with short O Mo bonds. In

contrast, bending of the Si-O-Mo hinge gives only one O-

lone pair the chance to overlap with the metal d orbital and

ꢁ

should therefore show up in longer Mo O bond lengths and

higher Lewis acidity at the central atom.

Figure 6. Metathetic conversion of methyl 2-(prop-1-ynyl)benzoate cata-

lyzed by 2 mol% of either ate-complex 28b (blue) or the neutral and

donor-free complex 32 (red) in toluene at 108C in the absence of molecu-

lar sieves to allow for equilibration rather than full conversion. When

molecular sieves are added to the mixture to absorb the released 2-

butyne, the conversion is quantitative in less than 5 min (see also Table 2,

entry 4).

Scheme 9. Angle dependence of the O!Mo donation of a silanolate vi-

sualized by the OACTHNUTRGNEUNG

(sp³) and O(sp) hybridization extremes.

Our experimental data allow these notions to be checked.

First, the variability of the bond angles V is striking, which

range from almost linear to significantly bent; the ate-com-

plex 26a is representative (167.4(1), 148.4(1), 138.8(1),

131.0(1)8).^[35]Secondly, the expected correlation between

tron-deficient substrate is significantly less reactive than 1-

phenyl-1-propyne previously used; moreover, the reaction

was performed at 108C to allow for proper monitoring by

GC/MS. Nevertheless, 12% conversion were reached within

the first two minutes (ꢃ33% when the reaction was per-

formed at ambient temperature), and the reaction already

starts leveling off after 20 min. In terms of reactivity, the

donor-free complex 32 is by far the most active alkyne meta-

ꢁ

the observed bond angles V and the associated Mo O dis-

tances is confirmed: the more stretched a given hinge, the

ꢁ

shorter the Mo O bond length (1.9325(2), 1.962(2),

2.009(2), 2.022(2) ꢀ in 26a). In line with this finding, it is

the most pyramidalized oxygen atom in 26a which makes

the closest contact to the potassium counterion but the lon-

gest distance to Mo; because one of its lone pairs is hardly

engaged with the transition-metal center, it can serve to

keep the escorting cation in place.

thesis catalyst ever tested in our laboratory. It outperforms

[16]

ꢀ

the parent Schrock alkylidyne [(tBuO)₃W CCMe₃] by far,

which serves as the benchmark in the field, but gave no de-

tectable conversion of this substrate under the chosen exper-

imental conditions even after prolonged reaction times.

In the neutral adduct 29, the three Mo-O-Si bond angles

are also fairly uneven (V = 164.0(1), 154.1(1), 134.7(1)8)

and hence the O–M orbital overlap hardly maximized. A

similar argument holds true for the donor-free alkylidyne

complexes 31 and 32, in which all Mo-O-Si units clearly de-

viate from linearity. It is therefore reasonable to assume

that the molybdenum centers of these complexes remain

Lewis-acidic, which is essential for productive alkyne meta-

thesis.^[26]In electronic terms, triarylsilanolates hence seem to

be an excellent choice as ancillary ligands for this type of

transformation. Moreover, one can speculate that their in-

fluence may not be static: it is tempting to assume that the

facile bending of the Mo-O-Si hinges contributes to the effi-

ciency of the overall catalytic process in a subtle way, as it

may help the active species to adjust to the different steric

and electronic requirements along the productive catalytic

Structural discussion: A few comments on the structures of

these and related complexes are deemed appropriate. Al-

though one might perceive Ph₃SiO groups as very bulky, the

X-ray structures of 26a and 28a show that the Mo^VIcenter

readily accommodates up to four such units; the ease of for-

mation of such ate complexes also suggests that there is no

serious steric impediment to the uptake of the fourth silano-

late. Likewise, the structures of the neutral adducts 5^[1]and

29 imply that the metal center in neutral complexes of the

ꢀ

general type [(Ar₃SiO)₃Mo CR] (1) remains well accessible.

The ancillary triarylsilanolate ligand sphere is therefore un-

likely to impede binding of an alkyne (except for, maybe,

very bulky substrates). It is, however, crowded enough to

Chem. Eur. J. 2012, 00, 0 – 0

ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

&

7

&

ÞÞ

These are not the final page numbers!

A. Fꢁrstner et al.

cycle: While bending of the Si-O-Mo unit gently increases

the Lewis acidity of the Mo^VIcenter and hence favors

uptake of the alkyne and the subsequent metallacycle for-

mation, stretching may assist the release of the product.

This notion of a potentially adaptable electronic and steric

environment created by R₃SiO ligands about the active

metal center is subject to further investigations. In any case,

the crystal structures outlined in this paper suggest that the

silanolate ligand sphere is floppy and that this suppleness

contributes to a favorable reactivity profile.

of this notion. Only the 2,9-dimethyl-1,10-phenanthroline

complex 33 remains air sensitive, since the extra methyl sub-

stituents clash into the already crowded ligand sphere and

hence cause its destabilization. Compounds 8a,^[1]8c, 8d and

33 are chemically isomorphic,^[41]featuring a distorted coordi-

nation geometry about the metal center, with one N-atom

cis to the alkylidyne; the structure of complex 8d depicted

in Figure 7 is representative (for the structures of 8c and 33,

Stable adducts with chelating nitrogen-donor ligands: The

now readily available and unambiguously characterized sol-

ꢀ

vate-free ate-complex [(Ph₃SiO)₄Mo CPh]K (28a) is an ex-

cellent starting material for the formation of various adducts

with bidentate N-donor ligands (although the other ate com-

plexes described herein can also be used). Specifically, addi-

tion of 1,10-phenanthroline (phen) to a solution of 28a in

toluene results in the loss of Ph₃SiOK and formation of the

ꢀ

neutral adduct [(Ph₃SiO)₃Mo CPh]·ACTHNUTRGNEUNG(phen) (8a), which, in

solid form, turned out to be bench-stable for at least one

year (Scheme 10).^[1]Its lifetime compares well to that of the

ꢀ

corresponding nitrido adduct [(Ph₃SiO)₃Mo N]·

N

which seems to be indefinitely air stable in crystalline form

(see below, however, for the hydrolysis of such complexes in

wet solution). As the phenanthroline ligands can be pulled

off either adduct with metal salts such as MnCl₂or ZnCl₂,

complexes 3 and 8a serve as convenient, bench stable preca-

talysts for all kinds of alkyne metathesis reactions.^[1]

Gratifyingly, the concept of stabilizing molybdenum alky-

lidynes and -nitrides with the aid of chelating nitrogen

donor ligands is quite general (Scheme 10).^[40]Compounds

8b–d containing substituted benzylidyne groups bear witness

Figure 7. Structure of complex 8d in the solid state. Selected bond

ꢁ

lengths [ꢀ] and angles [8]: Mo1 C13 1.768(2), Mo1 N1 2.403(2), Mo1

N2 2.266(2), Mo1-C13-C14 170.6(2), N1-Mo1-C13 165.5(1).

ꢁ

ꢀ

see the Supporting Information). In all cases, the N Mo C

angle deviates from linearity and the lengths of the two

Mo···N bonds are significantly different. These structural

features suggest that the nitrogen atom trans to the alkyli-

dyne is rather weakly bound; therefore moderately Lewis

acidic additives such as MnCl₂or ZnCl₂suffice to remove

the chelating phenanthroline from such adducts and, in so

ꢀ

doing, release the catalytically competent [(Ph₃SiO)₃Mo

CR] (1) unit.

As expected, 2,2’-bipyridine (bipy) can also be used as

a stabilizing ligand for molybdenum alkylidyne (Scheme 10)

and -nitrido complexes (Scheme 4). Since many substituted

bipyridine derivatives are commercially available, a certain

fine-tuning of the solubility and stability of the resulting ad-

ducts is possible; most of them are nicely crystalline and

bench-stable for extended periods of time. The structures of

representative members of this series in the solid state are

shown in Figures 8 and 9 and additional examples (34a, 34c

and 21b) can be found in the Supporting Information.

These bipyridine adducts are largely similar to the corre-

sponding phenanthroline complexes, with strongly distorted

octahedral coordination geometries and unequal Mo···N dis-

tances. The individual rings of the bipyridine units are slight-

ly twisted out of coplanarity (6.428 in case of the parent

ꢀ

Scheme 10. Molybdenum alkylidyne complexes stabilized by different N-

donor chelate ligands. All reactions were performed by adding the corre-

sponding phenanthroline or bipyridine derivative to a solution of com-

plex 28 in CH₂Cl₂or toluene at ambient temperature. The following

yields were obtained: 90% (8a), 84% (8b), 71% (8c), 73% (8d), 78%

(33), 75% (34a), 78% (34b), 71% (34c), 69% (35a), 75% (35b).

adduct [(Ph₃SiO)₃Mo CPh]·ACTHNUTRGNE(NUG bipy) (34a)].

&

8

&

www.chemeurj.org

ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 0000, 00, 0 – 0

ÝÝ

These are not the final page numbers!

Silyloxy-Based Alkyne Metathesis Catalysts

FULL PAPER

30 min before the substrate is added (Scheme 11); under

these conditions, a host of inter- and intramolecular meta-

thesis reactions were accomplished with good to excellent

yields (see ref. [1] and Table 2). However, MnCl₂is only

sparingly soluble in toluene; therefore, we suppose that only

ꢀ

a small fraction of the active species [(Ph₃SiO)₃Mo CR] (1)

is actually released from 8a after 30 min contact time. Un-

fortunately, however, it is difficult to obtain accurate kinetic

data for the decomplexation reaction with MnCl₂because

this salt is partly un-dissolved in [D₈]-toluene and causes

paramagnetic line-broadening in the NMR spectra.

Figure 8. Structure of complex 34b in the solid state. Selected bond

ꢁ

lengths [ꢀ] and angles [8]: Mo1 C13 1.766(1), Mo1 N1 2.422(1), Mo1

N2 2.244(9), Mo1-C13-C14 178.61(7), N1-Mo1-C13 166.51(3).

ꢀ

Figure 9. Structure of [(Ph₃SiO)₃Mo N]·

Selected bond lengths [ꢀ] and angles [8]: Mo1 N3 1.665(2), Mo1 N2

G

ꢁ

2.496(2), Mo1 N1 2.260(2), N2-Mo1-N3 161.4(1).

As already mentioned, the binding of the phenanthroline

or bipyridine to the metathesis-active alkylidyne unit

Scheme 11. Decomplexation of the stabilizing phenanthroline or bipyri-

dine ligands releases the catalytically active species 1; if molybdenum ni-

tride precatalysts are used, a further alkylidyne-for-nitride exchange on

contact with the substrate is necessary. Envisaged assisted decomplexa-

tion in the case of the bis-pyrimidine adduct 35 (see text).

ꢀ

[(Ph₃SiO)₃Mo CR] (1) is reversible; a variety of metal salts

is able to remove such chelating N-donor ligands from the

corresponding adducts in solution (MnCl₂, FeCl₂, FeCl₃,

CuCl₂, ZnCl₂, CoCl₂, MgCl₂), although the rate of this de-

complexation depends on the specific reagent combination.

In practice, we prefer to use cheap, non-toxic and benign

MnCl₂. This salt is non-hygroscopic and hence can be used

as obtained from the commercial suppliers. Since Mn^IIis

only weakly Lewis acidic, any excess that might be inadver-

tently introduced into the reaction mixture is unlikely to en-

danger even labile substrates or products. For preparative

purposes, it suffices to stir a mixture of, for example, 8a

(5 mol%) and MnCl₂(5 mol%) in toluene at 808C for

Therefore, we also studied the diamagnetic ZnCl₂as the

activating agent. ZnCl₂is cheap, benign and slightly more

Lewis-acidic, but needs to be dried prior to use because of

its hygroscopic nature. With this additive, the decomplexa-

tion of phenanthroline slowly proceeds even at ambient

temperature: it suffices to stir a given alkyne substrate in

ꢀ

the presence of [(Ph₃SiO)₃Mo CPh]·ACHTNUTRGNE(NUG phen) (8a) and ZnCl₂

in toluene to induce effective metathesis.^[42]When per-

Chem. Eur. J. 2012, 00, 0 – 0

ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

&

9

&

ÞÞ

These are not the final page numbers!

A. Fꢁrstner et al.

formed at 808C, a rapid onset of the decomplexation was

to meet the basic requirements for an effective ancillary

ligand for alkyne metathesis catalysts exceedingly well.

Moreover, many of the resulting alkylidyne complexes are

nicely crystalline, which facilitates their isolation and hand-

ling. Since Ph₃SiOH is cheap and commercial, it is a privi-

leged and a practical ligand alike.

Despite these advantageous attributes of the parent com-

pound, variation of the substituents R in R₃SiO- seemed

a worthwhile exercise. To this end, we surveyed the perfor-

mance of a representative set of silanolates by an “in situ”

screening protocol, in which four equivalents of a given

R₃SiOK salt were added to 2–5 mol% of the tribromoalkyli-

dyne complex 25a in toluene prior to the addition of 1-

phenyl-1-propyne as the model substrate. Since calibration

experiments using parent Ph₃SiOM (M = K, Li) resulted in

quantitative conversion (Table 1, entries 1 and 2), we were

confident that this assay would allow us to identify other

promising ancillary ligands for alkyne metathesis.

1

observed by H NMR, reaching ꢂ50% conversion after 1 h.

ꢀ

For a quantitative release of the active [(Ph₃SiO)₃Mo CPh]

template, however, stirring of the mixture for 3 h at 808C is

required. The reaction is clean and no signs of decomposi-

tion of the liberated alkylidyne were observed at this point

(when kept for extended periods of time, however, the

active species will eventually deteriorate; as described

below, a dimeric complex could be isolated from such a mix-

ture, the structure of which is informative in its own right).

In any case, these experiments show the remarkable stability

of the alkylidyne complexes carrying silanolate ligands to-

wards Lewis-acidic salts even at elevated temperatures. This

ꢁ

advantageous property is ascribed to the fairly robust Si O

3

ꢁ

bonds, whereas Lewis acid catalyzed rupture of the C

N

O bond of tertiary alkoxides (previously used as ancillary li-

gands, cf. Scheme 3) with formation of carbocations is argua-

bly more facile.

In an attempt to gain further control over the decomplex-

ation of the stabilizing N-donor ligands, we prepared the

bis-pyrimidine derivatives 35a,b (Scheme 10, Figure 10). It

was anticipated that binding to the “allosteric” site may fa-

The data compiled in Table 1 show a clear-cut correlation

between the substituents on the silanolate and the catalytic

performance of the resulting complex. All tested triarylsila-

nolates gave good to excellent results, whereas trialkylsila-

nolates performed less effectively or even failed completely.

Et₃SiOK and tBuMe₂SiOK, as the best examples of the

latter series, required much longer reaction times and/or

higher loadings to reach appreciable conversions (entries 7

and 9). The same reactivity trend is also manifested in the

order Ph₃SiO- ꢂ Ph₂MeSiO- > PhMe₂SiO- (cf. entries 1, 3

and 5).

Substituents on the phenyl groups are generally well toler-

ated and may offer advantages in practical terms. A notable

ꢀ

example is complex [(Ar₃SiO)₄Mo CPh][K] (37, Ar = p-

Table 1. Evaluation of the catalytic efficiency of complexes of the gener-

ꢀ

al type [(R₃SiO)₄Mo CPh][M] (M = K, Li) prepared in situ, which differ

only in the chosen silanolate ligands.^[a]

Entry

R₃SiOK

[Mo]/mol%^[b]

Yield [%]^[c]

1

2

3

4

5

6

7

8

Ph₃SiOK

Ph₃SiOLi

Ph₂MeSiOK

2

5

2

quant. (94)

quant.

51^[d]

Figure 10. Structure of complex 35a in the solid state. Selected bond

Ph₂ACTHNGUTER(NNUG tBu)SiOK

ꢁ

lengths [ꢀ] and angles [8]: Mo1 C9 1.766(2), Mo1 N1 2.263(1), Mo1 N4

2.390(1), Mo1-C9-C10 174.7(2), N4-Mo1-C9 167.8(1).

PhMe₂SiOK

Me₃SiOK

Et₃SiOK

33

0

56^[e]

T

5^[f]

cilitate the release of the catalytically active species

(Scheme 11). In fact, the characteristic NMR signals of 35b

disappear faster than those of the corresponding phenan-

throline from 8a when treated with ZnCl₂in toluene at

808C, although the rate difference was not as significant as

we had hoped (2 versus 3 h for complete conversion); more-

over, spectroscopic evidence suggests that the species re-

leased from 35b is not the donor free 1 itself; the exact

nature of this new catalytically competent derivative is sub-

ject to further investigation.

9

N

97 (95)

10

11

12

13

14

ACHTUNGTRENNUNG

9^[g]

(p-MeO-C₆H₄)₃SiOK

(m-Cl-C₆H₄)₃SiOK

(p-F₃C-C₆H₄)₃SiOK

quant.

93

A

73

ꢀ

[a] The catalysts were prepared by premixing [(Br₃Mo CPh)·ACHTUNGTRENNUNG(dme)]

(25a) with R₃SiOM (4 equiv relative to 25a) in toluene at ambient tem-

perature; after stirring for 1 h, 1-phenyl-1-propyne and MS 5 ꢀ were in-

troduced and the conversion checked by GC after 1 h. [b] Refers to the

amount of 25a used. [c] GC yield after 1 h reaction time; isolated yields

in brackets. [d] 96% conversion after 7 h. [e] 86% conversion after 7 h.

[f] The reaction proceeded very slowly and reached 50% conversion

after 22 h. [g] The reaction proceeds very slowly and reaches 89% con-

version after 22 h.

Variations of the silanolate ligands: As discussed above, the

electronic and steric properties of triphenylsilanolate seem

&

10

&

www.chemeurj.org

ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 0000, 00, 0 – 0

ÝÝ

These are not the final page numbers!

Silyloxy-Based Alkyne Metathesis Catalysts

FULL PAPER

MeO-C₆H₄-) endowed with tri(p-methoxyphenyl)silanolate

ligands, which seems to be more stable in solution than the

parent ate-complex 28a and also renders product isolation

more facile in certain applications (Scheme 12). Specifically,

we experienced that the silicon-containing by-products, re-

leased from complexes endowed with Ph₃SiO ligands upon

work-up of the reaction mixtures, sometimes co-elute with

the desired products, if the latter are fairly apolar. In such

cases, the use of 37 is recommended, because the silanole-

derived by-products are more polar and hence easily re-

moved by flash chromatography. For this very property, cat-

alyst 37 has been instrumental for a number of exigent ap-

plications, including the total syntheses of the ecklonialac-

tones, hybridalactone and tulearin C.^[7,10]

Figure 11. Structure of complex 38 in the solid state.^[38]Selected bond

ꢁ

lengths [ꢀ] and angles [8]: Mo2 C148 1.772(7), Mo2 N3 2.377(6), Mo2

N4 2.245(5), Mo2-C148-C149 178.3(5), N3-Mo2-C148 171.5(2).

Scheme 12. Variation of the silanolate ligands: a) (p-MeOC₆H₄)₃SiOK

(5 equiv), toluene, 64%; b) (p-MeOC₆H₄)₃SiOK (5 equiv), then 2,2’-bi-

pyridine, Et₂O, 36%; c) tBuMe₂SiOK, toluene, then 2,2’-bipyridine (see

text).

Figure 12. Structure of complex 39 in the solid state. One of the tert-

butyl-dimethylsilanolate ligands is disordered over two positions with rel-

ative occupancy 0.85:0.15, indicating some space for movement of the li-

The air-stable bipyridine adduct of this versatile complex

has also been prepared (Scheme 12). In combination with

ꢁ

gands. Selected bond lengths [ꢀ] and angles [8]: Mo1 C11 1.767(3),

ꢁ

Mo1 N1 2.395(3), Mo1 N2 2.288(3), Mo1-C11-C12 174.8(3), N1-Mo1-

C11 166.2(1).

ꢀ

MnCl₂or ZnCl₂, [(R₃SiO)₃Mo CPh)]·

G

MeOC₆H₄-) serves as a convenient precatalyst for various

synthetic applications (see below). The structure of 38 in the

solid state (Figure 11) resembles that of the parent complex

34a, with a distorted octahedral geometry, a visibly bent

Catalyst decomposition and hydrolysis: The stability of the

active species in solution determines the maximum turnover

number and hence the necessary catalyst loading. Previous

experiments had already shown that the alkylidyne complex

ꢀ

ꢁ

ꢀ

N Mo C axis, unequal Mo···N bond lengths and a slightly

twisted bipyridine unit being pertinent characteristics. The

same holds true for the structure of complex 39, containing

tBuMe₂SiO groups (Figure 12). Because of its high solubility

in pentane even at ꢁ788C, the isolation was more tedious.

However, despite a close structural similarity to 8 and 38,

complex 39 is catalytically less active. Therefore, subtle elec-

tronic and/or steric differences must be invoked to explain

why alkylidyne complexes supported by trialkylsilanolate li-

gands perform less effectively than their cousins carrying tri-

arylsilanolates.

[(Ph₃SiO)₄Mo CPh][KACHTNUTRGNE(NUG Et₂O)_n] (n = 1, 2) (27a) endowed

with triphenylsilanolate ligands outperforms the established

alkyne metathesis catalysts, because even a loading of

0.1 mol% (unoptimized) engendered metathesis of

1-phenyl-1-propyne (ꢃ95% GC) in short reaction times.^[1]

Qualitatively, we ascribed this favorable result to the in-

creased stability of the Si-O-M bonds (relative to the

C-O-M bonds of tert-alkoxides) as well as to the bulky sila-

nolate periphery about the active alkylidyne unit, which

Chem. Eur. J. 2012, 00, 0 – 0

ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

&

11

&

ÞÞ

These are not the final page numbers!

A. Fꢁrstner et al.

ꢀ

should disfavor any bimolecular decomposition pathway.

However, a better understanding of these aspects is desira-

ble.

that the tungsten methylidyne complex [(tBuO)₃W CH]

readily equilibrates with the ditungstatetrahedrane

[(tBuO)₆W₂ACTHNUTRGNEUNG

(m-C₂H₂)].^[44]Moreover, closely related species

To this end, the half-lifetime of the ate-complex 27a (sa-

turated solution in [D₈]-toluene) was determined by NMR

to be about 30 h at ambient temperature. GC inspection

showed that variable amounts of tolane and (E,Z)-stilbene

constituted the major organic products, suggesting that bi-

molecular decomposition pathways are indeed operative. To

were obtained in reactions of alkynes with dimolybdenum-

ꢀ

or ditungsten hexaalkoxide precursors containing M M

triple bonds.^[45–48]Dimetallatetrahedranes can be viewed as

m-bridging alkyne complexes; therefore, the isolation of

compound 40 from a 8b/ZnCl₂mixture lends credence to

the notion that the collision of two active catalyst entities

engenders coupling of their benzylidyne units with forma-

tion of (substituted) tolane. This dimerization process likely

represents an important pathway for the deactivation of

alkyne-metathesis catalysts of this sort.^[49]

ꢀ

clarify the fate of the molybdenum core, [(Ph₃SiO)₃Mo

CAr]·(phen) (8b, Ar = p-MeOC₆H₄) was first treated with

G

ZnCl₂in toluene for 3 h at 808C to liberate free

ꢀ

[(Ph₃SiO)₃Mo CAr] (1b); the precipitated phenanthroline–

zinc adduct was filtered off, the filtrate evaporated, the resi-

due taken up in the minimum amount of CH₂Cl₂, and the re-

sulting solution layered with pentane. After storage for sev-

eral days at ꢁ208C, small amounts of green crystals of

a new compound could be collected.

The dimetallatetrahedrane core of 40 is fairly symmetri-

ꢁ

cal, with the vector of the Mo Mo bond being almost or-

ꢁ

thogonal to the C C axis (92(1)8). The C C bond of the m-

bridging alkyne is significantly lengthened (1.390(3) ꢀ) and

the aryl rings bent away from linearity as a consequence of

the twofold metallacyclopropane substructure. The bond

lengths from the two metal centers to the bridging chloride

are uneven, whereas the bridging siloxide keeps equidis-

tance. The terminal Ph₃SiO-groups seem to engage in stron-

ger O!Mo p-donation than their bridging counterpart, as

evident from the significantly shorter bond lengths

(Figure 13).

Complex 40 thus obtained is characterized by a dimetalla-

tetrahedrane core formed from two alkylidyne species

(Figure 13). Such a dimerization has precedent in an early

report on the fate of a Fischer-carbyne complex,^[43]as well

as in the work of Chisholm and co-workers, who showed

Although the above data show that the new siloxide-

based catalysts are thermally quite robust and even fairly

stable on exposure to Lewis acids, anhydrous conditions are

mandatory for productive alkyne metathesis. It is not clear

however, if water simply hydrolyzes the triphenylsilanolate

ligands off or whether it also reacts with the alkylidyne (ni-

tride) unit. It is known, for example, that hydrolysis of

ꢀ

(tBuO)₃W CR (R = tBu, Ph) and subsequent treatment of

the crude material with Et₄NOH removes the alkoxides but

not the carbon-based ligand; rather, alkyl complexes of the

ꢁ

general type [Et₄N]

G

turned out to be surprisingly stable.^[50]

In no case, however, have we been able to obtain any hy-

drolysis product that retained an organometallic fragment

from the catalysts described here. Specifically, exposure of

ꢀ

the ate-complex [(Ph₃SiO)₄Mo CPh][KACTHNUTGRNEUGN(Et₂O)_n] (27a) to

moist solvents resulted in total loss of the benzylidyne unit

with formation of the dimeric species 41 (Scheme 13,

Figure 14). The oxo bridges of its Mo₂O₂core are highly un-

Figure 13. Structure of complex 40 in the solid state and schematic repre-

sentation of the bonding situation, highlighting the m-bridging alkyne unit

in red and the chloride bridge in green; Ar = p-MeOC₆H₄-. Selected

ꢁ

bond lengths [ꢀ] and angles [8]: Mo1 Mo2 2.5618(3), C1 C9 1.390(3),

Scheme 13. Schematic representation of the structure of a hydrolysis

product (41) derived from the unstabilized ate-complex 27a. The oxo

bridges of the Mo₂O₂core are highly unsymmetrical; the potassium coun-

terions also entertain strong cation–p interactions with the phenyl rings

of the peripheral silanolate ligands.

ꢁ

Mo1 C1 2.129(2), Mo2 C1 2.092(3), Mo1 C9 2.086(2), Mo2 C9

ꢁ

2.115(2), Mo1 Cl1 2.4948(7), Mo2 Cl1 2.5862(7), Mo1 O5 2.068(2),

Mo2 O5 2.031(2), Mo1 O3 1.919(2), Mo1 O4 1.891(2), Mo2 O6

1.900(2), Mo2 O7 1.952(2), C2-C1-C9 139.6(2), C1-C9-C10 138.5(2).

ꢁ

&

12

&

www.chemeurj.org

ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 0000, 00, 0 – 0

ÝÝ

These are not the final page numbers!

Silyloxy-Based Alkyne Metathesis Catalysts

FULL PAPER

Scheme 14. Partial hydrolysis of the molybdenum nitride 4 upon com-

plexation with 2,2-bipyridine, which had not been rigorously dried prior

to use.

Figure 14. Structure of the hydrolysis product derived from 41 in the

ꢁ

solid state. Selected bond lengths [ꢀ] and angles [8]: Mo1 Mo1*

ꢁ

2.4372(6), Mo1 O4 1.833(2), Mo1 O4* 2.083(2), Mo1 O1 1.950(2),

ꢁ

Mo1 O2 2.050(2), Mo1 O3 1.971(2), K1 O1 2.790(2), K1 O3* 2.692(2),

ꢁ

K1 O4 2.941(2), Mo1-O4-Mo1* 76.69(6).

Figure 15. Solid-state structure of the hydrolysis product 42 derived from

ꢁ

the nitrido complex 4. Selected bond lengths [ꢀ] and angles [8]: Mo1 O1

ꢁ

1.702(1), Mo1 O2 1.705(1), Mo1 O3 1.938(1), Mo1 O4 1.939(1), O3-

Mo1-O4 152.61(5), O1-Mo1-O2 106.15(7).

ꢁ

symmetrical (Mo1 O4 1.833(2), Mo1* O4 2.083(2) ꢀ); the

fate of the former alkylidyne group could not be clarified. It

should be noted, however, that this product is almost cer-

tainly only the least soluble of several hydrolysis products in

the crude mixture, judging from the low and variable

yields.^[51]

bench stability of 3, 8 and relatives in crystalline form is ki-

netic in origin. It needs to be seen in future investigations

whether the crafting of a tighter and more hydrophobic

ꢀ

In an attempt to determine whether the presence of a sta-

bilizing phenanthroline or bipyridine ligand changes the out-

come, the stability of four different alkylidyne adducts in

wet CD₂Cl₂solution (10 ppm water by Karl-Fischer titra-

tion) was checked by H NMR. By recording the integral of

the ortho-protons on the benzylidyne (relative to an internal

pocket allows the operative Mo CR unit to be further stabi-

lized and hence the lifetime of the catalysts in protic media

to be increased.

1

Decomposition by terminal alkynes: In addition to the de-

composition pathways by bimolecular collision or by hydrol-

ysis addressed above, Schrock alkylidyne complexes are also

known to degrade in attempted metathesis reactions of ter-

minal (rather than internal) alkynes as the substrates. The

critical step occurs after metallacycle formation; it consists

ꢀ

standard), the half-life time (t_1/2) of [(Ph₃SiO)₃Mo CPh]·-

(phen) (8a) could be estimated to be of the order of 5 d,

whereas

the

corresponding

bipyridine

adduct

ꢀ

[(Ph₃SiO)₃Mo CPh]·

U

ꢁ

ꢂ 16 h). As expected, adduct 34c comprising the better

donor 4,4’-dimethoxybipyridine is more robust (t_1/2ꢂ 48 h)

than the parent bipyridine complex 34a. The most stable

amongst the investigated adducts was 8d (t_1/2ꢂ 14 d). This

complex features a phenanthroline ligand and seems to ben-

efit from an additional shielding effect exerted by the two

methyl substituents on the benzylidyne unit.

of a transannular C H activation with formation of a depro-

tio-metallacyclobutadiene and concomitant loss of one al-

koxide.^[20a,53]The new alkylidynes supported by silanolate li-

gands did not allow us to overcome this limitation and

extend alkyne metathesis to terminal alkynes; rather, when

ꢀ

complex 32 was reacted with p-MeOC₆H₄C CH, the depro-

tio-molybdacyclobutadiene 43 was isolated after addition of

1,10-phenanthroline (Scheme 15). Figure 16 shows the struc-

ture of this peculiar complex in the solid state. The 1,3-di-

(methoxyphenyl)prop-2-ynylidene is bound symmetrically in

an h³-fashion to the Mo atom.

ꢀ

Attempts to prepare the stable adduct [(Ph₃SiO)₃Mo N]·-

(bipy) (21a) from complex 4 and 2,2’-bipyridine, which had

N

not been dried by careful sublimation prior to use, gave fur-

ther insights into the hydrolysis reactions. In this case, the

nitride unit was totally lost and the dioxomolybdenum disil-

oxide complex 42 was formed as the only isolable product

(Scheme 14 and Figure 15).^[52]The fact that all of the tested

complexes eventually react with water implies that the

Applications: Our previous investigations^[1,2]and the results

presented in this paper provide strong evidence that active

alkyne metathesis catalysts of the general type

Chem. Eur. J. 2012, 00, 0 – 0

ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

&

13

&

ÞÞ

These are not the final page numbers!

A. Fꢁrstner et al.

ꢀ

bipyridine adducts [(Ph₃SiO)₃Mo N]·(L) (3, L = phen;

21, L = bipy or substituted derivatives thereof) with

MnCl₂or ZnCl₂as preferred additives.

Since we have already investigated the reactivity of the

ꢀ

active species [(Ph₃SiO)₃Mo CR] (1) in some detail in our

earlier studies,^[1,7–10]it sufficed to compare the competence

of the new bipyridine adducts with that of their phenanthro-

line counterparts. Moreover, we surveyed the activity of

Scheme 15. Decompositon pathway by deprotiometallacyclobutadiene

formation upon reaction of the alkylidyne catalyst 32 with a terminal

alkyne substrate.

ꢀ

complex [(R₃SiO)₄Mo CPh][K] (37) (R = p-MeO-C₆H₄-)

and the bipyridine adduct 38 thereof, since the peripheral

p-methoxy substituents sometimes renders product isolation

more facile (see above).

As evident from Tables 2–4, the preparative results are

highly rewarding. Although not all substrates were tested

with each available precatalyst, the dataset is large enough

to conclude that their performance is remarkably uniform.

In line with our previous work,^[1,2]molybdenum nitride spe-

cies generally require higher reaction temperatures than

their alkylidyne counterparts. As expected, it makes no sig-

nificant difference if the active species is set free from a phe-

nanthroline- or from a bipyridine adduct. Complex 37 con-

taining p-methoxyphenyl groups on the silanolates is slightly

less reactive than the parent complex 28a with bare phenyl

rings and tends to give somewhat larger amounts of by-

products formed by (cyclo)dimerization of the substrate.

This minor drawback, however, is over-compensated by the

ease of separation of the silanole-derived residues formed

upon work-up, in cases in which the desired alkyne products

are fairly unpolar. This practical advantage was instrumental

for several advanced applications.^[7,10]

All molybdenum catalysts endowed with triarylsilanolate

ligands exhibit a remarkable and unparalleled compatibility

with functional groups. This includes a large variety of polar

substituents as well as moderately basic and/or strong donor

sites (pyridines, thioethers). Moreover, various fragile and

reactive entities subsist, including epoxides, skipped 1,4-

dienes, acid sensitive allyl- and dienyl esters, as well as cy-

clopropylcarbinol groups. It is also noteworthy that propar-

gylic alcohol derivatives were found to participate in ring

closing- as well as alkyne cross metathesis reactions

(Table 3, entry 2; Table 4, entries 5, 6). The resulting prod-

ucts are amenable to various post-metathetic transforma-

tions and hence promise access to a number of different im-

portant structural motifs. This and other aspects are current-

ly being investigated in our laboratory in more detail.

Figure 16. Structure of complex 43 in the solid state. Selected bond

ꢁ

lengths [ꢀ] and angles [8]: Mo1 C13 1.979(4), Mo1 C14 2.030(5), Mo1

ꢁ

C15 1.961(5), Mo1 N1 2.366(4), Mo1 N2 2.336(4), Mo1 O1 1.949(3),

ꢁ

Mo1 O2 1.980(3), C13 C14 1.374(6), C14 C15 1.371(7), O1-Mo1-O2

155.3(1), C13-C14-C15 135.2(4).

ꢀ

[(Ph₃SiO)₃Mo CR] (1) can be generated in at least three

different ways (see Scheme 1):

i) Loss of one Ph₃SiO residue from an ate precursor such

ꢀ

as [(Ph₃SiO)₄Mo CAr][K·

ACHTUNGTRENN(GNU dme)] (26), [(Ph₃SiO)₄Mo

ꢀ

CAr][K(Et₂O)_n] (27) or [(Ph₃SiO)₄Mo CAr][K] (28);

ACHTUNGTRENNUNG

this step may be spontaneous (equilibration) or induced

by the alkyne substrate and/or added MS 5 ꢀ.

ii) Removal of the phenanthroline or bipyridine ligand

Conclusion

ꢀ

from a stable adduct such as [(Ph₃SiO)₃Mo CAr]·(L) (8,

L = phen; 34, L = bipy) with the aid of MnCl₂or

ZnCl₂.

Triarylsilanolates impart a remarkable reactivity and unri-

valed compatibility with functional groups on molybdenum

alkylidyne and -nitride complexes, which, therefore, serve as

excellent catalysts for alkyne metathesis reactions of highly

functionalized and sensitive substrates. This outstanding ap-

plication profile derives from the favorable electronic and

steric features of the chosen Ar₃SiO groups. They ensure

iii) Alkylidyne-for-nitride exchange during the first turn-

ꢀ

over of a nitride complex such as [(Ph₃SiO)₃Mo N] (4)

on exposure to an alkyne substrate. Like in the alkyli-

dyne series, such complexes can be conveniently re-

leased in situ from the corresponding phenanthroline or

&

14

&

www.chemeurj.org

ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 0000, 00, 0 – 0

ÝÝ

These are not the final page numbers!

Silyloxy-Based Alkyne Metathesis Catalysts

FULL PAPER

ꢀ

Table 2. Intermolecular metathesis reactions of alkynes of the general type RC CMe effected by different

precatalysts; all data refer to isolated yields of product, unless stated otherwise.

tion sphere of the high-valent

molybdenum centers as well as

in the presence of Lewis acidic

additives. Most notably, com-

plexes of the general type

ꢀ

Entry Product

Nitride complexes [%]

Alkylidyne complexes [%]

3^[a]

21a^[a]

27a^[b]8a^[c]34a^[c]37^[b]38^[c]

1

99

97

99

97

99

97

98

95

[(Ar₃SiO)₃Mo X] (X

= N,

CR) can be rendered air stable

by reversible adduct formation

with 1,10-phenanthroline, 2,2’-

bipyridine or derivatives there-

of. First insights into the ther-

mal or hydrolytic decomposi-

tion pathways of such com-

plexes were obtained, which

may guide further catalyst op-

timization. Studies along these

lines together with applications

to advanced organic synthesis

are underway and will be re-

ported in due course.

2

3

96

94

98

99^[h]93^[l]

87

98^[d]

96^[d]95^[d]

4

5

72^[e]

95

97^[f]

94

94^[j]

94^[k]

6

7

94

93

84

95

84

94

85

<40^[e]

Experimental Section

8

99^[j]

93^[k]

Outlined below is the preferred and

scalable synthesis of complexes 28b,

9

94^[d]

69^[g]

86^[i]

32 and 8b (Ar

= p-MeOC₆H₄-),

which serve as standard catalysts for

all kinds of alkyne metathesis reac-

tions. All other experimental details

are found in the Supporting Informa-

tion. The details include description

of the experimental procedures, com-

pound characterization, and crystallo-

graphic abstracts. The Supporting In-

formation also contains the structures

of additional complexes in the solid

state (8c, 21b, 33, 34a, 34c, 39 and

44^[51]) not shown in the main text, as

well as the picture of the entire unit

cells of complexes 28 and 38.

10

11

61

12

13

14

76^[e]

86

90^[d]

88

88^[d]

87

62^[i]

93

83

90

92

98^[h]

General methods: The reactions were

carried out under Ar in flame-dried

glassware. The solvents used were pu-

rified by distillation over the drying

agents indicated and were transferred

under Ar: THF, Et₂O, (Mg/anthra-

cene), CH₂Cl₂, DME (CaH₂), pen-

tane, hexane, toluene (Na/K), MeOH

(Mg). NMR: Spectra were recorded

on Bruker DPX 300, AMX 300, AV

400, and AVIII 600 spectrometers in

the solvents indicated; chemical shifts

(d) are given in ppm relative to TMS,

coupling constants (J) in Hz. The sol-

vent signals were used as references

and the chemical shifts converted to

85

89

91

15

81

87

89

[a] 3 or 21a (10 mol%), MnCl₂(10 mol%), MS 5 ꢀ, toluene, 808C, 30 min, then addition of the substrate and

reaction at 808C unless stated otherwise. [b] 27a or 37 (2 mol%), toluene, ambient temperature, MS 5 ꢀ.

[c] 8a or 34a or 38 (5 mol%), MnCl₂(5 mol%), toluene, 808C, 30 min; then addition of the substrate and MS

5 ꢀ, and reaction at ambient temperature. [d] At 508C. [e] At 1008C. [f] With 8b (5 mol%), ZnCl₂

(10 mol%), toluene, MS 5 ꢀ, ambient temperature, quant. conversion after 6 h. [g] At 808C. [h] With 5 mol%

of 37. [i] With 10 mol% of 37 at 1108C. [j] Complex 28b was used instead of 27a. [k] Complex 8b was used.

[l] Using 10 mol% of the catalyst.

ꢀ

the TMS scale (CD₂Cl₂: d_C

53.8 ppm; residual ¹H: d_H

ꢀ

a well-balanced level of Lewis acidity at molybdenum, pro-

5.32 ppm). IR: Spectrum One (Perkin-Elmer) spectrometer, wavenum-

bers (n˜) in cm^ꢁ1. Elemental analyses: H. Kolbe, Mꢁlheim/Ruhr.

tect the operative alkylidyne unit against bimolecular de-

composition, but do not hinder substrate binding. Moreover,

triarylsilanolates are reasonably stable within the coordina-

1,10-Phenanthroline and 2,2’-bipyridine were purified and dried by subli-

mation of the commercial sample (Aldrich, ꢃ99%) under vacuum

Chem. Eur. J. 2012, 00, 0 – 0

ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

&

15

&

ÞÞ

These are not the final page numbers!

A. Fꢁrstner et al.

Table 3. Comparison of the performance of different alkyne metathesis catalysts in alkyne cross metathesis

(ACM) reactions.

2H), 4.03 (brs, 4H), 3.92 (s, 3H),

3.90 (s, 3H), 3.88 ppm (s, 3H); ¹³C

NMR (100 MHz, CD₂Cl₂): d = 332.8,

163.4, 135.6, 135.0, 113.2, 78.7, 73.0,

70.3, 60.5, 55.9 ppm; elemental analy-

sis calcd (%) for C₁₂H₁₇Br₃MoO₃: C

26.45, H 3.14; found: C 26.28, H 3.09.

Entry

Substrates

Product

27a^[a]/%

34a^[b]/%

37^[c]/%

1

5-decyne

tolane

65

75

65

ꢀ

2

62

71

[Mo

G

CAr)

G

= 4-methoxyphenyl): A solution of

Ph₃SiOK (18.40 g, 58.5 mmol) in tolu-

ene (200 mL) was added over 30 min

ꢀ

[a] 27a (2 mol%), toluene, ambient temperature, MS 5 ꢀ. [b] 34a (5 mol%), MnCl₂(5 mol%), toluene, 808C,

30 min, then addition of the substrate and MS 5 ꢀ followed by reaction at ambient temperature. [c] 37

(5 mol%), toluene, ambient temperature, MS 5 ꢀ.

to

a suspension of [MoCATHUNGTNRUEGN( CAr)Br₃-

AHCTUNGTRENGN(UN dme)] (25b, 6.37 g, 11.7 mmol) in

toluene (200 mL) under argon at

08C, causing color change from

a

(10^ꢁ3mbar). All other commercially available compounds (Fluka, Lan-

caster, Aldrich) were used as received.

yellow-brown to purple. The mixture was then stirred at room tempera-

ture for 2 h before it was filtered through a pad of Celite under argon

and the filtrate was evaporated. The residue was suspended in diethyl

ether (120 mL) and the mixture stirred for 15 min before pentane

(120 mL) was added and the precipitate filtered off under argon. The

purple solid was dissolved in CH₂Cl₂(30 mL) and the resulting turbid

mixture slowly filtered under argon into a flask containing vigorously

stirred pentane (300 mL). The resulting precipitate was filtered off under

ꢀ

A

CHTUNGTRENNUNG

4-methoxyphenyllithium (63.2 mmol) in diethyl ether (250 mL) was

added over 30 min to a suspension of [Mo(CO)₆] (16.68 g, 63.2 mmol)

and diethyl ether (600 mL) under argon, causing an immediate color

change to orange-brown and a slight exotherm. Once the addition was

complete, the resulting orange-brown solution was stirred at ambient

temperature for 4 h before the solvent was distilled off at room tempera-

argon and dried in vacuum to give [MoACHTNUTRGENN(UG CAr)ACHUTGNTREN(NUGN OSiPh₃)₄][K] (28b) as

a lilac solid (10.82 g, 68%). ¹H NMR (400 MHz, CD₂Cl₂): d = 7.51 (d, J

= 7.5 Hz, 24H), 7.20 (t, J = 7.5 Hz, 12H), 6.95 (t, J = 7.5 Hz, 24H), 6.34

(d, J = 8.8 Hz, 2H), 6.13 (d, J = 8.8 Hz, 2H), 3.68 ppm (s, 3H); ¹³C

NMR (100 MHz, CD₂Cl₂): d = 283.1, 158.6, 139.7, 138.9, 136.4, 135.5,

132.2, 130.3, 129.1, 128.1, 127.7, 112.4, 55.5 ppm; elemental analysis calcd

(%) for C₈₀H₆₇KMoO₅Si₄: C 70.87, H 4.98; found: C 70.47, H 5.04.

ture under reduced pressure (ca. 15 mbar).

A solution of NMe₄Br

(14.62 g, 94.9 mmol) in deionized water (150 mL) was added to the re-

maining dark yellow residue, causing the formation of an orange solid.

The resulting suspension was vigorously stirred for 15 min before the pre-

cipitate was filtered off in air and dried in vacuo to give the title complex

as a pale orange microcrystalline solid (24.24 g, 84%). The NMR spectra

show two distinct signal sets (ca. 90:10): ¹H NMR (400 MHz, CD₂Cl₂):

major isomer: d = 7.58 (d, J = 8.8 Hz, 2H), 6.86 (d, J = 8.8 Hz, 2H),

3.81 (s, 3H), 3.29 ppm (s, 12H); minor isomer: d = 7.63 (d, J = 8.3 Hz,

2H), 6.57 (d, J = 8.3 Hz, 2H), 3.68 (s, 3H), 3.29 ppm (s, 12H); ¹³C NMR

(125 MHz, CD₂Cl₂): d = 294.5, 218.3, 212.9, 212.0, 201.6, 160.5, 148.5,

ꢀ

[Mo

(

CAr)

G

(phen)] (8b, Ar = 4-methoxyphenyl): 1,10-Phenan-

throline (sublimed, 586 mg, 3.25 mmol) was added to

a solution of

ꢀ

[Mo

( CAr)

(OSiPh₃)₄][K] (28b) (4.53 g, 3.34 mmol) in toluene (100 mL)

under argon and the resulting mixture stirred for 5 h at room tempera-

ture before it was transferred via cannula into a flask containing vigo-

rously stirred pentane (300 mL). While stirring for 1 h, the color of the

resulting suspension changed from beige to pale green. The crude prod-

uct was filtered off, washed with pentane (50 mL) and dried in vacuum

before it was re-dissolved in CH₂Cl₂(40 mL). The resulting turbid solu-

tion was slowly filtered into a flask containing vigorously stirred pentane

(300 mL), causing the precipitation of the product. The precipitate is col-

ꢀ

146.4, 129.8, 128.1, 120.9, 114.2, 112.7, 56.6 (t, J_CN

= 4 Hz), 55.6,

55.4 ppm; IR (film): n˜ = 3033, 2045, 1952, 1842, 1594, 1516, 1483, 1446,

1299, 1253, 1122, 1029, 947, 851, 835, 775 cm^ꢁ1

.

ꢀ

[Mo

[NMe₄][Mo(CO)

(

CAr)Br

G

= 4-methoxyphenyl): A solution of

was stirred under argon for 15 min at ꢁ788C before a solution of oxalyl

bromide (3.92 mL, 41.9 mmol) in CH₂Cl₂(15 mL) was added over

10 min, causing a color change of the solution from orange to an intense

purple/violet with concomitant evolution of gas. Once the addition was

complete, stirring was continued for 15 min at ꢁ788C before the cooling

bath was removed and the mixture allowed to warm until a sudden color

change to yellow was observed (at ca. ꢁ308C). At this point, the flask

was immersed again into the dry ice/acetone cooling bath (ꢁ788C) and

the cold mixture quickly filtered at this temperature under argon through

a pad of Celite (5 cm), using a glass filter jacketed with a cooling mantle

connected to a cryostat set to ꢁ788C. 1,2-Dimethoxyethane (DME,

20.9 mL, 203 mmol) was added to the resulting pale yellow filtrate at

ꢁ788C before a solution of Br₂(2.12 mL, 42 mmol) in CH₂Cl₂(10 mL)

was introduced at this temperature over a period of 10 min, causing

a gentle evolution of gas. After stirring for 15 min, the then intensely

orange colored mixture was allowed to reach ambient temperature over

the course of 1 h, causing a color change from orange to yellow-brown.

Stirring was continued for 1 h at room temperature before the solvent

was distilled off under reduced pressure until a total volume of ca.

100 mL remained in the flask. Pentane (500 mL) was added to the stirred

mixture, causing the precipitation of a brownish solid. This precipitate

was filtered off under argon, washed with pentane (50 mL) and dried in

vacuum (10^ꢁ3mbar). The crude material was dissolved in CH₂Cl₂

(150 mL) under argon and the resulting turbid solution slowly filtered

into a flask containing vigorously stirred pentane (500 mL). The resulting

precipitate was filtered off under argon and dried in vacuum (10^ꢁ3mbar)

to give complex 25b as a bronze-colored solid (18.10 g, 80%). ¹H NMR

(400 MHz, CD₂Cl₂): d = 7.44 (d, J = 8.6 Hz, 2H), 6.98 (d, J = 8.6 Hz,

lected by filtration and dried in vacuum to give [MoACTHNURGTENN(UG CAr)ACHTUNGTRENNUNG(OSiPh₃)₃-

AHCTUNGTERNNG(UN phen)] as a

pale-green solid (2.77 g, 84%). ¹H NMR (400 MHz,

CD₂Cl₂): d = 9.28 (dd, J = 4.6, 1.6 Hz, 1H), 8.92 (dd, J = 5.0, 1.6 Hz,

1H), 8.32 (dd, J = 8.2, 1.6 Hz, 1H), 7.95 (dd, J = 8.0, 1.4 Hz, 6H), 7.83

(dd, J = 8.2, 1.5 Hz, 1H), 7.73 (d, J = 8.81 Hz, 1H), 7.62 (dd, J = 8.2,

4.6 Hz, 1H), 7.54 (d, J = 8.8 Hz, 1H), 7.22 (tt, J = 7.5, 1.4 Hz, 3H), 7.10

(tt, J = 7.5, 1.4 Hz, 6H), 7.08 (dd, J = 8.0, 1.4 Hz, 12H), 6.99 (t, J =

7.7 Hz, 6H), 6.84 (t, J = 7.7 Hz, 12H), 6.81 (dd, J = 8.2, 5.0 Hz, 1H),

6.52 (d, J = 8.9 Hz, 2H), 6.39 (d, J = 8.9 Hz, 2H), 3.78 ppm (s, 3H); ¹³C

NMR (100 MHz, CD₂Cl₂): d = 292.1, 158.6, 155.4, 147.5, 143.2, 142.7,

139.6, 139.0, 138.1, 137.9, 137.3, 136.3, 135.4, 133.0, 130.0, 129.7, 129.1,

128.6, 127.7, 127.3, 127.2, 127.1, 124.3, 123.9, 112.5, 55.5 ppm; IR (film): n˜

= 3046, 1590, 1565, 1502, 1483, 1427, 1244, 1110, 1033, 1017, 998, 917,

839, 735, 698 cm^ꢁ1; elemental analysis calcd (%) for C₇₄H₆₀MoN₂O₄Si₃: C

72.76, H 4.95, N 2.29; found: C 72.75, H 4.94, N 2.28.

[(Ph₃SiO)₃Mo CAr] (Ar

=

4-methoxyphenyl) (32):

A

solution of

Ph₃SiOK (10.5 g, 33.4 mmol) in toluene (100 mL) was added over 1 h to

a solution of complex 25b (6.07 g, 11.1 mmol) in toluene (200 mL) under

argon. The mixture was stirred for 1 h before the solvent was distilled off

under reduced pressure until a total volume of ca. 50 mL remained in the

flask. The mixture was filtered under argon through a pad of Celite and

the filtrate was evaporated. The residue was recrystallized from a minimal

amount of diethyl ether. The solid was collected under argon and dried

in vacuo (10^ꢁ3mbar) to give complex 32 as a yellow solid (7.66 g, 66%).

¹H NMR (400 MHz, CD₂Cl₂): d = 7.55 (d, J = 6.8 Hz, 18H), 7.35 (t, J =

7.5 Hz, 9H), 7.19 (t, J = 7.5 Hz, 18H), 6.06 (d, J = 8.8 Hz, 2H), 5.19 (d, J

= 8.8 Hz, 2H), 3.62 ppm (s, 3H); ¹³C NMR (100 MHz, CD₂Cl₂): d =

&

16

&

www.chemeurj.org

ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 0000, 00, 0 – 0

ÝÝ

These are not the final page numbers!

Silyloxy-Based Alkyne Metathesis Catalysts

FULL PAPER

Table 4. Evaluation of different alkyne metathesis catalysts in representative RCAM reactions; yields of isolated products [%].

Entry

Substrate

Product

21a^[a]

27a^[b]

34a^[d]

37^[b]

38^[f]

1

69

97

90

78^[e]

69^[e]

2

81

86

73

92

81

3

4

91

73

95

50^[f,g]

86 (R = Ac)^[f,h]

5

6

79 (R = MOM)^[f,h]

7

80^[7]

94^[7]

8

quant.^[c]

98

9

87^[9]

[a] 21a (10 mol%), MnCl₂(10 mol%), toluene, 808C, 30 min, then addition of the substrate and reaction at 808C in the absence of MS 5 ꢀ. [b] 27a

(2 mol%) or 37 (5 mol%), toluene, ambient temperature, MS 5 ꢀ. [c] Quantitative conversion but the isolated product contained silicon containing im-

purities derived from the Ph₃SiO ligands which could not be removed by flash chromatography. [d] 34a (5 mol%), MnCl₂(5 mol%), toluene, 808C, 30

min, then addition of the substrate and reaction in the presence of MS 5 ꢀ at ambient temperature. [e] Plus ca. 10% of a dimeric product. [f] Using

10 mol% of the catalyst. [g] Plus 18% of a dimeric product. [h] Complex 28b at 808C.

Chem. Eur. J. 2012, 00, 0 – 0

ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

&

17

&

ÞÞ

These are not the final page numbers!

A. Fꢁrstner et al.

300.5, 158.6, 140.0, 136.2, 135.6, 130.9, 130.2, 128.2, 112.0, 55.4 ppm; ele-

mental analysis (%) calcd for C₆₂H₅₂MoO₄Si₃: C 71.51, H 5.03; found: C

71.38, H 4.97.

[18] For complex 11 and related alkylidynes, see: a) S. Beer, K. Brand-

horst, C. G. Hrib, X. Wu, B. Haberlag, J. Grunenberg, P. G. Jones,

M. Tamm, Organometallics 2009 , 28 , 153 4 –1 545; b) S. Beer, C. G.

Hrib, P. G. Jones, K. Brandhorst, J. Grunenberg, M. Tamm, Angew.

Chem. 2007 , 119 , 904 7 –9 051; c) B. Haberlag, X. Wu, K. Brandhorst,

J. Grunenberg, C. G. Daniliuc, P. G. Jones, M. Tamm, Chem. Eur. J.

2010 , 16 , 886 8 –8 877; d) X. Wu, C. G. Daniliuc, C. G. Hrib, M.

Tamm, J. Organomet. Chem. 2011 , 696 , 414 7 –4 151.

Acknowledgements

[19] For complex 12 and related alkylidynes, see: a) I. A. Weinstock,

R. R. Schrock, W. M. Davis, J. Am. Chem. Soc. 1991 , 113 , 13 5 –1 44;

b) M. Chabanas, A. Baudouin, C. Copꢄret, J.-M. Basset, J. Am.

Chem. Soc. 2001 , 123 , 206 2 –2 063.

[20] For complex 13 and related alkylidynes, see: a) L. G. McCullough,

R. R. Schrock, J. C. Dewan, J. C. Murdzek, J. Am. Chem. Soc. 1985 ,

107 , 598 7 –5 998; b) L. G. McCullough, R. R. Schrock, J. Am. Chem.

Soc. 1984 , 106 , 406 7 –4 068.

Generous financial support by the MPG, the Fonds der Chemischen In-

dustrie and SusChemSys (Ziel 2 Programm NRW 2007–2013) is gratefully

acknowledged. We thank Dr. M. Bindl, Dr. E. Heilmann, Dr. D. A.

Clark, Dr. V. Hickmann for preliminary studies, Dr. K. Micoine, Dr. R.

Mariz, K. Lehr, K. Radkowski and A. Martꢃn-Lasanta for selected appli-

cations, and all analytical departments of our Institute for their excellent

support of our program.

[21] For the use of complex 14 in alkyne metathesis, see: a) A. Fꢁrstner,

C. Mathes, C. W. Lehmann, J. Am. Chem. Soc. 1999 , 121 , 945 3 –

9454; b) A. Fꢁrstner, C. Mathes, C. W. Lehmann, Chem. Eur. J.

2001 , 7 , 529 9 –5 317.

[1] J. Heppekausen, R. Stade, R. Goddard, A. Fꢁrstner, J. Am. Chem.

Soc. 2010 , 132 , 1104 5 –1 1057.

[2] M. Bindl, R. Stade, E. K. Heilmann, A. Picot, R. Goddard, A. Fꢁrst-

ner, J. Am. Chem. Soc. 2009 , 131 , 946 8 –9 470.

[3] Reviews on alkyne metathesis: a) A. Fꢁrstner, P. W. Davies, Chem.

Commun. 2005 , 230 7 –2 320; b) W. Zhang, J. S. Moore, Adv. Synth.

Catal. 2007 , 349 , 9 3 –1 20; c) R. R. Schrock, C. Czekelius, Adv. Synth.

Catal. 2007 , 349 , 5 5 –7 7; d) A. Mortreux, O. Coutelier, J. Mol. Catal.

A Chem. 2006 , 254 , 9 6 –1 04; e) X. Wu, M. Tamm, Beilstein J. Org.

Chem. 2011 , 7 , 8 2 –9 3.

[4] a) A. Fꢁrstner, Chem. Commun. 2011 , 47 , 650 5 –6 511; b) A. Fꢁrst-

ner, Angew. Chem. 2000 , 112 , 314 0 –3 172; Angew. Chem. Int. Ed.

2000, 39, 3012–3043.

[22] For selected applications of 14 to advanced organic synthesis, see:

a) A. Fꢁrstner, K. Grela, C. Mathes, C. W. Lehmann, J. Am. Chem.

Soc. 2000 , 122 , 1179 9 –1 1805; b) A. Fꢁrstner, K. Radkowski, J. Gra-

bowski, C. Wirtz, R. Mynott, J. Org. Chem. 2000 , 65 , 875 8 –8 762;

c) A. Fꢁrstner, C. Mathes, K. Grela, Chem. Commun. 2001 , 105 7 –

1059; d) A. Fꢁrstner, L. Turet, Angew. Chem. 2005 , 117 , 352 8 –3 532;

Angew. Chem. Int. Ed. 2005 , 44 , 346 2 –3 466; e) A. Fꢁrstner, M.

Bindl, L. Jean, Angew. Chem. 2007 , 119 , 943 5 –9 438; Angew. Chem.

Int. Ed. 2007 , 46 , 927 5 –9 278; f) A. Fꢁrstner, D. De Souza, L. Turet,

M. D. B. Fenster, L. Parra-Rapado, C. Wirtz, R. Mynott, C. W. Leh-

mann, Chem. Eur. J. 2007 , 13 , 11 5 –1 34; g) A. Fꢁrstner, D. Kirk,

M. D. B. Fenster, C. Aꢅssa, D. DeSouza, C. Nevado, T. Tuttle, W.

Thiel, O. Mꢁller, Chem. Eur. J. 2007 , 13 , 13 5 –1 49; h) A. Fꢁrstner,

O. Larionov, S. Flꢁgge, Angew. Chem. 2007 , 119 , 564 1 –5 644;

Angew. Chem. Int. Ed. 2007 , 46 , 554 5 –5 548; i) A. Fꢁrstner, S.

Flꢁgge, O. Larionov, Y. Takahashi, T. Kubota, J. Kobayashi, Chem.

Eur. J. 2009 , 15 , 401 1 –4 029; j) M. Bindl, L. Jean, J. Herrmann, R.

Mꢁller, A. Fꢁrstner, Chem. Eur. J. 2009 , 15 , 1231 0 –1 2319.

[23] For complex 15 and related alkylidynes, see: a) W. Zhang, S. Kraft,

J. S. Moore, J. Am. Chem. Soc. 2004 , 126 , 32 9 –3 35; b) W. Zhang, S.

Kraft, J. S. Moore, Chem. Commun. 2003 , 83 2 –8 33; c) K. Jyothish,

W. Zhang, Angew. Chem. 2011 , 123 , 349 7 –3 500; Angew. Chem. Int.

Ed. 2011 , 50 , 343 5 –3 438; d) for a study revealing the subtle influ-

ence of various phenols additives on the activity of such alkylidynes

in living ring-opening alkyne polymerizations, see: F. R. Fischer, C.

Nuckolls, Angew. Chem. 2010 , 122 , 741 5 –7 418; Angew. Chem. Int.

Ed. 2010 , 49 , 725 7 –7 260.

[5] B. J. Smith, G. A. Sulikowski, Angew. Chem. 2010 , 122 , 164 3 –1 646;

Angew. Chem. Int. Ed. 2010 , 49 , 159 9 –1 602.

[6] M. G. Nilson, R. L. Funk, Org. Lett. 2010 , 12 , 491 2 –4 915.

[7] a) V. Hickmann, M. Alcarazo, A. Fꢁrstner, J. Am. Chem. Soc. 2010 ,

132 , 1104 2 –1 1044; b) V. Hickmann, A. Kondoh, B. Gabor, M. Al-

carazo, A. Fꢁrstner, J. Am. Chem. Soc. 2011 , 133 , 1347 1 –1 3480.

[8] K. Micoine, A. Fꢁrstner, J. Am. Chem. Soc. 2010 , 132 , 1406 4 –1 4066.

[9] S. Benson, M.-P. Collin, A. Arlt, B. Gabor, R. Goddard, A. Fꢁrstner,

Angew. Chem. 2011 , 123 , 889 8 –8 903; Angew. Chem. Int. Ed. 2011 ,

50 , 873 9 –8 744.

[10] K. Lehr, R. Mariz, L. Leseurre, B. Gabor, A. Fꢁrstner, Angew.

Chem. 2011 , 123 , 1157 5 –1 1579; Angew. Chem. Int. Ed. 2011 , 50 ,

1137 3 –1 1377.

[11] M. Fouchꢄ, L. Rooney, A. G. M. Barrett, J. Org. Chem. 2012, 77,

3060–3070.

[12] T. J. Katz, J. McGinnis, J. Am. Chem. Soc. 1975 , 97 , 159 2 –1 594.

[13] a) A. Fꢁrstner, G. Seidel, Angew. Chem. 1998 , 110 , 175 8 –1 760;

Angew. Chem. Int. Ed. 1998 , 37 , 173 4 –1 736; b) A. Fꢁrstner, O.

Guth, A. Rumbo, G. Seidel, J. Am. Chem. Soc. 1999 , 121 , 1110 8 –

11113.

[14] For comprehensive treatises of Schrock alkylidynes, see: a) R. R.

Schrock, Chem. Rev. 2002 , 102 , 14 5 –1 79; b) R. R. Schrock, Angew.

Chem. 2006 , 118 , 383 2 –3 844; Angew. Chem. Int. Ed. 2006 , 45 , 374 8 –

3759; c) R. R. Schrock, J. Chem. Soc. Dalton Trans. 2001 , 254 1 –

2550.

[24] For complex 16 and related alkylidynes, see: a) Y.-C. Tsai, P. L. Di-

aconescu, C. C. Cummins, Organometallics 2000 , 19 , 526 0 –5 262;

b) J. M. Blackwell, J. S. Figueroa, F. H. Stephens, C. C. Cummins,

Organometallics 2003 , 22 , 335 1 –3 353.

[25] Phenolates likely serve as ligands to the structurally still elusive

alkyne metathesis catalyst generated in situ in the Mortreux proto-

col from [Mo(CO)₆] (or related molybdenum sources) and (substi-

tuted) phenol at high temperatures. For leading references, see:

a) A. Mortreux, M. Blanchard, J. Chem. Soc. Chem. Commun. 1974 ,

78 6 –7 87; b) A. Mortreux, F. Petit, M. Blanchard, Tetrahedron Lett.

1978 , 19 , 496 7 –4 968; c) D. Villemin, P. Cadiot, Tetrahedron Lett.

1982 , 23 , 513 9 –5 140; d) N. Kaneta, T. Hirai, M. Mori, Chem. Lett.

1995 , 62 7 –6 28; e) J. A. K. Du Plessis, H. C. M. Vosloo, J. Mol. Catal.

1991, 65, 51–54; f) N. G. Pschirer, U. H. F. Bunz, Tetrahedron Lett.

1999 , 40 , 248 1 –2 484; g) A. Fꢁrstner, A. Rumbo, J. Org. Chem. 2000 ,

65 , 260 8 –2 611; h) A. Fꢁrstner, F. Stelzer, A. Rumbo, H. Krause,

Chem. Eur. J. 2002 , 8 , 185 6 –1 871; i) V. Sashuk, J. Ignatowska, K.

Grela, J. Org. Chem. 2004 , 69 , 774 8 –7 751; j) V. Maraval, C. Lepetit,

A.-M. Caminade, J.-P. Majoral, R. Chauvin, Tetrahedron Lett. 2006 ,

47 , 215 5 –2 159.

[15] For a general discussion of alkoxides as privileged ligands in alkyne

and alkene metathesis, see: R. R. Schrock, Polyhedron 1995 , 14 ,

317 7 –3 195.

[16] For complex 9 and related alkylidynes, see: a) J. H. Wengrovius, J.

Sancho, R. R. Schrock, J. Am. Chem. Soc. 1981 , 103 , 393 2 –3 934;

b) R. R. Schrock, D. N. Clark, J. Sancho, J. H. Wengrovius, S. M.

Rocklage, S. F. Pedersen, Organometallics 1982 , 1 , 164 5 –1 651;

c) S. F. Pedersen, R. R. Schrock, M. R. Churchill, H. J. Wasserman,

J. Am. Chem. Soc. 1982 , 104 , 680 8 –6 809; d) J. H. Freudenberger,

R. R. Schrock, M. R. Churchill, A. L. Rheingold, J. W. Ziller, Orga-

nometallics 1984 , 3 , 156 3 –1 573; e) M. L. Listemann, R. R. Schrock,

Organometallics 1985 , 4 , 7 4 –8 3.

[17] For complex 10 and related alkylidynes, see: Z. J. Tonzetich, Y. C.

Lam, P. Mꢁller, R. R. Schrock, Organometallics 2007 , 26 , 47 5 –4 77.

[26] For theoretical studies, see: a) J. Zhu, G. Jia, Z. Lin, Organometal-

lics 2006 , 25 , 181 2 –1 819; b) T. Woo, E. Folga, T. Ziegler, Organo-

&

18

&

www.chemeurj.org

ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 0000, 00, 0 – 0

ÝÝ

These are not the final page numbers!

Silyloxy-Based Alkyne Metathesis Catalysts

FULL PAPER

metallics 1993 , 12 , 128 9 –1 298. According to these studies, molybde-

num complexes have higher barriers to metallacycle formation than

the analogous tungsten complexes; the barriers can be lowered by

decreasing the electron-donor abilities of the ancillary ligands.

[27] For previous alkyne metathesis catalysts modified by different sila-

nols, see: a) H. M. Cho, H. Weissman, S. R. Wilson, J. S. Moore, J.

Am. Chem. Soc. 2006, 128, 14742–14743; b) D. Villemin, M.

Hꢄroux, V. Blot, Tetrahedron Lett. 2001 , 42 , 370 1 –3 703.

[28] For alkyne metathesis catalysts grafted onto silica, see: a) O. Coute-

lier, R. M. Gauvin, G. Nowogrocki, J. Trꢄbosc, L. Delevoye, A. Mor-

treux, Eur. J. Inorg. Chem. 2007 , 554 1 –5 547; b) H. M. Cho, H.

Weissman, J. S. Moore, J. Org. Chem. 2008 , 73 , 425 6 –4 258; c) ref.

[19b]; d) H. Weissman, K. N. Plunkett, J. S. Moore, Angew. Chem.

2006 , 118 , 59 9 –6 02; Angew. Chem. Int. Ed. 2006 , 45 , 58 5 –5 88;

e) R. M. Gauvin, O. Coutelier, E. Berrier, A. Mortreux, L. Dele-

voye, J.-F. Paul, A.-S. Mamꢆde, E. Payen, Dalton Trans. 2007 , 312 7 –

3130; f) N. Merle, M. Taoufik, M. Nayer, A. Baudouin, E. Le Roux,

R. M. Gauvin, F. Lefebvre, J. Thivolle-Cazat, J.-M. Basset, J. Orga-

nomet. Chem. 2008 , 693 , 173 3 –1 737.

[29] Following our studies on silanolates as superior ancillary ligands for

alkyne metathesis, other authors adapted this concept, see: a) E. S.

Wiedner, K. J. Gallagher, M. J. A. Johnson, J. W. Kampf, Inorg.

Chem. 2011 , 50 , 593 6 –5 945; b) A. D. Finke, J. S. Moore, Chem.

Commun. 2010 , 46 , 793 9 –7 941; c) S. Lysenko, B. Haberlag, C. G.

Daniliuc, P. G. Jones, M. Tamm, ChemCatChem 2011 , 3 , 11 5 –1 18.

[30] For pioneering studies on the conversion of nitride complexes en-

dowed with fluorinated alkoxides as ancillary ligands into the corre-

sponding alkylidynes, see: a) A. M. Geyer, E. S. Wiedner, J. B. Gary,

R. L. Gdula, N. C. Kuhlmann, M. J. A. Johnson, B. D. Dunietz, J. W.

Kampf, J. Am. Chem. Soc. 2008 , 130 , 898 4 –8 999; b) R. L. Gdula,

M. J. A. Johnson, J. Am. Chem. Soc. 2006 , 128 , 961 4 –9 615; c) A. M.

Geyer, M. J. Holland, R. L. Gdula, J. E. Goodman, M. J. A. Johnson,

J. W. Kampf, J. Organomet. Chem. 2012 , 70 8 –7 09 , 1 –9.

[31] a) A. Mayr, G. A. McDermott, J. Am. Chem. Soc. 1986 , 108 , 54 8 –

549; b) A. Mayr, G. A. McDermott, A. M. Dorries, Organometallics

1985 , 4 , 60 8 –6 10; c) G. A. McDermott, A. M. Dorries, A. Mayr, Or-

ganometallic 1987 , 6 , 92 5 –9 31; d) A. Mayr, H. Hoffmeister, Adv.

Organomet. Chem. 1991, 32, 227–324.

[32] E. O. Fischer, A. Maasbçl, Chem. Ber. 1967 , 100 , 244 5 –2 456.

[33] We assume that adventitious moisture was the proton source for the

formation of free Ph₃SiOH during the extended period it took for

the crystals to be formed from the crude reaction mixture.

[34] a) O. W. Steward, D. R. Fussaro, J. Organomet. Chem. 1977 , 129 ,

C2 8 –C 32; b) it is of note, however, that the pK_avalues of silanols

determined by titration methods differ quite significantly from those

determined by NMR; for a comprehensive discussion, see: P. D.

Lickiss, Adv. Inorg. Chem. 1995, 42, 147–262.

[35] A crystallographic analysis of complex 26a has already been report-

ed in ref. [1]. Although the structure was largely similar to the one

described herein, the Mo-O-Si bond angles V (177.8(3), 151.8(2),

149.0(3), 134.5(2)8) differ from the ones observed in the current

study (167.42(11), 148.36(11), 138.77(10), 130.95(10)8). These subtle

differences support the notion that the Mo-O-Si units are rather

floppy.

[40] The concept also pertains to molybdenum alkylidene complexes

serving as catalysts for alkene metathesis, which can also be ren-

dered stable upon complexation with phenanthroline or bipyridine,

see: J. Heppekausen, A. Fꢁrstner, Angew. Chem. 2011 , 123 , 797 5 –

7978; Angew. Chem. Int. Ed. 2011 , 50 , 782 9 –7 832.

[41] However, the alkylidyne unit of the 2,6-dimethylbenzylidyne com-

ꢀ ꢁ

plex is significantly bent (Mo C Ar 170.6(2)8), whereas it is almost

ꢀ ꢁ

linear in its p-CF₃-substituted cousin (Mo C Ar 178.0(3)8).

[42] For example, the reaction of methyl 2-(propynyl)benzoate (shown in

Table 2, entry 4) with complex 8b (5 mol%) and ZnCl₂(10 mol%)

in the presence of MS 5 ꢀ in toluene at ambient temperature was

complete after 6 h reaction time.

[43] E. O. Fischer, A. Ruhs, P. Friedrich, G. Huttner, Angew. Chem. Int.

Ed. Engl. 1977 , 16 , 46 5 –4 66.

[44] a) M. H. Chisholm, K. Folting, D. M. Hoffman, J. C. Huffman, J.

Am. Chem. Soc. 1984 , 106 , 679 4 –6 805; b) M. H. Chisholm, B. K.

Conroy, B. W. Eichhorn, K. Folting, D. M. Hoffman, J. C. Huffman,

N. S. Marchant, Polyhedron 1987 , 6 , 78 3 –7 92.

[45] a) M. H. Chisholm, K. Folting, J. C. Huffmann, I. P. Rothwell, J. Am.

Chem. Soc. 1982 , 104 , 438 9 –4 399; b) M. H. Chisholm, D. M. Hoff-

man, J. C. Huffman, J. Am. Chem. Soc. 1984 , 106 , 680 6 –6 815;

c) M. H. Chisholm, Acc. Chem. Res. 1990 , 23 , 41 9 –4 25.

[46] See ref. [16e] and the following: a) R. R. Schrock, M. L. Listemann,

L. G. Sturgeoff, J. Am. Chem. Soc. 1982 , 104 , 429 1 –4 293; b) I. A.

Latham, L. A. Sita, R. R. Schrock, Organometallics 1986 , 5 , 150 8 –

1510.

[47] a) S. Fantacci, N. Re, M. Rosi, A. Sgamellotti, M. F. Guest, P. Sher-

wood, C. Floriani, J. Chem. Soc. Dalton Trans. 1997 , 384 5 –3 852;

b) M. H. Chisholm, E. R. Davidson, M. Pink, K. B. Quinlan, Chem.

Commun. 2002, 2270–2271.

[48] For the formation of other complexes from ditungsten precursors

and acetylenes, see: a) F. A. Cotton, W. Schwotzer, E. S. Sham-

shoum, Organometallics 1983 , 2 , 116 7 –1 171; b) F. A. Cotton, W.

Schwotzer, E. S. Shamshoum, Organometallics 1983 , 2 , 134 0 –1 343.

[49] It is of note, however, that dimerization of an alkylidyne to a dime-

tallatetrahedrane complex is not necessarily a deactivation pathway

since it may be reversible as suggested by the results of Chisholm et

ꢀ

al.^[44]Moreover, the cleavage of [(tBuO)₃W W

ACTHNUTRGNE(UNG OtBu)₃] by alkynes

is an efficient way to prepare the catalytically active Schrock alkyli-

[16e]

ꢀ

dyne complexes such as [(tBuO)₃W CtBu],

most likely via di-

tungstatetrahedranes as the key intermediates as suggested by theo-

retical studies.^[47]In the present case, however, the dimerization was

accompanied by loss of one siloxide and incorporation of a chloride

instead, which will likely impede the retro-reaction.

[50] a) I. Feinstein-Jaffe, S. F. Pedersen, R. R. Schrock, J. Am. Chem.

Soc. 1983 , 105 , 717 6 –7 177; b) I. Feinstein-Jaffe, J. C. Dewan, R. R.

Schrock, Organometallics 1985 , 4 , 118 9 –1 193.

[51] In fact, crystals of two other complexes could be collected in tiny

amounts (<5%), in which oxo-bridged dimeric units are further ag-

gregated by K⁺counterion. As the crystal data were of insufficient

quality and had to be treated with the SQUEEZE program, we re-

frain from a further discussion. However, the connectivity is unam-

biguous and the gross structure of one such example (complex 44)

depicted in the Supporting Information.

[52] Complexes of this type have previously been prepared by entirely

different routes, see the following and references therein: a) A.

Thapper, J. P. Donahue, K. B. Musgrave, M. W. Willer, E. Nordland-

er, B. Hedman, K. O. Hodgson, R. H. Holm, Inorg. Chem. 1999 , 38 ,

410 4 –4 114; b) H. Arzoumanian, R. Bakhtchadjian, G. Agrifoglio,

R. Atencio, A. BriceÇo, Transition Met. Chem. 2008 , 33 , 94 1 –9 51.

[53] For deprotiometallacyclobutadiene complexes of tungsten, see:

a) L. G. McCullough, M. L. Listemann, R. R. Schrock, M. R.

Churchill, J. W. Ziller, J. Am. Chem. Soc. 1983 , 105 , 672 9 –6 730;

b) M. R. Churchill, J. W. Ziller, J. Organomet. Chem. 1985 , 281 ,

23 7 –2 48.

[36] The bias of molybdenum alkylidynes for ate-complex formation has

ꢀ

early precedent in the preparation of [Et₄N][Cl₄Mo CtBu] described

by Schrock and coworkers, see ref. [20a].

[37] Note that the complex has only phenyl protons; as the number of

coordinated Et₂O molecules seems variable, the integral of the aro-

matic region cannot be securely calibrated.

[38] Only one of the two independent molecules in the unit cell is

shown. The complete unit cell is depicted in the Supporting Infor-

mation.

[39] For general discussions, see: a) C. Krempner, Eur. J. Inorg. Chem.

2011 , 168 9 –1 698; b) P. T. Wolczanski, Polyhedron 1995 , 14 , 333 5 –

3362.

Received: February 24, 2012

Published online: && &&, 0000

Chem. Eur. J. 2012, 00, 0 – 0

ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

&

19

&

ÞÞ

These are not the final page numbers!

A. Fꢁrstner et al.

Active yet tamed: Molybdenum benzy-

lidyne complexes endowed with triar-

ylsilanolates as ancillary ligands are

superbly active and exquisitely selec-

tive alkyne metathesis catalysts, but

can be rendered air-stable by reversi-

ble complexation with 1,10-phenan-

throline or 2,2’-bipyridine (see, e.g.,

figure). A systematic molecular editing

of their basic structural motif furnished

a “library” of such catalysts and gave

detailed insights into the parameters

determining their favorable application

profile.

Ring-Closing Metathesis

J. Heppekausen, R. Stade, A. Kondoh,

G. Seidel, R. Goddard,

A. Fꢀrstner* ................... &&&& — &&&&

Optimized Synthesis, Structural Inves-

tigations, Ligand Tuning and Synthetic

Evaluation of Silyloxy-Based Alkyne

Metathesis Catalysts

&

20

&

www.chemeurj.org

ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 0000, 00, 0 – 0

ÝÝ

These are not the final page numbers!

Article Doi

Optimized synthesis, structural investigations, ligand tuning and synthetic evaluation of silyloxy-based alkyne metathesis catalysts

DOI: 10.1002/chem.201200621

Source and publish data:

Authors:

Article abstract of DOI:10.1002/chem.201200621

Full text of DOI:10.1002/chem.201200621

Products guided by the article

R&D Labs maybe for 1335098-34-4

Relevant to this article

Hot Product