S-Block cooperative catalysis: Alkali metal magnesiate-catalysed cyclisation of alkynols

Article abstract of DOI:10.1039/c9sc01598a

Mixed s-block metal organometallic reagents have been successfully utilised in the catalytic intramolecular hydroalkoxylation of alkynols. This success has been attributed to the unique manner in which these reagents can overcome the challenges of the reaction: namely OH activation and coordination to and then addition across a CC bond. In order to optimise the reaction conditions and to garner vital catalytic system requirements, a series of alkali metal magnesiates were enlisted for the catalytic intramolecular hydroalkoxylation of 4-pentynol. In a prelude to the main investigation, the homometallic magnesium dialkyl reagent MgR₂ (where R = CH₂SiMe₃) was utilised. This reagent was unsuccessful in cyclising the alcohol into 2-methylenetetrahydrofuran 2a or 5-methyl-2,3-dihydrofuran 2b, even in the presence of multidentate Lewis donor molecules such as N,N,N′,N′′,N′′-pentamethyldiethylenetriamine (PMDETA). Alkali metal magnesiates M^IMgR₃ (when M^I = Li, Na or K) performed the cyclisation unsatisfactorily both in the absence/presence of N,N,N′,N′-tetramethylethylenediamine (TMEDA) or PMDETA. When higher-order magnesiates (i.e., M^I₂MgR₄) were employed, in general a marked increase in yield was observed for M^I = Na or K; however, the reactions were still sluggish with long reaction times (22-36 h). A major improvement in the catalytic activity of the magnesiates was observed when the crown ether molecule 15-crown-5 was combined with sodium magnesiate Na₂MgR₄(TMEDA)₂ furnishing yields of 87% with 2a:2b ratios of 95:5 after 5 h. Similar high yields of 88% with 2a:2b ratios of 90:10 after 3 h were obtained combining 18-crown-6 with potassium magnesiate K₂MgR₄(PMDETA)₂. Having optimised these systems, substrate scope was examined to probe the range and robustness of 18-crown-6/K₂MgR₄(PMDETA)₂ as a catalyst. A wide series of alkynols, including terminal and internal alkynes which contain a variety of potentially reactive functional groups, were cyclised. In comparison to previously reported monometallic systems, bimetallic 18-crown-6/K₂MgR₄(PMDETA)₂ displays enhanced reactivity towards internal alkynol-cyclisation. Kinetic studies revealed an inhibition effect of substrate on the catalysts via adduct formation and requiring dissociation prior to the rate limiting cyclisation step.

Full text of DOI:10.1039/c9sc01598a

Chemical
Science
View Article Online
View Journal
EDGE ARTICLE
s-Block cooperative catalysis: alkali metal
magnesiate-catalysed cyclisation of alkynols†
Cite this: DOI: 10.1039/c9sc01598a
b
Michael Fairley,a Laia Davin,a Alberto Hernan-Gomez,a Joaquın Garcıa-A´lvarez,
*
All publication charges for this article
have been paid for by the Royal Society
of Chemistry
´
a
´
´
´
a
*
*
Charles T. O'Hara
and Eva Hevia
Mixed s-block metal organometallic reagents have been successfully utilised in the catalytic intramolecular
hydroalkoxylation of alkynols. This success has been attributed to the unique manner in which these
reagents can overcome the challenges of the reaction: namely OH activation and coordination to and
then addition across a C^C bond. In order to optimise the reaction conditions and to garner vital
catalytic system requirements, a series of alkali metal magnesiates were enlisted for the catalytic
intramolecular hydroalkoxylation of 4-pentynol. In a prelude to the main investigation, the homometallic
magnesium dialkyl reagent MgR2 (where R ¼ CH2SiMe3) was utilised. This reagent was unsuccessful in
cyclising the alcohol into 2-methylenetetrahydrofuran 2a or 5-methyl-2,3-dihydrofuran 2b, even in the
presence of multidentate Lewis donor molecules such as N,N,N0,N00,N00-pentamethyldiethylenetriamine
(PMDETA). Alkali metal magnesiates MIMgR3 (when MI ¼ Li, Na or K) performed the cyclisation
unsatisfactorily both in the absence/presence of N,N,N0,N0-tetramethylethylenediamine (TMEDA) or
PMDETA. When higher-order magnesiates (i.e., MI2MgR4) were employed, in general a marked increase
in yield was observed for MI ¼ Na or K; however, the reactions were still sluggish with long reaction
times (22–36 h). A major improvement in the catalytic activity of the magnesiates was observed when
the crown ether molecule 15-crown-5 was combined with sodium magnesiate Na2MgR4(TMEDA)2
furnishing yields of 87% with 2a : 2b ratios of 95 : 5 after 5 h. Similar high yields of 88% with 2a : 2b ratios
of 90 : 10 after
3
h
were obtained combining 18-crown-6 with potassium magnesiate
K2MgR4(PMDETA)2. Having optimised these systems, substrate scope was examined to probe the range
and robustness of 18-crown-6/K2MgR4(PMDETA)2 as a catalyst. A wide series of alkynols, including
terminal and internal alkynes which contain a variety of potentially reactive functional groups, were
cyclised. In comparison to previously reported monometallic systems, bimetallic 18-crown-6/
K2MgR4(PMDETA)2 displays enhanced reactivity towards internal alkynol-cyclisation. Kinetic studies
revealed an inhibition effect of substrate on the catalysts via adduct formation and requiring dissociation
prior to the rate limiting cyclisation step.
Received 1st April 2019
Accepted 26th April 2019
DOI: 10.1039/c9sc01598a
rsc.li/chemical-science
Mg–H or Mg–X exchange processes) as well as anionic transfer
agents to unsaturated organic molecules.9,10 Some key break-
Introduction
Since rst reported by Wittig in 1951,1 alkali metal magnesiates
have evolved from mere curiosities to a new family of versatile
organometallic reagents which nds widespread applications in
organic synthesis.2–7 By engaging metal–metal cooperativities,
these bimetallic systems can offer superior chemo- and regio-
selectivities and/or functional group tolerances to those of their
monometallic counterparts.8 Most reactivity studies have
focused on using these reagents as metallating reagents (via
throughs in the eld which highlight the synergistic power of
these bimetallic partnerships include Knochel's Turbo Grignard
reagents RMgCl$LiCl (R ¼ alkyl) which allow the functionali-
sation of a wide range of organic molecules via Mg–halogen
exchange2 or Mulvey and O'Hara's template metallation, where
mixed Na/Mg macrocyclic bases enable remote di-magnesiation
of aromatic molecules.4,5 Despite the increasing interest that s-
block metal catalysis is currently attracting,9,11–30 catalytic
applications of these bimetallic systems have hardly been
researched.31 Breaking new ground in this area, we recently
aWestCHEM, Department of Pure and Applied Chemistry, University of Strathclyde,
32
reported sodium tris(alkyl)magnesiate NaMg(CH2SiMe3)3 as
Glasgow, G1 1XL, UK. E-mail: charlie.ohara@strath.ac.uk
an efficient precatalyst for hydroamination of a variety of car-
bodiimides and isocyanates.33,34 Operating synergistically, this
bimetallic precatalyst displays an enhanced catalytic ability
compared to those found for its homometallic components
b
´
´
´
´
Departamento de Quımica Organica e Inorganica, Facultad de Quımica, Universidad
de Oviedo, E-33071 Oviedo, Spain
† Electronic supplementary information (ESI) available: General experimental
procedures, additional kinetic plots, NMR spectra of catalytic reactions, and
characterization data for new compounds (PDF). See DOI: 10.1039/c9sc01598a
This journal is © The Royal Society of Chemistry 2019
Chem. Sci.

View Article Online
Chemical Science
Edge Article
NaCH2SiMe3 and Mg(CH2SiMe3)2, allowing hydroaminations to
take place under very mild reaction conditions (mostly at
ambient temperature) in nearly quantitative yields.
Going beyond these proof of concept studies, here we
investigate the catalytic ability of alkali metal magnesiates to
promote intramolecular hydroalkoxylation reactions of alkynyl
alcohols. Being 100% atom efficient,35 this is a highly desirable
route to provide straightforward access to a series of syntheti-
cally signicant O-heterocycles such as hydrofurans, pyrans,
benzofurans, etc. that are fundamentally important to many
natural products36 and pharmaceutical37 compounds. However,
the large bond enthalpy of typical O–H bonds and the modest
reactivity of electron-rich olens with nucleophiles can make
these transformations especially challenging.38
A catalogue of catalysts has been developed for the cyclisa-
tion of alkynols, including various transition metal complexes
as well as alkaline earth and f-block compounds. Transition
metals previously used to promote this transformation are
mostly precious metals39–48 but a few others such as copper,49
iron50 and tungsten51,52 have also been investigated. Different
metal complexes have been found to offer different product
selectivities, as there are two distinct cyclisation routes for
hydroalkoxylation of alkynols (Scheme 1). For instance, tung-
sten, ruthenium and molybdenum complexes have been used to
produce endocyclised enol ether products. This occurs due to
the proposed formation of an h2-vinylidene metal complex
intermediate (Scheme 1, I), which is then susceptible to nucle-
ophilic addition to generate a Fischer oxacarbene (Scheme 1, II)
that ultimately yields the endocyclised product (Scheme 1,
III).53–56 However more commonly, where no electronically
biased intermediate is formed, exocyclisation occurs due to the
more favourable approach of the nucleophile with respect to the
p-system.57,58
Scheme
2
Hydroalkoxylation of 4-pentynol catalysed by heavier
alkaline earth metal amides.62
study, several papers focusing on alkali metal and alkaline earth
catalysts have been published. Liu has shown that potassium
tert-butoxide is an effective and selective catalyst for cyclisation
of aromatic alkynylamines and alkynols,63 albeit with the need
for elevated temperatures and highly polar solvents. Utilising
similar reaction conditions, Wang employed potassium
carbonate to prepare indole/pyrrole-fused 1,4-oxazines.64 Baire
has also recently reported that cycloisomerisation of cis-6-
hydroxy- and cis-6-acyloxyhex-2-en-4-ynals to 2-acylfurans and 2-
(1-acyloxyalkenyl)furans is achievable using calcium catalysis.65
In contrast, magnesium-based reagents have not followed suit
by showing very little promise to catalyse these types of trans-
formations; whereas the catalytic properties of s-block bime-
tallic complexes remain unexplored in this context.
Expanding the scope of s-block cooperative catalysis, here we
report the rst catalytic applications of alkali metal magnesiates
to promote cyclisation reactions, focusing on intramolecular
hydroalkoxylation of alkynols. Benetting from the enhanced
metallating and nucleophilic abilities of these synergistic
systems, we envisaged that they could play a dual role in over-
coming the main challenges encountered in these trans-
formations, facilitating not only OH activation, but also the
required addition across C^C bonds. Combining kinetic
experiments with reactivity studies, here we provide informative
mechanistic insights on how cooperative effects can be maxi-
mised as well as on the key role that each metal plays in these
novel magnesiate-catalysed transformations.
Besides the numerous examples of transition metal catalysts,
the use of f- and s-block catalysts for this reaction has been
investigated. In terms of f-block chemistry, Marks has utilised
lanthanide amide catalysts,59,60 while Otero and Lara-Sanchez
´
Results and discussion
have employed heteroscorpionate rare earth catalysts.61 For s-
block catalysis, Hill62 has used calcium, strontium and barium Intrigued by the lack of a magnesium amide-catalysed reaction
amides (Scheme 2).
from the series of alkaline earth metals investigated by Hill, our
These particular studies highlight that transformations rst reaction used the neutral bis(alkyl)magnesium reagent
proceed selectively via exocyclisation, although with alkaline Mg(CH2SiMe3)2, containing thermally stable monosilyl groups,
earth metal amides two different product isomers were as a potential pre-catalyst. Using 4-pentynol (1) (as previously
observed, an internal and external alkene. Since Hill's initial employed by Marks and Hill) for benchmarking with a catalytic
quantity of the magnesium reagent, no reaction was observed
ꢀ
even when the mixture was heated to 75 C for 36 h (Table 1,
entry 1).
This raises the question, “why is magnesium unable to
mediate this cyclisation whereas calcium can?” Major contrib-
utory factors are likely due to the signicant difference in ionic
radii between the two metal cations and the low p-philicity of
magnesium.66 DFT calculations were carried out by Marks
investigating the quantity of alkynol molecules which could
coordinate to the active lanthanide catalyst when La(HMDS)3 is
added as a pre-catalyst.67 Interestingly these calculations
showed that the most likely catalytically active species is a La
Scheme 1 Conversion of 4-pentynol to respective five-membered
exo- and six-membered endocyclised products. The latter conversion
proceeds via vinylidene (I) and Fischer oxacarbene (II) intermediates.
ion coordinated to three alkynoxide ligands, binding via the
Chem. Sci.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Edge Article
Chemical Science
Table 1 Optimisation of the catalyst for cyclisation of 4-pentynol (1)
Isomer ratio
(2a : 2b)
Entry
Pre-catalyst (R ¼ CH2SiMe3)
Time, h
Yieldb (%)
a
1
2
3
4
5
6
7
8
MgR2
36
36
36
16
36
36
36
36
36
10
36
36
22
5
0
0
—
—
MgR2 + 18-c-6 (ꢁPMDETA)a,d
KRa
11
70
3
12
13
16
20
77
3
83
91
87
88
94
99 : 1
93 : 7
KR + 18-c-6c
LiMgR3
a
d
—
a
NaMgR3
97 : 3
96 : 4
98 : 2
94 : 6
81 : 19
NaMgR3(TMEDA)a
a
KMgR3
9
KMgR3(PMDETA)a
KMgR3 + 18-c-6a,d
10
11
12
13
14
15
16
b
e
Li2MgR4(TMEDA)2
Na2MgR4(TMEDA)2
—
b
91 : 9
86 : 14
95 : 5
90 : 10
90 : 10
b
K2MgR4(PMDETA)2
Na2MgR4(TMEDA)2 + 15-c-5b,d
K2MgR4(PMDETA)2 + 18-c-6b,d
K2MgR4(TMEDA)2 + 18-c-6b,d
3
3
a
Reactions were performed in a Young's cap NMR tube, using 0.5 mmol substrate (4-pentynol) (1) and 0.025 mmol (5 mol%) pre-catalyst.
b
Reactions were performed in a Young's cap NMR tube, using 0.6 mmol (1.2 eq.) substrate (4-pentynol) (1) and 0.025 (5 mol%) pre-catalyst.
This additional 0.1 mmol (20 mol%/0.2 eq.) 1 was employed to convert the pre-catalyst to the ‘active catalyst’ (vide infra). Calculated from 1H
NMR spectroscopic data by integration against an internal standard (10 mol% 1,2,3,4-tetraphenylnaphthalene). A stoichiometric quantity of
crown ether co-catalyst used according to the alkali metal [i.e., 10 mol% 18-c-6 for 5 mol% K2MgR4(PMDETA)2]. Due to low yields obtained
c
d
e
and the inherent error within the measurement, an isomer ratio is not reported.
oxide anion as well as forming a close-range p-interaction with promote a dual-activation where a magnesium alkoxide forms,
the alkyne unit. An additional substrate molecule coordinates thus preparing one end of the substrate for cyclisation while the
to the metal via its O atom, affording a coordinatively saturated alkali metal could then interact with the p-system of the
species. Furthermore, it appears that these La–p-interactions carbon–carbon triple bond. As alluded to previously, this latter
activate the alkyne for insertion into the lanthanum–oxygen feature is essential as the magnesium alone cannot provide this
bond and additionally position the alkynoxide perfectly for interaction. LiMg(CH2SiMe3)3 (Table 1, entry 5) was the rst
cyclisation. Bearing this in mind, the lack of reactivity observed magnesiate tested but it struggled to promote the reaction,
when using Mg(CH2SiMe3)2 could be due to the magnesium presumably again due to size limitations of the lithium centre
centre not being large enough to simultaneously bind to the (i.e., it is sterically blocked by the ligand set). Moving to sodium
oxide anion and be in close proximity to the p-density of the (Table 1, entry 6), disappointingly, this sodium magnesiate only
carbon–carbon triple bond.
promoted the reaction modestly [about as effectively as
We next focused on testing homometallic alkali metal K(CH2SiMe3)]. When the potassium magnesiate KMg(CH2-
organometallics, which, being signicantly more polar than SiMe3)3 (Table 1, entry 8) was employed, the yield (16%) was
their Mg counterparts, have recently found applications in comparable to that observed for K(CH2SiMe3) (11%). So, in
hydrophosphination catalysis.68 Li(CH2SiMe3) and Na(CH2- summary, lower order magnesiate reagents are generally poor at
SiMe3) were employed in the same reaction; however, the catalysing the hydroalkoxylation/cyclisation reaction.
conversions were poor (<5%) (see ESI†). Using the heavier
In the hope of improving this situation, higher-order (2 : 1,
congener K(CH2SiMe3) (Table 1, entry 3) also results in low alkali metal : Mg ratio) magnesiate reagents were considered.
conversion over a 36 h period (11%). Although the conversion is The Lewis basic proligands, TMEDA (N,N,N0,N0-tetramethyle-
again low, the fact that the reaction was more successful than thylenediamine)
and
PMDETA
(N,N,N0,N00,N00-pentam-
those with Li(CH2SiMe3), Na(CH2SiMe3) and Mg(CH2SiMe3)2 ethyldiethylenetriamine), were employed to completely ll the
can perhaps be attributed to the larger ionic radius of potas- coordination sphere of the alkali metals. Again, starting with
sium versus the other metals allowing it to interact with the the lithium derivative, the higher-order magnesiate (Table 1,
carbon–carbon triple bond, activating it for cyclisation.
entry 11) performed as poorly as the lower-order lithium mag-
Next, we investigated the catalytic ability of a lower-order nesiate in the reaction (3%). However, when Na2MgR4(TMEDA)2
(1 : 1, alkali metal : Mg ratio) magnesiate. It was envisaged was employed (Table 1, entry 12), pleasingly a substantially
that a bimetallic magnesiate could function cooperatively, i.e., more successful conversion of the alcohol to the cyclised
This journal is © The Royal Society of Chemistry 2019
Chem. Sci.

View Article Online
Chemical Science
Edge Article
ꢂ
product (83%) was obtained within 36 h. The reaction efficiency cooperative), that the remaining anionic Mg(OR)3
or
2ꢂ
can be improved further when the potassium reagent K2- Mg(OR)4 fragment, of the now solvent-separated complex,
MgR4(PMDETA)2 is used (91% aer 22 h, Table 1, entry 13). was not able to perform the catalysis on its own. Surprisingly, in
These two higher-order magnesiates display greatly enhanced contrast, the results obtained in these reactions showed
catalytic activity compared to that of homometallic K(CH2- a dramatic improvement in reactivity!
SiMe3). In addition, they continue to have a high isomer selec-
Adding catalytic quantities of 15-crown-5 or 18-crown-6 to
tivity with up to 90% of the external alkene isomer being the sodium and potassium systems respectively dramatically
produced. The remarkable alkali metal effect observed between increased the activity of the magnesiate catalysts, with the
Li2MgR4(TMEDA) and K2MgR4(PMDETA)2 is even more striking sodium higher-order magnesiate now nearing completion in 5 h
when considering that both mixed-metal species are iso- (Table 1, entry 14) and the potassium analogue in 3 h (Table 1,
structural.69,70 At this juncture we reasoned that addition of entry 15), in reactions retaining good selectivity. This reactivity
TMEDA or PMDETA in these reactions inuenced the marked enhancement was also seen with both the monometallic and
increase in reactivity. To test whether these multidentate Lewis lower order potassium magnesiate (Table 1, entries 4 and 10
bases were the key drivers in increasing the yields, we tested the respectively).
lower-order NaMgR3(TMEDA) and KMgR3(PMDETA) (Table 1,
Due to the signicant difference in completion rates of these
entries 7 and 9), but found that adding these multidentate two reactions (5 h vs. 3 h; Table 1, entries 14 and 15), the idea of
2ꢂ
ligands did not enhance reaction performance. With the higher- a completely solvent-separated Mg(OR)4 dianion can be dis-
order magnesiates, the reaction rate using K2MgR4(PMDETA)2 counted as it would be expected that similar rates would be
is unaffected when the bidentate Lewis base TMEDA is used in observed if this dianion was a catalytically active species.
place of its tridentate analogue PMDETA (Table 2, entry 16). Although the crown ether is oen expected to sequester the
Hence K2MgR4(TMEDA)2 displayed the same reaction time and alkali metal there are examples of structures where the alkali
selectivity as its PMDETA variant, demonstrating the reaction metal is still able to maintain p-interactions with the anionic
rate to be unaffected by the choice of Lewis base.
moiety whilst coordinated by the crown ether.71–78 Thus it
To establish whether or not both metals need to be present appears that under these conditions, the addition of the crown
for the magnesiate to function, a catalytic amount of crown ether must tune the coordination environment of K, modifying
ether was added as a co-catalyst. As a crown ether of appropriate its Lewis acidity to generate a more catalytically competent
size is oen able to sequester an alkali metal, the intention here species (vide infra). Therefore, these results suggest that the
was to ascertain, if the alkali metal was captured (non- alkali metal is intimately involved and plays a key role in the
cyclisation reaction. To ascertain whether the 18-crown-6 is
catalytically active, when it is combined with only Mg(CH2-
SiMe3)2 the reaction fails (entry 2).
Summarising Table 1, it is clear that the organomagnesium
compound used in this study is unable to catalytically promote
Table 2 Catalytic reactivity of K2MgR4(PMDETA)2 with the 18-crown-
6 additive towards various structurally diverse alkynols
Entrya,c Substrate (5 mol% cat., C6D6)
Yieldb
cyclisation of alkynols. Homometallic alkali metal compounds
can only poorly promote the reaction; however, when working
cooperatively with a Mg compound the two metals can perform
more efficiently by dual-activating the substrate. There are also
clear trends that higher-order magnesiates perform better than
lower-order magnesiates and that when comparing alkali
metals K > Na [ Li presumably due to a sequential decrease in
size and reduction in p-philicity. Perhaps the most surprising
outcome of the optimisation study was that introducing a crown
ether into higher-order magnesiates appears to greatly enhance
the rate of the cyclisation, whilst maintaining a high selectivity.
The reaction times and selectivity of our system (Table 1, entry
15) compare favourably with those of other catalytic systems. In
terms of typical reaction times (3 h; Table 1, entry 15), this is
similar to that observed by Marks (2.5 h) where only isomer 2a is
produced by employing La[N(SiMe3)2]3.38 Perhaps most perti-
nent to our study, Hill has employed heavy group 2 amides [Ae
99%
(14 : 9 : 77)
1
2
3
86%
89%
4
90%
a
Reactions were performed in a Young's cap NMR tube, using 0.6 mmol {N(SiMe3)2}2 where Ae ¼ Ca, Sr, Ba], where reaction times
(1.2 eq.) substrate (3, 5, 7, or 9) and 0.025 mmol (5 mol%) pre-catalyst
ranged from 2.5–6 h, exhibiting a 2a isomer selectivity of 90–
with 0.1 mmol (20 mol%/0.2 eq.) 1 consumed in converting the pre-
97% (Scheme 2).62 Therefore to re-emphasize, our results show
catalyst to the ‘active catalyst’ (vide infra). Calculated from 1H NMR
b
spectroscopic data by integration against an internal standard that by pairing a magnesium complex with a potassium one, it
c
(10 mol% 1,2,3,4-tetraphenylnaphthalene). A stoichiometric quantity
has been possible to overcome the severe limitations imposed
of crown ether co-catalyst used according to the alkali metal [i.e.,
by a magnesium homometallic catalyst on its own. Through
10 mol% 18-c-6 for 5 mol% K2MgR4(PMDETA)2].
cooperativity, the surprising effect was the generation of
Chem. Sci.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Edge Article
Chemical Science
a potassium magnesiate catalyst which is even more active and functionalised substrate (entry 1, Table 3). This reduction in
equally as selective as the homometallic heavier group 2 reaction rate is also observed, albeit to a lesser extent, with the
congeners.
weaker donating t-butyl group (entry 7, Table 3). When mildly
Having established the optimal conditions for cyclisation of withdrawing halides are incorporated within the substrate
4-pentynol, the scope of the reaction was evaluated. Terminal (entries 3 and 4, Table 3) the reaction is complete in less than
and internal alkynes were tested to assess the range and 1 h. At this juncture, it was deemed necessary to consider
ꢀ
robustness of the catalytic system (Tables 2 and 3). Interest- reaction times at 40 C to ascertain whether any difference in
ingly, by increasing the alkynol chain length by one carbon relative rates was observed. With a Cl substituent (entry 3, Table
atom (5-hexynol, 3, Table 2, entry 1), the major product is iso- 3), the yield was 51% aer 36 h compared to 6% for the non-
merised alkyne (4c), where the carbon–carbon triple bond has substituted alkynol (entry 1, Table 3). Using a F substituent,
moved from a terminal to an internal location (Table 2, entry 1). this yield can be increased to 82% in 24 h (entry 4, Table 3).
It can be envisaged that isomerisation occurs due to the angle of Moving to stronger electron-withdrawing groups, incorporating
ring closure being unfavourable, and hence it will occur prior to triuoromethyl (yield, 93%) brings the reaction time down to
ꢀ
a cyclisation process. Related to these ndings, Tsurugi has only 2 h at 40 C (entry 5, Table 3), and if a cyano substituent
reported that organomagnesium complexes can promote alkyne (yield, 76%) is utilised this is further reduced to 30 min (entry 6,
isomerisations of related compounds.79 Where the alkynol Table 3).
substrate is already held in a manner that is partially aligned to
This set of substrates based on 5-phenylpent-4-yn-1-ol (11)
that of the cyclised product, the cyclisation progresses much show nicely the benet of using a bimetallic system for these
more readily (Table 2, entry 2). This occurs even at ambient cyclisation reactions. With the heavier alkaline earth metal
temperature in 1 h. Interesting selectivity is observed when it amides and lanthanide amides, substrate 11 is more chal-
comes to cyclisation of Z-enynols (7 and 9). In Table 2 entry 3, lenging to cyclise, taking 16 h at 90 ꢀC with Ca(HMDS)2 and
with a terminal enynol (7) employing the optimised conditions about 15 h at 120 C with La(HMDS)2. Marks38 has suggested
ꢀ
we see the production of an aromatic furan product (8). This that the reason for this sluggish reaction time with La(HMDS)3
product differs from the product previously reported using is sterically driven by the phenyl substituent that prevents
heavier alkaline earth59,60,80 or lanthanide amides.38,57,58 With interaction between the metal and the internal alkyne. Using
these catalytic systems the non-aromatic product, 3-methyl-2- a bimetallic system, it appears that this steric clash is somewhat
methylene-2,5-dihydrofuran, is observed, and it is only when circumvented giving rise to a faster reaction time than the other
we move to a transition metal catalyst,44 we see formation of the two systems mentioned.
aromatic furan product.
In addition to reaction scope, kinetic studies have been
In order to further probe the functional group tolerance of carried out on the system. Prime motivations were to determine
the catalytic magnesiate system, a range of internal alkynes the reaction order of the alkynol substrate and catalyst present
were synthesised with different pendant functional groups, in the reaction, with the aim of providing insights into the
including halogens, and electron-donating and electron- reaction mechanism. The substrate chosen for these studies
withdrawing groups. It transpired that no side reactions took was 4-pentynol (1), as this was used for the initial parameter-
place when MeO, F, Cl, CF3 or CN functional groups were isation studies. A plot of concentration of 1 against time
incorporated into the substrate. Looking at 5-phenylpent-4-yn- displays a linear t until half conversion when a rate accelera-
1-ol (11) (Table 3, entry 1) the reaction reached completion in tion is observed [Fig. 1(a)].
1 h at 75 ꢀC yielding only two product isomers. The major
Consistent with substrate inhibition, the plot of [1]2 versus
product being the external Z-alkene 12b and the minor external time shows a good t [Fig. 1(b)]. Further evidence was gathered
E isomer 12a. In the aforementioned studies by Hill,62 a mixture by plotting initial reaction rates at varied alkynol concentrations
of three product isomers was obtained, the external alkene (Fig. S1†), where greater rates were observed as the concentra-
isomers (12a, 12b) and an internal alkene (12c) (note these three tion of 1 was decreased. This negative rst order was observed
products are still exocyclic). In this work, it was also noted that by Hill62 for the transformation employing heavier group 2
the product ratio appeared to have a degree of temperature amide catalysts, where its origin is proposed to be the decoor-
dependence and required signicantly harsher reaction condi- dination of a substrate molecule from the active catalyst being
ꢀ
tions (90–110 C, 16–40 h).
required before cyclisation can occur (Scheme 3).
Despite altering the reaction temperature (reactions per-
When coordination occurs (IV), it appears that the cyclisa-
formed at both 40 and 75 ꢀC), the same isomer ratios were tion process is hindered. As the reaction is inverse rst order in
obtained for all the reactions which tended towards completion the substrate (vide infra) it is assumed that the incoming alkynol
(entries 4–6, Table 3). For entries 1–3, Table 3, the low coordinates to the single Mg centre rather than the K centres,
temperature reactions (40 ꢀC) were relatively less efficient which which are already highly coordinated by bonding to the crown
precluded accurate determination of the product ratio at this ether. In previous studies, the addition of 18-crown-6 has had
temperature. Focusing on the specic substrates reacting at both benecial78,81 and detrimental77 impacts on synthetic
75 ꢀC, when the electron donating methoxy group (entry 2, performance. Here we observe an improvement in catalytic
Table 3) is incorporated onto the 5-phenylpent-4-yn-1-ol scaf- activity, which could perhaps be explained by the 18-crown-6
fold, the reaction rate is dramatically reduced with only a 64% that sufficiently satises the coordination environment of the
yield achieved in 36 h, compared to 1 h for the non-
This journal is © The Royal Society of Chemistry 2019
Chem. Sci.

View Article Online
Chemical Science
Edge Article
Table 3 Functional group tolerance of magnesiate-catalysed exocyclisation of internal alkynols
Conditions [5 mol%
K2MgR4(PMDETA)2, C6D6]
Entrya,c
Substrate [Ar–C^C–(CH2)3OH]
Product(s)
Yield (and product ratio)b
75 ꢀC, 1 h
84% (10 : 90 : 0)
6%
40 ꢀC, 36 h
1
75 ꢀC, 36 h
40 ꢀC, 36 h
64% (31 : 69)
1%
2
3
75 ꢀC, 1 h
99% (9 : 91)
51%
40 ꢀC, 36 h
75 ꢀC, 0.5 h
40 ꢀC, 24 h
95% (8 : 92)
82%
4
75 ꢀC, 0.5 h
40 ꢀC, 2 h
97% (4 : 96)
93%
5
6
75 ꢀC, 0.5 h
40 ꢀC, 0.5 h
84% (6 : 94)
76%
75 ꢀC, 6 h
87% (11 : 89)
2%
40 ꢀC, 36 h
7
a
Reactions were performed in a Young's cap NMR tube, using 0.5 mmol substrate (11, 13, 15, 17, 19, 21, and 23) and 0.025 mmol (5 mol%) pre-
catalyst with 0.1 mmol (20 mol%/0.2 eq.) 1 consumed in converting the pre-catalyst to the ‘active catalyst’ (vide infra). b Calculated from 1H NMR
spectroscopic data by integration against an internal standard (10 mol% ferrocene). A stoichiometric quantity of crown ether co-catalyst used
c
according to the alkali metal [i.e., 10 mol% 18-c-6 for 5 mol% K2MgR4(PMDETA)2].
metal, preventing further excess alcohol coordination from
occurring which would cause inhibition.
Steady state approximation
Pre-equilibrium approximation
First step fast
k1k2½IVꢃ
dP=dt ¼
(3)
First step slow
k ½1ꢃ
ꢂ1
k1k2½IVꢃ
dP=dt ¼
(1)
(2)
k ½1ꢃ þ k2
ꢂ1
Considering this catalytic process as initial dissociation of
1 (k1 and kꢂ1) followed by cyclisation reaction (k2) and using
the steady-state approximation eqn (1) can be deduced. This
equation can be simplied to eqn (2) at low concentrations of
dP/dt ¼ k1[IV]
Chem. Sci.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Edge Article
Chemical Science
Fig. 2 Observed rate constant at varied substrate concentrations
versus inverse substrate concentration (1).
intercept according to eqn (1), and hence consistent with
a slow ligand dissociation step followed by a rapid cyclisation
process.
Looking at the catalyst order, a good t was seen when the
deduced rates for the different catalyst loadings (Fig. S2†) were
plotted against the concentration of 2 (Fig. 3), indicating that
the reaction has a rst order dependence on the catalyst. This is
similar to that found by Marks,38 but not the second order
dependence that was found by Hill;62 however, in previous work
by Hill using alkaline earth metal amides, a second order
dependence was observed. This is said to be due to the pre-
catalysts existing as dimers. The rst order dependence on
the magnesiate pre-catalyst observed here has also been
observed in hydroamination reactions.62
In terms of the mechanism, there has been a previous
proposition by Hill62 built upon the work of Marks.38 It
reasons that there are two catalytic cycles at work simulta-
neously to allow for the production of the two product
isomers observed. One cycle involves the alkyne passing
through an allene intermediate, where it was demonstrated
that starting from an allenyl alcohol instead of the alkynyl
alcohol yielded the same products in a similar yield and ratio.
In this work utilizing alkali metal magnesiates in combina-
tion with substrate 3, we detected the presence of an allene
intermediate (by 1H NMR spectroscopy, at d 5.19 and 4.63
ppm) during the course of the reaction (Fig. 4). As such we
Fig. 1 (a) Plot of the consumption of 4-pentynol (1) versus time
(0.040 M cat., C6D6, 343 K). (b) Plot of the [4-pentynol (1)]2 versus time
(0.040 M cat., C6D6, 343 K).
Scheme 3 Reaction showing the coordination/decoordination of an
additional alkynol molecule.
1 (k2 [ kꢂ1[1]), while at high concentrations of 1 (kꢂ1[1] [
k2) eqn (3) rules this process. Alternatively, a pre-equilibrium
approximation in which dissociation of 1 from magnesium
does not inuence the reaction rate leads exclusively to eqn
(3), and hence to a good inverse rst order relationship in 1 at
any concentration of this ligand. The plot of the observed
initial rates vs. 1/[1] (Fig. 2) reveals a linear t with a non-zero
Fig. 3 Plot of the observed rate constant at varied catalyst loadings
versus catalyst concentration.
This journal is © The Royal Society of Chemistry 2019
Chem. Sci.

View Article Online
Chemical Science
Edge Article
Fig. 4 Evidence for the allene intermediate (via 1H NMR spectroscopy,
C6D6, 400 MHz) produced during the cyclisation of 5-hexynol (3)
(Table 2, entry 1) as proposed by Hill.62
believe that a similar two-cycle reaction mechanism is likely
to be at play. The ndings from the kinetics were also
included with the other results to form a proposed mecha-
nism for this alkali metal magnesiate-mediated catalysis
(Scheme 4).
Scheme 4 Proposed mechanism for the alkali metal magnesiate +
crown ether co-catalyst mediated catalytic cyclisation of alkynols via
dual activation.
Overall the deduced mechanism starts with formation of
the active catalyst from the magnesiate pre-catalyst. This
involves the deprotonation of four alcohol substrate molecules
to form a magnesiate alkoxide, liberating tetramethylsilane.
This ‘active catalyst’ is then involved in a coordination/
decoordination process with an additional substrate mole-
cule as suggested by the kinetic studies. This additional
molecule of 1 occupies the coordination sphere of the
magnesium inhibiting cyclisation, giving rise to an inverse
rst order term in the substrate. Cyclisation (insertion of the
carbon–carbon multiple bond into the magnesium oxygen
bond) only occurs upon its decoordination.
In cycle A the carbon–carbon triple bond of the alkyne is
directly inserted into the metal–oxygen bond, leading to the
formation of only one product upon protonolysis. Cycle B on the
other hand involves an equilibrium isomerisation from alkyne
to allene which upon insertion into the metal–oxygen bond, and
subsequent protonolysis, can yield two product isomers. Pro-
tonolysis releases the cyclised products and reforms the active
catalyst completing the cycle.
di-anionic (magnesiate) {Mg(OR)4}2ꢂ component, which is
signicantly more nucleophilic than a charge-neutral magne-
sium compound, facilitating intermolecular ring-closure to
furnish the relevant oxygen-heterocycle.
Experimental
General considerations
All reactions were carried out under a protective atmosphere of
argon using standard Schlenk techniques. Non-deuterated solvents
were dried by heating to reux over sodium ketyl radicals under
nitrogen and deuterated solvents were degassed and stored over
molecular sieves. Pre-catalysts [Mg(CH2SiMe3)2, K(CH2SiMe3),
LiMg(CH2SiMe3)3, NaMg(CH2SiMe3)3, KMg(CH2SiMe3)3, Li2-
Mg(CH2SiMe3)4(TMEDA)2,
Na2Mg(CH2SiMe3)4(TMEDA)2,
K2-
Mg(CH2SiMe3)4(TMEDA)2 and K2Mg(CH2SiMe3)4(PMDETA)2] were
prepared following literature procedures2 and handled in a glove-
box. Li(CH2SiMe3), 1 and 5 were obtained from Sigma-Aldrich; 1
and 3 from Alfa-Aesar; and 7 from Fluorochem. Substrates 11, 13,
15, 17, 19 and 21 were synthesised based on literature procedures82
(characterisation data provided for new substrates and products
13–22 can be found in the ESI†). NMR spectra were recorded on
a Bruker AV III 400 MHz spectrometer operating at 400.1 MHz for
1H, 100.6 MHz for 13C or 128.4 MHz for 19F.
Disclosing a unique cooperative behaviour under catalytic
conditions, each of the metals plays a key role in this process, by
simultaneously activating the O–H and C^C bonds of the
substrate. Thus, coordination of the C^C bond to the larger
more p-philic potassium centre enables further activation of
this unsaturated group and brings it into close proximity to the
Chem. Sci.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Edge Article
Chemical Science
placed in a Young's tap NMR tube with C6D6 (0.6 mL), 1,2,3,4-
tetraphenylnaphthalene (0.022 g, 0.05 mmol) and 18-crown-6
(0.010 g, 0.04 mmol). To this K2Mg(CH2SiMe3)4(PMDETA)2
(0.016 g, 0.02 mmol) was added. The reaction was maintained at
70 ꢀC in the NMR spectrometer and was monitored by 1H NMR
spectroscopy. Yields were calculated using NMR spectroscopic
integrals characteristic of 2a relative to the internal standard,
1,2,3,4-tetraphenylnaphthalene. This procedure was repeated
with 0.90, and 1.20 mmol alkynol (1) to provide the data pre-
sented in Fig. S1† which was subsequently used in Fig. 2.
The same procedure was applied to the investigation of the
dependence of the initial rates of the reaction on the catalyst.
The same concentration of 1 was used while varying the
concentration of the catalyst. In terms of the pre-catalyst, 0.025,
0.030 and 0.035 mmol quantities of K2Mg(CH2SiMe3)4(-
PMDETA)2 were employed, necessitating 0.05, 0.06 and
0.07 mmol 18-crown-6 co-catalyst. The data from these experi-
ments with varying catalyst concentration are presented in
Fig. S2† and are subsequently used in Fig. 3.
Procedure for pre-catalyst synthesis
Pre-catalysts were prepared and isolated prior to being
employed in reactions. Na(CH2SiMe3),83,84 K(CH2SiMe3)84,85 and
Mg(CH2SiMe3)2 (ref. 69) were prepared from literature proce-
dures. TMEDA and PMDETA were distilled over CaH2 before
use. K2Mg(CH2SiMe3)4(PMDETA)2 was prepared by suspending
K(CH2SiMe3) (0.25 g, 2 mmol) and Mg(CH2SiMe3)2 (0.20 g, 1
mmol) in hexane (10 mL). The mixture was stirred at ambient
temperature for 1 h. PMDETA (0.43 mL, 2 mmol) was added to
this white suspension and the mixture was gently heated until
ꢀ
homogeneity was achieved. Aer storage at ꢂ26 C overnight,
the mixture was ltered to yield a light yellow solid (typical yield;
0.62 g, 78%).
Li2Mg(CH2SiMe3)4(TMEDA)2, Na2Mg(CH2SiMe3)4(TMEDA)2,
and K2Mg(CH2SiMe3)4(TMEDA)2 were prepared using similar
methods to K2Mg(CH2SiMe3)4(PMDETA)2 with the following
modications: substitution of TMEDA (0.30 mL, 2 mmol);
Li(CH2SiMe3) (2 mL, 2 M in pentanes); and/or Na(CH2SiMe3)
(0.22 g, 2 mmol) where appropriate.
69
LiMg(CH2SiMe3)3,70 NaMg(CH2SiMe3)3,32 and KMg(CH2-
SiMe3)3 were prepared using half the molar quantity of
M(CH2SiMe3)3 (1 mL/0.11 g/0.13 g respectively), and toluene (5
mL) was added in place of the multidentate amine in order to
achieve dissolution upon heating.
Conclusions
In summary, alkali metal magnesiates have been shown to
successfully promote the catalytic intramolecular hydro-
alkoxylation of alkynols through cooperative bimetallic catal-
ysis. The roles of both magnesium and potassium components
are crucial for the success of the process, affording a unique
type of substrate activation that is not possible in conventional
single-metal systems. Through cooperativity, the utilisation of
an alkali metal magnesiate has overcome the inherent prob-
lems associated with homometallic magnesium systems. The
optimised catalyst system, (PMDETA)2K2Mg(CH2SiMe3)4 paired
with 18-crown-6 has been shown to function well with both
terminal and internal alkyne substrates having a range of
functional groups. Kinetic studies have revealed an inhibition
effect of substrate on the catalyst under standard conditions
(high concentrations of alkynol) by the formation of a coordi-
nation adduct which requires dissociation prior to the cyclisa-
tion step. Initial reactivity studies suggest that coordination of
the 18-crown-6 to K nely tunes the reactivity of the bimetallic
system, probably minimising the coordination of additional
substrate molecules to the catalyst.
Procedure for NMR scale reactions
In the catalytic cyclisation of 4-pentynol (1) with K2Mg(CH2-
SiMe3)4(PMDETA)2 + 18-c-6, 4-pentynol (56 mL, 0.6 mmol, 1.2
eq.) (0.2 eq./20 mol% excess required to form the ‘active cata-
lyst’) was added to a Young's tap NMR tube alongside C6D6 (0.54
mL), 1,2,3,4-tetraphenylnaphthalene (0.022 g, 0.05 mmol, 0.1
eq.) and 18-crown-6 (0.013 g, 0.05 mmol, 0.1 eq.). To this K2-
Mg(CH2SiMe3)4(PMDETA)2 (0.020 g, 0.025 mmol, 0.05 eq.) was
added and the tube was placed in an oil bath at 75 ꢀC. The
1
reaction was periodically monitored by H NMR spectroscopy
and the yields obtained were calculated using NMR spectro-
scopic integrals and were relative to the internal standard.
Optimisation and substrate scope reactions were carried out
using similar procedures, employing 10 mol% of an appropriate
internal standard [1,2,3,4-tetraphenylnaphthalene (0.022 g) or
ferrocene (0.009 g)] and with the oil bath temperature being
changed where necessary based on observed reaction times.
0.5 mmol alkynol substrate was used with the monometallic
species and lower-order magnesiates, and 0.6 mmol with
higher-order magnesiate pre-catalysts. 0.025 mmol pre-catalyst
was employed in all cases, quantities shown in Table 3. In
general, the reactions remain homogeneous until nearing
completion. At this point an unidentied solid/gel forms, which
is presumably a species akin to the ‘K2Mg(OAlk)4(18-c-6)2
resting-state’.
Conflicts of interest
There are no conicts to declare.
Acknowledgements
All high-resolution mass spectrometry (HRMS) was performed
at the EPSRC UK National Mass Spectrometry Facility at Swan-
sea University. We would like to thank the European Research
Council (ERC-STG grant to EH), EPSRC UK (EP/N011384/1 grant
to EH) and Leverhulme Trust (research grant RPG-2016-281 to
Procedure for kinetic studies
In the kinetic studies to determine the rate dependence of CTOH) for the generous sponsorship of this research. We are
alkynol in the cyclisation of 4-pentynol (1) with K2Mg(CH2- also grateful to Professor Robert Mulvey for insightful discus-
SiMe3)4(PMDETA)2 + 18-c-6, 4-pentynol (56 mL, 0.6 mmol) was sions and Cristian Vidal for synthesising substrates 7 and 9. J.
This journal is © The Royal Society of Chemistry 2019
Chem. Sci.

View Article Online
Chemical Science
Edge Article
G.-A. is indebted to the MINECO of Spain (CTQ2016-81797- 20 M. S. Hill, D. J. Liptrot and C. Weetman, Chem. Soc. Rev.,
REDC and CTQ2016-75986-P) and the Gobierno del Princi- 2016, 45, 972–988.
pado de Asturias (project GRUPIN14-006) for nancial support. 21 S. Kobayashi and Y. Yamashita, Acc. Chem. Res., 2011, 44, 58–
´
J. G.-A. also thanks: (i) the Fundacion BBVA for the award of
“Beca Leonardo a Investigadores y Creadores Culturales 2017”; 22 R. J. Less, R. L. Melen and D. S. Wright, RSC Adv., 2012, 2,
and (ii) PhosAgro/UNESCO/IUPAC for the award of “Green 2191–2199.
Chemistry for Life Grant”. The Fundacion BBVA accepts no 23 D. J. Liptrot, M. S. Hill, M. F. Mahon and A. S. S. Wilson,
responsibility for the opinions, statements and contents Angew. Chem., Int. Ed., 2015, 54, 13362–13365.
included in the project and/or the results thereof, which are 24 B. Liu, T. Roisnel, J. F. Carpentier and Y. Sarazin, Angew.
entirely the responsibility of the author. Data used within this Chem., Int. Ed., 2012, 51, 4943–4946.
publication can be accessed at https://doi.org/10.15129/ 25 D. Mukherjee, H. Osseili, T. P. Spaniol and J. Okuda, J. Am.
71.
´
5d0b6ec6-569c-4c8b-ae12-69d703ae3d.
Chem. Soc., 2016, 138, 10790–10793.
26 J. Penael, L. Maron and S. Harder, Angew. Chem., Int. Ed.,
2015, 54, 201–206.
27 P. P. Power, Nature, 2010, 463, 171–177.
28 D. Schuhknecht, C. Lhotzky, T. P. Spaniol, L. Maron and
J. Okuda, Angew. Chem., Int. Ed., 2017, 56, 12367–12371.
29 C. Weetman, M. D. Anker, M. Arrowsmith, M. S. Hill,
G. Kociok-Kohn, D. J. Liptrot and M. F. Mahon, Chem. Sci.,
2016, 7, 628–641.
Notes and references
1 G. Wittig, F. J. Meyer and G. Lange, Justus Liebigs Ann. Chem.,
1951, 571, 167–201.
¨
2 T. Klatt, J. T. Markiewicz, C. Samann and P. Knochel, J. Org.
Chem., 2014, 79, 4253–4269.
3 A. Krasovskiy, V. Krasovskaya and P. Knochel, Angew. Chem., 30 C. Weetman, M. S. Hill and M. F. Mahon, Chem. Commun.,
Int. Ed., 2006, 45, 2958–2961. 2015, 51, 14477–14480.
4 A. J. Martınez-Martınez, S. Justice, B. J. Fleming, 31 F. M. Younis, S. Krieck, H. Gorls and M. Westerhausen,
´
´
¨
A. R. Kennedy, I. D. H. Oswald and C. T. O'Hara, Sci. Adv.,
2017, 3, e1700832.
Organometallics, 2015, 34, 3577–3585.
´
´
32 S. E. Baillie, W. Clegg, P. Garcıa-Alvarez, E. Hevia,
A. R. Kennedy, J. Klett and L. Russo, Chem. Commun.,
2011, 47, 388–390.
´
´
5 A. J. Martınez-Martınez, A. R. Kennedy, R. E. Mulvey and
C. T. O'Hara, Science, 2014, 346, 834–837.
´
´
6 R. L.-Y. Bao, R. Zhao and L. Shi, Chem. Commun., 2015, 51, 33 M. De Tullio, A. Hernan-Gomez, Z. Livingstone, W. Clegg,
˜
´
6884–6900.
A. R. Kennedy, R. W. Harrington, A. Antinolo, A. Martınez,
F. Carrillo-Hermosilla and E. Hevia, Chem.–Eur. J., 2016,
22, 17646–17656.
7 S. D. Robertson, M. Uzelac and R. E. Mulvey, Chem. Rev.,
2019, DOI: 10.1021/acs.chemrev.1029b00047.
´
´
8 R. E. Mulvey and S. D. Robertson, in Organo-di-Metallic 34 A. Hernan-Gomez, T. D. Bradley, A. R. Kennedy,
Compounds (or Reagents): Synergistic Effects and Synthetic
Applications, ed. Z. Xi, Springer International Publishing,
Cham, 2014, pp. 129–158.
Z. Livingstone, S. D. Robertson and E. Hevia, Chem.
Commun., 2013, 49, 8659–8661.
35 B. M. Trost, Science, 1991, 254, 1471–1477.
9 F. Mongin and A. Harrison-Marchand, Chem. Rev., 2013, 113, 36 D. J. Faulkner, Nat. Prod. Rep., 1984, 1, 251–280.
7563–7727.
37 T. L. B. Boivin, Tetrahedron, 1987, 43, 3309–3362.
10 D. Tilly, F. Chevallier, F. Mongin and P. C. Gros, Chem. Rev., 38 S. Y. Seo, X. H. Yu and T. J. Marks, J. Am. Chem. Soc., 2009,
2014, 114, 1207–1257. 131, 263–276.
11 M. R. Crimmin, I. J. Casely and M. S. Hill, J. Am. Chem. Soc., 39 B. Seiller, C. Bruneau and P. H. Dixneuf, J. Chem. Soc., Chem.
2005, 127, 2042–2043.
Commun., 1994, 493–494.
12 S. Harder, Chem. Rev., 2010, 110, 3852–3876.
40 P. Pale and J. Chuche, Eur. J. Org. Chem., 2000, 1019–1025.
´
´
13 J. Spielmann, F. Buch and S. Harder, Angew. Chem., Int. Ed., 41 A. Varela-Fernandez, C. Garcıa-Yebra, J. A. Varela,
´
2008, 47, 9434–9438.
M. A. Esteruelas and C. Saa, Angew. Chem., Int. Ed., 2010,
14 M. D. Anker, M. S. Hill, J. P. Lowe and M. F. Mahon, Angew.
Chem., Int. Ed., 2015, 54, 10009–10011.
49, 4278–4281.
42 E. Benedetti, A. Simonneau, A. Hours, H. Amouri, A. Penoni,
G. Palmisano, M. Malacria, J.-P. Goddard and
L. Fensterbank, Adv. Synth. Catal., 2011, 353, 1908–1912.
¨
15 M. Arrowsmith, M. S. Hill and G. Kociok-Kohn,
Organometallics, 2014, 33, 206–216.
16 M. Arrowsmith, W. M. S. Shepherd, M. S. Hill and G. Kociok- 43 V. Diachenko, M. J. Page, M. R. D. Gatus, M. Bhadbhade and
Kohn, Chem. Commun., 2014, 50, 12676–12679.
B. A. Messerle, Organometallics, 2015, 34, 4543–4552.
´
¨
´
17 H. Bauer, M. Alonso, C. Farber, H. Elsen, J. Pahl, A. Causero, 44 C. Vidal, L. Merz and J. Garcıa-Alvarez, Green Chem., 2015,
G. Ballmann, F. De Pro and S. Harder, Nature Catalysis,
2018, 1, 40–47.
18 M. J. Butler, A. J. P. White and M. R. Crimmin,
Organometallics, 2018, 37, 949–956.
17, 3870–3878.
45 J. L. Arbour, H. S. Rzepa, A. J. P. White and K. K. Hii, Chem.
Commun., 2009, 46, 7125–7127.
46 A. G. Nair, R. T. McBurney, M. R. D. Gatus, S. C. Binding and
B. A. Messerle, Inorg. Chem., 2017, 56, 12067–12075.
¨
19 H. Elsen, C. Farber, G. Ballmann and S. Harder, Angew.
Chem., Int. Ed., 2018, 57, 7156–7160.
Chem. Sci.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Edge Article
Chemical Science
`
47 R. J. Harris, R. G. Carden, A. N. Duncan and 67 A. Motta, I. L. Fragala and T. J. Marks, Organometallics, 2010,
R. A. Widenhoefer, ACS Catal., 2018, 8, 8941–8952. 29, 2004–2012.
48 V. Cadierno, J. Dıez, J. Garcıa-Alvarez, J. Gimeno, N. Nebra 68 N. T. Coles, M. F. Mahon and R. L. Webster, Chem. Commun.,
´
´
´
´
and J. Rubio-Garcıa, Dalton Trans., 2006, 5593–5604.
2018, 54, 10443–10446.
49 M. J. Pouy, S. A. Delp, J. Uddin, V. M. Ramdeen, 69 S. E. Baillie, T. D. Bluemke, W. Clegg, A. R. Kennedy, J. Klett,
N. A. Cochrane, G. C. Fortman, T. B. Gunnoe,
T. R. Cundari, M. Sabat and W. H. Myers, ACS Catal., 2012,
2, 2182–2193.
L. Russo, M. de Tullio and E. Hevia, Chem. Commun., 2014,
50, 12859–12862.
´
´
70 S. E. Baillie, W. Clegg, P. Garcıa-Alvarez, E. Hevia,
A. R. Kennedy, J. Klett and L. Russo, Organometallics, 2012,
31, 5131–5142.
50 X. Xu, J. Liu, L. Liang, H. Li and Y. Li, Adv. Synth. Catal., 2009,
351, 2599–2604.
51 Y. Sheng, D. G. Musaev, K. S. Reddy, F. E. McDonald and 71 C. Lichtenberg, T. P. Spaniol, L. Perrin, L. Maron and
K. Morokuma, J. Am. Chem. Soc., 2002, 124, 4149–4157. J. Okuda, Chem.–Eur. J., 2012, 18, 6448–6452.
52 S. B. Moilanen and D. S. Tan, Org. Biomol. Chem., 2005, 3, 72 P. B. Hitchcock, M. F. Lappert and A. V. Protchenko, J. Am.
798–803. Chem. Soc., 2001, 123, 189–190.
53 F. E. McDonald and C. C. Schultz, J. Am. Chem. Soc., 1994, 73 I. Koehne, S. Bachmann, R. Herbst-Irmer and D. Stalke,
116, 9363–9364. Angew. Chem., Int. Ed., 2017, 56, 15141–15145.
54 F. E. McDonald, K. S. Reddy and Y. Dıaz, J. Am. Chem. Soc., 74 J. Kretsch, A. Kreyenschmidt, R. Herbst-Irmer and D. Stalke,
´
2000, 122, 4304–4309.
Dalton Trans., 2018, 47, 12606–12612.
´
55 J. A. Varela and C. Saa, Synthesis, 2016, 48, 3470–3478.
56 T. Cai, W.-W. Li, W.-B. Qin and T.-B. Wen, Chem.–Eur. J.,
2017, 24, 1606–1618.
75 S. Neander, F. E. Tio, R. Buschmann, U. Behrens and
F. Olbrich, J. Organomet. Chem., 1999, 582, 58–65.
`
´
76 A. Imberdis, G. Lefevre, P. Thuery and T. Cantat, Angew.
Chem., Int. Ed., 2018, 57, 3084–3088.
57 M. Mellor, A. Santos, E. G. Scovell and J. K. Sutherland, J.
Chem. Soc., Chem. Commun., 1978, 528–529.
58 M. D. Weingarten and A. Padwa, Tetrahedron Lett., 1995, 36,
4717–4720.
¨
¨
77 S. M. Harling, S. Krieck, H. Gorls and M. Westerhausen,
Inorg. Chem., 2017, 56, 9255–9263.
´
78 H. Abas, J. Mas-Rosello, M. M. Amer, D. J. Durand,
59 S. D. Wobser and T. J. Marks, Organometallics, 2013, 32,
2517–2528.
R. R. Groleau, N. Fey and J. Clayden, Angew. Chem., Int.
Ed., 2019, 58, 2418–2422.
60 X. H. Yu, S. Seo and T. J. Marks, J. Am. Chem. Soc., 2007, 129, 79 R. Rochat, K. Yamamoto, M. J. Lopez, H. Nagae, H. Tsurugi
7244–7245. and K. Mashima, Chem.–Eur. J., 2015, 21, 8112–8120.
61 J. Martınez, A. Otero, A. Lara-Sanchez, J. A. Castro-Osma, 80 S. Seo and T. J. Marks, Chem.–Eur. J., 2010, 16, 5148–5162.
´
´
´
´
´
J. Fernandez-Baeza, L. Sanchez-Barba and A. M. Rodrıguez, 81 M. Shigeno, K. Hanasaka, K. Sasaki, K. Nozawa-Kumada and
Organometallics, 2016, 35, 1802–1812. Y. Kondo, Chem.–Eur. J., 2019, 25, 6077–6081.
62 C. Brinkmann, A. G. M. Barrett, M. S. Hill, P. A. Procopiou 82 G. Liu and S. S. Stahl, J. Am. Chem. Soc., 2007, 129, 6328–
and S. Reidt, Organometallics, 2012, 31, 7287–7297. 6335.
63 D. Y. Li, K. J. Shi, X. F. Mao, Z. L. Zhao, X. Y. Wu and P. N. Liu, 83 P. B. Hitchcock, M. F. Lappert, W.-P. Leung, L. Diansheng
Tetrahedron, 2014, 70, 7022–7031.
64 J. K. Vandavasi, W.-P. Hu,
and T. Shun, J. Chem. Soc., Chem. Commun., 1993, 1386–
1387.
G.
C.
Senadi,
S. S. K. Boominathan, H.-Y. Chen and J.-J. Wang, Eur. J. 84 W. Clegg, B. Conway, A. R. Kennedy, J. Klett, R. E. Mulvey
Org. Chem., 2014, 6219–6226. and L. Russo, Eur. J. Inorg. Chem., 2011, 721–726.
65 S. Gandhi and B. Baire, ChemistrySelect, 2018, 3, 4490–4494. 85 P. B. Hitchcock, A. V. Khvostov and M. F. Lappert, J.
66 L. Garcia, M. D. Anker, M. F. Mahon, L. Maron and M. S. Hill,
Organomet. Chem., 2002, 663, 263–268.
Dalton Trans., 2018, 47, 12684–12693.
This journal is © The Royal Society of Chemistry 2019
Chem. Sci.