Nanogold-catalyzed cis -silaboration of alkynes with abnormal regioselectivity

Article abstract of DOI:10.1021/ol500225x

The first example of gold-catalyzed silaboration of alkynes with PhMe ₂SiBpin is documented in the presence of supported gold nanoparticles. In the case of terminal alkynes, the reaction proceeds at ambient conditions in very good yields and th

Full text of DOI:10.1021/ol500225x

Letter
pubs.acs.org/OrgLett
Nanogold-Catalyzed cis-Silaboration of Alkynes with Abnormal
Regioselectivity
Charis Gryparis and Manolis Stratakis*
Department of Chemistry, University of Crete, Voutes 71003 Iraklion, Greece
S
* Supporting Information
ABSTRACT: The first example of gold-catalyzed silaboration of alkynes
with PhMe2SiBpin is documented in the presence of supported gold
nanoparticles. In the case of terminal alkynes, the reaction proceeds at
ambient conditions in very good yields and the regioselectivity is opposite to
that observed in the presence of Pd or Pt catalysts. The abnormal
regioselectivity is attributed to steric factors imposed by the Au nanoparticle
during the 1,2-addition of silylborane to the alkyne.
he reaction among σ silylboranes and alkynes under
catalysis by Pd(0) and Pt(0) provides cis-1,2-addition
addition of Au/TiO2 (1 mol %) the smooth consumption of
T
alkyne within 1 h at 25 °C, with primarily formation of the 1,2-
silylboration product 1a in 76% relative yield and its regiosomer
1b in 10% relative yield (Scheme 1). In addition, 14% of the
products,1 useful precursors for the synthesis of stereodefined
functionalized alkenes2 via Hiyama and Suzuki cross-coupling
reactions at the C−Si or C−B functionalities, respectively. The
accepted reaction mechanism involves oxidative insertion of the
Si−B bond on metal (M0) forming a (silyl)MII(boryl)
intermediate that undergoes insertion from the alkyne. Finally,
reductive elimination of the metal affords the cis-1,2-addition
products. Products with opposite configuration (trans-addition)
can be obtained by tuning the stoichiometry of the reactants in
favor of the silylborane.3 Typically, in the case of terminal
alkynes, the boryl functionality is attached on the terminal
carbon atom of the alkyne, while the silyl group is on the
internal one (1-boryl-2-silyl-1-alkenes).4 A remarkable switch in
regioselectivity of silaboration to yield 2-boryl-1-silyl-1-alkenes
was observed by using a very bulky phosphine as ligand [o-
biphenyl(t-Bu)2P] and (n3-C3H5)Pd(L)Cl as catalyst.5
Scheme 1. Reaction of PhMe2SiBpin with 1-Heptyne (1) and
Its Hydrolysis in the Presence of Au/TiO2
As it has been recently shown that 1,2-disilanes6 or
distannanes7 undergo oxidative insertion on Au(I), forming
bis(silyl) or stannylgold(III) complexes, it would be compelling
to uncover the ability of gold to catalyze the addition of
silylboranes to alkynes, just as Pd(0) or Pt(0) do. For instance,
the stereoselective syn insertion of the gold−silicon bond from
Ph3PAu-SiR3 to alkynes has been recently documented.8 The
interest in the so far unknown gold-catalyzed silaboration
originates from the recent results from our group regarding the
activation of σ disilanes by gold nanoparticles (Au NPs). While
typical homogeneous Au(I) catalysts seem to be unreactive, Au
NPs supported on metal oxides9 are efficient catalysts. Thus,
Au/TiO2 catalyzes the hydrolysis/alcoholysis10 of Si−Si linkage
and the cis-selective addition of 1,2-disilanes to alkynes.11 The
mechanism possibly involves oxidative insertion of the Si−Si
moiety on gold nanoparticles.12 This precedent urged us to
examine the possible reaction among silylboranes and alkynes
in the presence of Au/TiO2. The readily available and fairly
stable PhMe2SiBpin [pin = pinacolato] was chosen. Upon
mixing 1.0 equiv of heptyne (1) with 1.2 equiv of PhMe2SiBpin
in DCM under strictly dry conditions, we observed after
1,2-disilylation product 1c11 was seen, as well as equimolar
amounts of pinB-Bpin [bis(pinacolato)diboron] relative to 1c.
It is remarkable that the observed regioselectivity is opposite to
what has been seen under typical Pd or Pt catalyst conditions,
except a specific case commented earlier5 utilizing an extremely
bulky phosphine as ligand. Increasing the amount of
PhMe2SiBpin to 3 equiv results to a faster consumption of 1.
Received: January 22, 2014
Published: February 17, 2014
© 2014 American Chemical Society
1430
dx.doi.org/10.1021/ol500225x | Org. Lett. 2014, 16, 1430−1433

Organic Letters
Letter
Yet, the relative yield of side product 1c increases to 30%. The
amount of 1c can be reduced to <10% by slow addition of
PhMe2SiBpin into the reaction mixture over the reaction time
period (∼1 h). Under this modification, silaboration product 1a
can be obtained in 72% isolated yield after column
chromatography. Au/Al2O3 provides even better regioselectiv-
ity (1a/1b = 94/6), yet the reaction rate is slow and the
amount of disilyl side product 1c increases substantially to 25%.
In the presence of Au/ZnO, the reaction rate is even slower,
the ratio 1a/1b = 85/15, and the isolated yield of 1a below
50%. Notably, the Au/TiO2-catalyzed reaction can be
performed in nondried DCM under open air conditions, at
the expense of PhMe2SiBpin (typically 1.5 equiv) to
compensate its partial hydrolysis, yielding PhMe2Si-O-Bpin.
The formation of side disilylation adduct 1c along with pinB-
Bpin arises from a competing pathway, as by mixing
PhMe2SiBpin with nondried DCM and subsequent treatment
with Au/TiO2 (1%) resulted after 2 h in the disappearance of
silylborane and formation of the fairly unstable hydrolysis
product PhMe2Si-O-Bpin,13 pinB-Bpin, and an equimolar
diborane mixture of PhMe2SiOH + (PhMe2Si)2O which varies
with time (Scheme 1). Traces of PhMe2SiH were also detected.
The relative ratio PhMe2Si-O-Bpin/pinB-Bpin was ∼1/1.
Apparently, a metal-catalyzed silylborane methathesis is taking
place, forming diborane and disilane; the latter undergoes
hydrolysis under the reaction conditions, forming silanol and
1,3-disiloxane via the transient participation of PhMe2SiH10
(see also the mechanistic analysis in Scheme 2). Pd-catalyzed
analogous element−element σ bond metatheses in disilanes14
and silylstannanes15 are already known. The reaction is
heterogeneous as leaching of gold into the supernatant solution
under of the reaction conditions is below parts per billion level
(ICP analysis) as also observed in the disilylation11 reaction.
Among homogeneous Au(I) catalysts tested in the reaction of 1
with PhMe2SiBpin, only Ph3PAuNTf2 (3 mol %) provided
traces of products after heating at 70 °C in DCE for 12 h (∼1%
yield; 1a/1b = 50/50). In the meantime, ∼4% of PhMe2SiBpin
had been disproportionated to pinB-Bpin and ∼20% into
PhMe2SiOBpin via hydrolysis by residual water.
Table 1. Addition of PhMe2SiBpin to Alkynes Catalyzed by
Au/TiO2
The scope and limitations of the Au/TiO2-catalyzed
silaboration were examined with various functionalized alkynes.
The results are presented in Table 1.16 In our experiments, a
series of terminal alkynes reacted smoothly with excess of
PhMe2SiBpin (typically 1.5 equiv) in nondried DCM (1.0 mol
% of gold from Au/TiO2). The reaction proceeds in good to
excellent yields in favor of 2-boryl-1-silyl-1-alkenes (products
a). The regioselectivity (a/b) varies from 85/15 to >99/1,
while the stereochemistry of all silaboration adducts is >97% cis
(E configuration) as proven by NOE experiments. In all cases
under the reaction conditions, cis-1,2-disilylation adducts are
also formed in 2−15% relative ratio, which can be removed
with column chromatography. Alkynes with pronounced steric
hindrance adjacent to the internal acetylenic carbon atom
exhibit the highest regioselectivity of silaboration, providing
essentially only one regioisomer (e.g., products 7a, 9a, and
11a). Notably, CF3-substituted aryl alkyne 15, which is
unproductive under Pd catalysis conditions,5 forms product
15a in good yield. Also, enyne 11 which is extremely
unreactive17 provides silaboration adduct 11a in low yield
due to an extensive disilylation as a single regioisomer. No
cyclization product(s) were seen in the crude reaction mixture
as occurs5 in the presence of (n3-C3H5)Pd[o-biphenyl(t-
Bu)2P]Cl. In general, increasing the molar equivalents of
a
b
Isolated yield of the major regioisomer a. With 2.5 equiv of
PhMe2SiBpin. The lower yield of 7a is due to partial cleavage of 4-
methoxyphenol under chromatographic conditions. With 6 equiv of
PhMe2SiBpin, 36 h. The low isolated yield of 11a is due to the
c
competing formation of disilylation product in ∼70% relative yield.
d
With 4 equiv of PhMe2SiBpin, 12 h. The low yield of 16a,b is due to
e
the formation of disilylation products in ∼55% relative yield. Isolated
yield.
1431
dx.doi.org/10.1021/ol500225x | Org. Lett. 2014, 16, 1430−1433

Organic Letters
Letter
PhMe2SiBpin results in gradual increase of the 1,2-disilylation
products, lowering the yield of the desired silaboration pathway
(e.g., enyne 11).
terminal alkynes and type II only when an extremely bulky
phosphine is ligated on Pd.5 In our case, we suggest that III is
destabilized by the steric interactions among the Au NP-bound
silane and the germinal R group. In other words, the metal
preferentially binds on the terminal carbon atom of alkyne. This
concept explains the lack of reactivity of internal alkynes, as
steric interactions among Au NP and the alkyne substituents
are developing at any binding approach, not counting the steric
hindrance among the substituents of the internal alkyne who
become cis to each other in the transition state of the addition.
In the case of ester 16, surprisingly, there is not any electronic
preference for the formation of the regioisomeric adducts (16a
and 16b are formed in almost equal amounts),19 but solely
steric factors operate, as the hindrance imposed to the C-bound
Au moiety by a methyl group (formation of product 16b) or an
ester functionality (formation of 16a) is more or less the same.
A similar regioselectivity profile has been previously seen by us
in the Au/TiO2-catalyzed hydrosilylation of 16.20 An additional
competing pathway also operates, which involves σ bond
metathesis among intermediate I and a neutral PhMe2SiBpin
molecule that generates PhMe2Si-Au-SiMe2Ph (IV) and pinB-
Bpin. The diborane is unreactive21 under gold nanoparticle
catalysis conditions toward triple-bond 1,2-diboration. Re-
cently, it was reported that pinB-Bpin diborylates alkynes in the
presence of a nanoporous gold material as catalyst.22 Similar σ
bond metatheses are known in 1,2-disilanes14 and 1,2-
silylstannanes.15 Intermediate IV undergoes cis-addition to
alkynes, forming the 1,2-disilylation side products,11 or is
captured by H2O10 to form eventually (PhMe2Si)2O. In
general, increasing the molar equivalents of PhMe2SiBpin
relative to the alkyne results in an increase of the relative yield
of 1,2-disilyl side products due to the operation of the σ bond
metathesis pathway (see, for example, the cases of alkynes 11
and 16). The preferential delivery of the Bpin moiety, relative
to the silyl group on the alkyne in the proposed intermediates
II and III, has already been rationalized in terms of
thermodynamics.23 The C−B bond is more stable compared
to the C−Si by approximately 24 kcal/mol, which is a larger
difference upon comparing the dissociation energies among the
Au−B and Au−Si bonds (87.8 and 74.8 kcal/mol, respectively).
Au−Si insertion from intermediate I to the alkynes cannot be
excluded, but is unlike on the basis of thermodynamics as
analyzed above and the lack of cyclization products in the
silaboration of 1,6-enyne 11.
A series of internal alkynes examined herein are unproduc-
tive. The possible reasons for this unreactivity, as well as the
importance of steric hindrance of alkyne in controlling the
regioselectivity of silaboration, will be analyzed in the
accompanying mechanistic discussion. An exception among
internal alkynes is acetylenic ester 16, which yields excess of
PhMe2SiBpin (4 equiv) an almost equimolar mixture of
regioisomers 16a and 16b (16a/16b = 48/52) in relatively
low yield due to competing 1,2-disilylation. Addition product
16b appears as a mixture of double bond isomers (E/Z ∼ 4/1).
The enhanced reactivity of this specific substrate has already
been shown in its dehydrogenative cyclization with a series of
1,n-dihydro 1,n-oligosilanes17 and ascribed to its low-energy
LUMO.
The Au-catalyzed addition of PhMe2SiBpin to the deuterium-
labeled alkyne 8-d (95% D, Table 1) revealed almost complete
deuterium retention in product 8a-d (94% D), which excludes
the formation of gold acetylides as intermediates. Also, as Au/
TiO2 is inefficient in catalyzing hydroboration of alkynes,18
a
pathway involving cross-coupling among a terminal alkyne and
PhMe2SiBpin to form silyl alkyne + hydroborane (pinBH)
followed by hydroboration of the internal silyl alkyne with
pinBH is also excluded. In Scheme 2, we present a mechanistic
Scheme 2. Proposed Mechanism for the Regioselective Au-
Catalyzed Reaction of Silylboranes with Terminal Alkynes
In conclusion, we have presented herein the first example of
a Au-catalyzed silaboration of alkynes with PhMe2SiBpin. The
reaction proceeds under mild conditions, no external additives
or ligands are necessary, while the regioselectivity is abnormal,
providing 2-boryl-1-silyl-1-alkenes which can be formed under
Pd catalysis conditions only utilizing an extremely bulky
phosphine as additive.5 Additionally, PhMe2SiBpin undergoes
gold-catalyzed disproportionation to disilane and diborane via σ
bond metathesis. Once more, the ability of gold to catalyze
reactions so far feasible with Pd, Pt, and other late transition
metals is expanding its significance in synthetic organic
chemistry.
rationalization, postulating cationic gold9,12 as the reactive
catalytic sites of Au/TiO2, capable of undergoing oxidative
insertion from the Si−B bond (intermediate I), as already
proposed in the case of σ disilanes.11 Intermediate I adds to the
alkyne, forming either II or III. Intermediates of type III have
been proposed4 in the Pd- or Pt-catalyzed silaboration of
ASSOCIATED CONTENT
■
S
* Supporting Information
1
Copies of H, 13C NMR of all products, and selected NOE
experiments for key compounds. This material is available free
of charge via the Internet at http://pubs.acs.org.
1432
dx.doi.org/10.1021/ol500225x | Org. Lett. 2014, 16, 1430−1433

Organic Letters
Letter
(16) General procedure for the Au/TiO2-catalyzed silaboration of
alkynes: To a vial containing the alkyne (0.2 mmol), PhMe2SiBpin
(0.3 mmol), and 1 mL of DCM is added 40 mg of Au/TiO2 (1.0 mol
% in Au). After a certain period of time (see Table 1) at 25 °C, the
reaction is complete (TLC, GC-MS). The slurry is filtered with the aid
of DCM under low pressure through a short pad of silica gel or Celite,
and the filtrate is evaporated to afford the silaboration products which
are purified by column chromatography.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: stratakis@chemistry.uoc.gr.
Notes
The authors declare no competing financial interest.
(17) A structurally similar 1,6-enyne also reacts with 1,1,3,3-
tetramethyldisiloxane in the presence of Au/TiO2 at surprisinly slow
rate compared to other terminal alkynes: Kotzabasaki, V.; Lykakis, I.
N.; Gryparis, C.; Psyllaki, A.; Vasilikogiannaki, E.; Stratakis, M.
Organometallics 2013, 32, 665.
(18) In our hands, Au/TiO2 does not catalyze hydroboration of
alkynes with pinBH. Only homogeneous Au(I or III) catalysts are
effective, yet in moderate yields: Leyva, A.; Zhang, X.; Corma, A.
Chem. Commun. 2009, 4947.
(19) The formation of (Z)-16b could be rationalized in terms of
isomerization of intermediate V to VI, by analogy to the lack of
stereoselectivity during the 1,2-disilylation of silyl-substituted
acetylenes by σ disilanes.11 MM2 calculations reveal that (Z)-16b is
less stable than (E)-16b by ∼2 kcal/mol.
ACKNOWLEDGMENTS
■
This work was supported by the project “IRAKLEITOS II-
University of Crete” of the Operational Programme for
Education and Lifelong Learning 2007-2013 (E.P.E.D.V.M.)
of the NSRF (2007-2013), which is cofunded by the European
Union (European Social Fund) and National Resources. We
thank ProFI (ITE, Heraklion, Greece) for obtaining the HRMS
spectra of unknown compounds.
REFERENCES
■
(1) (a) Oestreich, M.; Hartmann, E.; Mewald, M. Chem. Rev. 2013,
113, 402. (b) Ohmura, T.; Suginome, M. Bull. Chem. Soc. Jpn. 2009,
82, 29. (c) Beletskaya, I.; Moberg, C. Chem. Rev. 2006, 106, 2320.
(d) Suginome, M.; Ito, Y. Chem. Rev. 2000, 100, 3221. (e) Beletskaya,
I.; Moberg, C. Chem. Rev. 1999, 99, 3435.
(2) (a) Jiao, J.; Nakajima, K.; Nishihara, Y. Org. Lett. 2013, 15, 3294.
(b) Ohmura, T.; Oshima, K.; Suginome, M. Organometallics 2013, 32,
2870. (c) Saito, N.; Saito, K.; Sato, H.; Sato, Y. Adv. Synth. Catal. 2013,
355, 853. (d) Zhou, H.; Moberg, C. J. Am. Chem. Soc. 2012, 134,
15992.
(3) Ohmura, T.; Oshima, K.; Suginome, M. Chem. Commun. 2008,
1416.
(4) (a) Suginome, M.; Nakamura, H.; Ito, Y. Chem. Commun. 1996,
2777. (b) Suginome, M.; Matsuda, T.; Nakamura, H.; Ito, Y.
Tetrahedron 1999, 55, 8787. (c) Da Silva, J. C. A.; Birot, M.; Pillot,
J.-P.; Petraud, M. J. Organomet. Chem. 2002, 646, 179.
(5) Ohmura, T.; Oshima, K.; Taniguchi, H.; Suginome, M. J. Am.
Chem. Soc. 2010, 132, 12194.
(6) (a) Gualco, P.; Ladeira, S.; Miqueu, K.; Amgoune, A.; Bourissou,
D. Angew. Chem., Int. Ed. 2011, 50, 8320. (b) Joost, M.; Gualco, P.;
Coppel, Y.; Miqueu, K.; Kefalidis, C. E.; Maron, L.; Amgoune, A.;
Bourissou, D. Angew. Chem., Int. Ed. 2014, 53, 747.
(20) Psyllaki, A.; Lykakis, I. N.; Stratakis, M. Tetrahedron 2012, 68,
8724.
(21) Grirrane, A.; Corma, A.; Garcia, H. Chem.Eur. J. 2011, 17,
2467.
(22) Chen, Q.; Zhao, J.; Ishikawa, Y.; Asao, N.; Yamamoto, Y.; Jin, T.
Org. Lett. 2013, 15, 5766.
(23) Sakaki, S.; Biswas, B.; Musashi, Y.; Sugimoto, M. J. Organomet.
Chem. 2000, 611, 288.
(7) Lassauque, N.; Gualco, P.; Mallet-Ladeira, S.; Miqueu, K.;
Amgoune, A.; Bourissou, D. J. Am. Chem. Soc. 2013, 135, 13827.
(8) Joost, M.; Gualco, P.; Mallet-Ladeira, S.; Amgoune, A.; Bourissou,
D. Angew. Chem., Int. Ed. 2013, 52, 7160.
(9) Stratakis, M.; Garcia, H. Chem. Rev. 2012, 112, 4469.
(10) Gryparis, C.; Stratakis, M. Chem. Commun. 2012, 48, 10751.
(11) Gryparis, C.; Kidonakis, M.; Stratakis, M. Org. Lett. 2013, 15,
6038.
(12) We believe that the active catalytic sites are cationic gold species
at the Au nanoparticle−metal oxide support interface region:
(a) Fierro-Gonzalez, J. C.; Gates, B. C. Chem. Soc. Rev. 2008, 37,
2127. (b) Brown, M. A.; Fujimori, Y.; Ringleb, F.; Shao, X.; Stavale, F.;
Nilius, N.; Sterrer, M.; Freund, H.-J. J. Am. Chem. Soc. 2011, 133,
10668. (c) Hong, S.; Rahman, T. S. J. Am. Chem. Soc. 2013, 135, 7629.
(d) Wu, Y. Y.; Mashayekhi, N. A.; Kung, H. H. Catal. Sci. Technol.
2013, 3, 2881.
(13) PhMe2Si-O-Bpin is chromatographically unstable but was
undoubtedly characterized in crude reaction mixtures by comparison
to known spectroscopic data: Kleeberg, C.; Cheung, M. S.; Lin, Z.;
Marder, T. B. J. Am. Chem. Soc. 2011, 133, 19060. PhMe2Si-O-Bpin is
the main product in the attempted silaboration of alkynes catalyzed by
Au/Al2O3 in nondried DCM (see the NMR spectra in page S11 of
Supporting Information).
(14) (a) Tamao, K.; Hayashi, T.; Kumada, M. J. Organomet. Chem.
1976, 114, C19. (b) Suginome, M.; Oike, H.; Ito, Y. J. Am. Chem. Soc.
1995, 117, 1665.
(15) Chenard, B. L.; van Zyl, C. M. J. Org. Chem. 1986, 51, 3561.
1433
dx.doi.org/10.1021/ol500225x | Org. Lett. 2014, 16, 1430−1433