Base-Metal-Catalyzed Regiodivergent Alkene Hydrosilylations

Article abstract of DOI:10.1002/anie.201601197

A complementary set of base metal catalysts has been developed for regiodivergent alkene hydrosilylations: iron complexes of phosphine-iminopyridine are selective for anti-Markovnikov hydrosilylations (linear/branched up to >99:1), while the cobalt complexes bearing the same type of ligands provide an unprecedented high level of Markovnikov selectivity (branched/linear up to >99:1). Both systems exhibit high efficiency and wide functional group tolerance. Regiodivergent alkene hydrosilylation has been accomplished with high efficiency using a newly developed set of complementary base metal catalyst systems. An inversion of regioselectivity (linear/branched) from >99:1 to <1:99 is obtained when the iron version of the catalyst is exchanged for a cobalt-containing analogue.

Full text of DOI:10.1002/anie.201601197

Angewandte
Communications
Chemie
International Edition: DOI: 10.1002/anie.201601197
German Edition: DOI: 10.1002/ange.201601197
Markovnikov Hydrosilylation
Base–Metal-Catalyzed Regiodivergent Alkene Hydrosilylations
Xiaoyong Du+, Yanlu Zhang+, Dongjie Peng, and Zheng Huang*
Abstract: A complementary set of base metal catalysts has
been developed for regiodivergent alkene hydrosilylations:
iron complexes of phosphine-iminopyridine are selective for
anti-Markovnikov hydrosilylations (linear/branched up to
> 99:1), while the cobalt complexes bearing the same type of
ligands provide an unprecedented high level of Markovnikov
selectivity (branched/linear up to > 99:1). Both systems exhibit
high efficiency and wide functional group tolerance.
A
lkene hydrosilylation is one of the largest volume reactions
conducted with homogeneous catalysts.[1] Platinum-based
Speier and Karstedtꢀs complexes are the most widely used
hydrosilylation catalysts because of their high activities
and ease of handling.[2] However, over the last decade the
low abundance, high cost, and environmental concerns asso-
ciated with precious metals have spurred extensive study
toward developing earth-abundant base metal alternatives.[3]
To date, a few iron,[4] cobalt,[4j,5] and nickel[6] catalysts have
been developed for anti-Markovnikov alkene hydrosilyla-
tions.
While anti-Markovnikov alkene hydrosilylations are well-
established, reactions demonstrating Markovnikov selectivity
are rare. Several lanthanum metal catalysts are known for
Markovnikov hydrosilylation of activated alkenes, such as
styrene and acrylonitrile.[7] Markovnikov hydrosilylation of
unactivated alkenes using nickel[8] and palladium[9] catalysts
have been reported, but the existing systems suffer from
moderate to low regioselectivities and limited substrate scope.
Thus, a general method for selective Markovnikov alkene
hydrosilylation requires development. Driven by our interest
in developing base metal catalysts for alkene functionaliza-
tions,[4e, 10] herein we report the synthesis of iron and cobalt
complexes of phosphine-iminopyridine (PCNN) ligands. The
iron catalysts described are highly selective for anti-
Markovnikov hydrosilylations, whereas the corresponding
cobalt catalysts offer the opposite regioselectivity.
Scheme 1. Synthesis of phosphine-iminopyridine (PCNN) ligands and
iron and cobalt complexes, and the ORTEP structures for 5a and 6c.
backbone with a methene (CH2) unit.[11] The synthesis of the
PCNN ligands is outlined in Scheme 1.[12] The acetyl group of
1-(6-methylpyridin-2-yl)ethanone (1) was first protected with
ethylene glycol. Deprotonation of the Me group at the
2-pyridyl position in 2 with LDA, followed by addition of
tBu2PCl or iPr2PCl and subsequent deprotection of the acetyl
group with HCl, generated phosphine-acetylpyridines (3a,
3b). Condensations of 3a and 3b with arylamines bearing iPr
or Me substituents at the 2,6-aryl positions gave the PCNN
ligands 4a–4c in 88–90% yield. The FeII and CoII dichloride
complexes 5a–c and 6a–c were formed in high yields by
reaction of 4a–c with FeCl2 or CoCl2. Single-crystal X-ray
diffraction studies of complexes 5a and 6c revealed distorted
square pyramidal geometries around the metal centers
(Scheme 1).[13]
Initial studies of the catalytic activities of these new iron
and cobalt complexes focused on the hydrosilylation of
a simple aliphatic alkene, 1-octene, with PhSiH3. The results
are summarized in Table 1. In situ activation of iron and
cobalt dihalides with NaBHEt3 has been reported for catalytic
alkene hydrofunctionalizations.[4e, 10c,14] Using NaBHEt3
(2 mol %) as the activator, the most sterically hindered iron
complex (tBuPCNNiPr)FeCl2 (5a; 1 mol %) is active for anti-
Markovnikov hydrosilylation. The reaction of 1-octene with
PhSiH3 in THF at 258C formed the linear product 8a in 98%
yield with excellent regioselectivity (entry 1). Reducing the
During the course of our earlier studies on the phosphin-
ite-iminopyridine (PONN) iron catalysts for chemoselective
anti-Markovnikov alkene hydrosilylations,[4e] we found that
O
À
the P NN ligands are prone to degradation by O P bond
cleavage. To this end, we sought to prepare more robust
phosphine-iminopyridine (PCNN) ligands by replacing the
O-linker connecting the phosphorus atom and the pyridyl
[*] X. Du,[+] Dr. Y. Zhang,[+] Dr. D. Peng, Prof. Dr. Z. Huang
State Key Laboratory of Organometallic Chemistry,
Shanghai Institute of Organic Chemistry
345 Lingling Road, Shanghai 200032 (China)
E-mail: huangzh@sioc.ac.cn
[+] These authors contributed equally to this work.
Supporting information for this article can be found under:
http://dx.doi.org/10.1002/anie.201601197.
Angew. Chem. Int. Ed. 2016, 55, 6671 –6675
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
6671

Angewandte
Communications
Chemie
Table 1: Iron- and cobalt-catalyzed 1-octene hydrosilylation with
1-octene and PhSiH3 gave 9a in quantitative yield after 6 h
(TON = 2000, entry 13). Thus, an additional additive is not
required for activation of the catalyst in the cobalt-catalyzed
hydrosilylation reactions.[15] In a recent patent, Boyer et al.
PhSiH3.[a]
reported that
(2,4,6-Me3C6H2N CMe)2C5H3N equiv) and
a
=
combination of bis(imino)pyridine
(
MesPDI;
3
Entry Cat.
[mol %]
NaBHEt3
[mol %]
T
t
Yield [%]
CoCl2 catalyzes Markovnikov alkene hydrosilylation with
PhSiH3. It was proposed that MesPDI acts as a promoter for
catalyst activation.[16] However, using the conditions de-
scribed in the patent, the reaction gave only 5% of 9a and
4% of 8a (entry 14) in the presence of CoCl2 (0.35 mol %)
and MesPDI (1.1 mol %). To explore the role of our iPrPCNNMe
ligand 4c in the hydrosilylation process, an experiment in the
presence of additional ligand was conducted. The reaction
with 6c (0.05 mol%) and 4c (10 equiv relative to 6c) was less
effective than the reaction without 4c, implying that there is
an inhibitory effect of the additional ligand on the catalysis
(entries 15 vs. 13). Finally, a simple combination of 4c and
CoCl2 (1 mol % each) was effective for Markovnikov hydro-
silylation (entry 16), albeit with reduced catalytic efficiency in
comparison with the reaction employing 6c.
[8C] [h] 8a 9a
1
5a (1)
5b (1)
5c (1)
5a (1)
5a (0.1)
5a (0.1)
5a (0.02)
6a (1)
6b (1)
6c (1)
2
2
2
0
0.2
0.2
0.08
2
2
2
0
0
0
0
0
0
25
25
25
60
25
25
25
25
25
25
25
25
60
60
60
60
3
3
3
3
98
72
50
0
1
2[b]
3[c]
4
1
6
0
1
5
24 96
24 93 <1
24 79 <1
3
3
3
3
6[d,e]
7[d,f]
8
3
6
2
23
49
98
9
10
11
12
6c (1)
6c (1)
<1 62
12 <1 98
13[d,e] 6c (0.05)
6
6
6
<1 >99
14[g]
CoCl2 (0.35)/MesPDI (1.1)
4
5
15[d,e] 6c (0.05)/4c (0.5)
<1 78
16
CoCl2/4c (1)
24 <1 79
Utilizing the base metal catalysts, we examined the
substrate scope with respect to alkenes for regiodivergent
hydrosilylations (Scheme 2). The iron catalyst mediates
hydrosilylation of a diverse array of alkenes with PhSiH3,
furnishing linear products in high isolated yield with excellent
regioselectivities. Most reactions employed 0.5 mol % of 5a
and 1 mol % NaBHEt3. Functional groups including chloride
(8g), protected alcohol (8i), ether (8k), acetal (8m), and
gem-disubstituted olefin (8q) were tolerated. Hydrosilylation
of 1,5-hexadiene with 2 equiv of PhSiH3 produced
1,6-disilylhexane (8r) in high yield. Allylarenes gave the
desired products in moderate yields (8s, 60%; 8sa, 65%) with
the formation of a small amount of dehydrogenative silylation
products, (E)-allylsilanes (ca. 20%). Styrene was hydrosily-
lated in high yield (8t). Furthermore, using the least crowded
complex 5c (1 mol %) as the precatalyst, hydrosilylations of
styrene with various secondary and tertiary silanes in neat
conditions occurred smoothly (8ta–8te). Notably, reaction of
1-octene with (EtO)3SiH[17] or Ph2SiH2 gave a mixture of the
hydrosilylation (8u or 8v) and the dehydrogenative silylation
products (allylsilane, 10u or 10v).[18]
Awide variety of functionalized alkenes were subjected to
selective cobalt-catalyzed Markovnikov hydrosilylation
(9 f–q). Most reactions were performed at 608C with very
low catalyst loadings (0.05–0.5 mol % 6c). Reactions of
substrates bearing bromide (9h, 9sc), ester (9o), amide (9p),
and 1,1-disubstituted olefin (9q) functional groups proceeded
in a chemo- and regioselective manner. Markovnikov addi-
tions to the two terminal double bonds in 1,5-hexadiene
formed the branched disilyl product (9r). Allylbenzene
and its derivatives, which bear both electron-donating
and withdrawing groups, were efficiently hydrosilylated
(9s–9sd). An exception was the reaction of styrene (9t),
which gave the branched and linear products in an approx-
imately 1:1 ratio. Markovnikov hydrosilylations with secon-
dary and tertiary silanes are more difficult than that with
primary silane. The catalyst generated from 6b and NaBHEt3
was identified as the optimal system for Markovnikov hydro-
[a] Conditions: 7a (0.6 mmol), PhSiH3 (0.6 mmol) in THF (1 mL). The
yields were determined by GC with mesitylene as an internal standard.
[b] 2% (E)-allylsilane was detected. [c] 7% (E)-allylsilane was detected.
[d] In neat conditions. [e] 3 mmol scale. [f] 10 mmol scale. [g] 7a
(1.6 mmol), PhSiH3 (1.6 mmol) in THF (1 mL).
steric hindrance of the PCNN ligands resulted in inferior
activity and selectivity (entries 2, 3). A control experiment in
the absence of NaBHEt3 revealed that the activator is
necessary to achieve catalytic conversion with the iron
complex 5a (entry 4). The reactions using 0.1 mol % 5a,
with or without solvent, gave 8a in nearly quantitative yield
after 24 h (entries 5, 6). In the presence of 0.02 mol % of 5a
(51 ppm iron metal), the reaction in neat conditions gave 8a
in 79% yield (TON = 3950, entry 7). The results indicate that
complex (tBuPCNNiPr)FeCl2 5a is significantly more active than
the precatalyst (tBuPONNiPr)FeCl2 bearing an O-linker; the
hydrosilylation reaction using the latter required 1 mol % of
catalyst to attain good conversion.[4e] The selective formation
of the anti-Markovnikov product is consistent with earlier
reports using iron catalyst systems.[4]
Remarkably, regioselectivity could be reversed when
cobalt analogues were employed. All cobalt complexes 6a–c
gave the Markovnikov product 9a as the major product
(entries 8–10). In particular, the least crowded precatalyst
(
iPrPCNNMe)CoCl2 6c was very effective for Markovnikov
hydrosilylation. The reaction in the presence of 1 mol % of 6c
and 2 mol % of NaBHEt3 at 258C formed 9a in 98% yield
after 3 h (entry 10). Surprisingly, control experiments with 6c
in the absence of NaBHEt3, afforded the branched product 9a
in 62 and 98% yield after 3 and 12 h, respectively (entries 11,
12). These runs gave regioselectivity similar to that observed
in the reaction using a combination of 6c and NaBHEt3, albeit
with a relatively slow reaction rate (entries 11 vs. 10). At
elevated temperature (608C), in the presence of 1100 ppm of
6c (0.05 mol %, 130 ppm cobalt metal), the reaction in neat
6672 www.angewandte.org
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 6671 –6675

Angewandte
Communications
Chemie
Scheme 2. Iron-catalyzed anti-Markovnikov and cobalt-catalyzed Markovnikov hydrosilylation of various alkenes. General conditions: 7 (1.2 mmol),
PhSiH3 (1.2 mmol) in THF (0.5–1 mL). Yield of isolated products, unless noted otherwise. Product ratios were determined by 1H NMR
spectroscopy. [a] With cat. (0.1 mol %). [b] With cat. (1 mol %). [c] With cat. (2 mol %). [d] 25% (E)-allylsilane detected. [e] 23% (E)-allylsilane
detected. [f] With 5c (1 mol %) and NaBHEt3 (2 mol %), 1.8 mmol scale, in neat conditions. [g] With 5c (2 mol %) and NaBHEt3 (4 mol %), yield
determined by NMR. [h] With 6c (0.05 mol %). [i] Neat conditions. [j] With 6c (2 mol %) and NaBHEt3 (4 mol %). [k] Yield determined by NMR.
[l] With 6b (1 mol %) and NaBHEt3 (2 mol %) in toluene (1 mL), yield determined by NMR. 10u (18%) and 10v (15%) detected.
silylations of 1-octene with Ph2SiH2 (9u, 53%) and (EtO)3SiH
(9v, 60%), though a small amount of allylsilanes (10u, 18%;
10v, 15%) was also detected.
Potential applications of Markovnikov hydrosilylations
include the conversion of readily available a-olefins into
synthetically versatile secondary alcohols. Satisfactorily, treat-
ment of various Markovnikov hydrosilylation products with
HBF4·Et2O, followed by oxidation with H2O2 under basic
conditions, produced secondary alcohols 11 in useful yields
(Supporting Information, Scheme S1).
For iron-catalyzed anti-Markovnikov alkene hydrosilyla-
tions, mechanistic proposals involving oxidative addition of
silane, followed by olefin insertion and reductive elimination
of linear alkylsilane, have previously been elucidated.[4b]
Although additional experimental and theoretical studies
are ongoing in an attempt to gain a full understanding of the
mechanism of cobalt-catalyzed Markovnikov hydrosilyla-
tions, our preliminary experiments provide interesting results.
A radical mechanism is not very likely, as the reactions of
dienes 7q and 7r did not form any cyclized products
(Scheme 2). More importantly, a diallylether (12) known to
undergo rapid radical cyclization under conditions that allow
the generation of a radical,[19] reacted with 2 equiv of PhSiH3
to form a dual hydrosilylation product 13 in 86% yield; no
cyclized products were detected during the catalysis [Eq. (1)].
Lastly, addition of 10 mol % of TEMPO (10 equiv relative to
cobalt) as a radical trap imposed no detrimental effects,
although increasing the amount of TEMPO further to
100 mol %, led to complete inhibition of catalysis (see the
Supporting Information).[20]
To provide insight into the catalyst activation mode and
the nature of the cobalt intermediacy, stoichiometric and
deuterium-labeling experiments were conducted. Treatment
of
(
iPrPCNNMe)CoCl2 (6c) with 1 equiv of PhSiH3 in
[D6]benzene at 608C for 4 h formed a diamagnetic, putative
CoI monochloride (iPrPCNNMe)CoCl (14) in 49% yield, along
with PhSiH2Cl in 27% yield [Eq. (2)].[21] To conclusively
determine the identity of the CoI species, 14 was independ-
ently prepared using a procedure known for transforming CoII
dichloride into CoI monochloride.[22] The reaction of 6c with
1 equiv of LiMe in C6D6 at room temperature gave 54% of
the monochloride (14) and another diamagnetic species 15 in
10% yield. Further addition of 1.5 equiv of LiMe to the
mixture resulted in the disappearance of 14 and build-up of
15, which was identified as
(
iPrPCNNMe)CoMe (71%;
1
[Eq. (3)]). The NMR spectra of 15 display a diagnostic H
doublet at À0.62 ppm and a 13C doublet at À58.1 ppm
corresponding to the Co-Me moiety. Importantly, both CoI
complexes 14 and 15 were active for catalytic Markovnikov
hydrosilylation [Eq. (4)]. The data indicates that the CoII
Angew. Chem. Int. Ed. 2016, 55, 6671 –6675
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 6673

Angewandte
Communications
Chemie
dichloride can be reduced/activated with PhSiH3 to form
a catalytically active CoI species.
ciency, mild conditions, and broad functional group compat-
ibility combined with a divergent and high level of regio-
chemistry.
The reaction of 1-dodecene-d2 with PhSiH3 in the
presence of 6c (1 mol %) gave 9c-d2 selectively [Eq. (5)].
Additionally, the reaction of 1-octene with the deuterated
silane PhSiD3 yielded 9a-d3 [Eq. (6)]. No deuterium incor-
poration into the a-carbon of the products was observed in
either case. These results argue strongly against the formation
of a cobalt deuteride or hydride intermediate in the catalytic
cycle, because facile H/D exchange between Co-D and Ca-H
Acknowledgements
Financial support was provided by the National Basic
Research Program of China (2015CB856600) and the
National Natural Science Foundation of China (21422209,
21432011, and 21421091).
would otherwise occur through reversible olefin insertion into
[10d,18]
À
the Co D bond.
Keywords: alkenes · cobalt · homogeneous catalysis ·
hydrosilylation · iron
How to cite: Angew. Chem. Int. Ed. 2016, 55, 6671–6675
Angew. Chem. 2016, 128, 6783–6787
[1] a) A. F. Noels in Industrial Applications of Homogeneous
Catalysis (Ed: A. Mortreux), Kluwer, Amsterdam, 1985,
pp. 80 – 91; b) A. K. Roy in Adv. Organomet. Chem., Vol. 55
(Eds.: A. F. H. Robert West, J. F. Mark), Academic Press, San
Diego, 2007, pp. 1 – 59.
[2] a) J. L. Speier, J. A. Webster, G. H. Barnes, J. Am. Chem. Soc.
1957, 79, 974; b) B. D. Karstedt, US patent 3,715,334, 1973.
[3] a) R. M. Bullock, Catalysis Without Precious Metals, Wiley-
VCH, Weinheim, 2010, pp. 83 – 42; b) K. Junge, K. Schrçder, M.
Beller, Chem. Commun. 2011, 47, 4849; c) B. D. Sherry, A.
Fürstner, Acc. Chem. Res. 2008, 41, 1500; d) I. Bauer, H.-J.
Knçlker, Chem. Rev. 2015, 115, 3170; e) M. D. Greenhalgh, A. S.
Jones, S. P. Thomas, ChemCatChem 2015, 7, 190; f) J. Sun, L.
Deng, ACS Catal. 2016, 6, 290.
[4] For examples: a) J. C. Mitchener, M. S. Wrighton, J. Am. Chem.
Soc. 1981, 103, 975; b) S. C. Bart, E. Lobkovsky, P. J. Chirik, J.
Am. Chem. Soc. 2004, 126, 13794; c) A. M. Tondreau, C. C.
Atienza, K. J. Weller, S. A. Nye, K. M. Lewis, J. G. Delis, P. J.
Chirik, Science 2012, 335, 567; d) C. C. Hojilla Atienza, A. M.
Tondreau, K. J. Weller, K. M. Lewis, R. W. Cruse, S. A. Nye, J. L.
Boyer, J. G. P. Delis, P. J. Chirik, ACS Catal. 2012, 2, 2169; e) D.
Peng, Y. Zhang, X. Du, L. Zhang, X. Leng, M. D. Walter, Z.
Huang, J. Am. Chem. Soc. 2013, 135, 19154; f) M. D. Green-
halgh, D. J. Frank, S. P. Thomas, Adv. Synth. Catal. 2014, 356,
584; g) J. Y. Wu, B. N. Stanzl, T. Ritter, J. Am. Chem. Soc. 2010,
132, 13214; h) R. N. Naumov, M. Itazaki, M. Kamitani, H.
Nakazawa, J. Am. Chem. Soc. 2012, 134, 804; i) D. Noda, A.
Tahara, Y. Sunada, H. Nagashima, J. Am. Chem. Soc. 2016, 138,
2480.
[5] For examples: a) A. J. Chalk, J. F. Harrod, J. Am. Chem. Soc.
1965, 87, 1133; b) C. L. Reichel, M. S. Wrighton, J. Am. Chem.
Soc. 1981, 103, 7180; c) M. Brookhart, B. E. Grant, J. Am. Chem.
Soc. 1993, 115, 2151; d) Z. Mo, Y. Liu, L. Deng, Angew. Chem.
Int. Ed. 2013, 52, 10845; Angew. Chem. 2013, 125, 11045; e) C.
Chen, M. B. Hecht, A. Kavara, W. W. Brennessel, B. Q. Mer-
cado, D. J. Weix, P. L. Holland, J. Am. Chem. Soc. 2015, 137,
13244.
On the basis of our preliminary results and reported
precedent regarding cobalt-catalyzed alkene functionaliza-
tion,[10d,18] we propose a silyl migration pathway involving
a
CoI silyl intermediate (Supporting Information,
Scheme S4).[23] For apolar olefins, 1,2-insertion of the C C
=
À
double bond into the Co Si bond prevails (presumably
because of steric effects),[24] furnishing the Markovnikov
product. An alternative, but less likely pathway, involves
2,1-insertion of the olefin, which leads to formation of the
anti-Markovnikov product. For vinylarenes, 2,1-insertion may
compete with the 1,2-insertion pathway because there is
marked electronic discrimination between the two olefinic
carbon atoms. This proposal is consistent with the data
obtained from the hydrosilylation of styrene (9t; Scheme 2).
In summary, we have prepared a series of iron and cobalt
alkene hydrosilylation catalysts bearing phosphine-iminopyr-
idine ligands. The regioselectivity of alkene hydrosilylations
can be inverted from > 99:1 to < 1:99 by changing the metal
center of the catalyst from iron to cobalt. The advantages
these non-precious metal catalysts offer include, high effi-
[6] I. Buslov, J. Becouse, S. Mazza, M. Montandon-Clerc, X. Hu,
Angew. Chem. Int. Ed. 2015, 54, 14523; Angew. Chem. 2015, 127,
14731.
[7] For examples: a) P.-F. Fu, L. Brard, Y. Li, T. J. Marks, J. Am.
Chem. Soc. 1995, 117, 7157; b) G. A. Molander, E. D. Dowdy,
B. C. Noll, Organometallics 1998, 17, 3754; c) T. I. Gountchev,
T. D. Tilley, Organometallics 1999, 18, 5661; d) Y. Chen, D.
Zargarian, Can. J. Chem. 2009, 87, 280.
[8] For examples: a) M. Kumada, K. Sumitani, Y. Kiso, K. Tamao, J.
Organomet. Chem. 1973, 50, 319; b) L. F. Groux, D. Zargarian,
6674 www.angewandte.org
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 6671 –6675

Angewandte
Communications
Chemie
Organometallics 2003, 22, 4759; c) Y. Chen, C. Sui-Seng, S.
[17] Caution: HSi(OEt)3 can potentially form explosive gases, and
Boucher, D. Zargarian, Organometallics 2005, 24, 149.
[9] a) N. Komine, M. Abe, R. Suda, M. Hirano, Organometallics
2015, 34, 432; b) T. Hayashi, Y. Uozumi, Pure Appl. Chem. 1992,
64, 1911.
thus should be handled with the proper precautions!
[18] C. C. H. Atienza, T. Diao, K. J. Weller, S. A. Nye, K. M. Lewis,
J. G. P. Delis, J. L. Boyer, A. K. Roy, P. J. Chirik, J. Am. Chem.
Soc. 2014, 136, 12108.
[19] a) B. Kopping, C. Chatgilialoglu, M. Zehnder, B. Giese, J. Org.
Chem. 1992, 57, 3994; b) A. L. J. Beckwith, I. Blair, G. Phillipou,
J. Am. Chem. Soc. 1974, 96, 1613; c) P. Burkhard, E. Roduner, J.
Hochmann, H. Fischer, J. Phys. Chem. 1984, 88, 773.
[20] Inhibition by TEMPO is not a definitive indication of a radical
mechanism because TEMPO may halt catalysis by reaction with
PhSiH3 or the cobalt alkyl group. For example, see: A. C.
AlbØniz, P. Espinet, R. López-Fernµndez, A. Sen, J. Am. Chem.
Soc. 2002, 124, 11278.
[10] a) L. Zhang, D. Peng, X. Leng, Z. Huang, Angew. Chem. Int. Ed.
2013, 52, 3676; Angew. Chem. 2013, 125, 3764; b) L. Zhang, Z.
Zuo, X. Wan, Z. Huang, J. Am. Chem. Soc. 2014, 136, 15501;
c) L. Zhang, Z. Zuo, X. Leng, Z. Huang, Angew. Chem. Int. Ed.
2014, 53, 2696; Angew. Chem. 2014, 126, 2734; d) L. Zhang, Z.
Huang, J. Am. Chem. Soc. 2015, 137, 15600; e) X. Jia, Z. Huang,
Nat. Chem. 2016, 8, 157.
[11] Y. Cao, Y. Zhang, L. Zhang, D. Zhang, X. Leng, Z. Huang, Org.
Chem. Front. 2014, 1, 1101.
[21] Treatment of 6c with PhSiH3 (2 equiv) at 608C after 2 h gave 14
in 76% yield (See the Supporting Information for a proposed
pathway).
[22] T. M. Kooistra, Q. Knijnenburg, J. M. M. Smits, A. D. Horton,
P. H. M. Budzelaar, A. W. Gal, Angew. Chem. Int. Ed. 2001, 40,
4719; Angew. Chem. 2001, 113, 4855.
[12] During the preparation of this manuscript, Milstein et al.
reported the synthesis of a similar PCNN ligand containing
a mesitylimino group. A different procedure was used, which
required five steps starting from commercially available materi-
als, and furnished the ligand in low yield (< 10%). See: B.
Butschke, K. L. Fillman, T. Bendikov, L. J. W. Shimon, Y.
Diskin-Posner, G. Leitus, S. I. Gorelsky, M. L. Neidig, D.
Milstein, Inorg. Chem. 2015, 54, 4909.
[13] The supplementary crystallographic data for this paper are
available at CCDC 1415554 (5a) and 1415555 (6c). This data
can be obtained free of charge from The Cambridge
Crystallographic Data Centre.
[14] M. W. Bouwkamp, A. C. Bowman, E. Lobkovsky, P. J. Chirik, J.
Am. Chem. Soc. 2006, 128, 13340.
[15] Activation of Co(acac)3 and Co(OAc)2 without additional
reagents has been reported for catalytic alkene hydrogenation,
see: M. R. Friedfeld, M. Shevlin, J. M. Hoyt, S. W. Krska, M. T.
Tudge, P. J. Chirik, Science 2013, 342, 1076.
[23] Note that CoI silyl complexes have been documented. See
Ref. [18] and Z. Mo, J. Xiao, Y. Gao, L. Deng, J. Am. Chem. Soc.
2014, 136, 17414. Here the CoI monochloride 14 may react with
PhSiH2Cl derived from the reaction of 6c with PhSiH3 [see
Eq. (2)] to yield a CoI hydride. The CoI hydride then reacts with
PhSiH3 to give the CoI silyl complex (See the Supporting
Information for further details).
[24] P. Wucher, L. Caporaso, P. Roesle, F. Ragone, L. Cavallo, S.
Mecking, I. Gçttker-Schnetmann, Proc. Natl. Acad. Sci. USA
2011, 108, 8955.
Received: February 3, 2016
Revised: March 16, 2016
Published online: April 25, 2016
[16] J. L. Boyer, A. K. Roy, PCT Int. Appl. WO 2014/186513 A1,
2014.
Angew. Chem. Int. Ed. 2016, 55, 6671 –6675
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
6675