Titanium-catalyzed reductive umpolung reactions with a metal-free terminal reducing agent

Article abstract of DOI:10.1002/chem.201500102

A new method for titanium-catalyzed reductive umpolung reactions is reported that overcomes the traditional requirement for a stoichiometric metallic reductant. With N,N′-disilylated tetramethyldihydropyrazine as a potent organic reducing agent, reductive carbonyl-nitrile, enone-acrylonitrile and pinacol coupling reactions can be achieved in good yields and stereoselectivities. [Cp₂TiI₂] is a superior catalyst to [Cp₂TiCl₂], which is rationalized by a faster generation of the active catalyst [Cp₂TiI]. A mechanism is proposed that is in agreement with the experimental results. Replacing zinc: A protocol for titanium(III)-catalyzed reductive umpolung reactions is presented that enables the title reactions in the presence of an N,N′-disilylated tetramethyldihydropyrazine as an organic sacrificial reducing agent. It is successfully applied to carbonyl-nitrile, enone-acrylonitrile and pinacol coupling reactions. A remarkable effect of the titanocene counterion renders titanocene diiodide a superior catalyst.

Full text of DOI:10.1002/chem.201500102

DOI: 10.1002/chem.201500102
Communication
&
Umpolung Reactions |Very Important Paper|
Titanium-Catalyzed Reductive Umpolung Reactions with a Metal-
Free Terminal Reducing Agent
Georg Frey, J. Niklas Hausmann, and Jan Streuff*[a]
Abstract: A new method for titanium-catalyzed reductive
umpolung reactions is reported that overcomes the tradi-
tional requirement for a stoichiometric metallic reductant.
With N,N’-disilylated tetramethyldihydropyrazine as
a potent organic reducing agent, reductive carbonyl–ni-
trile, enone–acrylonitrile and pinacol coupling reactions
can be achieved in good yields and stereoselectivities.
[Cp2TiI2] is a superior catalyst to [Cp2TiCl2], which is ration-
alized by a faster generation of the active catalyst [Cp2TiI].
A mechanism is proposed that is in agreement with the
experimental results.
Scheme 1. Concept of a titanium-catalyzed reductive ketone–nitrile coupling
using metal-free reductant 3.
Metal-catalyzed reductive umpolung reactions such as the pi-
nacol coupling allow unconventional CÀC bond formations be-
tween similarly polarized functionalities.[1,2] However, a strong
limitation of these methodologies has been the requirement
of a stoichiometric metallic reducing agent, which usually
leads to inhomogeneous reaction mixtures and the formation
of undesired Lewis-acid salts as byproducts. Therefore, it has
been a longstanding goal to replace the stoichiometric metal
reductant with a metal-free alternative. We report herein tita-
nium(III)-catalyzed reductive umpolung reactions in the pres-
ence of an organic reducing agent and [Cp2TiI2] as catalyst.
Recently, our group explored the concept of reductive um-
polung towards cross-coupling reactions that had little litera-
ture precedence.[3–5] This way, the direct synthesis of a-hy-
droxyketones was achieved by a coupling of ketones and ni-
triles in the presence of an in situ-generated low-valent titani-
um catalyst and zinc dust (Scheme 1).[4a] In search of a substi-
tute for the stoichiometric zinc metal and of simplified,
homogeneous reaction conditions, we became interested in
disilylated 1,4-cyclohexadiene 1 and dihydropyrazine 2, which
were reported in 1962 and 1971, respectively.[6] Both represent
strong organic reductants and silylating agents. Mashima and
co-workers recently pioneered the application of such reagents
in the reduction of transition metal complexes and catalysts.[7]
Furthermore, a tetramethylated dihydropyrazine derivative 3
was developed that exhibited enhanced reducing properties
and a lower tendency to inhibit catalysts by coordination.[8]
The reagent was found suitable for the generation of low-
valent tantalum and titanium species and was employed in
stoichiometric experiments as well as catalytic Reformatsky-
type reactions.[8]
We therefore decided to investigate 3 as an organic reduc-
tant for titanium-catalyzed ketone–nitrile couplings and related
reactions. The optimization of reaction conditions was carried
out using the cross-coupling of acetophenone 4a and benzyl
cyanide 5a to a-hydroxyketone 6a (Table 1).[4b] Initially, only
small amounts of product were formed with 10 mol% of tita-
nocene dichloride ([Cp2TiCl2]) and 2 equivalents of 3 and tri-
Table 1. Optimization of catalyst and conditions.
Entry
Catalyst
Equiv 3
Base·HX
Yield [%]
1
2
3
4
5
6
7
8
[Cp2TiCl2]
[(EtCp)2TiCl2]
[Cp2TiBr2]
[Cp2TiI2]
none
[Cp2TiI2]
2.0
2.0
2.0
2.0
2.0
2.0
1.0
1.0
Et3N·HCl
Et3N·HCl
Et3N·HCl
Et3N·HCl
Et3N·HCl
Et3N·HI
18
16
33
65
0
46
5
68
[a] G. Frey, J. N. Hausmann, Dr. J. Streuff
Institut fꢀr Organische Chemie
Albert-Ludwigs-Universitꢁt Freiburg
Albertstraße 21, 79104 Freiburg (Germany)
Fax: (+49)761-203-8715
E-mail: jan.streuff@ocbc.uni-freiburg.de
[Cp2TiCl2]
[Cp2TiI2]
Et3N·HI
Et3N·HCl
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201500102.
Chem. Eur. J. 2015, 21, 1 – 5
1
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Communication
ethylamine hydrochloride each (Table 1, entry 1). Although al-
kylation of the Cp ligand did not lead to an improvement, tita-
nocene dibromide and diiodide both gave significantly higher
yields (Table 1, entries 2–4). No reaction took place in absence
of the titanium catalyst and an exchange of the triethylamine
hydrochloride with its hydroiodide led to lower yields (Table 1,
entries 5 and 6). Interestingly, a combination of [Cp2TiCl2] and
Et3N·HI gave almost no yield of 6a (Table 1, entry 7). It should
be noted that Et3N·HI dissolved significantly better in THF than
Et3N·HCl. Finally, the amount of 3 could be reduced to 1.0
equivalent, affording the desired product in 68% yield (Table 1,
entry 8). Further changes in concentration, reaction time, tem-
perature, or hydrochloride amount did not have a positive
effect on the reaction outcome.[9] The ammonium hydrochlo-
ride additive was required for the reaction to occur and addi-
tion of chlorotrimethylsilane or tetramethylpyrazine had no
effect.
(Table 2, entry 4). The direct synthesis of a-aminoketone 8 was
achieved in a good yield of 70% by reductive coupling of
imine 7 with benzyl cyanide (Table 2, entry 5).[4b] The combina-
tion of [Cp2TiI2] and 3 was also successful in cyclization reac-
tions to give products 10, 12, and pyrrolidin-3-one 14, in 57–
68% yields (Table 2, entries 6–8).[4c] We further carried out an
enantioselective cyclization of ketonitrile 9 with Brintzinger’s
ansa-titanocene catalyst [(R,R)-(ebthi)TiCl2] (Scheme 2; ebthi=
These conditions were then applied to a number of different
carbonyl-nitrile coupling reactions (Table 2).[4b] Acetophenone
was successfully coupled to 4-methoxy benzyl cyanide 5b and
butyronitrile 5c (Table 2, entries 2 and 3). 4-Bromoacetophe-
none could also be employed as substrate and hydroxyketone
6d was obtained without loss of the halide in 60% yield
Scheme 2. Enantioselective ketonitrile cyclization with organic reductant 3.
ethylenebis(tetrahydroindenyl)).[4d,10] Since the corresponding
diiodide was not accessible by simple chloride–iodide meta-
thesis, the dichloride was employed. Nevertheless, the desired
product could be isolated in 34% yield with a high 88% ee,
the remaining material being substrate 9 (previously enantio-
selectivity with Zn: 91% ee).[4d] In general, all reactions de-
scribed herein were chemoselective with unreacted starting
material completing the mass balance.
Table 2. Reductive cross-couplings between ketones/imines and nitriles
using 3 as reductant.[a]
Entry Carbonyl
Nitrile
Product
Yield [%]
68
The metal-free reductant was then employed in our reduc-
tive cross-coupling reactions between Michael acceptors, yield-
ing 1,6-ketonitriles (Scheme 3).[4a,c] With cyclohexenone 15 and
1
2
3
4
4a
4a
4a
5a
56
38
60
5a
5a
5
6
7
70
68
61
Scheme 3. Reductive umpolung of electron-deficient alkenes.
acrylonitrile as coupling partners, the b-cyanoalkylated cyclo-
hexanone 16 was isolated in 53% yield after workup with
tetra-n-butylammonium fluoride (TBAF). The 4-quinolone deriv-
ative 17 reacted in a similar fashion and product 18 was ob-
tained in 73% yield as a single diastereoisomer.
The feasibility of Mashima’s reagent 3 was further probed in
titanium-catalyzed pinacol couplings of aromatic aldehydes.[11]
Using various precursors typically moderate to good yields
were achieved (Table 3). A good 85:15 diastereomeric ratio
(rac/meso) was obtained for benzaldehyde (Table 3, entry 1a).
It improved slightly with [Cp2TiCl2] as catalyst (88:12) but again
8
57
[a] Reactions on a 0.5 mmol scale with 14 h reaction time. For conditions,
see Table 1, entry 8.
&
&
Chem. Eur. J. 2015, 21, 1 – 5
www.chemeurj.org
2
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Communication
showed characteristics of both the electrochemical and the
metallic reduction. A more detailed study will be reported in
due course to further elucidate this behavior. As an important
property of reagent 3, we found it to be stable in solution
over prolonged periods of time upon exclusion of oxygen. This
allowed the preparation of stock solutions for convenient han-
dling of the reagent, which were used in the cyclic voltamme-
try experiments.[9]
Table 3. Diastereoselective pinacol coupling of aromatic aldehydes with
3 as terminal reductant.
Entry
Major product
rac/meso
Yield [%][a]
A simplified reaction mechanism is depicted in Scheme 4,
showing the pinacol coupling of benzaldehyde 19a. Reagent 3
reduces the titanocene dihalogenide to a titanium(III) species,
1a
1b
85:15
88:12
70
40[b]
2
92:8
66
3
4
93:7
47
73
85:15
[a] Combined yield of isolated products; [b] reaction with [Cp2TiCl2] as
catalyst.
the yield was decreased (40%; Table 3, entry 1b). With 4-chlo-
robenzaldehyde, the 92:8 d.r. was in range of the previously re-
ported value (Table 3,entry 2).[11d] The reaction worked well
with further aromatic aldehydes, such as 4-methoxybenzalde-
hyde and 2-thiophenecarboxaldehyde (2-thenaldehyde;
Table 3, entries 3 and 4). Overall, the diastereoselectivities were
comparable to the values reported with zinc as reductant and
MgBr2 as additive or under buffered protic conditions with Mn
and collidine hydrochloride.[11] Still, the origin of the stereose-
lectivity could not be disclosed since the previously proposed
transition states were in disagreement with our observations.
The formation of an intermediate M2+-bridged titanium dimer
Scheme 4. Proposed catalytic cycle for the pinacol coupling of 19a with
[Cp2TiI2] and 3 as reductant.
releasing one equivalent of iodotrimethylsilane per molecule
of the reduced catalyst.[9] In analogy to recent studies with
[Cp2TiIIICl] solutions, the reduced titanium(III) species is likely in
equilibrium with the ammonium chloride coordination com-
plex 22.[15] This stabilizes the catalyst in solution and prevents
premature deactivation by decomposition pathways. Aldehyde
coordination to the monomeric [Cp2TiIIII] followed by radical
combination yields the bis-titanium(IV)-coordinated pinacol
product. Based on our observations, we conclude that one
iodide remains at the titanium center in the course of the reac-
tion. Silylation of the titanium-oxygen bonds closes the catalyt-
ic cycle and instantly consumes the trimethylsilyl iodide (TMSI)
that is formed in the catalytic process. Thus, TMSI is not
formed as a byproduct. In comparison, metallic reductants pro-
duce stoichiometric quantities of a Lewis acid, which could
cause side reactions and required removal afterwards.
was ruled out due to absence of M2+ [11,12]
as well as the re-
,
vised pentavalent transition state proposed by Nicholas and
co-workers,[13] which did not match the tetravalent nature of
the titanium center.
As noted, the reactions were faster with [Cp2TiI2] than with
[Cp2TiCl2], which was an unprecedented counter ion effect.
Daasbjerg and co-workers reported that electrochemical reduc-
tion of [Cp2TiI2] initially gave [Cp2TiIIII2]À. Formation of the cata-
lytically active [Cp2TiIIII], by dissociation of IÀ, was significantly
faster in comparison to [Cp2TiIIICl2]À.[14] This would lead to
a higher concentration of the active catalyst in our case, ex-
plaining the observed effect. We confirmed the presence of ti-
tanium(III) species by cyclic voltammetry.[9] Interestingly, a mix-
ture of [Cp2TiIIII2]À, [Cp2TiIIII], [(Cp2TiIIII)2] and [Cp2Ti]+ was detect-
ed. Previously, the formation of [Cp2TiI2]À had only been de-
tected in electrochemical reduction experiments while the for-
mation of [Cp2Ti]+ had been an exclusive feature of the
reduction with metals.[14a] The reduction with 3, however,
In conclusion, we have established a procedure for titanium-
catalyzed reductive umpolung reactions in the presence of
a metal-free reducing agent. [Cp2TiI2] was found to be a superi-
or catalyst to [Cp2TiCl2] and preliminary studies confirmed the
generation of titanium(III) in the presence of tetramethylated
dihydropyrazine derivative 3. The protocol was successfully ap-
Chem. Eur. J. 2015, 21, 1 – 5
www.chemeurj.org
3
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Communication
2006, 8, 5433–5436; b) U. Groth, M. Jung, T. Vogel, Chem. Eur. J. 2005,
11, 3127–3135; c) M. Jung, U. Groth, Synlett 2002, 2015–2018; d) R. K.
Boeckman Jr., R. A. Hudack Jr., J. Org. Chem. 1998, 63, 3524–3525;
e) E. J. Corey, G. Z. Zheng, Tetrahedron Lett. 1997, 38, 2045–2048.
[6] a) R. A. Sulzbach, A. F. M. Iqbal, Angew. Chem. Int. Ed. Engl. 1971, 10, 127;
Angew. Chem. 1971, 83, 145; b) L. H. Toporcer (Dow Corning), Bissilylcy-
clohexadiꢄnes, prꢂparation et emploi. Fr. Patent FR1327544(A), May 17,
1963; c) D. R. Weyenberg, L. H. Toporcer, J. Am. Chem. Soc. 1962, 84,
2843–2844.
[7] a) H. Tsurugi, H. Tanahashi, H. Nishiyama, W. Fegler, T. Saito, A. Sauer, J.
Okuda, K. Mashima, J. Am. Chem. Soc. 2013, 135, 5986–5989; b) H. Tsur-
ugi, T. Saito, H. Tanahashi, J. Arnold, K. Mashima, J. Am. Chem. Soc.
2011, 133, 18673–18683; c) R. Arteaga-Mꢅller, H. Tsurugi, T. Saito, M. Ya-
nagawa, S. Oda, K. Mashima, J. Am. Chem. Soc. 2009, 131, 5370–5371.
[8] T. Saito, H. Nishiyama, H. Tanahashi, K. Kawakita, H. Tsurugi, K. Mashima,
J. Am. Chem. Soc. 2014, 136, 5161–5170.
plied to a number of cross-coupling reactions between carbon-
yls and nitriles as well as Michael acceptors. Furthermore, good
yields and high diastereoselectivities were obtained for the pi-
nacol coupling of various aromatic aldehydes. These results, to-
gether with the observed longevity in solution, suggest that 3
is a useful organic alternative for reductive umpolung reac-
tions. Details on the reaction mechanism and further applica-
tions will follow in due course.
Acknowledgements
The Fonds der Chemischen Industrie (Liebig Fellowship to J.S.)
and the Deutsche Forschungsgemeinschaft (DFG, grant STR
1150/3-1) are acknowledged for financial support. We thank
Selina Schierl for initial experimental contributions.
[9] See the Supporting Information for further details.
[10] F. R. W. P. Wild, L. Zsolnai, G. Huttner, H. H. Brintzinger, J. Organomet.
Chem. 1982, 232, 233–247.
[11] a) A. Gansꢁuer, D. Bauer, Eur. J. Org. Chem. 1998, 2673–2676; b) A. Gan-
sꢁuer, M. Moschioni, D. Bauer, Eur. J. Org. Chem. 1998, 1923–1927; c) A.
Gansꢁuer, Synlett 1997, 363–364; d) A. Gansꢁuer, Chem. Commun. 1997,
457–458.
[12] a) D. W. Stephan, Organometallics 1992, 11, 996–999; b) Y. Handa, J. Ina-
naga, Tetrahedron Lett. 1987, 28, 5717–5718; c) R. Jungst, D. Sekutow-
ski, G. Stucky, J. Am. Chem. Soc. 1974, 96, 8108–8109; d) C. G. Vonk, J.
Cryst. Mol. Struct. 1973, 3, 201–207; e) R. S. P. Coutts, P. C. Wailes, R. L.
Martin, J. Organomet. Chem. 1973, 47, 375–382; f) J.-J. Salzmann, Helv.
Chim. Acta 1968, 51, 526–529.
[13] M. S. Dunlap, K. M. Nicholas, J. Organomet. Chem. 2001, 630, 125–131.
[14] a) R. J. Enemærke, J. Larsen, G. H. Hjøllund, T. Skrydstrup, K. Daasbjerg,
Organometallics 2005, 24, 1252–1262; b) R. J. Enemærke, J. Larsen, T.
Skrydstrup, K. Daasbjerg, Organometallics 2004, 23, 1866–1874.
[15] a) A. Gansꢁuer, D. von Laufenberg, C. Kube, T. Dahmen, A. Michelmann,
M. Behlendorf, R. Sure, M. Seddiqzai, S. Grimme, D. V. Sadasivam, G. D.
Fianu, R. A. Flowers II, Chem. Eur. J. 2015, 21, 280–289; b) A. Gansꢁuer,
C. Kube, K. Daasbjerg, R. Sure, S. Grimme, G. D. Fianu, D. V. Sadasivam,
R. A. Flowers II, J. Am. Chem. Soc. 2014, 136, 1663–1671; c) A. Gansꢁuer,
M. Behlendorf, D. von Laufenberg, A. Fleckhaus, C. Kube, D. V. Sadasi-
vam, R. A. Flowers II, Angew. Chem. Int. Ed. 2012, 51, 4739–4742;
Angew. Chem. 2012, 124, 4819–4823.
Keywords: homogeneous catalysis
coupling · titanium · umpolung
· radicals · reductive
[1] a) J. Streuff, Synthesis 2013, 45, 281–307; b) T. Hirao, Top. Curr. Chem.
2007, 279, 53–75; c) A. Gansꢁuer, H. Bluhm, Chem. Rev. 2000, 100,
2771–2788. See also: d) D. A. Everson, D. J. Weix, J. Org. Chem. 2014,
79, 4793–4798; e) C. E. I. Knappke, S. Grupe, D. Gꢁrtner, M. Corpet, C.
Gosmini, A. Jacobi von Wangelin, Chem. Eur. J. 2014, 20, 6828–6842.
[2] For recent reviews on the pinacol coupling reaction, see: a) B. S. Terra,
F. Macedo Jr., ARKIVOC 2012, i, 134–151; b) A. Chatterjee, N. N. Joshi,
Tetrahedron 2006, 62, 12137–12158; c) T. Wirth, Angew. Chem. Int. Ed.
Engl. 1996, 35, 61–63; Angew. Chem. 1996, 108, 65–67.
[3] J. Streuff, Chem. Rec. 2014, 14, 1100–1113.
[4] a) P. Bichovski, T. M. Haas, D. Kratzert, J. Streuff, Chem. Eur. J. 2014, 21,
2339–2342; b) M. Feurer, G. Frey, H.-T. Luu, D. Kratzert, J. Streuff, Chem.
Commun. 2014, 50, 5370–5372; c) G. Frey, H.-T. Luu, P. Bichovski, M.
Feurer, J. Streuff, Angew. Chem. Int. Ed. 2013, 52, 7131–7134; Angew.
Chem. 2013, 125, 7271–7274; d) J. Streuff, M. Feurer, P. Bichovski, G.
Frey, U. Gellrich, Angew. Chem. Int. Ed. 2012, 51, 8661–8664; Angew.
Chem. 2012, 124, 8789–8792; e) J. Streuff, Chem. Eur. J. 2011, 17, 5507–
5510.
[5] For noteworthy examples, see: a) R. E. Estꢂvez, J. L. Oller-Lꢃpez, R.
Robles, C. R. Melgarejo, A. Gansꢁuer, J. M. Cuerva, J. E. Oltra, Org. Lett.
Received: January 10, 2015
Published online on && &&, 0000
&
&
Chem. Eur. J. 2015, 21, 1 – 5
www.chemeurj.org
4
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Communication
COMMUNICATION
&
Umpolung Reactions
G. Frey, J. N. Hausmann, J. Streuff*
&& – &&
Titanium-Catalyzed Reductive
Umpolung Reactions with a Metal-
Free Terminal Reducing Agent
Replacing zinc: A protocol for titani-
um(III)-catalyzed reductive umpolung
reactions is presented that enables the
title reactions in the presence of an
N,N’-disilylated tetramethyldihydropyra-
zine as an organic sacrificial reducing
agent. It is successfully applied to car-
bonyl–nitrile, enone–acrylonitrile and pi-
nacol coupling reactions. A remarkable
effect of the titanocene counterion ren-
ders titanocene diiodide a superior cata-
lyst.
Chem. Eur. J. 2015, 21, 1 – 5
www.chemeurj.org
5
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ