Catalysis via homolytic substitutions with C-O and Ti-O bonds: Oxidative additions and reductive eliminations in single electron steps

Article abstract of DOI:10.1021/ja907817y

In a combined theoretical and experimental study, an efficient catalytic reaction featuring epoxide opening and tetrahydrofuran formation through homolytic substitution reactions at C-O and Ti-O bonds was devised. The performance of these two key steps of

Full text of DOI:10.1021/ja907817y

Published on Web 11/04/2009
Catalysis via Homolytic Substitutions with C-O and Ti-O
Bonds: Oxidative Additions and Reductive Eliminations in
Single Electron Steps
Andreas Gansa¨uer,*,† Andre´ Fleckhaus,† Manuel Alejandre Lafont,† Andreas Okkel,†
Konstantinos Kotsis,‡ Anakuthil Anoop,‡ and Frank Neese*,‡
Kekule´-Institut fu¨r Organische Chemie und Biochemie der UniVersita¨t Bonn, Gerhard Domagk
Str. 1, 53121 Bonn, Germany and Institut fu¨r Physikalische und Theoretische Chemie der
UniVersita¨t Bonn, Wegeler Str. 12, 53115 Bonn, Germany
Received September 15, 2009; E-mail: andreas.gansaeuer@uni-bonn.de; neese@thch.uni-bonn.de
Abstract: In a combined theoretical and experimental study, an efficient catalytic reaction featuring epoxide
opening and tetrahydrofuran formation through homolytic substitution reactions at C-O and Ti-O bonds
was devised. The performance of these two key steps of the catalytic cycle was studied and could be
adjusted by modifying the electronic properties of the catalysts through introduction of electron-donating or
-withdrawing substituents to the titanocene catalysts. By regarding both steps as single electron versions
of oxidative addition and reductive elimination, a mechanism-based platform for the design of catalysts
and reagents for electron transfer reactions evolved that opens broad perspectives for further investigations.
Introduction
The microscopic reversal of these reactions, homolytic
substitutions with C-O bond cleavage, has only recently
attracted substantial synthetic interest.6 They were first discussed
in the context of epoxide deoxygenations through Cr(II)
complexes by Kochi.7 With the introduction of titanocene(III)
complexes for epoxide opening by Nugent and RajanBabu,8 the
resulting ꢀ-titanoxy radicals could for the first time be reduced
by hydrogen atom transfer8c and employed in cyclizations8a,d
with carbonyl compounds and nitriles9 or intermolecular addition
reactions to R,ꢀ-unsaturated olefins.8b The recently developed
C-O bonds can be formed or cleaved via homolytic
substitutions of carbon spin centers (radicals) or metal spin
centers with M-O or C-O bonds, respectively. These reactions
are discussed as key steps in a number of topical transformations.
Homolytic substitutions that lead to the formation of C-O bonds
include the ring closing step of ꢀ-metaloxy radicals in olefin
epoxidations mediated by chromium complexes1 or catalyzed
by manganese salen complexes,2 the hydroxylation of radicals
by organometallic complexes,3 and the titanocene catalyzed or
mediated tetrahydrofuran-synthesis via alkoxylation of carbon
centered radicals.4 Even the oxygen rebound step in the
hydroxylation reactions of hydrocarbons by the P450 enzymes
may be considered as another example of this type of transfor-
mation.5 The formation of the C-O bonds in these reactions is
accompanied by the cleavage of the M-O bond and a decrease
in the oxidation state of the metal by one. Formally, these
elementary steps therefore constitute single electron versions
of reductive eliminations.
(4) (a) Gansa¨uer, A.; Rinker, B.; Pierobon, M.; Grimme, S.; Gerenkamp,
M.; Mu¨ck-Lichtenfeld, C. Angew. Chem., Int. Ed. 2003, 42, 3687–
3690. (b) Trost, B. M.; Shen, H. C.; Surivet, J. P. Angew. Chem., Int.
Ed. 2003, 42, 3943–3947. (c) Leca, D.; Fensterbank, L.; Lacoˆte, E.;
Malacria, M. Angew. Chem., Int. Ed. 2004, 43, 4220–4222. (d)
Gans¨auer, A.; Rinker, B.; Ndene-Schiffer, N.; Pierobon, M.; Grimme,
S.; Gerenkamp, M.; Mu¨ck-Lichtenfeld, C. Eur. J. Org. Chem. 2004,
2337–2351. (e) Trost, B. M.; Shen, H. C.; Surivet, J. P. J. Am. Chem.
Soc. 2004, 126, 12565–12579. (f) Leca, D.; Song, K.; Albert, M.;
Goncalves, M. C.; Fensterbank, L.; Lacoˆte, E.; Malacria, M. Synthesis
2005, 1405–1420.
† Kekule´-Institut fu¨r Organische Chemie und Biochemie der Universita¨t
Bonn.
(5) For some recent reviews, see: (a) Newcomb, M.; Toy, P. H. Acc. Chem.
Res. 2000, 33, 449–455. (b) Shaik, S.; Cohen, S.; de Visser, S. P.;
Sharma, P. K.; Kumar, D.; Kozuch, S.; Ogliaro, F.; Danovich, D. Eur.
J. Inorg. Chem. 2004, 207–226. (c) Meunier, B.; de Visser, S. P.;
Shaik, S. Chem. ReV. 2004, 104, 3947–3980. (d) Shaik, S.; Kumar,
D.; de Visser, S. P.; Thiel, W. Chem. ReV. 2005, 105, 2279–2328. (e)
Denisov, I. G.; Makris, T. M.; Sligar, S. G.; Schlichting, I. Chem.
ReV. 2005, 105, 2253–2277.
‡ Institut fu¨r Physikalische und Theoretische Chemie der Universita¨t
Bonn.
(1) (a) Samsel, E. G.; Srinivasan, K.; Kochi, J. K. J. Am. Chem. Soc.
1985, 107, 7606–7617. (b) Srinivasan, K.; Michaud, P.; Kochi, J. K.
J. Am. Chem. Soc. 1986, 108, 2309–2320.
(2) (a) Zhang Loebach, J. L.; Wilson, S. R.; Jacobsen, E. N. J. Am. Chem.
Soc. 1990, 112, 2801–2803. (b) Irie, R.; Noda, K.; Ito, Y.; Katsuki,
T. Tetrahedron Lett. 1990, 31, 7345–7348. (c) Daly, A. M.; Renehan,
M. F.; Gilheany, D. G. Org. Lett. 2001, 3, 663–666.
(6) (a) Gansa¨uer, A. Synlett 1998, 801–809. (b) Gansa¨uer, A.; Bluhm, H.
Chem. ReV. 2000, 100, 2771–2788. (c) Gansa¨uer, A.; Narayan, S. AdV.
Synth. Catal. 2002, 344, 465–475. (d) Gansa¨uer, A.; Lauterbach, T.;
Narayan, S. Angew. Chem., Int. Ed. 2003, 42, 5556–5573. (e) Cuerva,
J. M.; Justicia, J.; Oller-Lo´pez, J. L.; Oltra, J. E. Top. Curr. Chem.
2006, 264, 63–92. (f) Barrero, A. F.; Qu´ılez del Moral, J. F.; Sa´nchez,
E. M.; Arteaga, J. F. Eur. J. Org. Chem. 2006, 1627–1641. (g)
Gansa¨uer, A.; Justicia, J.; Fan, C.-A.; Worgull, D.; Piestert, F. Top.
Curr. Chem. 2007, 279, 25–52.
(3) (a) Groves, J. T.; Viski, T. J. Org. Chem. 1990, 55, 3628–3634. (b)
Larrow, J. F.; Jacobsen, E. N. J. Am. Chem. Soc. 1994, 116, 12129–
12130. (c) Hamachi, K.; Irie, R.; Katsuki, T. Tetrahedron Lett. 1996,
37, 4979–4982. (d) Hamada, T.; Irie, R.; Mihara, J.; Hamachi, K.;
Katsuki, T. Tetrahedron 1998, 54, 10017–10028. (e) Katsuki, T. In
ComprehensiVe Asymmetric Catalysis II; Jacobsen, E. N., Pfaltz, A.,
Yamamoto, H., Eds.; Springer: Berlin, 1999, 791-799. (f) Katsuki,
T. Synlett 2003, 281–297.
(7) Kochi, J. K.; Singleton, D. M.; Andrews, L. J. Tetrahedron 1968, 24,
3503–3515.
9
10.1021/ja907817y CCC: $40.75  2009 American Chemical Society
J. AM. CHEM. SOC. 2009, 131, 16989–16999 16989

A R T I C L E S
Gansa¨uer, et al.
catalytic conditions10 have expanded the scope of this reaction
even further and have resulted in a number of useful synthetic
applications, such as the stereoselective generation of radicals,11
unusual cyclizations,12 epoxypolyene cyclizations via radicals,13
radical tandem processes,14 and polymerization reactions.15
These C-O bond cleavages are accompanied by the formation
of the M-O bond and an increase of the oxidation state of the
metals by one and hence constitute single electron versions of
oxidative additions.16
able to realize exactly such a Cp2TiCl (1) catalyzed process,
shown in Scheme 1.4
While the reaction is potentially of high synthetic value, bi-
and tricyclic tetrahydrofurans can be prepared atom-economi-
cally18 from simple starting materials; the low 67% yield of 3
suggests that 1 is by no means the most suitable catalyst.
Here, we describe a combined experimental and theoretical
approach to the development of much more efficient conditions
for this transformation. It is based on controlling the performance
of the single electron oxidative addition to the epoxide and the
reductive elimination leading to tetrahydrofuran formation
through carefully adjusting the electronic properties of the
titanocene catalysts. This can be achieved in an experimentally
straightforward manner by the introduction of electron-donating
or -withdrawing substituents to one or both of the cyclopenta-
dienyl ligands.
Due to its easy shuttling between the oxidation states three
and four,17 the Cp2TiCl2/Cp2TiCl couple is to date unique in
being able to efficiently induce both single electron oxidative
addition and single electron reductive elimination. This particular
characteristic is highly attractive for the development of catalytic
reactions featuring both processes as key-steps. We have been
To the best of our knowledge, this concept has not been
described in the literature for single electron transfer catalysis.
Thus, our results will be of relevance for the design of other
reagents or catalysts for reactions involving single electron
transfer steps.19 Potential applications include the highly
important transformations of ketyl radicals,20 anion radical
[2 + 2] cycloadditions,21 or reactions proceeding via SOMO
catalysis.22
(8) (a) Nugent, W. A.; RajanBabu, T. V. J. Am. Chem. Soc. 1988, 110,
8561–8562. (b) RajanBabu, T. V.; Nugent, W. A. J. Am. Chem. Soc.
1989, 111, 4525–4527. (c) RajanBabu, T. V.; Nugent, W. A.; Beattie,
M. S. J. Am. Chem. Soc. 1990, 112, 6408–6409. (d) RajanBabu, T. V.;
Nugent, W. A. J. Am. Chem. Soc. 1994, 116, 986–997.
(9) (a) Ferna´ndez-Mateos, A.; Martin de la Nava, E.; Pascual Coca, G.;
Ramos Silva, A.; Rubio Gonza´lez, R. Org. Lett. 1999, 1, 607–609.
(b) Ferna´ndez-Mateos, A.; Mateos Buro´n, L.; Rabanedo Clemente,
R.; Ramos Silva, A. I.; Rubio Gonza´lez, R. Synlett 2004, 1011–1014.
(c) Ferna´ndez-Mateos, A.; Herrero Teijo´n, P.; Rabanedo Clemente,
R.; Rubio Gonza´lez, R.; Sanz Gonza´lez, F. Synlett 2007, 2718–2722.
(d) Ferna´ndez-Mateos, A.; Herrero Teijo´n, P.; Mateo´s Buron, L.;
Rabanedo Clemente, R.; Rubio Gonza´lez, R. J. Org. Chem. 2007, 72,
9973–9982.
Results and Discussion
This investigation is firmly based on a close interaction
between theoretical and experimental studies for the identifica-
tion of an efficient catalyst for our reaction. The prime virtue
of this combined study is that theory provides the knowledge
of reaction and activation energies as well as the determina-
tion of the structures of pertinent intermediates of the oxidative
addition and reductive elimination. In this manner, the crucial
data for the evaluation of our concept will be provided. The
synthetic study will establish which of the complexes are stable
under the reaction conditions and which of these two elementary
reactions is more critical for an efficient catalytic reaction.
Moreover, only an experimental study can establish which
titanocene(III) complexes can effectively suppress undesired
intermolecular side-reactions, e.g., interception of the pivotal
radical intermediates by a second equivalent of the titanoce-
(10) (a) Gansa¨uer, A.; Pierobon, M.; Bluhm, H. Angew. Chem., Int. Ed.
1998, 37, 101–103. (b) Gansa¨uer, A.; Bluhm, H.; Pierobon, M. J. Am.
Chem. Soc. 1998, 120, 12849–12859. (c) Gansa¨uer, A.; Bluhm, H.
Chem. Commun. 1998, 2143–2144. (d) Barrero, A. F.; Rosales, A.;
Cuerva, J. M.; Oltra, J. E. Org. Lett. 2003, 5, 1935–1938.
(11) (a) Gansa¨uer, A.; Lauterbach, T.; Bluhm, H.; Noltemeyer, M. Angew.
Chem., Int. Ed. 1999, 38, 2909–2910. (b) Gansa¨uer, A.; Bluhm, H.;
Pierobon, M.; Keller, M. Organometallics 2001, 20, 914–919. (c)
Gansa¨uer, A.; Bluhm, H.; Lauterbach, T. AdV. Synth. Catal. 2001,
343, 785–787. (d) Gansa¨uer, A.; Bluhm, H.; Rinker, B.; Narayan, S.;
Schick, M.; Lauterbach, T.; Pierobon, M. Chem.sEur. J. 2003, 9,
531–542. (e) Gansa¨uer, A.; Fan, C.-A.; Keller, F.; Keil, J. J. Am. Chem.
Soc. 2007, 129, 3484–3485. (f) Gansa¨uer, A.; Fan, C.-A.; Keller, F.;
Karbaum, P. Chem.sEur. J. 2007, 13, 8084–8090.
(12) (a) Gansa¨uer, A.; Lauterbach, T.; Geich-Gimbel, D. Chem.sEur. J.
2004, 10, 4983–4990. (b) Friedrich, J.; Dolg, M.; Gansa¨uer, A.; Geich-
Gimbel, D.; Lauterbach, T. J. Am. Chem. Soc. 2005, 127, 7071–7077.
(c) Friedrich, J.; Walczak, K.; Dolg, M.; Piestert, F.; Lauterbach, T.;
Worgull, D.; Gansa¨uer, A. J. Am. Chem. Soc. 2008, 130, 1788–1796.
(13) (a) Justicia, J.; Rosales, A.; Bun˜uel, E.; Oller-Lo´pez, J. L.; Valdivia,
M.; Ha¨ıdour, A.; Oltra, J. E.; Barrero, A. F.; Ca´rdenas, D. J.; Cuerva,
J. M. Chem.sEur. J. 2004, 10, 1778–1788. (b) Justicia, J.; Oller-
Lopez, J. L.; Campan˜a, A. G.; Oltra, J. E.; Cuerva, J. M.; Bun˜uel, E.;
Ca´rdenas, D. J. J. Am. Chem. Soc. 2005, 127, 14911–14921. (c)
Cuerva, J. M.; Justicia, J.; Oller-Lo´pez, J. L.; Bazdi, B.; Oltra, J. E.
Mini-ReV. Org. Chem. 2006, 3, 2335, 89. (d) Gansa¨uer, A.; Rosales,
A.; Justicia, J. Synlett 2006, 927–929. (e) Gans¨; auer, A.; Justicia, J.;
Rosales, A.; Worgull, D.; Rinker, B.; Cuerva, J. M.; Oltra, J. E. Eur.
J. Org. Chem. 2006, 4115–4127.
(17) (a) Enemærke, R. J.; Hjøllund, G. H.; Daasbjerg, K.; Skrydstrup,
T. C. R. Acad. Sci. 2001, 4, 435–438. (b) Enemærke, R. J.; Larsen,
J.; Skrydstrup, T.; Daasbjerg, K. Organometallics 2004, 23, 1866–
1874. (c) Enemærke, R. J.; Larsen, J.; Skrydstrup, T.; Daasbjerg, K.
J. Am. Chem. Soc. 2004, 126, 7853–7864. (d) Enemærke, R. J.; Larsen,
J.; Hjøllund, G. H.; Skrydstrup, T.; Daasbjerg, K. Organometallics
2005, 24, 1252–1262. (e) Larsen, J.; Enemærke, R. J.; Skrydstrup,
T.; Daasbjerg, K. Organometallics 2006, 25, 2031–2036.
(18) (a) Trost, B. M. Science 1991, 254, 1471–1477. (b) Trost, B. M.
Angew. Chem., Int. Ed. 1995, 34, 259–281.
(19) For a review on catalytic radical reactions, see: Ford, L.; Jahn, U.
Angew. Chem., Int. Ed. 2009, 48, 6386–6389.
(14) (a) Gansa¨uer, A.; Pierobon, M.; Bluhm, H. Angew. Chem., Int. Ed.
2002, 41, 3206–3208. (b) Cuerva, J. M.; Campan˜a, A. G.; Justicia, J.;
Rosales, A.; Oller-Lo´pez, J. L.; Robles, R.; Ca´rdenas, D.; Bun˜uel, E.;
Oltra, J. E. Angew. Chem., Int. Ed. 2006, 45, 5522–5526. (c) Gansa¨uer,
A.; Fan, C.-A.; Piestert, F. J. Am. Chem. Soc. 2008, 130, 6916–6917.
(d) Campan˜a, A. G.; Bazdi, B.; Fuentes, N.; Robles, R.; Cuerva, J. M.;
Oltra, J. E.; Porcel, S.; Echavarren, A. Angew. Chem., Int. Ed. 2008,
47, 7515–7519.
(20) (a) Emonds, D. J.; Johnston, D.; Procter, D. J. Chem. ReV. 2004, 104,
3371–3403. (b) Nicolaou, K. C.; Ellery, S. P.; Chen, J. S. Angew.
Chem., Int. Ed. 2009, 48, 7140–7165.
(21) (a) Baik, T.-G.; Luiz, A. L.; Wang, L.-C.; Krische, M. J. J. Am. Chem.
Soc. 2001, 123, 6716–6717. (b) Wang, L.-C.; Jang, H.-Y.; Roh, Y.;
Lynch, V.; Schultz, A. J.; Wang, X.; Krische, M. J. J. Am. Chem.
Soc. 2002, 124, 9448–9453. (c) Yang, J.; Felton, G. A. N.; Bauld,
N. L.; Krische, M. J. J. Am. Chem. Soc. 2004, 126, 1634–1635. (d)
Ischay, M. A.; Anzovino, M. E.; Du, J.; Yoon, T. P. J. Am. Chem.
Soc. 2008, 130, 12886–12887.
(15) (a) Asandei, A. D.; Moran, I. W. J. Am. Chem. Soc. 2004, 126, 15932–
15933. (b) Asandei, A. D.; Moran, I. W. J. Polym. Sci., Part A: Polym.
Chem. 2006, 44, 6028–6038. (c) Asandei, A. D.; Moran, I. W. J.
Polym. Sci., Part A: Polym. Chem. 2005, 43, 6039–6047. (d) Asandei,
A. D.; Moran, I. W. J. Polym. Sci., Part A: Polym. Chem. 2006, 44,
1060–1070. (e) Asandei, A. D.; Chen, Y.; Saha, G.; Moran, I. W.
Tetrahedron 2008, 64, 11831–11838.
(22) (a) Beeson, T. D.; Mastracchio, A.; Hong, J. B.; Aston, K.; MacMillan,
D. C. W. Science 2007, 316, 582–585. (b) Conrad, J. C.; Kong, J.;
MacMillan, D. C. W. J. Am. Chem. Soc. 2009, 131, 11640–11641.
(c) Nagib, D. A.; Scott, M. E.; MacMillan, D. C. W. J. Am. Chem.
Soc. 2009, 131, 10875–10877.
(16) Halpern, J. Acc. Chem. Res. 1970, 3, 386–392.
9
16990 J. AM. CHEM. SOC. VOL. 131, NO. 46, 2009

Catalysis via Homolytic Substitution
A R T I C L E S
Scheme 1. Cp2TiCl (1) Catalyzed Tandem Sequence Featuring Homolytic C-O Bond Cleavage and Formation (Coll ) 2,4,6-Trimethyl
Pyridine) and Its Mechanism
ne(III) reagent. Thus, the combination of these complementary
methods is unique in providing a mechanism-based platform
for efficient catalyst design.
frequency calculations that were performed by two-sided
numerical differentiation of analytic gradients. Zero Point
Vibrational Energy (ZPVE) corrections were computed from
the calculated harmonic frequencies and are included in the
reaction energies and activation energies throughout. All
calculations on open-shell species employed the spin-unrestricted
formalism.
In order to avoid a possible bias introduced into the modeling
through a specific method, density functional as well as wave
function-based correlation methods were investigated. The
methods used include the “gold standard” B3LYP method,28
the more efficient “pure” BP functional,24 and the spin-
component scaled second-order many body perturbation theory
(SCS-MP2).29
The tetrahydrofuran solvent was simulated by a polarizable
continuum model (COSMO).30 Molecular dynamics runs were
performed with the CAChe software31 and the MM3 force
field.32 A few of the lowest energy structures (typically 3-5)
from a 100-ps trajectory were further optimized using den-
sity functional theory (BP86/TZVP) and the lowest energy
conformer was regarded as the most significant for this study.
Epoxide Complexation and Oxidative Addition. We chose
the complexation of oxirane and oxidative addition to this
Theoretical Studies
Methods. The ORCA23 electronic structure program package
was used for all calculations. Structures and transition states
were optimized with the BP86 functional24 and Ahlrichs’
TZVP25 basis set, making use of the resolution-of-the-identity
RI approximation (in the Split-RI-J variant26 with the appropriate
Coulomb fitting sets).27 The recently implemented transition
state searcher of the ORCA package is based on eigenvector
following and is very efficient due to the use of partial Hessians.
Transition states were found by first performing a relaxed surface
scan along the bonds to be broken. The maximum energy
structures of these scans were taken as a guess for the transition
state search. Transition states and minima were verified through
(23) ORCAsan ab initio, density functional and semiempirical program
package. Written by F. Neese with contributions from Ute Becker,
Dmitry Ganyushin, Dimitrios Liakos, Simone Kossmann, Taras
Petrenko, Christoph Riplinger, and Frank Wennmohs. Version 2.6,
revision 71, Universita¨t Bonn, 2008.
(24) (a) Becke, A. D. Phys. ReV. A 1988, 38, 3098–3100. (b) Perdew, J. P.
Phys. ReV. B 1986, 33, 8822–8824.
(25) Scha¨fer, A.; Huber, C.; Ahlrichs, R. J. Chem. Phys. 1994, 100, 5829–
5835.
(28) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648–5652. (b) Lee, C. T.;
Yang, W. T.; Parr, R. G. Phys. ReV. B 1988, 37, 785–789.
(29) Grimme, S. J. Chem. Phys. 2003, 118, 9095–9102.
(30) Klamt, A.; Schuurmann, G. J. Chem. Soc., Perkin Trans. 1993, 2,
793–805.
(26) (a) Dunlap, B. I.; Connolly, J. W. D.; Sabin, J. R. J. Chem. Phys.
1979, 71, 3396–3402. (b) Baerends, E. J.; Ellis, D. E.; Ros, P. Chem.
¨
Phys. 1973, 2, 41–51. (c) Eichkorn, K.; Treutler, O.; Ohm, H.; Ha¨ser,
M.; Ahlrichs, R. Chem. Phys. Lett. 1995, 242, 652–660.
(27) (a) Eichkorn, K.; Treutler, O.; o¨hm, H.; Ha¨ser, M.; Ahlrichs, R. Chem.
Phys. Lett. 1995, 240, 283–289. (b) Eichkorn, K.; Weigend, F.;
Treutler, O.; Ahlrichs, R. Theor. Chem. Acc. 1997, 97, 119–124.
(31) CAChe, CAChe Scientific, 18700 N. W. Walker Rd., Bldg. 92-01,
Beaverton, OR 97006, 1-800-544-6634.
(32) Allinger, N. L.; Yuh, Y. H.; Lii, J.-H. J. Am. Chem. Soc. 1989, 111,
8551–8565.
9
J. AM. CHEM. SOC. VOL. 131, NO. 46, 2009 16991

A R T I C L E S
Gansa¨uer, et al.
Scheme 2. Model System for the Study of the Complexation of and the Oxidative Addition to Epoxides
substrate as our model system for epoxide opening (Scheme
2). Since steric interactions between this epoxide and the ligands
are minimal and the question of the regioselectivity of ring-
opening does not arise, the electronic effects of the substituents
can be studied with the smallest possible interference by other
factors. The reaction proceeds in two discrete steps via com-
plexation of the epoxide to yield B and inner-sphere electron
transfer from titanium in B to yield the ꢀ-titanoxy radical A.
The inner sphere electron transfer in B, that yields the pivotal
ꢀ-titanoxy radical A, constitutes the single electron oxidative
addition and is the critical step of epoxide opening. The reaction
energies of this reaction are summarized in Table 2.
As for the complexation, the absolute values of the reaction
energies of the oxidative additions depend to some extent on
the computational method employed. Interestingly, the SCS-
MP2 values are quite similar to the B3LYP results and are
uniformly about 6 kcal mol-1 less exothermic than those
predicted with the BP functional. However, the trends along
the series of catalysts obtained by each of the three methods
are again quite similar. Since only these differences are essential
for the discussion of the relative reactivity of the catalysts, our
data provide a most satisfactory basis for further discussions.
For all methods, the oxidative addition to ethylene oxide is
thermodynamically most favorable for the methyl substituted
catalyst 10. The importance of steric interactions is highlighted
by the bulky tert-butyl substituent in 11 that renders the
oxidative addition less exothermic for all methods by about 4
kcal mol-1. Electron withdrawing substituents disfavor the
oxidative addition. This is a reflection of their reduced electron
transfer ability that seems to be more pertinent than the stronger
Ti-O bonds formed with more Lewis acidic titanocenes. The
ester-substituted complexes 7 and 8 are leading to the least
exothermic reaction. It is surprising that complex 9, containing
a nitro group and the most powerful electron withdrawing
substituent studied here, results in more exothermic oxidative
additions than 7 and 8.
Finally, we note that in solution, where a fast exchange
between the solvent (usually THF) and the epoxide substrates
at titanium is occurring,17 the oxidative addition constitutes the
decisive contribution in the generation of the pivotal ꢀ-titanoxy
radicals. In the gas phase, however, the complexation is an
essential step of the complete sequence because of the absence
of solvation. These overall reaction energies that are obtained
by addition of the energies of both steps are summarized in
Table 3.
For all complexes, the overall gas phase reaction energies
are exothermic. However, they follow no clear tendency. The
trends for the reaction energies of the complexation and the
oxidative addition are opposing and since neither of the steps
is dominating, no clear-cut dependency of the overall gas phase
reaction energy on the substitution pattern of the cyclopenta-
dienyl ligand is observed.
Table 1. Reaction Energies (∆E) of the Formation of B in kcal
mol-1 (def2-TZVPP basis set)a
cat
R, R′
BP
B3LYP
SCS-MP2
1
4
5
6
7
8
9
10
11
H, H
H, Cl
H, OMe
H, CN
H, CO2Me
CO2Me, CO2Me
H, NO2
H, Me
H, t-Bu
-1.2
-2.1
-0.8
-2.3
-3.9
-2.4
-5.0
0.3
-2.2
-3.0
-2.0
-3.2
-4.1
-2.8
-4.5
-0.7
0.2
-10.9
-12.2
-11.5
-12.3
-13.3
-11.9
-13.0
-9.9
1.0
-9.9
a ZPVE calculated by using BP86/TZVP is included.
The results for the formation of B are summarized in Table
1. Whereas both DFT methods give similar results for the
complexation energies of ethylene oxide, the SCS-MP2 method
predicts a substantially more exothermic reaction. However, the
trends of the complexation energies are similar for all methods
employed.
In agreement with intuition, epoxide binding is most favorable
for the more Lewis-acidic titanocenes 9, 7, and 6 containing
the electron withdrawing -NO2, -CO2Me, and -CN group,
respectively. Complex 8, the only disubstituted titanocene
investigated here, behaves differently for sterical reasons. The
epoxide ligand is positioned parallel to the Ti-Cl bond and
thus rotated by 90° compared to the other titanocenes where
the epoxide is oriented perpendicular to the Ti-Cl bond, as
shown in Figure 1. The tert-butyl substituted 11 displays the
lowest binding energies due to the combination of the bulkiness
of its ligand and its reduced Lewis-acidity.
Finally, we note that the Ti-O distances are shortest for the
complexes with electron withdrawing groups due to their
increased Lewis-acidity (233 pm for 1, 228 pm for 7, and 225
pm for 8). This particular point is of prime importance for the
design of novel catalysts for stereoselective ring-opening
reactions11 because the steric interactions between the complex
and the substrates can be fine-tuned by electronically modulating
the cyclopentadienyl ligands.
Activation Energy. For an understanding of the reactivity of
the titanocene complexes, knowledge of the activation energies
9
16992 J. AM. CHEM. SOC. VOL. 131, NO. 46, 2009

Catalysis via Homolytic Substitution
A R T I C L E S
Figure 1. Structures of the oxirane complexes of 1, 7, and 8.
Table 2. Reaction Energies (∆E) of the Ring-Opening of B in kcal
mol-1 (def2-TZVPP basis set) via Oxidative Addition to Ethylene
Oxidea
mark calculations on open-shell transition states indicate that
the results obtained with the BP functional tend to be signifi-
cantly too low, while those resulting from SCS-MP2 are mostly
too high.33 Hence, in absolute terms, preference should probably
be given to the B3LYP values that are generally intermediate
between BP and SCS-MP2. Gratifyingly, and similar to what
was observed for the reaction energies, the trends of the
activation energies are nearly identical for all methods.
For all methods, the oxidative addition is predicted to be
slowest for the electron deficient complexes, especially 8, and
fastest for 4, 10, and 11. The high Ea for 8 is most likely also
due to the already observed steric effects in this complex. In
the monosubstituted complexes, steric effects are less pertinent.
This is because the presence of the tert-butyl group in 11 does
not lead to a higher activation energy than for 1 or 10 for all
methods employed.
cat
R, R′
BP
B3LYP
SCS-MP2
1
4
5
6
7
8
9
10
11
H, H
H, Cl
H, OMe
H, CN
H, CO2Me
CO2Me, CO2Me
H, NO2
H, Me
H, t-Bu
-7.0
-6.7
-5.8
-6.5
-4.7
-5.2
-6.2
-10.5
-6.5
-1.1
-0.5
0.3
-0.6
0.7
-0.8
-0.4
-1.2
-0.3
1.1
-0.5
-0.6
-4.3
0.0
0.1
-1.5
-4.3
-0.3
a ZPVE calculated by using BP86/TZVP is included.
Table 3. Overall Reaction Energies (∆E) of the Formation of A
from the Titanocenes and Ethylene Oxide via B in kcal mol-1
(def2-TZVPP basis set) through Complexation and Oxidative
Additiona
cat
R, R′
BP
B3LYP
SCS-MP2
1
4
5
6
7
8
9
10
11
H, H
H, Cl
H, OMe
H, CN
H, CO2Me
CO2Me, CO2Me
H, NO2
H, Me
H, t-Bu
-8.2
-8.8
-3.3
-3.5
-1.7
-3.8
-3.4
-2.7
-5.0
-3.6
-0.1
-11.7
-12.6
-12.7
-12.6
-12.1
-12.4
-13.6
-14.2
-9.9
-6.6
-8.8
-10.4
-7.6
-11.2
-10.2
-5.5
a ZPVE calculated by using BP86/TZVP is included.
Table 4. Activation Energies (Ea) in kcal mol-1 (def2-TZVPP basis
set) of Oxidative Addition to Ethylene Oxide from Ba
Figure 2. Transition states of oxidative addition to ethylene oxide by 1
and 7.
cat
R, R′
BP
B3LYP
SCS-MP2
1
4
5
6
7
8
9
10
11
H, H
H, Cl
H, OMe
H, CN
H, CO2Me
CO2Me, CO2Me
H, NO2
H, Me
H, t-Bu
7.4
5.3
7.9
8.6
8.4
9.4
8.9
7.5
7.2
14.3
12.9
14.7
14.9
14.4
15.9
14.9
14.4
13.5
20.6
19.0
21.5
22.6
22.9
24.6
23.4
19.6
19.7
In the transition state, the epoxide is largely cleaved and hence
product-like,34 as indicated by the C-O bond lengths of 181
pm TS1 and 184 pm for TS7, respectively (Figure 2). The C-O
bond length in ethylene oxide is 143 pm. In accordance with
the Hammond postulate34 TS7 is “later” than TS1. Because the
transition structures are product-like, the same factors affecting
the reaction energies should also dominate the activation
energies although to a lesser extent. In general, this trend is
indeed observed and TS6, TS7, TS8, and TS9 with the electron
deficient titanocenes are therefore less favorable than TS1,
TS10, and TS11 featuring the electron rich complexes, even
though the differences are small.
a ZPVE calculated by using BP86/TZVP is included.
of the oxidative additions is essential. In this manner, relative
rates can be predicted that are essential for issues of selectivity
when undesired side-reactions are to be avoided.
The activation energies of the opening of ethylene oxide are
summarized in Table 4. They were obtained by locating the
transition states for each catalyst system. As observed for the
reaction energies, the absolute values of the activation energies
differ significantly depending on the method employed. Bench-
The B3LYP values for Ea, that are most reliable in an absolute
sense, are high compared to typical radical reactions, such as
(33) Schwabe, T.; Grimme, S. Acc. Chem. Res. 2008, 41, 569–579.
(34) Hammond, G. S. J. Am. Chem. Soc. 1955, 77, 334–338.
9
J. AM. CHEM. SOC. VOL. 131, NO. 46, 2009 16993

A R T I C L E S
Gansa¨uer, et al.
Scheme 3. Model System for the Study of the Tetrahydrofuran Forming Reductive Elimination
Table 5. Reaction Energies (∆E) in kcal mol-1 (def2-TZVPP basis
set) of the Reductive Elimination of 12 to 13a
radical reduction by a second equivalent of the titanocene,8,35
or the 5-exo cyclization.36 However, in our tandem sequence,
epoxide deoxygenation after the undesired trapping by a second
equivalent of the titanocene(III) should not be critical, because
the 5-exo cyclization ensuing oxidative addition is much faster
than intermolecular radical trapping.
In summary, the computational study demonstrates that the
oxidative addition to ethylene oxide is a thermodynamically
possible process for all complexes investigated. However, the
B3LYP and SCS-MP2 methods predict the reaction energy of
ring-opening with the electron deficient 7 and 8 as well as the
bulky 11 to be close to zero or even slightly endothermic. Also
the oxidative addition is considerably slower for titanocenes with
electron withdrawing substituents than for titanocenes with
electron donating substituents. This is in line with the experience
from oxidative additions proceeding in two electron steps that
are generally faster for more electron rich complexes.
Reductive Elimination. In order to devise efficient catalysts
for the reductive elimination leading to tetrahydrofuran forma-
tion, it is crucial to establish if this step is thermodynamically
favorable. Moreover, it is essential to know the activation
energies of the process in order to devise conditions suppressing
undesired side-reactions. This may be especially critical for an
intermolecular radical reduction by the titanocene (III) reagents
before the single electron reductive elimination. We chose model
system 12, lacking only the ester groups of the synthetic
substrate, for studying these issues (Scheme 3). The calculations
were carried out with the BP and B3LYP methods because they
had delivered results similar to the SCS-MP2 method in the
study of the oxidative addition. After the reductive elimination,
the titanocene complex is generated in an unsolvated form.
Potentially interactions with the solvent could significantly
influence the reaction energy. Therefore, COSMO-calculations
mimicking the THF solvent of the reaction were also included.
These values are summarized in Tables 5 and 6.
cat
R, R
BP
B3LYP
B3LYP (COSMO)
1
H, H
cis
trans
cis
trans
cis
trans
cis
trans
cis
trans
-12.6
-1.4
-11.4
-0.3
-10.8
-0.3
-12.5
-1.4
-18.9
-7.6
-18.1
-6.5
-17.7
-6.7
-19.1
-7.5
-21.4
-9.3
-20.3
-8.8
-20.7
-9.8
-22.6
-11.7
-22.9
-13.6
6
H, CN
7
H, CO2Me
CO2Me, CO2Me
H, t-Bu
8
11
-12.5
-3.9
-19.5
-10.2
a ZPVE calculated by using BP86/TZVP is included.
Table 6. Activation Energies (kcal mol-1; def2-TZVPP basis set) of
the Reductive Elimination of 12 to 13a
cat
R, R
BP
B3LYP
1
H, H
cis
trans
cis
2.4
9.9
3.3
5.1
11.1
5.6
6
H, CN
trans
cis
trans
cis
trans
cis
12.8
2.9
7.9
3.0
8.1
13.0
5.8
9.0
5.6
9.0
7
H, CO2Me
CO2Me, CO2Me
H, t-Bu
8
11
5.6
8.2
trans
10.1
11.2
a ZPVE calculated by using BP86/TZVP is included.
thermically). Thus, the solvation of the respective substrates and
products is almost equally favorable. We suggest that the more
favorable interaction of THF with the more Lewis-acidic Ti(IV)
in 12 compensates for the solvation of the two products of the
reductive elimination to some extent because the Ti(III) center
formed constitutes a weaker Lewis-acid. For the purposes of
this study, solvation effects are therefore surprisingly small.
It is nevertheless clear, that 7 constitutes a much stronger
Lewis-acid than 1, because the computational solvent exchange
between 1*THF and 7 that leads to 1 and 7*THF is exothermic
by 16.6 kcal mol-1 (Scheme 4).
For an understanding of the overall catalytic cycle, the
activation energies (Table 6) and transition structures (Figure
3) of the reductive elimination, that are discussed next, are
probably even more relevant than the reaction energies. This is
because they give an indication of the relative rates of the
elementary steps and provide the basis for a discussion of
potential side-reactions and their prevention.
All methods employed predict the formation of the cis-13
system to be thermodynamically highly favored over the
generation of trans-fused 13 by about 10 kcal mol-1. In
agreement with chemical intuition, this reflects the higher strain
of systems containing two trans-fused five-membered rings.37
Surprisingly, the COSMO-calculations resulted in only small
changes in reaction energies (1.7-4.2 kcal mol-1 more exo-
(35) (a) Gansa¨uer, A.; Rinker, B. Tetrahedron 2002, 58, 7017–7026. (b)
Daasbjerg, K.; Svith, H.; Grimme, S.; Gerenkamp, M.; Mu¨ck-
Lichtenfeld, C.; Gansa¨uer, A.; Barchuk, A.; Keller, F. Angew. Chem.,
Int. Ed. 2006, 45, 2041–2044. (c) Gansa¨uer, A.; Barchuk, A.; Keller,
F.; Schmitt, M.; Grimme, S.; Gerenkamp, M.; Mu¨ck-Lichtenfeld, C.;
Daasbjerg, K.; Svith, H. J. Am. Chem. Soc. 2007, 129, 1359–1371.
(36) Chatgilialoglu, C.; Ingold, K. U.; Scaiano, J. C. J. Am. Chem. Soc.
1981, 103, 7739–7742.
For all titanocenes and methods the activation energy leading
to the formation of cis-13 is substantially lower than that for
the formation of trans-13. Thus, some of the strain of the product
trans-bicyclo [3.3.0] system is already present in the transition
state of the reductive elimination. However, the difference is
(37) Gordon, H. L.; Freeman, S.; Hudlicky, T. Synlett 2005, 2911–2914.
9
16994 J. AM. CHEM. SOC. VOL. 131, NO. 46, 2009

Catalysis via Homolytic Substitution
A R T I C L E S
Scheme 4. Solvent Exchange between 1 and 7 by COSMO
smaller than for the reaction energies, because in agreement
with the Hammond postulate,34 the transition states are “early”.
Within the series leading to the formation of cis-13, the
activation energies are lowest for 1 and highest for 11. The
introduction of the tert-butyl substitution leads to steric interac-
tions that deteriorate the approach of the radical to the Ti-O
bond. The electron deficient titanocenes 6, 7, and 8 result in
somewhat higher activation energies for the formation of both
cis- and trans-13 than 1. In the cis-transition states, the C-O
distance is shortest for TS1-cis-13 (221 pm) and slightly longer
for TS6-cis-13 and TS7-cis-13 (226 and 227 pm, respectively)
and longest for TS8-cis-13 (233 pm). As expected from the
Hammond postulate, the C-O distances in the trans-transition
structures are substantially shorter. Two examples are shown
in Figure 3.
In agreement with the experiments, this should be less critical
or even uncritical for the much faster formation of the
cis-disubstituted tetrahydrofuran derivatives.4
In summary, the computational results are of prime impor-
tance for the development of an efficient overall catalytic cycle.
Both the oxidative addition and the reductive elimination are
predicted to be feasible for all titanocenes investigated. Of these
complexes, the electron rich titanocenes should react fastest in
the absence of competing radicals pathways. The most critical
side-reaction is predicted to be the trapping of radical of the
type trans-12. The rather complex interplay of the undesired
side-reactions with the desired reductive elimination will depend
on a number of parameters of the concrete practical setup and
is beyond the scope of the theoretical study. It is therefore
experimentally investigated.
Because of the high activation energies for the formation of
trans-13 the interception of trans-12 by a second equivalent of
the titanocene(III) complexes will most likely be the most
efficient undesired side-reaction of the entire catalytic sequence.
Synthetic Investigations. Of course, the ultimate goal of any
study aiming at the design of efficient catalysts must be the
realization of the overall reaction and, if possible, an under-
standing of the experimental factors affecting the efficacy of
Figure 3. Transition states of the tetrahydrofuran forming reductive elimination of 12 to 13 by 1 and 7.
J. AM. CHEM. SOC. VOL. 131, NO. 46, 2009 16995
9

A R T I C L E S
Gansa¨uer, et al.
Scheme 5. Test Reactions for the Oxidative Additions and Radical Trappings ([Ti] ) (RC5H4)(R′C5H4)TiCl)
Scheme 6. Interception of ꢀ-Metaloxy Radicals by Titanocene(III) Reagents ([Ti] ) (RC5H4) (R′C5H4)TiCl)
the pertinent steps. Also, only a synthetic study can establish
the stability and thus the practical applicability of the compu-
tationally investigated complexes under the experimental
conditions.
We have addressed these issues in a two step approach: First,
the oxidative addition to epoxides was studied with a special
emphasis on understanding the characteristics of undesired
radical trapping. Second, the complete catalytic cycle was
implemented. This was done with the aim of avoiding any
undesired side reactions identified above.
Of the titanocenes computationally investigated, we have
synthesized the titanocene(IV) dichlorides of 1, 4, 7, 8, 10, and
11 according to known procedures.38 The titanocene(IV) dichlo-
ride of 6 was prepared for the first time. For details of the
preparation see Supporting Information. Complexes 5 and 9
were not employed. This is because the calculations did not
suggest particular properties for 5 and under the experimental
conditions, the -NO2 group in 9 would most likely be reduced
either by the metal powders employed or by the titanium(III)
center. The titanocene(III) derivatives were prepared from their
corresponding dichlorides by in situ reduction with Mn powder.8,17
efficient competition between oxidative addition followed by
H-atom abstraction from 16 to yield 17 and 19, respectively,
and undesired radical trapping that yields 18 and 20, respectively
(Scheme 5). By using two substrates with different substitution
patterns at the epoxide, radical trapping by the titanocenes can
be examined as a function of radical substitution in a straight-
forward manner. The novel nitrile substituted 6 (for details of
the preparation, see Supporting Information) could not be used
due to its relatively low solubility that is about 0.01 M in THF.
This renders a comparison of the results with the other
complexes pointless.
The mechanisms of the formation of 18 and 20 are depicted
in Schemes 6. Olefin 18 is formed via trapping of ꢀ-titanoxy
radical 21 and elimination of an oxotitanium-species. Allylic
alcohol 20 is generated from 22 by interception with a second
equivalent of the titanocene(III) reagent, ꢀ-hydride elimination,
and protic workup of 23 as shown in Scheme 6.
The results of the epoxide opening are summarized in Tables
7 and 8.
All of the titanocenes investigated oxidatively add to epoxide
14. However, the substitution of the complexes has a significant
influence on the fate of radical 21. Complex 10 and especially
11 lead to high isolated yields of the deoxygenation product 18
and hardly any 17. Hence, these complexes trap 21 much more
efficiently than even γ-terpinene (16). The use of 1 and 4 does
not alter the picture substantially. However, the situation changes
significantly with the introduction of electron withdrawing
groups. For 7, hydrogen atom abstraction can already compete
with radical trapping. Only the use of 8 leads to the preferred
formation of 17 (17: 18 ≈ 4:1) in about 50% yield. Thus, 8 is
rather inefficient in the reductive trapping of secondary radicals.
However, 8 slowly decomposes under the reaction conditions
and hence the conversion to the desired products is incomplete.
Epoxide Opening. We examined the oxidative addition to
epoxides 14 and 15 in order to establish the feasibility of this
process. Our reaction conditions [2.2 equiv. of the titanocene(III)
reagent, 4.5 equiv. γ-terpinene (16)] were chosen to ensure an
(38) (a) Complex 4: Conway, B.; Rausch, M. Organometallics 1985, 4,
688–693. (b) Complex 7: Rausch, M. D.; Lewison, J. F.; Hart, W. P.
J. Organomet. Chem. 1988, 358, 161–168. (c) Complex 8: Bruce,
M. R. M.; Sclafani, A.; Tyler, D. R. Inorg. Chem. 1986, 25, 2546–
2549. (d) Complex 10: Ott, K. C.; deBoer, E.; J, M.; Grubbs, R. H.
Organometallics. 1984, 3, 223–230. (e) Complex 11: Hart, S. L.;
Duncalf, D. J.; Hastings, J. J.; McCamley, A.; Taylor, P. C. J. Chem.
Soc., Dalton Trans. 1996, 284, 3–2849.
9
16996 J. AM. CHEM. SOC. VOL. 131, NO. 46, 2009

Catalysis via Homolytic Substitution
A R T I C L E S
Scheme 7. Performance of the Titanocene (R, R′ ) H) Catalyzed Tandem Sequence at 95 °C (Coll ) 2,4,6-trimethyl pyridine)
Table 7. Oxidative Addition Reactions to 14
Scheme 8. Critical Issues for the Performance of Tetrahydrofuran
Formation via Reductive Elimination ([Ti] ) (RC5H4) (R′C5H4)TiCl)
cat
R, R′
yield 17 + 18/ %
17/ %
18/ %
1
4
7
8
10
11
H, H
H, Cl
H, CO2Me
CO2Me, CO2Me
H, Me
97
84
89
ca. 60
87
19
20
35
ca. 50
13
78
64
54
12
74
81
H, t-Bu
86
5
Table 8. Oxidative Addition Reactions to 15
cat
R, R′
19 + 20/ %
19/ %
20/ %
1
4
7
8
10
11
H, H
H, Cl
H, CO2Me
CO2Me, CO2Me
H, Me
52
67
70
46
66
61
45
56
67
43
51
31
7
11
3
3
15
30
relatively high outer sphere character. In this case, the redox-
potential represents the crucial property of the reductant. It is
well established that alkyl-substituted titanocenes constitute
significantly more powerful electron transfer reagents.40 In
agreement with the experimental results, in the oxidative
additions to both 14 and 15 the use of 11 should therefore lead
to the highest amount of the undesired reductive trapping of
the intermediate radicals 21 and 22.
Of the titanocenes studied here, 7 is best for mediating the
epoxide opening because it is stable under the reaction condi-
tions and does not promote undesired radical trapping, even
though it is intrinsically not the most reactive complex for
oxidative addition. In order to investigate the importance of these
effects on the efficiency of the tetrahydrofuran forming reductive
elimination, the performance of the overall catalytic sequence
was studied next.
H, t-Bu
Moreover, the dimer of the cyclopentadienyl ligand, that is
formed via a Diels-Alder reaction, could not be completely
removed from 17 via column chromatography. Therefore, only
an approximate yield of 17 is indicated in Table 7. Nevertheless,
these findings amply highlight the power of the electronic fine-
tuning of the titanocene(III) properties in oxidative addition
reactions to epoxides. To obtain a more general understanding
of the interplay of this effect with substitution pattern of the
epoxide we studied the reaction of 15 next (Table 8).
Gratifyingly, all complexes investigated also oxidatively add
to 15. As above, the substituents on the cyclopentadienyl ligand
exert a large influence on the outcome of the reaction. With
the ester-substituted complex 7, 19 is obtained in 67% yield
together with hardly any of the undesired 20 (3%) being formed.
This constitutes the highest combined yield of products and
suggests that 7 is the best titanocene(III) reagent for the opening
of 1,1-disubstituted epoxides without interference of reductive
radical trapping. The introduction of a second ester group in 8
results in a marked decline of the isolated yields due to the low
stability of 8. All other complexes also gave lower combined
yields of 19 and 20 than 7. Moreover, 20 was formed in higher
relative amounts. The use of tert-butyl substituted 11 gave the
worst result in this respect. Alcohols 19 and 20 were formed in
almost equal proportions. Surprisingly, the undesired radical
reduction is also most efficient for tertiary radicals with the
bulkiest complex 11 even though titanocene mediated or
catalyzed reactions are usually rather sensitive to steric effects.39
Implementation of the Complete Catalytic Sequence. To this
end, we investigated the transformation of 2 to 3 in some detail.
Before employing the substituted titanocenes, we focused on
the identification of reactions conditions favoring the desired
intramolecular course of the catalytic reaction over the undesired
intermolecular reductive radical trapping (Scheme 7).
Table 9. Transformation of 2 to 3 by Substituted Titanocene(III)
Complexes
cat
R, R′
3/ %
cis/trans
1
4
6
7
8
10
11
H, H
H, Cl
H, CN
H, CO2Me
CO2Me, CO2Me
H, Me
78
72
80
95
24
69
61
89/11
86/14
78/22
83/17
82/18
90/10
94/6
H, t-Bu
It has been reported that for the titanocene mediated reduc-
tions the inner sphere character of electron transfer decreases
with decreasing complexation ability of the organic substrates.17
The radical centers in 21 and 22 that possess no coordination
site for titanium, should therefore be reduced in a process with
Two essential parameters in this respect are the reaction
temperature and the concentration of the substrates and catalysts.
No significant difference in the isolated yield was observed when
running the reaction at room temperature or 70 °C (67% yield
at both temperatures). Quite surprisingly, we found that a further
(39) Gansa¨uer, A.; Rinker, B.; Barchuk, A.; Nieger, M. Organometallics
2004, 23, 1168–1171.
(40) Johnston, R. F.; Borjas, R. E.; Furila, J. L. Electrochim. Acta 1995,
40, 473–477.
9
J. AM. CHEM. SOC. VOL. 131, NO. 46, 2009 16997

A R T I C L E S
Gansa¨uer, et al.
Scheme 9. Titanocene Catalyzed Tetrahydrofuran Synthesis (Coll ) 2,4,6-trimethyl pyridine)
increase of the oil bath temperature to 95 °C increased the
isolated yield of 3 to 78%. Moreover, trans-3 was obtained for
the first time (Scheme 7). The concentration (0.1 M in THF)
was not decreased further in order to avoid the use of larger
amounts of solvent and extended reaction times. It has been
shown however, that this may be necessary in cases where
sterically demanding radical substitution is encountered.4a,d
Collidine hydrochloride was added to the reaction medium to
enforce catalyst regeneration after undesired radical trapping.
The conditions shown were employed in the further catalyst
screening experiments. The nitrile substituted 6 could be readily
employed in these investigations, because the 10 mol % of 6
used were readily soluble under these conditions (Table 9).
The ester substituted complex 7 gave an almost quantitative
yield of 3 and is therefore substantially superior to all other
complexes. Introduction of the second ester group in 8 results
in a much lower yield of 3, which is due to the instability of
this complex. Extensive decomposition of 2 was also observed
in this particular case. Of the other complexes, the alkyl
substituted titanocenes 10 and 11 gave the worst results, while
1 and the novel 6 gave almost identical yields of 3.
THF formation constitutes the critical aspect of the catalytic
sequence. We suggest that the prevention of radical reduction
is most relevant here, because the calculated activation energies
for tetrahydrofuran formation are rather high for radical reac-
tions. Thus, the decreased capability of 7 for radical reduction
results in an increased lifetime of the pivotal radical intermedi-
ates cis- and trans-C (Scheme 8), that is essential for the
occurrence of the reductive elimination. Radical trapping is not
critical after the oxidative addition since the radical translocating
5-exo cyclization is a very fast reaction in both titanocene-
mediated or -catalyzed radical reactions.
The quick interception of trans-C is also responsible for the
apparently higher diastereoselectivity of the overall sequence
in the presence of 10 and 11. Both complexes are more potent
SET reagents than 7 and hence reduce higher portions of trans-A
before the reductive elimination takes place. This also accounts
for the lower yields of 3 obtained with these catalysts. The
alternative explanation, a radical translocating 5-exo cyclization
proceeding with higher diastereoselectivity in the presence of
10 and 11 is less likely.41 This is because it has been
demonstrated that the influence of alkyl substitution on the
diastereoselectivity of titanocene mediated or catalyzed cycliza-
The superiority of 7 over the alkyl substituted 10 and 11
implies that the relative performances of radical trapping leading
to undesired side-products and reductive elimination leading to
9
16998 J. AM. CHEM. SOC. VOL. 131, NO. 46, 2009

Catalysis via Homolytic Substitution
A R T I C L E S
tions in such simple systems is rather low. Moreover, this simple
analysis does not take the lower yields of 3 with 10 and 11 into
account.
of sterically and electronically difficult reductive eliminations
for an efficient overall catalytic sequence.
Conclusions
We investigated the performance of 7 for a number of other
substrates that lead to relatively low yields of the desired
products with 1. In all cases, the reductive elimination is even
more critical than for 2 because the radicals attacking the Ti-O
bond are sterically either more demanding (from 24, 26, 28,
and 30) or electronically stabilized (from 32). These results are
summarized in Scheme 9.
In summary, we have demonstrated by a combined theoretical
and experimental study how the performance of the titanocene
catalyzed homolytic C-O bond cleavage and homolytic C-O
bond formation can be controlled by adjusting the electronic
and steric properties of the cyclopentadienyl ligands. The
knowledge of the activation and reaction energies and the
structures of pertinent intermediates and transition states
obtained as well as the ability to suppress undesired side
reactions provides a mechanism-based platform for the develop-
ment of efficient tandem sequences. This was demonstrated in
the development of a much more efficient catalyst for a tandem
sequence featuring a single electron oxidative addition and a
single electron reductive elimination for the synthesis of bi- and
polycyclic tetrahydrofurans.
The concentration and catalyst loading for each substrate are
identical to those that have turned out to be most suitable for 1.
With 1 it is often essential to employ high dilution (0.02 M in
THF) or to use ethyl acetate (EtOAc) as solvent, because
reduction of Cp2TiCl2 is slower in this solvent and hence
reductive radical trapping is additionally disfavored. With 7,
the use of EtOAc as solvent is not necessary.
Most gratifyingly, catalyst 7 is superior to 14a,d in all examples
investigated. This is most striking for the reaction of 30, where
no product 31 is formed at all in the presence of 1. The effect
is less pronounced but still highly important in the other cases,
also. Even though for 32 the yield is clearly not ideal, it is still
much higher than the 17% obtained for 1. These relatively low
yields with the highly stabilized benzylic radicals are most likely
due to a substantially reduced exothermicity of tetrahydrofuran
generation compared to the reactions of the alkyl substituted
radicals. Thus, side reactions such as the formation of polymers
were observed. Complex 7 is also markedly superior to Cp2TiCl
for the preparation of 35 featuring a 6-exo cyclization. In this
case, the diastereoselectivity is only slightly reduced from 87:
13 to 83:17.
In a broader perspective, our study highlights that the same
concepts that have proven highly successful in the development
of classical catalytic systems, featuring organometallic key
reactions that proceed in two electron steps, can be applied to
the development of metal-catalyzed reactions occurring in single
electron steps. This should open broad perspectives for the
design of novel reactions, such as tandem sequences featuring
slow elementary steps, e.g., titanocene catalyzed epoxypolyene
cyclizations, or transformations with the highly important ketyl
radicals.
Acknowledgment. We are indebted to the Deutsche For-
schungsgemeinschaft (SFB 813 “Chemistry at Spin Centers”) and
the Fonds der Chemischen Industrie (Sachbeihilfen) for generous
financial support.
These findings amply highlight the general significance of
effectively suppressing reductive radical trapping in the case
Supporting Information Available: The computational struc-
tures and experimental details. This material is available free
of charge via the Internet at http://pubs.acs.org.
(41) (a) Gansa¨uer, A.; Pierobon, M. Synlett 2000, 1357–1359. (b) Gansa¨uer,
A.; Pierobon, M.; Bluhm, H. Synthesis 2001, 2500–2520.
JA907817Y
9
J. AM. CHEM. SOC. VOL. 131, NO. 46, 2009 16999