Sustainable radical reduction through catalytic hydrogen atom transfer

Article abstract of DOI:10.1021/ja801232t

A system with coupled catalytic cycles is described that allows radical reduction by hydrogen atom abstraction from rhodium hydrides. These intermediates are generated from H₂ activation by Wilkinson's catalyst. Radical generation is carried out by titanocene-catalyzed electron transfer to epoxides. Copyright

Full text of DOI:10.1021/ja801232t

Published on Web 05/06/2008
Sustainable Radical Reduction through Catalytic Hydrogen Atom Transfer
Andreas Gans a¨ uer,* Chun-An Fan, and Frederik Piestert
Kekul e´ -Institut f u¨ r Organische Chemie and Biochemie der UniVersit a¨ t Bonn, Gerhard Domagk Strasse 1,
5
3121 Bonn, Germany
Received February 22, 2008; E-mail: andreas.gansaeuer@uni-bonn.de
Owing to their high versatility, selectivity, and compatibility with
Scheme 1. Sustainable Radical Reduction through Catalytic
Hydrogen Atom Transfer
densely functionalized substrates, radical reactions are frequently
1
employed in the synthesis of complex molecules. However,
limitations also exist. One of the unresolved problems is constituted
by catalytic and environmentally benign reduction of carbon-
2
centered radicals. Water or alcohols have been introduced as
3
stoichiometric reductants to address the second issue. Here, we
2
demonstrate that H constitutes an attractive hydrogen atom source
for at least two reasons. First, the reduction will proceed under
4
complete atom economy. Second, after activation by a transition
metal complex, the powerful arsenal of catalytic hydrogenation
5
methods can be made accessible for radical reduction. The use of
stoichiometric amounts of stannanes, silanes, or cyclohexadienes
2
can be avoided.
Here, we describe a process exploiting these features. It is based
on the combination of the catalytic cycles outlined in Scheme 1.
The radical intermediates are generated by titanocene-catalyzed
Table 1. Optimization of the Catalytic Hydrogen Atom Transfer in
THF (See Supporting Information for Details) at 4 bar H2
6
epoxide opening via electron transfer. Radical reduction proceeds
through hydrogen atom transfer from Wilkinson’s catalyst, one of
7
the most common hydrogenation catalysts. In this manner, the well-
established advantages of both radical chemistry and hydrogenation
methods can be combined for efficient radical reduction. Through
our decoupling of radical generation and reduction, the kinetic
entry
4/mol %
5/mol %
T/°C
concn/ M
reductant
6/%
a
b
1
2
3
4
5
6
7
8
9
10
25
25
25
25
25
25
25
25
25
50
0
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.4
0.2
0.2
Mn
<20
0
constraints of the CpCr(CO)
using H are circumvented.
2
3
-catalyzed radical chain cyclizations
5
5
5
5
2.5
2.5
1
1
1
8
10
10
10
10
5
5
5
5
10
Mn
Mn
Zn
Mn
Mn
Mn
Mn
Mn
Mn
68
84
81
81
72
73
55
46
34
To be effective, the radical reducing hydrogen atom abstraction
from both 1 and 2 should be exothermic and faster than epoxide
opening. It is reasonable to assume that both conditions can be
fulfilled. Usually, hydrogen atom abstractions by carbon-centered
radicals from transition metal hydrides are highly exothermic due
to the weakness of MsH bonds compared to CsH bonds. In the
1
1
0
1
5
-
1 9
case of rhodium(III) hydrides, these BDEs are about 58 kcal mol
.
1
0
9
a
b
The rate constants k of these processes may be as high as 10
Obtained with two other unknown compounds. With 1 bar H2.
-
1 -1 11
1
5
M
s
.
Thus, even for sterically shielded ꢀ-titanoxy radicals,
1
2 2
). No formation of 6 was observed without Cp TiCl (entry 2),
these reactions will be much faster than titanocene-mediated epoxide
and 3 was reisolated in >85% yield. This rules out alernative
pathways, such as epoxide isomerization to an allylic alcohol and
hydrogenation or radical reduction by adventitious water. When
-
1 -1 12
openings (k ≈ 0.5-2 M
s ).
Our approach can only be successful if the cycles of Scheme 1
do not interfere. We were optimistic that this could be achieved
for two reasons. First, Wilkinson’s catalyst is stable toward strong
2
applying only 1 bar of H (entry 3), a 68% yield of 6 was obtained
compared to 84% with 4 bar pressure (entry 4). As desired, this
suggests that radical reduction is fast. Hydrogen activation hence
constitutes the slowest step in the central catalytic cycle. With 10
mol % of 4 and 5 mol % of 5, yields of more than 80% of 6 can
be obtained with either Mn or Zn as reductant (entries 4 and 5).
Decreasing the amount of 5 to 2.5 mol % resulted in a slightly
reduced yield of 6 (81%, entry 6). A similar trend was observed
when lowering the loading of both 4 to 5 mol % and 5 to 2.5 or 1
mol %, respectively (entries 7 and 8). This suggests that epoxide
opening constitutes the rate-determining step of our process. An
increase in the concentration results in a deterioration of the isolated
yield of 6 (entry 9). This is also the case for changing the
temperature (entries 10 and 11).
1
3
3 2
Lewis acids, such as BF *Et O. It should therefore tolerate the
mildly acid protic conditions of the titanocene-catalyzed oxirane
opening. Second, titanocenes are stable under typical hydrogenation
2
conditions and activate H only slowly after reaction with alkyl
1
4
lithium reagents. Hence, the titanium(III) complexes involved in
epoxide opening should not be affected by the any of the Rh species
2
or H and vice versa.
The relative rates of the two cycles were adjusted to avoid an
accumulation of reactive intermediates such as 2 or the titanocene-
bound radicals as summarized in Table 1.
Without RhCl(PPh
3 3
) , a complex mixture containing 6 and
unidentified side products was obtained in only 20% yield (entry
6
916 9 J. AM. CHEM. SOC. 2008, 130, 6916–6917
10.1021/ja801232t CCC: $40.75  2008 American Chemical Society

C O M M U N I C A T I O N S
Table 2. Functional Group Tolerance under Optimized Conditions
tion by the titanocene(III) reagents and radical reduction by the
rhodium hydrides are fully compatible. For both 4 and 19, the
isolated yields of the products are similar to the reactions with 5
equiv of the hydrogen atom donors such as cyclohexadienes. Thus,
radical reduction by rhodium hydrides is as efficient as titanocene-
catalyzed radical generation.
15
In summary, we have presented the first system of combined
catalytic cycles for a sustainable reduction of radicals. Our approach
unites titanocene-catalyzed reductive epoxide opening with the
2
powerful rhodium-catalyzed H activation and hydrogen atom
transfer. Because of their different affinities toward the substrates
and ligands, the early and late transition metal catalysts are mutually
compatible. Epoxide opening tolerates a number of functional
groups incompatible with nucleophilic ring opening by hydride
reagents. The regioselectivity of ring opening is complementary to
N
S 2 reactions. Moreover, opening of meso-epoxides occurs with
high enantioselectivity.
Acknowledgment. We thank the Alexander von Humboldt-
Stiftung (Forschungsstipendium to C.-A.F.) for financial support.
Supporting Information Available: Experimental details and
compound characterization. This material is available free of charge
via the Internet at http://pubs.acs.org.
a
References
See Supporting Information for details.
(
1) (a) Zard, S. Z. Radical Reactions in Organic Synthesis; Oxford University:
Oxford, 2003. (b) McCarroll, A. J.; Walton, J. C. Angew. Chem., Int. Ed.
Table 3. Catalytic Enantioselective Epoxide Opening
2
001, 40, 2225–2250.
(
2) (a) Baguley, P. A.; Walton, J. C. Angew. Chem. Int. Ed. 1998, 37, 3072–
3
2
082. (b) Gilbert, B. C.; Parsons, A. F. J. Chem. Soc., Perkin Trans. 2
002, 367–387. (c) Studer, A.; Amrein, S. Synthesis 2002, 835–849.
(
3) (a) Spiegel, D. A.; Wiberg, K. B.; Schacherer, L. N.; Medeiros, M. R.;
Wood, J. L. J. Am. Chem. Soc. 2005, 127, 12513–12515. (b) Renaud, P.
J. Am. Chem. Soc. 2005, 127, 14204–14205. (c) Cuerva, J. M.; Campa n˜ a,
A. G.; Justicia, J.; Rosales, A.; Oller-L o´ pez, J. L.; Robles, R.; C a´ rdenas,
D. J.; Bu n˜ uel, E.; Oltra, J. E. Angew. Chem., Int. Ed. 2006, 45, 5522–
5
526.
(
(
4) Trost, B. M. Science 1991, 254, 1471–1477.
5) (a) Knowles, W. S. Angew. Chem., Int. Ed. 2002, 41, 1998–2007. (b)
Noyori, R. Angew. Chem., Int. Ed. 2002, 41, 2008–2022. (c) Jang, H. Y.;
Krische, M. J. Acc. Chem. Res. 2004, 37, 653–661. (d) For a recent
compilation of reviews, see: Acc. Chem. Res., 2007, 40, 1237–1419.
6) (a) RajanBabu, T. V.; Nugent, W. A. J. Am. Chem. Soc. 1994, 116, 986–
997. (b) Gans a¨ uer, A.; Bluhm, H.; Pierobon, M. J. Am. Chem. Soc. 1998,
120, 12849–12859. (c) Gans a¨ uer, A.; Narayan, S. AdV. Synth. Catal. 2002,
a
entry
substrate
5/mol %
product
yield
(R):(S)
1
2
3
17
20
22
10
5
5
18
21
23
60
64
75
93:7
97:3
96:4
(
3
44, 465–475.
a
(7) (a) Osborn, J. A.; Jardine, F. H.; Young, J. F.; Wilkinson, G. J. Chem.
Soc. A 1965, 1711–1732.
(
Measured by GC analysis of the isolated products (see Supporting
Information).
8) Smith, D. M.; Pulling, M. E.; Norton, J. R. J. Am. Chem. Soc. 2007, 129,
7
70–771.
We next turned our attention to the compatibility of our reaction
conditions with different functional groups and epoxide substitution
patterns. These results are summarized in Table 2.
The reaction conditions tolerate sensitive functionality. Ac-
etophenone can be reisolated in 87% yield when added to the
reduction of 3, while the yield of 6 remains essentially unchanged
(9) (a) Drago, R. S.; Miller, J. G.; Hoseton, M. A.; Farris, R. D.; Desmond,
M. J. J. Am. Chem. Soc. 1983, 105, 444–449. (b) Martinho Sim o˜ es, J. A.;
Beauchamp, J. L. Chem. ReV. 1990, 90, 629–688.
(
10) (a) Eisenberg, D. C.; Lawrie, C. J. C.; Moody, A. E.; Norton, J. R. J. Am.
Chem. Soc. 1991, 113, 4888–4895. (b) Eisenberg, D. C.; Norton, J. R. Isr.
J. Chem. 1991, 31, 55–66.
(
11) (a) Bakac, A.; Thomas, L. M. Inorg. Chem. 1996, 35, 5880–5884. (b) Bakac,
A. Inorg. Chem. 1998, 37, 3548–3552.
(
12) (a) Daasbjerg, K.; Svith, H.; Grimme, S.; Gerenkamp, M.; M u¨ ck-
Lichtenfeld, C.; Gans a¨ uer, A.; Barchuk, A.; Keller, F. Angew. Chem., Int.
Ed. 2006, 45, 2041–2044. (b) Gans a¨ uer, A.; Barchuk, A.; Keller, F.; Schmitt,
M.; Grimme, S.; Gerenkamp, M.; M u¨ ck-Lichtenfeld, C.; Daasbjerg, K.;
Svith, H. J. Am. Chem. Soc. 2007, 129, 1359–1371.
(13) Shi, L.; Tu, Y.-Q.; Wang, M.; Fan, C.-A.; Zhao, Y.-M.; Xia, W.-J. J. Am.
Chem. Soc. 2005, 127, 10836–10837.
14) (a) Halterman, R. L.; Vollhardt, K. P. C.; Welker, M. E.; Bl a¨ ser, D.; Boese,
R. J. Am. Chem. Soc. 1987, 109, 8105–8107. (b) Broene, R. D.; Buchwald,
S. L. J. Am. Chem. Soc. 1993, 115, 12569–12570.
15) Gans a¨ uer, A.; Barchuk, A.; Fielenbach, D. Synthesis 2004, 2567–2573.
16) (a) Gans a¨ uer, A.; Bluhm, H.; Lauterbach, T. AdV. Synth. Catal. 2001, 343,
785–787. (b) Gans a¨ uer, A.; Bluhm, H.; Rinker, B.; Narayan, S.; Schick,
M.; Lauterbach, T.; Pierobon, M. Chem.—Eur. J. 2003, 9, 531–542.
17) (a) Cesarotti, E.; Kagan, H. B.; Goddard, R.; Kr u¨ ger, C. J. Organomet.
Chem. 1978, 162, 297–309. (b) Gans a¨ uer, A.; Bluhm, H.; Pierobon, M.;
Keller, M. Organometallics 2001, 20, 914–919.
(
(
81%, entry 1). Tosylates (entry 2), esters (entry 3), and chlorides
entry 4) remain unaffected. Silyl groups can also be submitted to
our conditions (entry 5). The reaction of 15 is especially noteworthy.
In titanocene-based methodology, the reduction of benzylic radicals
6
a
is notoriously difficult and requires thiols or selenols (entry 6).
(
Bulky ethers in the close vicinity of a secondary radical center are
tolerated (entry 7).
(
(
To probe the sensitivity of our reaction toward the substitution
pattern of the titanocene catalyst, we investigated the enantiose-
16
17
lective opening of three meso-epoxides by Kagan’s complex 19
Table 3). Alcohols 18, 21, and 23 were isolated with enantiose-
lectivities identical to those of the reactions performed with 5 equiv
(
(
17
of 1,4-cyclohexadiene as radical reductant. Thus, radical genera-
JA801232T
J. AM. CHEM. SOC. 9 VOL. 130, NO. 22, 2008 6917