Functionalized tetrahydrofurans from alkenols and olefins/alkynes via aerobic oxidation-radical addition cascades

Article abstract of DOI:10.1021/ja108403s

Aerobic oxidation of alkyl- and phenyl-substituted 4-pentenols (bishomoallyl alcohols), catalyzed by cobalt(II) complexes in solutions of γ-terpinene or cyclohexa-1,4-diene, stereoselectively gave tetrahydrofurylmethyl radicals. Cyclized radicals were tra

Full text of DOI:10.1021/ja108403s

ARTICLE
pubs.acs.org/JACS
Functionalized Tetrahydrofurans from Alkenols and Olefins/Alkynes
via Aerobic Oxidation-Radical Addition Cascades
Patrick Fries, Daniel Halter, Alexander Kleinschek, and Jens Hartung*
Fachbereich Chemie, Organische Chemie, Technische Universit €a t Kaiserslautern, Erwin-Schr €o dinger-Strasse, D-67663 Kaiserslautern,
Germany
S Supporting Information
b
ABSTRACT: Aerobic oxidation of alkyl- and phenyl-substituted
4
-pentenols (bishomoallyl alcohols), catalyzed by cobalt(II) com-
plexes in solutions of γ-terpinene or cyclohexa-1,4-diene, stereo-
selectively gave tetrahydrofurylmethyl radicals. Cyclized radicals
were trapped with monosubstituted olefins (e.g., acrylonitrile,
methyl acrylate), (E)- and (Z)-1,2-diacceptor-substituted olefins
(
e.g., dimethyl fumarate, fumarodinitrile, N-phenyl maleic imide),
and ester-substituted alkynes (e.g., ethyl propynoate). Oxidation-
addition cascades thus furnished side-chain-substituted (CN, CO R, COR, or SO R) di- and trisubstituted tetrahydrofurans in
2
2
stereoselective reactions (2,3-trans, 2,4-cis, and 2,5-trans). A diastereomerically pure bistetrahydrofuran was prepared in a cascade
consisting of two aerobic oxidations, one alkyne addition, and one final H-atom transfer.
1
. INTRODUCTION
To combine advantages of catalysis, stereoselective synthesis,
and radical chemistry for synthesis of tetrahydrofuran-derived
Carbon radical addition to alkenes has become a cornerstone
16-18
natural products,
bond in a sequence of polar and free radical reactions.
we chose to functionalize the alkenol double
of organic synthesis from the time methods to selectively generate
radicals and principles to control reactivity and selectivity became
available.
14,19,20
This
1-5
sequence of transformation steps is not available from oxidation
catalysis or radical chemistry alone. For our strategy, we selected
oxidation catalysis to construct the tetrahydrofuran ring. The
oxidation leaves a cyclized radical, which must be trapped by the
alkene. Termination of the sequence requires a reductant. To
maintain the catalytic cycle, the same reductant must convert
the oxidized form of the catalyst into the reduced form, which is
the active reagent (Scheme 1).
Useful carbon radical additions in synthesis are fast
and exothermic processes that proceed via early transition states.
In early transition states, according to frontier molecular orbital
theory, favorable interactions arise between the singly occupied
molecular orbital (SOMO) of the radical and a suitable orbital of
the alkene. Nucleophilic alkyl radicals (primary, secondary, or
tertiary) have high SOMO energies and therefore interact with the
lowest unoccupied molecular orbital (LUMO) of an electrophilic,
that is an acceptor-substituted, alkene. Addition of the nucleophilic
The results of our study show that substituted 4-pentenols
undergo stereoselective tetrahydrofuran ring closures if oxidized
with molecular oxygen in solutions containing cyclohexa-1,4-
diene (CHD). The oxidation, which is catalyzed by cobalt(II)
diketonate complexes, generates tetrahydrofurylmethyl radicals
which add to acceptor-substituted alkenes. By this approach,
tetrahydrofurans were stereoselectively prepared (2,3-trans, 2,4-
cis, and 2,5-trans) in up to 66% yield. The products were side-
chain substituted with CN, CO R, COR, and SO R groups,
6,7
radical thereby occurs at the terminal alkene position.
Radicals in synthesis must be generated from progenitors,
such as alkyl halides, xanthates, mixed anhydrides, chalcogenides,
or carbonyl compounds, to mention the most important product
classes. The dominating mechanism to transform a precursor
for synthetic applications into a radical is the chain reaction. In a
chain reaction, radical concentrations are kept low to prevent
unproductive radical/radical reactions, such as combination or
disproportionation, in order to favor productive reactions, such as
1
,3
2
2
originating from the alkene/alkyne (R = alkyl). We applied the
method to prepare a diastereomerically pure bistetrahydrofuran
in a cascade, consisting of two aerobic oxidations, one alkyne
addition, and one final H-atom transfer.
8,9
trapping with an olefin or a heteroatom donor (cf. Scheme 1).
Conducting a chain reaction requires a mediator. A mediator is
a reagent formally composed of a transferable group and a
chain-propagating radical. A typical mediator is Bu SnH, con-
3
2
. RESULTS AND DISCUSSION
•
sisting of Bu Sn , which is the chain-propagating radical, and
3
•
H , the reducing equivalent. By this approach, stoichiometric
2.1. Cobalt(II) Complexes. In earlier studies it was discovered
amounts of radical progenitor and mediator are consumed.
that complexes of cobalt(II), derived from trifluoromethyl-substituted
Catalytic sustainable methods for radical generation are surpris-
1
0-12
ingly rare,
in spite of the growing significance of radical
Received: September 17, 2010
Published: February 24, 2011
1
3-15
reactions in synthesis.
r 2011 American Chemical Society
3906
dx.doi.org/10.1021/ja108403s
|J. Am. Chem. Soc. 2011, 133, 3906–3912

Journal of the American Chemical Society
ARTICLE
Scheme 1. Mechanistic Concept for Heterobisfunctionaliza-
to prepare 59% of addition products 6 (44%) and 10 (15%),
28
tion of Alkenols via Aerobic Oxidation-Radical Addition
besides 16% of byproduct 5a (Table 2, entry 4). Yields of addition
products and the fraction of 6 versus 10 from cobalt-catalyzed
reactions are similar to references from radical additions to
a
Cascades
acrylonitrile, mediated by Bu SnH or substituted cyclohexa-
3
1
,29
1
,4-dienes.
Higher acrylonitrile concentrations increase
the yield of 2-fold addition product 10, whereas lower con-
centrations of the alkene favored formation of reduction
product 5a. Concentrations of 1.5-1.7 M (i.e., ∼5 equiv with
respect to 1a), depending on the solubility of the alkene in
toluene/CHD mixtures, were therefore used as standard to
compare the alkene reactivity (cf. section 2.2).
(
ii). Temperature Effects. The onset of cobalt-catalyzed
oxidative ring closure occurs at temperatures above 30 °C
Figure 1). Quantitative alkenol turnover is attainable within
(
2
9
7
5
4 h at 60 °C. An increase of the reaction temperature from 60 to
0 °C leads to a plateau for monoaddition product formation at
5 °C and for dinitrile formation at 60 °C (Figure 1). The yield of
-methyl-2-phenyltetrahydrofuran 5a gradually increases as the
a
2
H X = e.g., 1,4-cyclohexadiene (CHD) or γ-terpinene (γ-Ter), R =
e.g., alkyl or aryl; Z = e.g., CO CH or CN; Y = e.g., H or Z; HL = 1,3-
diketone; for reactions with alkynes, refer to section 2.4.
2
3
reaction temperature rises from 30 to 90 °C. We therefore
performed aerobic oxidations catalyzed by cobalt complex 3 in
the following sections at 60 °C, whereas phenyl-trifluoromethyl-
butanedione derivative 4 was more active at 75 °C.
1
,3-diketones, are able to activate molecular oxygen for oxidative
1
9-22
cyclization of 4-pentenols.
using donor- and acceptor-substituted diketonates as auxiliaries for
cobalt(II), we selected cobalt(II) complex 3, an established re-
agent that was available from a previous study, and derivative 4,
for performing oxidation-addition cascades. The improved sta-
bility and reactivity of cobalt(II) complex 4 was discovered in a late
phase of the project. It was applied for the most important
substrate permutations but not for all (vide infra). Analytical data
showed that the monoethanol adduct of bis{4-[3,5-bis(trifluoro-
methyl)phenyl]-(2-oxo-κO)-but-3-en-(4-olato-κO}cobalt(II) (3)
and the dihydrate of bis-[1,1,1-trifluoro-2-(oxo-κO)-4-phenylbut-
From benchmark reactivity tests
(
iii). Concentration and Chemical Nature of Reductant
23
H2X. Turnover in cobalt-catalyzed aerobic alkenol oxidation
requires a chemoselective reductant. We used CHD and γ-
terpinene as reductants for the oxidation-addition cascade.
Cyclohexa-1,3-diene was surprisingly less effective than CHD
24
(
cyclohexa-1,4-diene). A likewise performed oxidation of 1a
catalyzed by 3 provided 22% of 5a in the absence of acrylonitrile
a. If acrylonitrile 2a is added, turnover entirely stops. Assumed
2
Diels-Alder adducts between acrylonitrile and cyclohexa-1,3-
diene were not found (GC-MS).
3
-en-4-olato-κO]cobalt(II) (4) are formed from the synthesis.
30
γ-Terpinene (γ-Ter), a naturally occurring compound, is a
We used both complexes, the way they were obtained from the
synthesis, as oxidation catalysts (see Table 1.
very effective alternative to CHD. In aerobic oxidations catalyzed
by cobalt(II) compounds, γ-terpinene is oxidized to isopropyl-
4-methylbenzene (>200%). From results of Karl Fischer
titrations we concluded that notable amounts of water form
in aerobic alkenol oxidations in 1,4-dihydroarene solution. The
amount of water, however, was not systematically quantified.
2
.2. Cascades with Terminal Olefins. 2.2.1. Parameters for
19
Selective Alkylative Trapping. Acrylonitrile 2a served as a reporter
substrate for elucidating principles of aerobic oxidation-olefin
addition cascades because rate constants, regioselectivity,
and theoretical details of carbon radical addition to the alkene
are known in detail from Bu SnH- and alkylmercury hydride-
mediated reactions. From the results of mechanistic studies, we
derived that (i) alkenol, olefin, and cobalt concentrations, (ii)
reaction temperature, (iii) the concentration and chemical nature
of the reductant H X, (iv) the cobalt(II) reagent, and (v) olefin
reactivity are important parameters to adjust for conducting
25
26
27
A gradual increase of CHD concentration improved the yield
CHD
3
of tetrahydrofuryl butyronitrile 6. This trend leveled off at c0
=
8
3-4 M (Figure 2). The same trend is observed for dinitrile
formation (product 10), although at lower yields. The yield of
2-phenyl-5-methyltetrahydrofuran 5a correlates with CHD concen-
tration. We therefore concluded that CHD is involved in formation
of product 5a (vide infra). To achieve maximum selectivity for
monoaddition product formation, we chose a value ∼3.3 M as the
standard CHD concentration in oxidation-addition cascades.
(iv). Cobalt(II) Reagents. Aerobic oxidative cyclization of
1-phenyl-4-pentenol 1a is catalytic in cobalt(II) complexes 3
and 4. No oxidative transformation of 1a occurs if catalysts 3 and
2
oxidation-addition cascades.
(
i). Alkenol, Olefin, and Cobalt Concentrations. We con-
sidered 24 h as a reasonable time limit to achieve quantitative
conversion of substrate 1a. To meet this prerequisite, a solution
of 1-phenyl-4-pentenol 1a (0.33 M) in cyclohexa-1,4-diene
(CHD; 3.3 M)/toluenecontaining cobalt(II) complex 3 (5 mol %)
4 are replaced by Co(OAc) (as the tetrahydrate or in anhydrous
2
has to be stirred at a reaction temperature of 60 °C in an open
form). Cobalt(II) complex 4 in combination with γ-terpinene at
a temperature of 75 °C was more reactive than 3 (Supporting
Information), thus allowing us to reduce the amount of catalyst
by 40% and shorten reaction times by ∼70%. Catalyst 4, how-
ever, did not necessarily lead to the highest yields, so that both
catalysts, that is 3 and 4, were checked. The best yields obtained
are given in the tables and schemes.
flask equipped with a reflux condenser. This set up provides
1
4
8
5% of trans-5-methyl-2-phenyltetrahydrofuran 5a. In an
atmosphere of argon under otherwise similar conditions,
substrate 1a is virtually inert.
Alkenol and cobalt(II) concentrations are a critical parameter
for the oxidative part of the cascade. Their concentrations were
therefore kept constant in reactions performed in the presence of
acrylonitrile 2a. We found that 1.7 M acrylonitrile 2a is required
(v). Olefin Reactivity. Alkenes substituted by a CO CH
2
3
(2b), COCH (2c), and SO CH group (2d) give monoaddition
3
2
3
3
907
dx.doi.org/10.1021/ja108403s |J. Am. Chem. Soc. 2011, 133, 3906–3912

Journal of the American Chemical Society
ARTICLE
Table 1. Preparation of Cobalt(II) Chelates from CF -Substituted 1,3-Diketones
3
n
1
2
3
19
a
CdO/cm-1b
c
entry
HL
R = R
R
3/4/%
δ F/ppm
ν
λmax/nm (lg ε/ε*)
1
2
e
1
2
HL
HL
CF
3
CH
3
3: 94
-55.4
1624
1609
231 (3.28), 305 (2.67), 429 sh
d
H
CF
3
4: 99
6.1
252 (3.33), 319 (3.55)
a
b
2
-1 c
2
-1 d
e
In acetone/CDCl
3
[50:50 (v/v)]. Pelletized in KBr. ε in m mol . ε* = 1 m mol . Formed as EtOH adduct (combustion analysis). Formed as
dihydrate (combustion analysis) that quantitatively loses H
2
O upon drying (IR).
Table 2. Selectivity in Aerobic Alkenol Oxidation-Olefin
a
Addition Cascades
n d
2
entry
2
Z
CoL
2
H X T/°C 5a/% 6-9/% 10-13/%
1
2
3
4
5
6
7
8
9
2a CN
2a CN
2a CN
2a CN
2a CN
2b CO
2b CO
3
3
4
4
4
3
3
3
4
4
CHD 60
γ-Ter 60
CHD 75
γ-Ter 60
γ-Ter 75
CHD 60
γ-Ter 75
CHD 60
γ-Ter 75
CHD 60
29
16
23
15
16
28
30
19
27
20
6: 34
6: 40
6: 35
6: 41
6: 44
7: 27
7: 32
8: 21
8: 31
9: 43
10: 11
10: 18
10: 12
10: 16
10: 15
11: 13
11: 13
12: 20
12: 13
Figure 1. Temperature profile of product selectivity in aerobic oxida-
tion of alkenol 1a in the presence of acrylonitrile 2a and CHD catalyzed
2a
CHD
1a
by 3 (c = 1.5 M; 5.0 equiv; c0
c0 = 0.02 M).
= 3.5 M in toluene; c0 = 0.30 M,
0
3
to the existing mechanistic and stereochemical analysis, a kinetic
2
CH
3
3
14
approach.
2
CH
Structure effects on rate constants of primary alkyl radicals in
additions or H-atom abstractions are small within the experi-
mental error. Therefore, we used the rate constants of the ethyl
radical (H-atom abstraction from CHD) and the 5-hexen-1-yl
radical (addition to acrylonitrile 2a and methyl acrylate 2b) for
describing the reactivity of assumed primary radical 14. To apply
reference data for ambient temperature reactions for comparison
with experimental values obtained at 60-75 °C, temperature
effects on homolytic substitution (from CHD) and addition
(to 2a-c) were assumed to be similar.
2c C(O)CH
3
3
2c C(O)CH
e
1
0
2d SO
2
CH
3
13: -
a
b
Quantitative alkenol conversion within 24 (for 3) or 8 h (for 4). Cis:
trans < 1:99 (GC and H NMR). 50/50 ratio of diastereomers with
respect to the stereocenter in the R-position to CN. 5 mol % of 3 and 3
mol % of 4 with respect to alkenol 1a. Not detected ( H NMR, GC).
1
c
d
e
1
products 7-9 besides 2-fold-addition products 10-12 if treated
under standard conditions (Table 2, entries 6-10). Surprisingly,
no 2-fold addition product was found in reactions starting from
methyl vinyl sulfone (2d). The oxidative ring closure of 1-phenyl-
-pentenol 1a occurs in all instances 2,5-trans selectively ( H
Since H-atom abstractions from CHD and primary alkyl
radical additions to acrylonitrile are irreversible, the ratio of 5a
versus the combined yield of addition products 6 and 10 directly
1
4
exp
NMR, GC). For stereoassignment, we used NOESY spectra and
chemical shift correlations ( H, C) obtained from a combined
leads to an experimental partial rate factor (f2 ; Table 3, eq 1).
1
13
The physical organic meaning of the partial rate factor is similar
to a relative rate constant. A relative rate constant, however, is
determined from a series of experiments, whereas the partial rate
factor is an approximation from only one data point.
31
0
NMR/X-ray diffraction study. The stereocenter 2 , which is in
proximity to the cyano group in dinitrile 10, is formed without
diastereoselectivity (dr =50:50). Despite considerable efforts, we
were not able to obtain useful yields for reactions between 1a and
crotononitrile (13% of monoaddition product; 58% of 5a) or
cinnamoyl nitrile [44% of 5a; traces of addition product (GC-
MS); not shown] using the available catalysts.
By inserting absolute rate constants from the literature and
considering respective concentrations from our experiments (3.0
M for CHD and 1.5 M for alkenes 2a-c; Table 2), partial rate
calcd
2
exp
factors were calculated according to eq 1 (f
f2
). Values f2 and
for olefins 2a-c nearly match. We therefore conclude that
the reactivity of 14 and of primary alkyl radicals used to reference
calcd
2
.2.2. Mechanistic Considerations. To verify the radical nat-
ure of oxidatively cyclized 4-pentenol 1a, we chose, in extension
3
908
dx.doi.org/10.1021/ja108403s |J. Am. Chem. Soc. 2011, 133, 3906–3912

Journal of the American Chemical Society
ARTICLE
Table 4. Oxidation-Radical Addition Cascades Starting
from Phenylpent-4-en-1-ols 1a-d and Fumarodinitrile 2e
a
Figure 2. Correlation of product selectivity and CHD concentration in
aerobic alkenol oxidations in the presence of acrylonitrile 2a (toluene as
2
a
1a
3
cosolvent; c0 = 1.5 M; c0 = 0.30 M; c0 = 0.02 M; T = 60 °C).
1
1
3
4
n b
2
c
c
entry
1
R
R
R
R
CoL
5/% (cis:trans) 17/% (cis:trans)
calcd
Table 3. Correlation of Calculated Rate Factors f2
Experimental Data from Aerobic Alkenol Oxidation in the
CHD/2 Competition System (cf. Scheme 2)
versus
1
2
3
4
5
1a Ph H
1a Ph H
1b Ph H
H
H
H
H
3
4
3
3
3
5a: 2 (<1:99) 17a: 66 (<1:99)
H
5a: 1 (<1:99) 17a: 66 (<1:99)
c
CH
3
5b: -
17b: 58 (<1:99)
17c: 58 (74:26)
17d: 53 (3:97)
calcd a
exp b
entry
2
Z
f
2
f
2
c
1c
H
Ph H
H
5c: -
c
1
2
3
2a
2b
2c
CN
CO
4.7
1.4
1.7
1d H
H
Ph H
5d: 8 (3:97)
c
a
2
CH
3
1.4
2.2
For reactant concentrations refer to the text and the Experimental
b
d
C(O)CH
3
(2.2)
Section; quantitative conversion of 1a-d. 5 mol % of 3 and 3 mol % of
c
d
a
H
4 with respect to alkenol 1a. Not detected (GC-MS). 50/50 mixture
of diastereomers with respect to the configuration of the substituents at
For 3.0 M CHD and 1.5 M olefin concentration according to eq 1; k =
4
-1 -1
•
32 b
5
.8 ꢁ 10 M
s
( C H þ CHD; 27 °C). For 60 °C (cf. Table 2,
2
5
0
0
c
add
5
-1 -1
•
the 1 and 2 positions.
entries 1, 6, and 8). k = 5.4 ꢁ 10 M
s
[H
2
CdCH(CH
2
2
)
3
CH
2
5
-1 -1
•
þ 2a] and 1.6 ꢁ 10 M
s
[H
2
CdCH(CH
2
)
3
CH
þ 2b] for 20 °C
d
in CH
2
Cl
2
.
Estimated on the basis of relative rate constants for
Table 5. Oxabicyclo[4.3.0]nonylmethylbutyrodinitrile
Synthesis from cis-Allylcyclohexanols 1e and 1f and Fumarodi-
nitrile 2e
5
-hexen-1-yl radical addition to methyl acrylate and methyl vinyl ketone
add add
33-35
(k2c /k2b = 1.6 at 69 °C).
Scheme 2. Elementary Reactions for Rate Factor (f ) Anal-
2
a
ysis in Aerobic Oxidation Olefin Addition Cascades (see text
and Table 3)
a
a
entry
1
R
5/% (cis:trans)
17/% (cis:trans)
1
2
1e
H
5e: 5 (<1:99)
5f: 24 (17:87)
17e: 67 (<1:99)
17f: 19 (25:75)
partial rate factor analysis is similar (Table 3).
b
1f
2 3
CO CH
kadd½2aꢀ
a
b
ð½6ꢀ þ ½10ꢀÞ
Refers to positions 6 and 8. Control in the absence of 2e:5f: 73%
(8:92).
¼
f2a ¼
ð1Þ
H
½
.3. Terminating Sequential Reactions with 1,2-Substi-
5aꢀ
k ½CHDꢀ
2
from 2-phenyl-4-pentenol 1c (2,4-cis), which is in agreement
1
4
tuted Alkenes. We chose fumarodinitrile 2e (c = 1.7 M) to
improve the selectivity of monoaddition product formation.
with the stereochemical guidelines of the cobalt method. The
use of catalysts 3 and 4 provided a similar yield for 17a but
different values for 17e (67% for catalyst 4 and 47% for catalyst
3). Compared to the acrylonitrile reactions, yields of reduction
0
Alkyl radicals (nucleophilic) add faster to the olefin 2e than to
6
acrylonitrile 2a. The set of substrates was extended to 1-, 2-, or
1
4,28,32
3
-phenyl-substituted 4-pentenols 1a-d and to cis-2-allyl cyclo-
products 5a-d
remained low (<1-8%). No 2-fold alkene
hexanol 1e to broaden the scope of the method. We performed
all experiments under standard conditions, leading to butyrodi-
nitriles 17a-b (2,5-trans), 17d (2,3-trans), and 17e (6,8-trans;
Tables 4 and 5). Cis selectivity is found for oxidations starting
addition products to cyclized radicals were found (GC-MS).
To probe an oxidation-cyclization cascade starting from an
acceptor-substituted alkenol, we subjected substrate 1f (R =
CO
CH
) and fumarodinitrile 2e to standard conditions. From
3
2
3
909
dx.doi.org/10.1021/ja108403s |J. Am. Chem. Soc. 2011, 133, 3906–3912

Journal of the American Chemical Society
ARTICLE
a
Table 6. CoL -Catalyzed Oxidation of Phenylpent-4-en-1-ols 1a-c in the Presence of Dimethyl Fumarate (E)-2f
2
1
1
3
n b
2
c,d
entry
1
R
R
R
CoL
T/°C
H
2
X
5/% (cis:trans)
18/% (cis:trans)
1
2
3
4
1a
1a
1c
1d
Ph
Ph
H
H
H
Ph
H
H
H
H
Ph
3
4
3
3
60
75
60
60
CHD
γ-Ter
CHD
CHD
5a: 15 (1:99)
18a: 51 (<1:99)
18a: 60 (<1:99)
18c: 57 (73:27)
18d: 47 (5:95)
5a: 28 (<1:99)
e
5c: -
H
5d: 27 (5:95)
a
b
For reactant concentrations refer to the text and the Experimental Section; quantitative conversion of 3. 5 mol % of 3 and 3 mol % of 4 with respect to
alkenol 1a. Cis/trans ratios refer to the relative configuration of the substituents at tetrahydrofuran. 50/50 mixture of diastereomers with respect to the
configuration at position 2 . Not detected (GC).
c
d
0
e
Scheme 3. Alkenol Oxidation-Alkylation Sequences Start-
ing from (Z)-1,2-Diacceptor-Substituted Olefins
Table 7. Substituent Effects in Oxidative Cyclization-Alkyne
a
Addition Cascades
1
2
a
a
entry
R
R
20
5a/%
21/22/% (E:Z)
1
2
H
Et
20a
20b
41
28
21: 47 (53:47)
22: 50 (62:38)
CO
2
CH
3
Me
a
1
Quantitative conversion of 1a; cis:trans < 2:98 for 5a, 21, and 22 ( H
NMR); 3 mol % of 4.
this experiment, reduction product 5f and butyrodinitrile 17f
were obtained in low-yields (Table 5, entry 2). This observation
is in line with the reduced nucleophilicity of the radical that is
formed from oxidative cyclization of 1f and thus the lower rate for
addition to alkene 2e. From previous studies on acceptor-
Attempts to apply maleic acid anhydride to conduct oxida-
tion-cyclization cascades lead to an irreversible deactivation of
catalysts 3 and 4. N-Phenyl maleic imide, on the other hand,
provided 65% of addition product 19 if subjected to standard
conditions (Scheme 3, bottom).
14
substituted substrates, we were prepared to find lower trans
selectivity for oxidative cyclization of 1f (66% for 5f) compared
to 1e (>99% for 5e). Treatment of alkenol 1f with a potassium
tert-butoxide in toluene on the other hand provides a 55/45
mixture of 5f (81%, not shown), whereas Lewis-acidic cobalt(II)
complex 4 did not induce cyclization of the substrate.
2.4. Trapping with Alkynes. Mono- and 2-fold ester-sub-
stituted alkynes gave R,β-unsaturated esters 21 and 22 under
standard conditions (Table 7). We transformed the yields into
relative reactivity and thereby used methyl acrylate (2b) as a
reference. The value f /f = 1.4 (75 °C) compares reasonably
2
b 20a
well with the relative rate constant for cyclohexyl radical addition to
rel
6
Starting from alkenols 1a and 1c-d and dimethyl fumarate
E)-2f, we prepared tetrahydrofurans 18a and 18c-d in yields
methyl acrylate 2b and methyl propynoate (k = 3.0, 20 °C). We
therefore explain the chemistry of alkyne trapping in oxidation-
radical addition cascades in extension to the mechanism outlined
for reactions with alkenes (Schemes 1 and 2). Support for this inter-
pretation comes from stereochemical analysis of fumarate formation
(
between 47% and 60% (Table 6). Product and stereoselectivity
thereby follow the same trends as outlined for fumarodinitrile
reactions (cf. Table 4). The use of dimethyl maleate (Z)-2f
provided only 23% dimethyl [2-(2-phenyltetrahydrofuryl-5-
methyl)] succinate 18a but 54% of reduction product 5a
33
from 20b. Syn-selective H-atom transfer onto vinyl radicals having
small substituents attached in the β position to the radical center
generally is favored, thus providing an explanation for the (E)-
selectivity in the synthesis of fumarate 22 (Table 7, entry 2).
Aerobic oxidation of alkenol rel-(1R,3S)-1f and ethyl propio-
late 20a in CHD/toluene catalyzed by cobalt(II) complex 4
furnished bistetrahydrofuran 23 as the single diastereomer
(Scheme 3, top). The data and a rate factor analysis [f(E)-2f/
f(Z)-2f = 5.1 for 75 °C] show that different performances of (E)-2f
and (Z)-2f in the cobalt method correlate with the relative
rel
reactivity of cyclohexyl radical addition to the alkenes (k
k(E)-2f/k(Z)-2f = 10, 20 °C).
=
6
3
910
dx.doi.org/10.1021/ja108403s |J. Am. Chem. Soc. 2011, 133, 3906–3912

Journal of the American Chemical Society
ARTICLE
Scheme 4. Bistetrahydrofuran Formation in Oxidation-Alkyne Addition-Oxidation Cascade
without the necessity to apply a hydroxyl protecting group (Scheme
). The total yield of 1f-derived products added to 48% at a
conversion of 68%. Substrate turnover stopped in a reproducible
manner at this point and could not be taken to completion, even
upon addition of further aliquots of catalyst 4 and CHD.
products will be minimized to improve the yield of 1/1 addition
5
products (e.g., 6). Although this selectivity is not perfect for
6,8,34
Bu SnH and CHD chemistry as well,
the cobalt(II) method
3
could add another dimension to this issue.
To explain this chemistry, we assumed that alkyne trapping by
the intermediate vinyl radical and subsequent reduction leads to
alkenol 25. In a second aerobic alkenol cyclization, alkenol 25 is
transformed into product 23. For two reasons, we consider the
second cyclization to occur via a cobalt-catalyzed reaction as well.
First, the fused tetrahydrofuran ring is formed with the trans
selectivity that is characteristic for the cobalt method (see text
associated with Table 5). Second, at one stage of the reaction a
carbon nucleophile must have existed, which added to ethyl
propynoate (20a) to give after H-atom abstraction from CHD
product 24 [12%; mixture of (E)/(Z)-isomers in favor of (E)-24]
4
. EXPERIMENTAL SECTION
4.1. General. For general laboratory practice and instrumentation
see ref 14 and the Supporting Information.
4.2. Reaction of 1-Phenylpent-4-enol (1a) with Fumaro-
dinitrile 2e. A suspension of alcohol 1a (164 mg, 1.01 mmol),
fumarodinitrile 2e (396 mg, 5.07 mmol), CHD (1.0 mL, 10.2 mmol),
and cobalt(II) reagent 3 (33.2 mg, 50.8 μmol) in toluene (1.6 mL) was
stirred for 21 h at 60 °C while being exposed to laboratory atmosphere.
The reaction mixture was cooled to 20 °C. Unspent fumarodinitrile was
removed by filtration. The filtrate was purified by column chromatog-
19
(Scheme 4).
2
raphy [SiO , acetone/pentane = 1:5-1:3 (v/v)].
2
8
trans-2-Methyl-5-phenyltetrahydrofuran (5a) . Yield: 3.1 mg
19.1 μmol, 2%).
(
3. CONCLUDING REMARKS
2-[(trans-5-Phenyltetrahydrofur-2-yl)-methyl] Butanedinitrile (17a).
The synthesis of side-chain-extended tetrahydrofurans from
Yield: 161 mg (66%); colorless oil [cis:trans < 1:99, 50/50 mixture of
alkenols and acceptor-substituted alkenes combines aspects of
aerobic oxidation catalysis and reductive alkyl radical chemistry.
The combination of a polar and a radical reaction for adding
chemically different entities at the two carbons of a π bond is not
available from oxidation catalysis or free radical chemistry alone.
The method is furthermore one of the few catalytic procedures
for radical generation in synthesis.
diastereoisomers with respect to C ]. R = 0.22 for acetone/pentane =1:5
R
f
(v/v). δ (CDCl , 400 MHz) 1.69-1.78 (1 H, m), 1.86-2.11 (3 H, m),
H
3
2.24-2.30 (1 H, m), 2.37-2.42 (1 H, m), 2.80 (2 H, d, J 6.5), 2.83-2.95
(2 H, m), 3.21 (1 H, quint, J 6.5), 3.26-3.31 (1 H, m), 4.38-4.43 (1 H,
m), 4.99-5.04 (1 H, m), 7.27-7.36 (5 H, m). δ (CDCl , 100 MHz)
C
3
2
8
0.3/21.2, 25.8/26.5, 32.36/32.39, 34.7/35.0, 36.2/37.9, 75.7/76.6, 80.5/
0.7, 115.7/116.0, 118.9/119.1, 125.4, 127.4, 128.3, 142.5/142.6. Anal.
Calcd for C15
H
16
N
2
O (240.30): C, 74.97; H, 6.71; N, 11.66. Found: C,
From a synthetic point of view, the cobalt method is expected
to offer a quite general solution for radical generation from
þ
74.63; H, 6.50; N, 11.62. MS (EI) m/z (%) 240 (31, M ), 223 (6), 183
(
12), 146 (14), 129 (9), 117 (32), 105 (100), 91 (44), 77 (37).
.3. Reaction of Alkenol 1a with Dimethyl Butynedioate
20b). A solution of alcohol 1a (164 mg, 1.01 mmol), alkyne 20b (727
mg, 5.01 mmol), γ-terpinene (1.9 mL, 98% pure, 11.5 mmol), and
4
-pentenols. Also, other acceptor-substituted olefins that are
4
applied for carbon-carbon bond formation via radical addition
to alkenes, such as aryl vinyl sulfones or 1,1-dichloroalkenes, shall
be considered as candidates for broadening the scope of the homo-
logization step. In terms of efficiency, the cobalt method compares
(
2
[
CoL (4)] 2H O (15.8 mg, 30.0 μmol) in toluene (0.4 mL) was
2 2
3
stirred for 7 h at 75 °C while being exposed to laboratory atmosphere.
The reaction mixture was cooled to 20 °C and directly poured onto a
well with values from Bu SnH-mediated reactions that uses stoichio-
3
metric amounts of the progenitor (such as an alkyl halide, xanthate,
or carboxylic acid) and an acceptor-substituted olefin.
column (SiO
1:10-1:5 (v/v)].
trans-2-Methyl-5-phenyltetrahydrofuran (5a) . Yield: 45.1 mg
2
) for chromatographic purification [acetone/pentane
=
The major challenge for further improving selectivity in
oxidation-addition cascades certainly lies in the quest for a
smarter reductant, which is able to maintain the catalytic cycle
but also to more delicately respond to polarity differences
between cyclized radicals and more electrophilic carbon radicals
that result from addition to an acceptor-substituted alkene. By
this approach formation of 2-fold addition and reduction
28
(278 μmol, 28%).
Dimethyl 2-[(trans-5-phenyltetrahydrofur-2-yl)methyl] (Z)-Butene-
dioate ((Z)-22). Yield: 59.6 mg (19%), colorless oil. R = 0.42 for
f
acetone/pentane = 1:5 (v/v). δ (CDCl , 400 MHz) 1.72-1.85 (2 H,
H
3
m), 2.11-2.20 (1 H, m), 2.38-2.43 (1 H, m), 3.02 (1 H, dd, J 12.4, 4.8),
3.29 (1 H, dd, J 12.4, 8.4), 3.72 (3 H, s), 3.79 (3 H, s), 4.46 (1 H, quint,
3
911
dx.doi.org/10.1021/ja108403s |J. Am. Chem. Soc. 2011, 133, 3906–3912

Journal of the American Chemical Society
ARTICLE
J 6.3), 4.96 (1 H, t, J 6.6), 6.84 (1 H, s), 7.20-7.32 (5 H, m). δ
C
(CDCl
3
,
(7) Ramaiah, M. Tetrahedron 1987, 43, 3541–3676.
1
00 MHz) 31.4, 33.5, 34.6, 51.6 (CH ), 52.5 (CH ), 78.4, 79.9, 125.6,
(8) Giese, B. Angew. Chem., Int. Ed. Engl. 1985, 24, 553–565.
(9) Chatgilialoglou, C. In Organic Synthesis; Renaud, P., Sibi, M. P.,
Eds.; Wiley-VCH: Weinheim, 2001; Vol. 1, pp 28-49.
3
3
1
27.0, 127.7, 128.2, 143.4, 144.7, 166.2, 167.4. MS (EI) m/z (%) 304
þ
(4, M ), 272 (4), 244 (3), 185 (12), 147 (100), 129 (45), 120 (24), 105
22), 91 (72), 77 (16).
(
10) Andrews, R. S.; Becker, J. J.; Gagn ꢀe , M. R. Angew. Chem., Int. Ed.
010, 49, 7274–7276.
11) Shih, H.-W.; Vander Waal, M. N.; Grange, R. L.; MacMillan,
D. W. C. J. Am. Chem. Soc. 2010, 132, 13600–13603.
12) Tucker, J. W.; Nguyen, J. D.; Narayanam, J. M. R.; Krabbe,
S. W.; Stephenson, C. R. J. Chem. Commun. 2010, 46, 4985–4987.
13) Affo, W.; Ohmiya, H.; Fujioka, T.; Ikeda, Y.; Nakamura, T.;
(
2
Dimethyl 2-[(trans-5-Phenyltetrahydrofur-2-yl)methyl] (E)-Butene-
(
dioate ((E-22). Yield: 96.1 mg (31%), colorless oil. R
pentane = 1:5 (v/v). δ (CDCl , 400 MHz) 1.68-1.77 (1 H, m), 1.82-
.91 (1 H, m), 2.14-2.21 (1 H, m), 2.35-2.42 (1 H, m), 2.59-2.76
2 H, m), 3.73 (3 H, s), 3.82 (3 H, s), 4.39 (1 H, quint, J 6.6), 5.01 (1 H, t,
J 7.3), 6.00 (1 H, s), 7.22-7.35 (5 H, m). δ (CDCl , 100 MHz) 31.9,
f
= 0.21 acetone/
H
3
(
1
(
(
C
3
Yorimitsu, H.; Oshima, K.; Imamura, Y.; Mizuta, T.; Miyoshi, K. J. Am.
Chem. Soc. 2006, 128, 8068–8077.
34.9, 40.6, 51.8 (CH
3
), 52.3 (CH
3
), 77.1, 80.5, 121.8, 125.4, 127.1,
þ
128.3, 143.2, 146.8, 165.3, 169.0. MS (EI) m/z (%) 304 (3, M ), 272
(
14) Schuch, D.; Fries, P.; D €o nges, M.; Men ꢀe ndez-P ꢀe rez, B.; Har-
tung, J. J. Am. Chem. Soc. 2009, 131, 12918–12920.
15) Ishii, Y.; Sakaguchi, S. In Modern Oxidation Methods; B €a ckvall,
(
7
6), 244 (4), 185 (9), 147 (97), 129 (54), 120 (21), 105 (31), 91 (100),
7 (24).
.4. Reaction of rel-(1R,3S)-1-Phenylpent-4-en-1,3-diol
1f) with Ethyl Propynoate (20a). A solution of alkenol 1f (193
mg, 1.08 mmol), alkyne 20a (1.09 g, 10.8 mmol), CHD (1.5 mL, 15.3
(
4
J.-E., Ed.; Wiley-VCH: Weinheim, 2004; pp 119-163.
(16) Hartung, J.; Greb, M. J. Organomet. Chem. 2002, 661, 67–84.
(17) Cardillo, G.; Orena, M. Tetrahedron 1990, 46, 3321–3408.
(
2
(
(
18) Wolfe, J. P.; Hay, M. B. Tetrahedron 2007, 63, 261–290.
19) Men ꢀe ndez P ꢀe rez, B.; Schuch, D.; Hartung, J. Org. Biomol. Chem.
mmol), and [CoL
1.5 mL) was stirred at 60 °C for 16 h while being exposed to laboratory
atmosphere. The reaction mixture was cooled to 20 °C and directly
poured onto a column (SiO ) for chromatographic purification [acetone/
petroleum ether =1:5 (v/v)].
Ethyl 2-{rel-(2R,3aS,5R,6aS)-Hexahydro-2-phenylfuro[3,2-b]fur-5-
yl} Acetate (23). Yield: 108 mg (36%), colorless oil. R = 0.37 for
acetone/pentane =1:5 (v/v). δ (CDCl , 600 MHz) 1.28 (3 H, t, J 7.2),
.81 (1 H, ddd, J 13.6, 9.4, 5.1), 1.90 (1H, ddd, J 13.6, 10.4, 4.6), 2.38
2 2
(4)] 2H O (28.9 mg, 55.0 μmol) in toluene
3
(
2
008, 6, 3532–3541.
(
(
20) Inoki, S.; Mukaiyama, T. Chem. Lett. 1990, 67–70.
21) Wang, Z.-M.; Tian, S.-K.; Shi, M. Eur. J. Org. Chem. 2000, 349–
2
3
56.
(22) For a tutorial review on aerobic cobalt(II)-catalyzed oxidations,
f
see: Sam ꢀa ndi, L. I. In Advances in Catalytic Activation of Dioxygen by Metal
Complexes; Sim ꢀa ndi, L. I., Ed.; Kluwer Academic Publishers: Dordrecht,
H
3
1
2
003; pp 265-328.
23) Men ꢀe ndez P ꢀe rez, B.; Hartung, J. Tetrahedron Lett. 2009, 50,
60–962.
24) Yaqub, M.; Koob, R. D.; Morris, M. L. J. Inorg. Nucl. Chem.
(
(
1 H, dd, J 13.6, 5.1), 2.48-2.65 (3 H, m), 4.18 (2 H, q, J 7.2), 4.55-4.60
1 H, m), 4.82 (1 H, t, J 4.6), 4.92 (1 H, t, J 4.6), 5.08 (1 H, dd, J 10.4,
(
9
5
6
C 3
.1), 7.29-7.36 (5 H, m). δ (CDCl , 150 MHz) 14.2, 40.6, 40.9, 43.9,
(
0.6, 76.4, 81.3, 84.1, 84.2, 125.7, 127.5, 128.4, 141.7, 171.0. MS (EI)
1971, 33, 1944–1946.
þ
m/z (%) 276 (4, M ), 258 (6), 189 (22), 117 (25), 105 (100), 91 (15),
(25) Giese, B.; Kretzschmar, G. Chem. Ber 1984, 117, 3160–3164.
(26) Giese, B.; Carboni, B.; G €o bel, T.; Muhn, R.; Wetterich, F.
Tetrahedron Lett. 1992, 33, 2673–2676.
77 (25).
(
27) Giese, B.; Hartung, J.; Hasskerl, T.; Houk, K. N.; H €u ter, O.;
’
ASSOCIATED CONTENT
Zipse, H. J. Am. Chem. Soc. 1992, 114, 4067–4079.
S
Supporting Information. Experimental procedures and
(28) Hartung, J.; Hiller, M.; Schmidt, P. Chem.—Eur. J. 1996, 2,
b
1
014–1023.
29) Binmore, G.; Walton, J. C.; Gardellini, L. J. Chem. Soc., Chem.
Commun. 1995, 27–28.
30) Dewick, P. M. in Medicinal Natural Products, 2nd ed.; Wiley:
Chichester, 2002; Chapter 5, pp 172-186.
31) Hartung, J.; Kneuer, R.; Laug, S.; Schmidt, P.; Spehar, K.;
Svoboda, I.; Fuess, F. Eur. J. Org. Chem. 2003, 4033–4052.
32) Hawari, J. A.; Engel, P. S.; Griller, D. Int. J. Chem. Kinet. 1985,
17, 1215–1219.
33) Minisci, F.; Zammori, P.; Bernardi, R.; Cecere, M.; Galli, R.
Tetrahedron 1970, 26, 4153–4166.
34) Citterio, A.; Minisci, F.; Arnoldi, A.; Pagano, R.; Parravicini, A.;
Porta, O. J. Chem. Soc., Perkin Trans 2 1978, 519–524.
35) Citterio, A.; Arnoldi, A.; Minisci, F. J. Org. Chem. 1979, 44,
674–2682.
36) Hartung, J.; Hiller, M.; Schmidt, P. Liebigs Ann. Chem. 1996,
425–1436.
37) Giese, B. Angew. Chem., Int. Ed. Engl. 1989, 28, 969–980.
(38) Baguley, P. A.; Walton, J. C. Angew. Chem., Int. Ed. Engl. 1998,
37, 3073–3082.
spectral and analytical data of new compounds. This material is
available free of charge via the Internet at http://pubs.acs.org.
(
(
’
AUTHOR INFORMATION
ꢁ
(
Corresponding Author
hartung@chemie.uni-kl.de
(
(
’
ACKNOWLEDGMENT
(
This work was supported by the Schering-Stiftung
(
(
Scholarship for P.F.) and the Stiftung f €u r Innovation des Landes
Rheinland-Pfalz (Grant 898). It is part of the Ph.D. thesis of P.F.
2
1
(
’
REFERENCES
(
(
1) Giese, B. Radicals in Organic Synthesis-The Formation of Carbon
Carbon Bonds; Pergamon: Oxford, 1986.
2) Giese, B. In Radicals in Organic Synthesis; Renaud, P., Sibi, M. P.,
Eds.; Wiley-VCH: Weinheim, 2001; Vol. 1, pp 381-399.
3) Jasperse, C. P.; Curran, D. P.; Fevig, T. L. Chem. Rev. 1991, 91,
237–1286.
4) Curran, D. P. Synthesis 1988, 489–513. (b) Curran, D. P.
Synthesis 1988, 417–439.
5) Lee, E. In Radicals in Organic Synthesis; Renaud, P., Sibi, M. P.,
Eds.; Wiley-VCH: Weinheim, 2001; Vol. 2, pp 303-333.
6) Giese, B. Angew. Chem., Int. Ed. Engl. 1983, 22, 753–764.
(
(
1
(
(
(
3
912
dx.doi.org/10.1021/ja108403s |J. Am. Chem. Soc. 2011, 133, 3906–3912