Synthesis of gem-difluoromethylene building blocks through regioselective allylation of gem-difluorocyclopropanes

Article abstract of DOI:10.1021/ol500803r

gem-Difluorocyclopropane derivatives react with allyltributylstannane in the presence of 2,2′-azobis(isobutyronitrile) to afford 1,6-dienes with a gem-difluoromethylene moiety at the allylic position. The reaction proceeds regioselectively with high yield

Full text of DOI:10.1021/ol500803r

Letter
pubs.acs.org/OrgLett
Synthesis of gem-Difluoromethylene Building Blocks through
Regioselective Allylation of gem-Difluorocyclopropanes
Daisuke Munemori,† Kent Narita,† Toshiki Nokami,†,‡ and Toshiyuki Itoh*,†,‡
†Department of Chemistry and Biotechnology, Graduate School of Engineering, Tottori University, 4-101 Koyama-minami, Tottori
680-8552, Japan
‡Center for Research on Green Sustainable Chemistry, Tottori University, 4-101 Koyama-minami, Tottori 680-8552, Japan
S
* Supporting Information
ABSTRACT: gem-Difluorocyclopropane derivatives react with allyltributylstannane in
the presence of 2,2′-azobis(isobutyronitrile) to afford 1,6-dienes with a gem-difluoro-
methylene moiety at the allylic position. The reaction proceeds regioselectively with high
yields, and the 1,6-dinenes obtained are good precursors for cyclic systems containing a
gem-difluoromethylene moiety. Although S-methyl carbonodithioate also works as a
leaving group, rearrangement of the leaving group competes with the desired allylation,
depending on the amount of allyltributylstannane.
he incorporation of a fluorine atom into an organic
Tmolecule can alter the chemical reactivity of the resulting
compound due to the strong electron-withdrawing nature of
fluorine, thus making it possible to create a new molecule that
exhibits unique physical and biological properties.1 As a result,
Figure 1. Working hypothesis of the present synthetic project.
much attention has focused on the preparation of gem-
difluoromethylene derivatives as a source of novel functional
materials.1,2 Syntheses of such compounds have generally been
(1a; R = BnOCH2) was initially selected as the substrate for the
present study because it was established that decomposition of
achieved by difluorination of carbonyl or thiocarbonyl func-
tional groups.3 However, the number of fluorination reagents is
the cis isomer of 1,2-dialkyl-3,3-difluorocyclopropane took place
more easily than that of its trans isomer.7
The reaction was conducted as follows (Table 1): a mixture
limited and the reagents are generally very expensive; hence,
synthetic strategies that use building blocks containing a gem-
of 1a (R = BnOCH2),8 allylBu3Sn, and a catalytic amount of
difluoromethylene moiety have been recognized as an attractive
2,2′-azobis(isobutyronitrile) (AIBN) in benzene (1.0 M) was
alternative route through which to access gem-difluoro-
methylene compounds. Over the years, we synthesized a
range of gem-difluorocyclopropane derivatives and revealed
Table 1. Optimizations for Regioselective Allylation
their unique physical and biological properties; as a result of
these studies, numerous types of gem-difluorocyclopropane
compounds are now available.2 Kobayashi and co-workers
reported a radical-induced regioselective ring-opening reaction
of [2,2-difluoro-3-(iodomethyl)cyclopropyl]benzene and suc-
a
entry amount of allylBu3Sn (equiv) amount of AIBN (%) yield (%)
ceeded in preparing (2,2-difluorobut-3-en-1-yl)benzene deriv-
b
atives.4 Dolbier and co-workers reported that radical-type
cleavage of the gem-difluorocyclopropane ring took place very
quickly.5 More recently, Gurjar and co-workers reported the
preparation of a diallyl-substituted compound through a ring-
opening reaction of (halomethyl)cyclopropanes with allyltri-
butylstannane (allylBu3Sn).6 Inspired by these works, we
hypothesized that a novel gem-difluoromethylene compound
2 might be obtained from gem-difluorocyclopropane 1 through
radical-type allylation following regioselective ring opening
(Figure 1).
1
2
3
4
5
6
7
8
2.0
2.0
6.0
7.0
7.0
8.0
3.6
25
5
25
68
84
62
89
71
89
5
5
5
5
5
5
c
10
8.0
d
a
b
Isolated yield. NMR yield based on trioxane as an internal standard.
The reaction was carried out in toluene at 80 °C. The trans isomer
c
d
of 1a was used.
Here, we wish to report the preparation of gem-difluoro-
methylene compounds 2 by the radical-type ring-opening
reaction of gem-difluorocyclopropane 1. The cis isomer of 1-
bromomethyl-2-benzyloxymethyl-3,3-difluorocyclopropane
Received: March 18, 2014
Published: April 30, 2014
© 2014 American Chemical Society
2638
dx.doi.org/10.1021/ol500803r | Org. Lett. 2014, 16, 2638−2641

Organic Letters
Letter
stirred at 80 °C for 12 h and then treated with a mixed solvent
of ethyl acetate and a saturated aqueous potassium fluoride
(KF) solution at room temperature for 1 h to initiate
precipitation. The precipitate formed was removed by filtration,
the filtrate was evaporated, and subsequent purification of the
residue by silica gel thin-layer chromatography (TLC) afforded
2a (R = BnOCH2). By using various types of 1-alkyl-2-
bromomethyl-3,3-difluorocyclopropane 1, regioselective allyla-
tion under the above conditions gave a range of products, 2; the
results are summarized in Table 2.
Table 2. Scope of the Substrates
Initially, we conducted the reaction by using 1a in the
presence of 2.0 equiv of allylBu3Sn and 3.6% AIBN. The
desired product 2a was, however, obtained in only poor yield
under these conditions (Table 1, entry 1). Increasing the
amount of allylBu3Sn improved the chemical yield of the
product significantly, and 2a was obtained in 68%, 84%, and
89% yields when 6, 7, and 8 equiv of allylBu3Sn were used,
respectively (entries 3, 4, and 6). The yield dropped, however,
when 10 equiv of allylBu3Sn were employed in the reaction
(entry 7). A slight drop of the chemical yield of 2a was
recorded when the reaction was carried out in toluene (entry
5). Although we initially expected that the cis isomer of gem-
difluorocyclopropane would be more reactive than the trans
isomer, no difference in the reactivity was observed, and the
chemical yields of the products were similar (entries 6 and 8).
It was thus found that the amount of allylation reagent was
important to achieve the desired reaction. In particular, a large
excess of allylBu3Sn was required for the reaction of [3-
(bromomethyl)-2,2-difluorocyclopropyl]benzene (1b; R = Ph)
because of the relatively poor reactivity of the radical generated
by the ring-opening reaction. The desired product 2b was
obtained in 71% yield in the presence 16 equiv of allyBu3Sn,
whereas only a moderate yield (45%) was obtained when the
reaction was carried out with 8 equiv of allylBu3Sn (Table 2,
entries 1 and 2). No difference was observed between the trans
and cis isomers of 1b (entries 2 and 3). The presence of
electron-withdrawing substituents on the benzene ring, such as
fluorine, chlorine, or bromine, contributed to an improved
product yield, and the desired products 2c (R = 4-F−C6H4), 2d
(R = 4-Cl−C6H4), and 2e (R = 4-Br−C6H4) were obtained in
84%, 89%, and 84% yields, respectively (entries 4, 5, and 6).
Conducting the reaction with 2-arylcyclopropanes substituted
with an electron-donating group at the 4-position led to more
complex results. Whereas 1f (R = 4-Me−C6H4) gave the
desired product 2f in excellent yield (entry 7), 1g (R = 4-
MeO−C6H4) afforded 2g in poor yield (25%) together with the
formation of unidentified byproducts (entry 8). Reaction of
gem-difluorocyclopropane derivatives with aliphatic substituents
such as 1h (R = PhCH2CH2) and 1i (R = 4-MeO−C6H4−
CH2CH2) proceeded very smoothly with 2 equiv of allylBu3Sn
and gave the products 2h and 2i in 77% and 68% yields,
respectively (entries 9 and 10). Furthermore, the allylation was
applicable to bis-gem-difluorocyclopropane 1j, and the desired
product 2j was attained in 72% yield, although in this case the
reaction required an excess of allylBu3Sn to reach completion
(entry 11).
a
b
c
d
Isolated yield. AIBN (25%). AIBN (5%). The cis isomer of 1b was
e
used. 2 equiv of allylBu3Sn were used.
Xanthate is known to be a good leaving group in the
formation of a radical species,9 so we next attempted allylation
using xanthates 1k, 1l, and 1m. The desired product 2h was
indeed obtained in 50% yield by using 2.0 equiv of allyl-Bu3Sn
with O-[(2,2-difluoro-3-phenethylcyclopropyl)-methyl] S-
methyl carbonodithioate (1k). Compound 1l was also attained
in 70% yield as a mixture of E/Z isomers (83:17) when O-[1-
(2,2-difluoro-3-phenethylcyclopropyl)-ethyl] S-methyl car-
bonodithioate (1l) was used as a substrate, although the
reaction required 8.0 equiv of allylBu3Sn (Figure 2). On the
other hand, a mixture of two compounds, (E)-(5,5-difluoro-
nona-3,8-dien-1-yl)benzene (2m) and (E)-S-(2,2-difluoro-6-
phenylhex-3-en-1-yl) S-methyl carbonodithioate (3), was
obtained when O-[1-(2,2-difluorocyclopropyl)-3-phenylpropyl]
2639
dx.doi.org/10.1021/ol500803r | Org. Lett. 2014, 16, 2638−2641

Organic Letters
Letter
Table 3. Results of Radical Type Allylation of gem-
Difluorocyclopropane 1m
allyBu3Sn
(equiv)
amount of AIBN
(%)
yield of 2m
yield of 3
(%)
entry
(%)
a
a
1
2
3
4
5
6
7
4.0
8.0
16
10
10
10
25
25
40
50
35
a
a
55
22
b
b
75
15
0
0
trace
b
b
1.0
0.5
25
65
b
b
15
70
c
b
b
0.5
40
14
74
a
b
c
NMR yield. Isolated yield. 1,1′-Azobis(cyclohexanecarbonitrile)
(V-40) was used as a radical initiator.
the presence of 10% AIBN (entry 3). On the other hand,
compound 3 was obtained as the major product in 70% yield
when the reaction was carried out using 0.5 equiv of allylBu3Sn
in the presence of 40% AIBN (entry 6). It has been reported
that the rate of decomposition of the radical initiator is
important to achieve the desired radical trapping.10 In fact, a
slight increase in the yield of 3 was recorded when 1,1′-
azobis(cyclohexanecarbonitrile) (V-40) was used as the radical
initiator (entry 7).
Figure 2. Allylation of gem-difluorocyclopropanes equipped with the
S-methyl carbonodithioate as a leaving group.
S-methyl carbonodithioate (1m) was subjected to the reaction
conditions (Figure 2).
Figure 3 shows a plausible mechanism of formation for the
We then demonstrated a simple application of gem-
difluoromethylene building block 2h (Scheme 1). The ring-
two products 2m and 3, starting from 1m. Cyclopropane 1m
Scheme 1. Preparation of (2-(2,2-Difluorocyclopent-3-en-1-
yl)ethyl)benzene (6) Derived from 2h through the Ring-
Closing Metathesis Reaction
closing metathesis reaction proceeded smoothly, and cyclo-
pentene 6 was obtained in excellent yield (97%) when diene 2h
was treated with 5% Grubbs catalyst (first generation).11
In summary, we have accomplished the regioselective
allylation of gem-difluorocyclopropane derivatives through a
radical-type ring-opening reaction. Although the reaction
requires a relatively large amount of allylBu3Sn, unique gem-
difluoromethylene compounds were produced. We have also
demonstrated an application of one of the resultant gem-
difluoromethylene compounds. Because gem-difluorocyclo-
propane is easily prepared from relatively inexpensive 2-
chloro-2,2-difluoroacetic acid, the present method opens the
way to an economical synthesis of useful gem-difluoro-
methylene compounds. Fluorine-containing molecules are
now established as key compounds in medicinal and material
chemistry. Because product 2 has two olefin moieties with
differing reactivities, this molecule is expected to become a key
intermediate in the synthesis of many gem-difluoromethylene
compounds. Further investigations into the scope and
limitations of the present method are expected to expand the
potential applications of this approach.
Figure 3. A plausible mechanism of formation of two products 2m and
3 through ring opening of gem-difluorocyclopropane.
reacts with a tributylstannyl radical to give 2,2-difluorocyclo-
propylcarbinyl radical 4. It has been reported that this radical
species undergoes an extraordinarily fast ring-opening reac-
tion.5 Thus, a ring-opening reaction of radical 4 would take
place to rapidly produce 5, which would either be trapped by
allylBu3Sn to afford 2m or generate the rearranged product 3
by trapping with S-methyl S-(tributylstannyl) carbonodithioate
under low concentrations of allylBu3Sn.
According to the proposed mechanism, we concluded that
selective production might be possible by simply changing the
amount of allylBu3Sn (Table 3). As expected, it was found that
2m was indeed obtained in 75% yield as the major product
when a large excess of allylBu3Sn (16 equiv) was employed in
2640
dx.doi.org/10.1021/ol500803r | Org. Lett. 2014, 16, 2638−2641

Organic Letters
Letter
ASSOCIATED CONTENT
* Supporting Information
■
S
Experimental procedures and spectral data for the products.
This material is available free of charge via the Internet at
http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
■
*E-mail: titoh@chem.tottori-u.ac.jp.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
The authors are grateful to Dr. Tomoe Inoue of Arid Land
Research Center, Tottori University for HRMS (EI) analyses.
■
REFERENCES
■
(1) For review, see: Ojima, I., Ed.; Fluorine in Bioorganic and
Medicinal Chemistry; Wiley-Blackwell: London, 2009.
(2) For review, see: Itoh, T. gem-Difluorinatedcyclopropanes as key
building blocks for novel biologically active molecules. In Fluorine in
Bioorganic and Medicinal Chemistry; Ojima, I., Ed.; Wiley-Blackwell:
London, 2009; pp 313−355.
(3) For review, see: Fedorynski, M. Chem. Rev. 2003, 103, 1099.
(4) Morikawa, T.; Uejima, M.; Kobayashi, Y. Chem. Lett. 1988, 1407.
(5) (a) Dolbier, W. R., Jr.; Al-Sader, B. H.; Sellers, F.; Koroniak, H. J.
Am. Chem. Soc. 1981, 103, 2138. (b) Dolbier, W. R., Jr. Acc. Chem. Res.
1981, 14, 195. (c) Tian, S.; Lewis, S. B.; Bartberger, M. D.; Dolbier, W.
R., Jr.; Borden, W. T. J. Am. Chem. Soc. 1998, 120, 6187. (d) Tian, F.;
Bartberger, M. D.; Dolbier, R. W., Jr. J. Org. Chem. 1999, 64, 540.
(6) Gurjar, M. K.; Ravindranadh, S. V.; Sankar, K.; Karmakar, S.;
Cherian, J.; Chorghade, M. S. Org. Biomol. Chem. 2003, 1, 1366.
(7) (a) Mitsukura, K.; Korekiyo, S.; Itoh, T. Tetrahedron Lett. 1999,
40, 5739. (b) Itoh, T.; Ishida, N.; Mitsukura, K.; Uneyama, K. J.
Fluorine Chem. 2001, 112, 63. (c) Itoh, T.; Ishida, N.; Mitsukura, K.;
Hayase, S.; Ohashi, K. J. Fluorine Chem. 2004, 125, 775.
(8) We chose the benzyl protecting group to facilitate detection of
the products.
(9) For recent reviews, see: (a) Zard, S. Z. Chem. Soc. Rev. 2008, 37,
1603. (b) Quiclet-Sire, B.; Zard, S. Z. Beilstein J. Org. Chem. 2013, 9,
557.
(10) Fukuyama, T.; Kobayashi, M.; Rahman, Md. T.; Kamata, N.;
Ryu, I. Org. Lett. 2008, 10, 533.
(11) (a) Schwab, P.; France, M. B.; Ziller, J. W.; Grubbs, R. H. Angew.
Chem., Int. Ed. Engl. 1995, 34, 2039. For recent reviews, see:
(b) Deraedt, C.; d’Halluin, M.; Astruc, D. Eur. J. Inorg. Chem. 2013,
4881. (c) Li, H.; Seechurn, C. C. C. J.; Colacot, T. J. ACS Catal. 2012,
2, 1147.
2641
dx.doi.org/10.1021/ol500803r | Org. Lett. 2014, 16, 2638−2641