A practical synthesis of the PDE4 inhibitor, KW-4490

Article abstract of DOI:10.1021/op1001287

A practical and scalable synthesis of a PDE4 inhibitor KW-4490 (1) was developed. This improved synthesis features the construction of the 1-arylcyclohexene (9) by the Diels-Alder reaction followed by a newly established Brnsted acid-promoted hydrocyanation. Subsequent crystallization-induced dynamic resolution enabled the high-yield production of the desired cis-isomer (cis-8). The synthesis was achieved in seven steps in 37% overall yield.

Full text of DOI:10.1021/op1001287

Organic Process Research & Development 2010, 14, 1182–1187
A Practical Synthesis of the PDE4 Inhibitor, KW-4490
Arata Yanagisawa,*,† Koichiro Nishimura, Kyoji Ando, Tetsuya Nezu, Ayako Maki, Sachiko Kato, Wakako Tamaki,
Eiichiro Imai, and Shin-ichiro Mohri
Chemical Process Research and DeVelopment Laboratories, Kyowa Hakko Kirin Co., Ltd., 1-1-53, Takasu-cho,
Sakai-ku, Sakai, Osaka 590-8554, Japan
Scheme 1. Medical chemistry route
Abstract:
A practical and scalable synthesis of a PDE4 inhibitor KW-4490
(1) was developed. This improved synthesis features the construc-
tion of the 1-arylcyclohexene (9) by the Diels-Alder reaction
followed by a newly established Brønsted acid-promoted hydro-
cyanation. Subsequent crystallization-induced dynamic resolution
enabled the high-yield production of the desired cis-isomer (cis-
8). The synthesis was achieved in seven steps in 37% overall yield.
Introduction
Phosphodiesterase 4 (PDE4) is a cyclic adenosine mono-
phosphate (cAMP)-specific phosphodiesterase which is located
predominantly in inflammatory cells. High levels of cAMP
inhibit the production of cytokines and other molecules that
modulate the inflammatory response.1 Therefore, PDE4 inhibi-
tors have emerged as potential therapeutic agents in the
treatment of asthma and chronic obstructive pulmonary disease
(COPD).2 Since cis-4-cyano-4-(2,3-dihydro-8-methoxy-1,4-
benzodioxin-5-yl)cyclohexanecarboxylic acid (KW-4490, 1)
was identified in Kyowa Hakko Kirin as a potent PDE4
inhibitor,3 multikilogram quantities were thus required in order
to carry out both pharmacological profiling and clinical trials.
Structurally, the compound presents the interesting synthetic
challenges of constructing a tetra-substituted electron-rich
benzene, a tertiary benzylic nitrile, and cis stereochemistry of
a carboxylic acid on a 1,4-disubstituted cyclohexane.
In general, tertiary benzylic nitriles has been prepared by
the double alkylation of benzylic nitrile,4 arylations of secondary
nitrile anions with aryl halides,5 or the displacement of tertiary
benzylic alcohol with cyanide.6 The latest approach was applied
to our medicinal chemistry synthesis of 1, which was suitable
for the delivery of multigram quantities (Scheme 1).3b However,
the synthesis was not directly amenable to large-scale produc-
tion. The most problematic step is the addition of aryllithium 4
to ketone 5, because of its low yield and poor reproducibility.
Our deuterium oxide quenching experiments suggested that the
deprotonation of active protons on 5 was a significantly
competing side reaction, which caused both a low yield and
complex self-condensed byproducts. Furthermore, the resulting
alcohol 6 was a mixture of cis/trans diastereomers which did
not form crystals, but an oily mixture. This finding thereby ruled
out the option of purifying 6 by crystallization. Therefore,
column chromatography purification was unavoidable to obtain
a sufficient quality of the product usable for the following
reaction. We were forced to develop a practical, robust, and
scalable synthesis for the preparation of 1 while keeping out
the alcohol 6.
* Author to whom correspondence may be sent. E-mail: arata.yanagisawa@
kyowa-kirin.co.jp.
† Current address: Medicinal Chemistry Research Laboratories, Fuji Research
Park, 1188 Shimotogari, Nagaizumi-cho, Sunto-gun, Shizuoka 411-8731, Japan.
(1) Teixeira, M. M.; Gristwood, R. W.; Cooper, N.; Hellewell, P. G. Trends
Pharmacol. Sci. 1997, 18, 164–170.
(2) (a) Dyke, H. J.; Montana, J. G. Exp. Opin. InVestig. Drugs 2002, 11,
1–13. (b) Lipworth, B. J. Lancet 2005, 365 (9454), 167–175. (c)
Kroegel, C.; Foerster, M. Exp. Opin. InVestig. Drugs 2007, 16, 109–
124.
(3) (a) Yanagawa, K. The 26th Medicinal Chemistry Symposium, 2007,
2P-29. (b) Ohshima, E.; Yanagawa, K.; Manabe, H.; Miki, I.; Masuda,
Y. PCT Int. Appl. WO0164666 A1, 2001.
(4) Christensen, S. B.; Guider, A.; Forster, C. J.; Gleason, J. G.; Bender,
P. E.; Karpinski, J. M.; DeWolf, W. E., Jr.; Barnette, M. S.;
Underwood, D. C.; Griswold, D. E.; Cieslinski, L. B.; Burman, M.;
Bochnowicz, S.; Osborn, R. R.; Manning, C. D.; Grous, M.; Hillegas,
L. M.; O’Leary-Bartus, J.; Ryan, M. D.; Eggleston, D. S.; Haltiwanger,
R. C.; Torphy, T. J. J. Med. Chem. 1998, 41, 821–835.
(5) (a) Caron, S.; Vazquez, E.; Wojcik, J. M. J. Am. Chem. Soc. 2000,
122, 712–713. (b) Culkin, D. A.; Hartwig, J. F. Acc. Chem. Res. 2003,
36, 234–245. (c) You, J.; Verkade, J. G. Angew. Chem., Int. Ed. 2003,
42, 5051–5053.
Concerning other aspects of the medicinal chemistry syn-
thesis, a cyanide substitution of the alcohol 6 was a high-yield
(6) (a) Muratake, H.; Natsume, M. Tetrahedron 1990, 46, 6331–6342.
(b) Caron, S.; Vazquez, E. Org. Process Res. DeV. 2001, 5, 587–592.
1182
•
Vol. 14, No. 5, 2010 / Organic Process Research & Development
10.1021/op1001287  2010 American Chemical Society
Published on Web 07/08/2010

Scheme 3. Cyano group introduction
and robust reaction. On the basis of the notion that the
intermediate cation 7 should be stabilized by the oxygen
substituents, we envisioned that 1-arylcyclohexene (9) might
be a new precursor for nitrile 8 through protonation (Scheme
1), even though the hydrocyanation of such highly substituted
olefins has been very limited.7 Additionally, we planned an
alternative approach via the oxidative cyanation of 1-arylcy-
clohexane (10), in which case cation 7 also emerges as an
intermediate.8
In the current article, we report our approaches for the
synthesis of 1 featuring the scalable construction of arylcyclo-
hexene 9, trials for the introduction of the cyano group at the
benzylic position, which resulted in the discovery of Brønsted
acid-promoted hydrocyanation,9 and the crystallization-induced
dynamic resolution of the resulted nitrile 8 diastereomers.
Results and Discussion
Our study began with the preparation of 1-arylcyclohexene
(9) by Suzuki-Miyaura cross-coupling (Scheme 2). Both
boronic acid 11 and enol triflate 12 were prepared from the
above medicinal chemistry intermediates 3 and 5, respectively.
Using triisopropyl borate as a boron source,10 the assay yield
of the boronic acid was approximately 90%, but we were unable
to isolate 11 by crystallization due to the contamination by
impurities. Therefore, we used an extracted solution of the
boronic acid for the Suzuki-Miyaura coupling. Notably, the
electron-rich boronic acid was gradually decomposed via proto-
deboronation11 in the presence of a trace of acid, and therefore
the extract was stocked after neutralization by aqueous sodium
hydroxide.
commonly used effective non-nucleophilic base for enol triflate
synthesis,12 and we first employed this method. However, in
view of its high cost and low availability in large quantities,
we substituted a simpler method using pyridine instead of
DTBMP in the toluene solvent.13 It was essential to use a slight
excess of triflic anhydride over pyridine to achieve a good
conversion, affording a 89% yield of triflate 12 after a SiO2
batch treatment. This newly established enol triflate synthesis
will be reported elsewhere.
A study of the Suzuki-Miyaura cross coupling between 11
and 12 identified the optimal reaction conditions as a 1:9 molar
ratio of palladium on carbon (Pd/C) and PPh3 in refluxing
DME-H2O with K2CO3. The optimal ratio of palladium and
phosphine minimized byproduct 13 arising from an aryl-aryl
exchange, which might be enhanced by electron-donating
substituents between the palladium center and the phosphine
ligand.14 The assay yield of 9 was 93% according to an HPLC
analysis, and as expected, 9 was easily crystallized with a yield
of 82%. This synthesis was convenient for procuring small
amounts of 9 in the early stages of our study. However, there
were still several issues in the large-scale preparation of both
11 and 12, especially the SiO2 batch treatment of triflate 12
and the low temperature below -65 °C for the lithiation of 3.
Facing the key issue of cyano group introduction, we first
examined DDQ-mediated oxidative benzylic cyanation (Scheme
3).8 The hydrogenation of 9 afforded cyclohexane 10 as a 3/1
cis/trans mixture. The mixture was subjected to DDQ and
trimethylsilyl cyanide in dichloromethane at room temperature.
Scheme 2. Suzuki-Miyaura coupling route
(10) Brown, H. C.; Cole, T. E. Organometallics 1983, 2, 1316–1319.
(11) (a) Kuivila, H. G.; Nahabedian, K. V. J. Am. Chem. Soc. 1961, 83,
2159–2163. (b) Kuivila, H. G.; Nahabedian, K. V. J. Am. Chem. Soc.
1961, 83, 2164–2166. (c) Nahabedian, K. V.; Kuivila, H. G. J. Am.
Chem. Soc. 1961, 83, 2167–2174.
The preparation of the enol triflate 12 was achieved by
applying a 2,6-di-tert-butyl-4-methylpyridine (DTBMP)-free
condition as established in our laboratories. DTBMP is the
(12) (a) Stang, P. J.; Fisk, T. E. Synthesis 1979, 438–440. (b) Stang, P. J.;
Treptow, W. Synthesis 1980, 283–284. (c) Saulnier, M. G.; Kadow,
J. F.; Tun, M. M.; Langley, D. R.; Vyas, D. M. J. Am. Chem. Soc.
1989, 111, 8320–8321.
(7) North, M. In ComprehensiVe Organic Functional Group Transforma-
tions; Katritzky, A. R., Meth-Cohn, O., Rees, C. W., Eds.; Pergamon:
Oxford, 1995; Vol. 3, p 614.
(13) Kato, S.; Chujo, I. Suzuki, K. PCT Int. Appl. WO04000795 A1, 2004.
(14) (a) Miyaura, N.; Suzuki, A. Chem. ReV. 1995, 95, 2457–2483. (b)
O’Keefe, D. F.; Dannock, M. C.; Marcuccio, S. M. Tetrahedron Lett.
1992, 33, 6679–6680. (c) Segelstein, B. E.; Bulter, T. W.; Chenard,
B. L. J. Org. Chem. 1995, 60, 12–13. (d) Kong, K.-C.; Cheng, C.-H.
J. Am. Chem. Soc. 1991, 113, 6313–6315.
(8) (a) Lemaire, M.; Doussot, J.; Guy, A. Chem. Lett. 1988, 1581–1584.
(b) Guy, A.; Doussot, J.; Guette, J.-P.; Garreau, R.; Lemaire, M. Synlett
1992, 821–822. (c) Kurti, L.; Czako, B.; Corey, E. J. Org. Lett. 2008,
10, 5247–5250.
(9) Yanagisawa, A.; Nezu, T.; Mohri, S. Org. Lett. 2009, 11, 5286–5289.
Vol. 14, No. 5, 2010 / Organic Process Research & Development
•
1183

Scheme 4. Process route via Diels-Alder and CIDR
The reaction was slow and resulted in the formation of biphenyl
14 as the main product, while the desired nitrile was the minor
product. This is most likely due to the tendency of proton-loss
from the benzylic cation 7, and subsequent oxidation afforded
the aromatized product.
Therefore, we next examined the more direct approach via
hydrocyanation of cyclohexene 9. As reported previously, we
have examined various Brønsted acids. Triflic acid was found
to be the most effective in promoting the desired hydrocyanation
under mild conditions. Trimethylsilyl cyanide was optimal as
a cyanide source, and a slight excess acid equivalent was
required for promoting this reaction.9 Successful completion of
this reaction facilitated compound 1 synthesis, avoiding issues
around alcohol 6, and shifted our focus into the development
of a more practical synthesis of 9.
trap for the completion of the one-pot reaction, propyl acetate
was found to be optimal because its boiling point (101 °C) is
close to that of both water and ethyl acrylate (100 °C). Although
15% of regioisomer 19 was obtained in the reaction, it was
effectively removed during crystallization from ethanol/water
to afford pure 9 in the 72% isolated yield. This route appears
to be superior to the previous Suzuki-Miyaura coupling route
from the viewpoint of scale-up, because there are no low-
temperature reactions or column chromatography steps.
Through a retrosynthesis analysis of 9, we were attracted
by an approach featuring the Diels-Alder reaction to construct
the cyclohexene ring. We expected that the electron-donating
nature of the benzene ring should be advantageous in terms of
the reactivity and selectivity of the Diels-Alder reaction.
The optimized process route is shown in Scheme 4. Catechol
2 was treated with 1,2-dibromoethane followed by selective
Friedel-Crafts acetylation to give the methylketone 15. The
observed regioisomer 17 was only 1%. This high selectivity
can be explained by steric repulsion between the methyl group
and acylating reagent. Because the normal addition of vinyl
Grignard into a solution of 15 produced a considerable amount
of a self-aldol adduct, the optimal method was thus determined
to be the inverse addition of 15 into a vinylmagnesium bromide
solution at 5 °C, thereby giving tertiary alcohol 16 with a 97%
assay yield. Due to difficulty in crystallization, alcohol 16 was
used without purification. Dehydration proceeded smoothly at
80 °C in the presence of a catalytic amount of pyridinium
p-toluenesulfonate. Subsequent Diels-Alder reaction required
a relatively low temperature (only 100 °C) to allow for
completion within a reasonable time period of several hours
with ethyl acrylate to give the desired adduct 9 with ∼5/1
regioselectivity. However, the diene 18 was unstable during both
evaporation and storage. Therefore, we examined a one-pot
procedure which was successfully achieved by refluxing the
crude 16 in propyl acetate in the presence of ethyl acrylate and
pyridinium p-toluenesulfonate to afford 9 in a 77% assay yield.
In order to eliminate the generating water using the Dean-Stark
The optimal conditions for the hydrocyanation were identi-
fied as using 2.1 equiv of triflic acid and 2.0 equiv of
trimethylsilyl cyanide in trifluorotoluene at -20 °C. Arylalkene
9 was successfully converted to the tertiary nitrile 8 with an
89% assay yield in an approximately 62:38 cis/trans mixture.
This poor cis/trans selectivity remained among various condi-
tions including varying temperature, solvents, and CN sources.9
Therefore, the development of a recovery method via isomer-
ization of trans-8 was strongly required. Base-induced isomer-
ization was amenable by treatment with a base such as
potassium tert-butoxide, but was stopped at a 75/25 cis/trans
ratio. Isomerization from both pure cis-8 and pure trans-8 gave
the same result, which proved that the thermodynamic equi-
librium point is unfortunately 75/25. An observation that cis-8
was less soluble than trans-8 in ethanol led us to speculate
that crystallization-induced dynamic resolution (CIDR)15 could
be applied to solve this problem. In the presence of a catalytic
amount of potassium tert-butoxide, the ethanol suspension of
crude 8 (cis/trans ) 62/38) was stirred with simultaneous
heating, and then the poor solvent hexane was added portion-
wise in order to push cis-8 to precipitate. Continuous isomer-
ization in the solution state and simultaneous preferential
crystallization of cis-8 successfully increased the ratio to 99/1
with a 90% isolated yield. The last step, the basic hydrolysis
of cis-8, proceeded smoothly at a moderate temperature without
(15) Brands, K. M. J.; Davies, A. J. Chem. ReV. 2006, 106, 2711–2733.
1184
•
Vol. 14, No. 5, 2010 / Organic Process Research & Development

any isomerization. For the final recrystallization, the use of
alcoholic solvents was eliminated because of the contamination
of a small amount of the corresponding ester. The recrystalli-
zation was performed with acetone-water to give compound
1 in a 95% yield and a 99.9% purity.
Ethyl 4-(2,3-Dihydro-8-methoxy-1,4-benzodioxin-5-yl)-3-
cyclohexenecarboxylate (9): Suzuki-Miyaura coupling. Un-
der N2, a solution of boronic acid 119 (8.39 g, 40.0 mmol, 126
mL of DME/ethyl acetate solution), triflate 12 (14.5 g, 1.2
equiv), 5% Pd/C (1.98 g, 0.01 equiv), PPh3 (943 mg, 0.09
equiv), and K2CO3 (11.0 g, 2.0 equiv) in DME/water (46 mL:
84 mL) was heated under reflux for 3 h and then filtered through
pad of Celite. The filtrate was separated, and the aqueous layer
was extracted with ethyl acetate (30 mL). The combined organic
layers were washed with water (100 mL), concentrated under
reduced pressure, and recrystallized from ethanol/water (7:3 v/v,
42 mL) to afford 9 (10.4 g, 32.7 mmol, 82%) as a white solid:
mp 88 °C; 1H NMR (CDCl3) δ 6.65 (d, J ) 8.5 Hz, 1H), 6.44
(d, J ) 8.5 Hz, 1H), 5.80-5.75 (m, 1H), 4.32-4.24 (m, 4H),
4.16 (q, J ) 7.2 Hz, 2H), 3.86 (s, 3H), 2.50-2.41 (m, 1H),
2.31-2.19 (m, 4H), 2.01-1.90 (m, 1H), 1.67-1.54 (m, 1H),
1.27 (t, J ) 7.2 Hz, 3H); 13C NMR (CDCl3) δ 176.5, 147.7,
141.2, 135.1, 132.9, 124.9, 123.7, 119.1, 103.3, 64.2, 64.0, 60.2,
56.0, 39.1, 28.2, 28.0, 25.7, 14.2; IR (KBr) 2932, 1720, 1607,
1502, 1458, 1373, 1286, 1169, 1115 cm-1; HRMS ESI(+) calcd
for C18H23O5 [M + H]+ 319.1545, found 319.1540.
Conclusions
In summary, the successful short and practical synthesis of
PDE4 inhibitor KW-4490 (1) was developed. The synthesis was
achieved in seven steps and with a 37% overall yield, of which
four steps were carbon-carbon bond-forming reactions using
simple synthons. The key to the success of this new method
was the development of a novel hydrocyanation reaction of
cyclohexene 9, which was constructed by a one-pot dehydration
Diels-Alder sequence. Subsequent successful crystallization-
induced dynamic resolution of cis-8 enhanced the practicability
of the synthesis. Multikilogram quantities of KW-4490 (1) have
now been produced in our laboratories. We believe that this
synthesis highlights the benefits of a problem-driven approach
toward the development of new synthetic methodologies.
Experimental Section
Hydrogenation and Oxidative Cyanation of 9. A solution
of cyclohexene 9 (32.0 mg, 0.10 mmol) and 20% Pd(OH)2/C
(6 mg, 10 wt %) in ethanol/ethyl acetate (5 mL:5 mL) was
stirred under H2 (1 atm) at room temperature for 2 h. The
reaction mixture was filtered and concentrated to obtain
cyclohexane 10 (3:1 diastereomer mixture: 31.8 mg, 0.10 mmol,
99%) as a colorless oil. The crude cyclohexane 10 (6.4 mg,
0.02 mmol) was dissolved in CH2Cl2 (1 mL), and LiClO4 (0.4
mg, 0.004 mmol, 0.2 equiv), Me3SiCN (0.0063 mL, 0.05 mmol,
2.5 equiv), and DDQ (6.8 mg, 0.03 mmol, 1.5 equiv) were
added. The mixture was stirred at room temperature for 4 h,
and then Me3SiCN (6.3 µL, 0.05 mmol, 2.5 equiv), LiClO4 (10
mg, 0.09 mmol, 4.7 equiv), and DDQ (5.8 mg, 0.03 mmol, 1.3
equiv) were added, and the mixture was stirred for 66 h at room
temperature. The resultant mixture was extracted with ethyl
acetate, washed with saturated aqueous NaHCO3 and saturated
General Information. All reagents and solvents are com-
mercially available (Tokyo Kasei Kogyo Co., Ltd.; Sigma-
Aldrich Co.) and used without further purification. H NMR
1
spectra were obtained on a JEOL JNM-LA300 spectrometer
(300 MHz) or Bruker AVANCE 500 spectrometer (500 MHz).
Chemical shifts (δ) are reported in ppm. Multiplicities are given
as: s (singlet), d (doublet), t (triplet), q (quartet), dd (doublet of
doublets), m (multiplet), brs (broad singlet). Proton-decoupled
13C NMR spectra were obtained on a JEOL JNM-LA300
spectrometer (75 MHz) or Bruker AVANCE 500 spectrometer
(125 MHz). IR frequencies are given in cm-1, and spectra were
obtained on a Shimadzu FT-IR8700 spectrometer. Mass spectra
and high-resolution mass spectra were obtained on a Micromass
LCT spectrometer. HPLC analyses were performed on a Hitachi
L-4000 system. GC analyses were performed on a Shimadzu
GC-14A equipped with a capillary column TC-5. Melting points
(mp) were determined on a Mettler FP61.
Ethyl 4-Trifluoromethylsulfonyloxy-3-cyclohexenecar-
boxylate (12). To a solution of pyridine (10.4 mL, 12.9 mmol,
1.10 equiv) in toluene (350 mL) was added dropwise Tf2O (21.8
mL, 13.0 mmol, 1.11 equiv) over 30 min at 15 °C under N2.
Ketone 5 (20.0 g, 11.8 mmol) was added to the mixture with
toluene (10 mL) at room temperature. The mixture was heated
at 40 °C. Tf2O was added after 10 h (0.10 mL, 0.058 mmol,
0.005 equiv) and 12 h (0.10 mL, 0.058 mmol, 0.005 equiv).
After heating at 40 °C for additional 2 h, the reaction mixture
was assayed by GC to contain 12 (91%). Water (200 mL) was
added at 15 °C to the reaction mixture. After separation, SiO2
(60 g) was added to the organic layer and then filtered and
washed with toluene (20 mL). The filtrate was evaporated to
give 12 (31.6 g, 89%); 1H NMR (CDCl3) δ 5.79-5.77 (m, 1H),
4.16 (q, J ) 7.1 Hz, 2H), 2.64-2.53 (m, 1H), 2.48-2.38 (m,
4H), 2.18-2.09 (m, 1H), 1.99-1.86 (m, 1H), 1.27 (t, J ) 7.1
Hz, 3H); 13C NMR (CDCl3) δ 174.0, 148.5, 118.6 (q, J ) 319.6
Hz), 116.9, 60.9, 37.9, 26.6, 26.1, 25.1, 14.1; IR (film) 2984,
2941, 1732, 1695, 1418, 1248, 1209, 1142, 1057, 1036, 876
cm-1; LC/MS ESI(+) m/z 303 [M + H]+.
1
aqueous NaCl, dried over MgSO4, and evaporated. H NMR
spectrum of the residue indicated that the ratio of biphenyl
compound 14, cis-8, and trans-8 was 6:1:1. An analytical
sample of 14 was obtained by SiO2 column chromatography
1
(hexane/ethyl acetate; 3:1); H NMR (CDCl3) δ 8.06 (d, J )
8.2 Hz, 2H), 7.58 (d, J ) 8.2 Hz, 2H), 6.89 (d, J ) 8.5 Hz,
1H), 6.60 (d, J ) 8.5 Hz, 1H), 4.43-4.27 (m, 4H), 4.12 (q, J
) 7.2 Hz, 2H), 3.93 (s, 3H), 1.26 (t, J ) 7.2 Hz, 3H); 13C
NMR (CDCl3) δ 166.6, 149.0, 142.4, 141.4, 133.4, 129.3, 129.2,
128.6, 122.7, 121.4, 104.0, 64.3, 64.2, 60.8, 56.1, 14.4; IR (KBr)
2985, 1794, 1604, 1366, 1270, 1017, 772 cm-1; HRMS ESI(+)
calcd for C18H19O5 [M + H]+ 315.1232, found 315.1235.
2,3-Dihydro-5-methoxy-1,4-benzodioxin. Under N2 atmo-
sphere, a solution of 3-methoxycatechol 2 (25.3 g, 180 mmol),
K2CO3 (104.5 g, 756 mmol, 4.2 equiv), and 1,2-dibromoethane
(74.4 g, 396 mmol, 2.2 equiv) in DMF (100 mL) was heated
at 60 °C for 6 h. The mixture was quenched with water and
then extracted with ethyl acetate. The organic layer was washed
with saturated aqueous NaHCO3, dried over MgSO4, and
concentrated to afford the title compound (25.4 g, 153 mmol,
85%) as an amber oil; 1H NMR (CDCl3) δ 6.77 (dd, J ) 8.4,
Vol. 14, No. 5, 2010 / Organic Process Research & Development
•
1185

8.1 Hz, 1H), 6.54 (d, J ) 8.4, 1.5 Hz, 1H), 6.49 (d, J ) 8.1,
1.5 Hz, 1H), 4.34-4.24 (m, 4H), 3.88 (s, 3H); 13C NMR
(CDCl3) δ 149.1, 144.2, 133.3, 120.2, 110.1, 104.1, 64.5, 64.2,
56.1; IR (film) 2936, 2876, 1595, 1497, 1477, 1331, 1285, 1238,
1211, 1113, 1051, 952, 885, 768, 718 cm-1; LC/MS ESI(+)
m/z 167 [M + H]+.
aqueous NaHCO3 and saturated aqueous NaCl. The organic
layer was dried over Na2SO4 and evaporated to give diene 18
as a pale-yellow oil: 1H NMR (CDCl3) δ 6.65 (d, J ) 8.4 Hz,
1H), 6.58 (dd, J ) 17.2, 6.7 Hz, 1H), 6.50 (d, J ) 8.4 Hz, 1H),
5.40-5.39 (m, 1H), 5.16-5.14 (m, 1H), 5.12 (d, J ) 6.7 Hz,
1H), 4.93 (d, J ) 17.2 Hz, 1H), 4.33-4.22 (m, 4H), 3.89 (s,
3H); 13C NMR (CDCl3) δ 148.4, 144.8, 141.4, 138.1, 132.9,
121.6, 121.3, 118.6, 116.2, 103.5, 64.3, 64.2, 56.0; IR (film)
3086, 2970, 2931, 1609, 1504, 1443, 1373, 1323, 1277, 1211,
1115, 1076, 1053, 953, 891, 799, 760 cm-1; LC/MS ESI(+)
m/z 219 [M + H]+.
5-Acetyl-2,3-dihydro-8-methoxy-1,4-benzodioxin (15). 16
Under N2 atmosphere, to a solution of AlCl3 (12.0 g, 90 mmol,
1.5 equiv) in nitromethane (200 mL) was added acetyl chloride
(5.57 mL, 78 mmol, 1.3 equiv) at 0 °C. Then a solution of
2,3-dihydro-5-methoxy-1,4-benzodioxin (10.0 g, 60 mmol) in
CH3NO2 (100 mL) was added dropwise over 40 min. After
stirring at the same temperature for 3 h, 1 mol/L HCl was added.
The organic layer was washed with saturated aqueous NaCl
and dried over MgSO4 and concentrated. The residue was
dissolved in isopropyl alcohol (25 mL) at 70 °C, cooled to room
temperature, and stirred at 0 °C for 3 h. The precipitate was
filtered to afford 15 (10.1 g, 49 mmol, 81%) as a pale-yellow
Ethyl 4-(2,3-Dihydro-8-methoxy-1,4-benzodioxin-5-yl)-3-
cyclohexenecarboxylate (9) and Isomer (19): Diels-Alder
Reaction. Under N2 atmosphere, a solution of the above crude
16 (net 22.7 g, 96 mmol), ethyl acrylate (52 mL, 480 mmol, 5
equiv), and pyridinium p-toluenesulfonate (145 mg, 0.58 mmol,
0.006 equiv) in n-PrOAc (114 mL) was heated under reflux.
H2O (∼1.7 mL) was eliminated from the refluxing reaction
mixture using Dean-Stark water separator. After 8 h, the
reaction mixture was cooled, washed with a combined solution
of saturated aqueous NaHCO3 and saturated aqueous NaCl. The
organic layer was concentrated to afford a crude mixture of
cyclohexene 9 and isomer 19 (combined HPLC assay yield
92.1%; ratio )5.3:1). The residue was dissolved in ethanol (58
mL) and H2O (10 mL) at 70 °C and cooled to 3 °C. The formed
precipitates were collected by filtration to obtain 9 (22.0 g, 69
mol, 71.8% yield) as a pale-yellow solid. Analytical sample of
isomer 19 was obtained from concentrated residue of the above
filtrate through SiO2 column chromatography (hexane/ethyl
acetate, 10:1 f 3:1); 1H NMR (CDCl3) δ 6.65 (d, J ) 8.4 Hz,
1H), 6.44 (d, J ) 8.4 Hz, 1H), 5.77 (brs, 1H), 4.36-4.26 (m,
4H), 4.15 (q, J ) 7.2 Hz, 2H) 3.86 (s, 3H), 2.72-2.53
(m, 3H), 2.31-2.22 (m, 2H), 2.10-2.00 (m, 1H), 1.81-1.67
(m, 1H), 1.26 (t, J ) 7.2 Hz, 3H); 13C NMR (CDCl3) δ 175.9,
148.0, 141.4, 134.5, 132.9, 125.8, 125.4, 119.9, 103.4, 64.2,
64.1, 60.2, 56.0, 40.0, 31.2, 25.1, 24.8, 14.2; IR (film) 2932,
1728, 1605, 1504, 1443, 1377, 1284, 1173, 1115, 1034, 953,
891 cm-1; HRMS ESI(+) calcd for C18H23O5 [M + H]+
319.1545, found 319.1557.
Ethyl 4-Cyano-4-(2,3-dihydro-8-methoxy-1,4-benzodioxin-
5-yl)cyclohexanecarboxylate (8). Caution! When CF3SO3H
and Me3SiCN are mixed, in situ HCN formation occurs. Be
sure to seal the reaction flask until basic quench, and keep
handling under a well-ventilated fume hood. Prepare the CN
testing tube for leak check. To a solution of CF3SO3H (126.3
g, 841 mmol, 2.1 equiv) and Me3SiCN (79.5 g, 801 mmol, 2.0
equiv) in PhCF3 (383 mL) was added dropwise a solution of
cyclohexene 9 (127.5 g, 400 mmol) in PhCF3 (892 mL) over
60 min at -20 °C under N2 atmosphere. The reaction mixture
was allowed to warm to -2 °C over 4 h and quenched with
1.5 mol/L aqueous NaOH (1250 mL). After stirring for 30 min,
the organic layer was separated, washed with 1.5 mol/L aqueous
NaOH (1250 mL) and 1.0 mol/L aqueous NaCl (1000 mL ×
2), and dried over MgSO4(50 g), then concentrated to give a
cis/trans mixture of nitrile 8 (HPLC assay; cis/trans ) 62/38,
355 mmol, net 122.6 g, 88.7% yield). The analytical sample of
trans-8 was obtained by SiO2 column chromatography (hexane/
ethyl acetate, 3:1): mp 104 °C; 1H NMR (CDCl3) δ 6.82 (d, J
1
solid: mp 81 °C; H NMR (CDCl3) δ 7.42 (d, J ) 8.9 Hz,
1H), 6.54 (d, J ) 8.9 Hz, 1H), 4.39-4.32 (m, 4H), 3.91 (s,
3H), 2.56 (s, 3H); 13C NMR (CDCl3) δ 197.1, 152.4, 144.6,
132.9, 122.7, 121.3, 103.5, 64.2, 63.9, 56.2, 31.5; IR (KBr)
2982, 1676, 1597, 1501, 1468, 1441, 1373, 1356, 1281, 1207,
1178, 1136, 1113, 1069, 1042, 976, 935, 800, 746, 677, 633,
594, 546 cm-1; LC/MS ESI(+) m/z 209 [M + H]+. The
analytical sample of 17 was obtained by SiO2 column chro-
1
matography (hexane/ethyl acetate; 5/2): H NMR (CDCl3) δ
7.29 (d, J ) 8.8 Hz, 1H), 6.67 (d, J ) 8.8 Hz, 1H), 4.33-4.26
(m, 4H), 3.94 (s, 3H), 2.60 (s, 3H); LC/MS ESI(+) m/z 209
[M + H]+.
3-(2,3-Dihydro-8-methoxy-1,4-benzodioxin-5-yl)-1-buten-
3-ol (16). To a stirred solution of vinylmagnesium bromide (288
mL, 0.80 mol/L solution in THF, 231 mmol, 1.6 equiv) in THF
(240 mL), was added dropwise a solution of ketone 15 (30.0
g, 144 mmol) in THF (120 mL) over 30 min at 5 °C under N2
atmosphere. The reaction mixture was stirred at the same
temperature for 45 min and then quenched with a saturated
aqueous NH4Cl (300 mL), diluted with saturated aqueous NaCl
(100 mL), and extracted with ethyl acetate (300 mL). The
extract was washed with saturated aqueous NaCl (200 mL),
dried over Na2SO4 (50 g), and concentrated to afford crude 16
(HPLC assay 139 mmol; net 32.9 g, 97% yield) as a yellow
oil: 1H NMR (CDCl3) δ 6.83 (d, J ) 8.7 Hz, 1H), 6.47 (d, J )
8.7 Hz, 1H), 6.18 (dd, J ) 17.2, 10.6 Hz, 1H), 5.11 (dd, J )
17.2, 1.1 Hz, 1H), 5.04 (dd, J ) 10.6, 1.1 Hz, 1H), 4.36-4.24
(m, 4H), 3.87 (s, 3H), 1.64 (s, 3H); 13C NMR (CDCl3) δ 148.4,
144.8, 141.7, 133.4, 126.9, 117.2, 111.6, 103.3, 74.6, 64.1, 64.0,
56.0, 27.4; IR (film) 3541, 3086, 2978, 2978, 1609, 1504, 1443,
1373, 1281, 1115, 999, 953, 891, 829, 733 cm-1; HRMS
ESI(-) calcd for C13H15O4 [M - H]- 235.0970, found
235.0966.
2-(2,3-Dihydro-8-methoxy-1,4-benzodioxin-5-yl)-1,3-buta-
diene (18). Under N2 atmosphere, a solution of alcohol 16 (0.47
g, 2.0 mmol) and pyridinium p-toluenesulfonate (2.5 mg, 0.01
mmol, 0.005 equiv) in toluene (100 mL) was heated at 80 °C
for 2 h. The reaction mixture was washed with saturated
(16) (a) Baker, W.; Jukes, E. H. T.; Subrahmanyam, C. A. J. Chem. Soc.
1934, 1681–1684. (b) Dallacker, F.; Van Wersh, J. Chem. Ber. 1972,
105, 3301–3305.
1186
•
Vol. 14, No. 5, 2010 / Organic Process Research & Development

) 8.8 Hz, 1H), 6.48 (d, J ) 8.8 Hz, 1H), 4.36-4.33 (m, 4H),
4.15 (q, J ) 7.2 Hz, 2H), 3.87 (s, 3H), 2.75 (brs, 1H), 2.42-1.99
(m, 8H), 1.26 (t, J ) 7.2 Hz, 3H); 13C NMR (CDCl3) δ 174.3,
148.8, 142.2, 133.5, 122.1, 121.0, 116.9, 103.2, 64.2, 63.6, 60.3,
55.9, 39.7, 38.0, 31.0, 24.2, 14.2; IR (KBr) 2970, 2941, 2228,
1717, 1609, 1506, 1464, 1377, 1329, 1286, 1123, 1040, 953,
891, 793 cm-1; HRMS ESI(+) calcd for C19H24NO5 [M + H]+
346.1654, found 346.1631.
mL) was stirred at room temperature for 4 h. The resultant
mixture was diluted with water (102 mL), cooled to 5 °C, and
neutralized with 6 mol/L HCl. The precipitate was collected
by filtration and dried to give crude carboxylic acid 1 (18.2 g,
57.4 mmol). The crude 1 (18.0 g, 56.7 mmol) was dissolved in
acetone (170 mL) and water (30 mL) under reflux. The solution
was filtered hot and maintaining 55 °C during addition of water
(180 mL) to form precipitation. The suspension was cooled to
5 °C, stirred 3 h, and filtered to obtain 1 (17.2 g, cis/trans )
>99.99/<0.01, 54.2 mmol, 95% yield) as a white solid: mp 245
°C; 1H NMR (DMSO-d6) δ 12.24 (s, 1H), 6.79 (d, J ) 8.8 Hz,
1H), 6.60 (d, J ) 8.8 Hz, 1H), 4.33-4.23 (m, 4H), 3.75 (s,
3H), 2.36-2.25 (m, 3H), 2.06-1.98 (m, 2H), 1.86-1.64 (m,
4H); 13C NMR (DMSO-d6) δ 175.9, 148.6, 141.9, 133.5, 121.6,
120.5, 116.3, 103.9, 63.7, 63.4, 55.4, 41.0, 38.9, 32.9, 25.6; IR
(KBr) 3288, 2930, 2232, 1730, 1508, 1456, 804 cm-1; HRMS
ESI(-) calcd for C17H18NO5 [M - H]- 316.1185, found
316.1195.
Crystallization-Induced Dynamic Resolution of cis-8. To
a suspension of the cis/trans mixture 8 (net 28.5 g, 82.5 mmol,
cis/trans ) 61/39) in anhydrous ethanol (15 mL) and hexane
(60 mL), was added t-BuOK (0.93 g, 8.26 mmol, 0.1 equiv)
and heated under reflux. After 3 h, hexane (75 mL) was added
portionwise and the refluxing was continued for 1 h. The
resulting suspension was cooled to 20 °C over 1 h, stirred for
1 h at the same temperature, and filtered to afford crude cis-8
(27.5 g, cis/trans ) 99.6/0.4) as pale-yellow crystals. The crude
cis-8 was dissolved in ethanol (82.5 mL) at 65 °C. The solution
was cooled to 5 °C over 1 h and stirred for 1 h at the same
temperature. The formed precipitates were collected by filtration
to provide cis-8 (25.8 g, cis/trans ) 99.97/0.03, 74.7 mmol,
90.4% yield) as a white solid: mp 131 °C; 1H NMR (CDCl3)
δ 6.84 (d, J ) 8.9 Hz, 1H), 6.49 (d, J ) 8.9 Hz, 1H), 4.39-4.33
(m, 4H), 4.17 (q, J ) 7.1 Hz, 2H), 3.88 (s, 3H), 2.44 (brd, J )
12.6 Hz, 2H), 2.32 (tt, J ) 11.8, 3.8 Hz, 1H), 2.18-1.95 (m,
4H), 1.86 (dt, J ) 3.6, 12.6 Hz, 2H), 1.28 (t, J ) 7.1 Hz, 3H);
13C NMR (CDCl3) δ 174.5, 148.9, 142.2, 133.7, 121.7, 121.0,
116.8, 103.3, 64.2, 63.7, 60.5, 56.0, 42.3, 40.0, 33.6, 25.8, 14.2;
IR (KBr) 2953, 2228, 1722, 1607, 1504, 1460, 1381, 1325,
1281, 1117, 1043, 953, 787 cm-1; HRMS ESI(+) calcd for
C19H24NO5 [M + H]+ 346.1654, found 346.1675.
Acknowledgment
We thank Drs. Masaji Kasai, Masashi Yanase, and Yoshiaki
Masuda for helpful discussions, and Hikaru Nishimura for
HRMS analysis at former Sakai Research Laboratories of
Kyowa Hakko Kogyo.
Supporting Information Available
Copies of 1H and 13C NMR spectra for compounds 14, 16,
18, and 19. HPLC and GC methods. This material is available
free of charge via the Internet at http://pubs.acs.org/.
cis-4-Cyano-4-(2,3-dihydro-8-methoxy-1,4-benzodioxin-
5-yl)cyclohexanecarboxylic Acid (1). A mixture of ester cis-8
(20.0 g, 57.9 mmol), ethanol (100 mL), and 6 mol/L KOH (19
Received for review May 10, 2010.
OP1001287
Vol. 14, No. 5, 2010 / Organic Process Research & Development
•
1187