A practical, kilogram-scale implementation of the Wolff-Kishner reduction

Article abstract of DOI:10.1021/op9000274

A safe and practical strategy has been developed for the largescale preparation of imidazole 7. The key transformation involved the Wolff-Kishner reduction of the sterically demanding neopentyl- trifluoromethylcyclopropyl imidazole ketone 8. The described

Full text of DOI:10.1021/op9000274

Organic Process Research & Development 2009, 13, 576–580
A Practical, Kilogram-Scale Implementation of the Wolff-Kishner Reduction
Jeffrey T. Kuethe,* Karla G. Childers, Zhihui Peng, Michel Journet, and Guy R. Humphrey
Department of Process Research, Merck & Co., Inc., Rahway, New Jersey 07065, U.S.A.
Thomas Vickery, Donald Bachert, and Thientu T. Lam
Chemical and Engineering Research and DeVelopment, Merck & Co., Inc., Rahway, New Jersey 07065, U.S.A.
Scheme 1
Abstract:
A safe and practical strategy has been developed for the large-
scale preparation of imidazole 7. The key transformation involved
the Wolff-Kishner reduction of the sterically demanding neo-
pentyl-trifluoromethylcyclopropyl imidazole ketone 8. The de-
scribed process provided the desired product in 74% overall yield
without recourse to chromatographic purification. Safety consid-
erations which allowed for the reaction to be conducted safely on
kilogram scale are discussed.
Introduction
The deoxygenation of aldehydes and ketones to methyl or
methylene derivatives employing hydrazine/KOH first reported
by Kishner in 19111 and soon thereafter by Wolff in 19122 (The
Wolff-Kishner reduction) remains one of the most proven,
convenient, and synthetically useful processes available for this
important type of transformation.3 The use of higher boiling
solvents such as ethylene glycol, diethylene glycol (DEG), and
triethylene glycol (TEG) in order to obtain the higher reaction
temperatures necessary for successful conversion of carbonyl
carbons to methylene derivatives was first introduced by
Whitmore4 and further studied by Soffer.5 The serendipitously
discovered Huang-Minlon modification,6 which utilizes an
excess of safer and less expensive hydrazine hydrate in the
presence of alkali, followed by removal of water by distillation
subsequent to hydrazone formation, is now the most commonly
employed protocol. The generally accepted mechanism involves
reversible formation of hydrazone intermediate 2 followed by
base-induced proton abstraction to give 4. Alternatively, hy-
drazone 2 can react with a second mole of 1 to give azine 3;
however, this process is reversible in the presence of hydra-
Scheme 2
zine.3b,c The rate-determining step is transformation of 4 to 5
followed by rapid expulsion of nitrogen and protonation of the
resulting carbanion to give the reduced product 6 (see Scheme
1).3a It has been reported through mechanistic studies that a
minimum temperature of ∼190 °C was necessary to achieve
reasonable rates of reduction.7 While the Cram8 and Henbest9
modifications allow for the Wolff-Kishner reduction to be
carried out at lower temperatures, these methods require the
formation and isolation of sensitive hydrazone intermediates
of type 2 prior to their reduction. Often irreversible and
competitive formation of the corresponding azine 3 has limited
the scope and application of these methodologies.
Recently we required kilogram quantities of substituted
imidazole 7 in order to support preclinical development of a
number of active substrates (Scheme 2). The initial preparation
of 7 involved the microwave-mediated Wolff-Kishner reduc-
tion of ketone 8 in a sealed vessel at 180 °C and provided 7 in
* Author for correspondence. E-mail: Jeffrey_Kuethe@merck.com.
(1) Kishner, N. Zh. Russ. Fiz.-Khim. OVa., Chast. Khim. 1911, 43, 582.
(2) Wolff, L. Justus Liebigs Ann. Chem. 1912, 394, 86.
(3) For reviews, see: (a) Hutchins, R. O.; Hutchins, M. K. In Compre-
hensiVe Organic Synthesis; Trost, B. M. Ed.; Pergamon Press: Oxford,
U.K., 1991; Vol. 8, pp 327-362. (b) Reusch, W. In Reduction;
Augustine, R. L., Ed.; Dekker: New York, 1968; pp 171-211. (c)
Huang-Minlon Sci. Sin. 196110, 711. (d) Buu-Ho¨ı, N. P.; Hoa´n, N.;
Xuong, N. D. Recl. TraV. Chim. Pays-Bas 1952, 71, 285. (e) Todd,
D. Org. React. 1948, 4, 378–422.
(4) Herr, C. H.; Whitmore, F. C.; Schiessler, R. W. J. Am. Chem. Soc.
1945, 67, 2061.
(7) (a) Szmant, H. H.; Harnsberger, H. F.; Butler, T. J.; Barie, W. P. J. Am.
Chem. Soc. 1952, 74, 2724. (b) Szmant, H. H.; Harmuth, D. M. J. Am.
Chem. Soc. 1964, 86, 2909.
(5) (a) Soffer, M. D.; Soffer, M. B.; Sherk, K. W. J. Am. Chem. Soc.
1945, 67, 1435. (b) Sherk, K. W.; Augur, M. V.; Soffer, M. D. J. Am.
Chem. Soc. 1945, 67, 2239.
(8) Cram, D. J.; Sahyun, M. R. V.; Knox, G. R. J. Am. Chem. Soc. 1962,
88, 4034.
(6) (a) Huang-Minlon J. Am. Chem. Soc. 1946 68, 2487. (b) Huang-Minlon
J. Am. Chem. Soc. 1949, 71, 3301.
(9) Grundon, M. F.; Henbest, H. B.; Scott, M. D. J. Chem. Soc. 1963,
1855.
576
•
Vol. 13, No. 3, 2009 / Organic Process Research & Development
10.1021/op9000274 CCC: $40.75  2009 American Chemical Society
Published on Web 04/06/2009

Scheme 3
Scheme 4
60% yield after chromatography. This procedure presented
formidable challenges for scale-up and raised serious concerns
regarding safety. Consequently, an alternative protocol for the
large-scale synthesis of 7 was required. Key to the successful
preparation of 7 would involve conducting the key reduction
at lower temperatures, avoiding the use of a sealed tube due to
off-gassing, eliminating the use of a microwave, and having a
thorough understanding of the reaction in order to safely run it
on scale. There have been few reports describing large scale
Wolff-Kishner reductions and the issues describing the safe
execution of such reactions.10,11 In this paper we document a
practical implementation of a modified Wolff-Kishner reduc-
tion of 8 for the kilogram preparation of 7 and outline the key
safety concerns that arose from these studies.
Wolff-Kishner reduction, the direct deoxygenation of ketones
to methylene derivatives has been achieved by such methods
as Clemmensen reduction,14 LiAlH4-AlCl3,15 NaBH4-AlCl3,16
NaBH4-TFA,17 borane-BF3,18 phosphorus-HI,19 Et3SiH-BF3
or -TFA,20 Et3SiH-SnCl2,21 diphenylsilane,22 triphenylsilane,23
and catalytic hydrogenation.24 Due to the potentially reactive
trifluoromethylcyclopropyl moiety present in 8, reductions
which proceed through radical-based mechanisms such as
Clemmensen and silylhydride-mediated protocols were not
considered. Attempted reduction of 8 in the presence of Lewis
acids such as AlCl3 and BF3 ·OEt2 in the presence of a number
of reducing agents (LiAlH4, NaBH4, LiBH4, NaCNBH3) only
afforded alcohol 12 as the single identifiable product (Scheme
4). No detectable amounts of 7 were observed by either HPLC
or NMR of the crude reaction mixtures of these reactions. We
speculate that the hindered neopentyl alcohol 12 precluded
further reduction to 7 under these reaction conditions. Attempted
hydrogenation of 8 was also investigated; however, these
reactions were either plagued by over-reduction of the cyclo-
propyl group, gave exclusively alcohol 12, or resulted in no
reaction.
Results and Discussion
Our first goal was the development of an efficient synthesis
of ketone 8. Reaction of commercially available 1-trifluorom-
ethylcyclopropane-1-carboxylic acid 9 with 1.4 equiv of 1,1′-
carbonyldiimidazole (CDI) in CH2Cl2 was followed by the direct
addition of NEt3 and N,O-dimethylhydroxylamine hydrochloride
and furnished Weinreb amide 10 in 95% HPLC assay yield
(Scheme 3). Amide 10 was not isolated, and was used crude in
the next step without further purification following an aqueous
workup. Treatment of trityl-protected 4-iodoimidazole 1112 first
with 1.1 equiv of EtMgBr in CH2Cl2 followed by the addition
of 10 afforded the desired ketone 8 in 92% isolated yield as a
highly crystalline solid.
With the desired ketone 8 in hand our attention turned to
the preparation of 7. The ketone moiety of 8 is flanked by a
potentially reactive, sterically demanding neopentyl-trifluorom-
ethylcyclopropyl center and the electron-rich, trityl-protected
imidazole core. Reductions which would preserve these sensitive
functionalities were particularly attractive. Initial efforts were
focusedonalternativestothemicrowave-mediatedWolff-Kishner
reduction employed in the original route. In addition to the
(14) For reviews, see: (a) Yamamura, S.; Nishiyama, S. In ComprehensiVe
Organic Synthesis; Trost, B. M., Ed.; Pergamon Press: Oxford, U.K.,
1991; Vol. 8, pp 307-325. (b) Vedejs, E. Org. React. 1975, 22, 401.
(c) Martin, E. L. Org. React. 1942, 1, 155.
(15) (a) Blackwell, J.; Hickinbottom, W. J. J. Chem. Soc. 1961, 1405. (b)
Nystrom, R. F.; Berger, C. R. A. J. Am. Chem. Soc. 1958, 80, 2896.
(c) Brown, B. R.; White, A. M. S. J. Chem. Soc. 1957, 3755.
(16) Ono, A.; Suzuki, N.; Kamimura, J. Synthesis 1987, 736.
(17) (a) Gribble, G. R.; Lese, R. M.; Evans, B. E. Synthesis 1977, 172. (b)
Gribble, G. R.; Kelly, W. J.; Emery, S. E. Synthesis 1978, 763.
(18) Breuer, E. Tetrahedron Lett. 1967, 8, 1849.
(19) (a) Bradsher, C.; Vingiello, F. J. Org. Chem. 1948, 13, 786. (b)
Morrison, D. C. J. Org. Chem. 1958, 23, 1772. (c) Reimschneider,
R.; Kassahn, H. Chem. Ber. 1959, 92, 1705.
(10) For a previously reported large-scale Wolff-Kishner reduction, see:
Eisenbraun, E. J.; Payne, K. W.; Bymaster, J. S. Ind. Eng. Chem. Res.
2000, 39, 1119.
(20) (a) Fry, J. L.; Orfanopoulos, M.; Adlington, M. G.; Dittman, W. P.;
Silverman, S. B. J. Org. Chem. 1978, 43, 374. (b) West, C. T.;
Donnelly, S. J.; Kooistra, D. A.; Doyle, M. P. J. Org. Chem. 1973,
38, 2675. (c) Kursanov, D. N.; Parnes, Z. N.; Loim, N. M. Synthesis
1974, 633.
(11) Kampmann, D.; Stuhlmu¨eller, G.; Simon, R.; Cottet, F.; Leroux, F.;
Schlosser, M. Synthesis 2005, 1028.
(12) Cliff, M. D.; Pyne, S. G. Synthesis 1994, 681.
(13) The use of CH2Cl2 has been reported to be crucial for successful
transmetalation due to the greater covalent character of the organo-
magnesium intermediate, see: Turner, R. M.; Lindell, S. D. J. Org.
Chem. 1991, 56, 5739.
(21) Frainnet, E.; Calas, R.; Berthault, A. C. R. Acad. Sci. 1964, 258, 613.
(22) Diehl, J. M.; Gilman, H. Chem. Ind. (London) 1959, 1095.
(23) Patin, H.; Roullier, L.; Dolard, R. C. R. Acad. Sci., Ser. C. 1971, 272,
675.
(24) Ram, S.; Spicer, L. D. Tetrahedron Lett. 1988, 29, 3741.
Vol. 13, No. 3, 2009 / Organic Process Research & Development
•
577

Scheme 5
3-3.5 h. Analysis of the crude reaction mixture by HPLC
indicated complete consumption of both 14 and 15. Caution:
hydrazine is likely to be present and particular care should
be taken to avoid contact with any vapors or liquids.
Approximately one-half of the theoretical amount of water in
the reaction mixture based on the amount in hydrazine hydrate,
KOH, and water added to the reaction mixture was collected
in the Dean-Stark trap during the course of the reaction. The
majority (∼80%) of the water was removed during the initial
stages of the distillation between 143 and 155 °C, and distillation
completely ceased after 1 h at 155 °C. The safety aspects as
well as waste disposal are noted below. The head space
temperature above the reaction mixture did not exceed 127 °C
at any point during both the heatup and age at 155 °C. The
boiling point of hydrazine hydrate is 120 °C, and its flash point
is 75 °C. The flammability of aqueous hydrazine decreases, and
the flash point increases with increasing dilution with a 40%
solution being just ignitable. Due to these considerations, serious
concerns regarding the flammability of any vapors in the head
space were raised. Therefore, the concentration of hydrazine
in the distillate was monitored.30 The initial distillate collected
in the heat up from 143 to 155 °C was found to contain ∼1
mg/mL hydrazine. The distillate collected during the 3-3.5 h
age was found to contain ∼3.9 mg/mL hydrazine, where the
combined total distillate was found to be ∼2.1 mg/mL hydra-
zine. These numbers indicated that the majority of hydrazine
added to the initial reaction mixture remained in solution with
low concentrations in the head space.
Once the reaction was complete, it was cooled to 70-75
°C and the resulting slurry of 7 diluted with MeCN (3 L/kg
based on starting material) which helped solubilize unidentifi-
able polymeric impurities. The reaction mixture was further
cooled to 50-55 °C, and water (7 L/kg based on starting
material) was added. The reaction mixture was then cooled to
rt and filtered. The mother liquors were observed to have a
hydrazine concentration of ∼16 mg/mL and should be disposed
accordingly, see Waste Disposal section below. The wet cake
was washed thoroughly with MeCN/water and then with water
and dried under vacuum to give 7 in analytically pure form
and in 85% isolated yield.
Reduction of tosylhydrazones with borohydride reagents
(typically sodium borohydride,25 sodium cyanoborohydride,26
catecholborane,27 or bis(benzoyloxy)borane28) or the recently
reported reduction of N-tert-butyldimethylsilylhydrazones29 is
often employed as an attractive low-temperature alternative for
the deoxygenation of ketones. Unfortunately, all attempts to
convert 8 into tosylhydrazone 13 were unsuccessful at room
temperature and gave significant amounts of decomposition
when performed at higher temperatures, and this approach was
abandoned.
Having established that alternative methods for the conver-
sion of 8 to 7 were unsuccessful, our attention focused on scale-
up of the Wolff-Kishner reduction of 8 (Scheme 5). Prelimi-
nary experiments were conducted at 180 °C in diethylene glycol
in the presence of 5 equiv of powdered KOH and a large excess
of hydrazine hydrate (17 equiv) and furnished 7 in 70% isolated
yield after addition of water, extraction of the product, and
chromatography of the residue on silica gel. The reaction was
then optimized in terms of solvent, temperature, hydrazine
hydrate, and KOH charge. After extensive experimentation, it
was discovered that the optimal conditions for the conversion
of 8 to 7 involved slowly heating a slurry of 8 in diethylene
glycol (10 L/kg) in the presence of 8 equiv of hydrazine hydrate
(degree of hydration ∼1.5, 55% hydrazine), 4 equiv of 85%
powdered KOH, and water (∼0.23 g/g of hydrazine) to 143
°C (reflux) over 2 h. The addition of water to the reaction
mixture minimized the concentration of hydrazine in the head
space of the reaction (Vida infra) and had no detrimental effect
on the reaction. The internal reaction temperature was main-
tained at 143-145 °C for 30 min at which point the HPLC
analysis indicated complete consumption of the starting material
and formation of a mixture of hydrazone 14, azine 15, and
product 7. In order to drive the reaction to completion, the
refluxing reaction mixture was diverted to a Dean-Stark
apparatus, and the internal temperature was slowly increased
to 155 °C over 1 h and was maintained at this temperature for
Safety Considerations. Hydrazine is extremely toxic and
caution should be taken to avoid prolonged exposure. Hydrazine
is rapidly absorbed through the skin and other routes of exposure
and appropriate gloves31 and face shield should be utilized
during manipulation. Aqueous hydrazine containing >40 mol%
hydrazine is flammable and can support combustion even in
the absence of oxygen. While the ignition temperature of
hydrazine vapors is 270 °C in the presence of air, a nitrogen
atmosphere should be maintained during the heatup, age, and
cool-down of the reaction. As long as a ratio of 1.86 mol of
water to 1 mol of hydrazine is maintained in the liquid reaction
(30) Hydrazine concentrations were monitored by a combination of methods
found in Jogota, N. K.; Chetram, A. J.; Nair, J. B. J. Pharm. Biomed.
Anal. 1998, 16, 1083 and Larson, S.L.; Strong, A. B. Ion Chroma-
tography with Electrochemical Detection for Hydrazine Quantitation
in EnViornmental Samples Technical Report IRRP-96-3, March 1996.
See Experimental Section for exact method employed.
(31) Safeskin Purple Nitrile Xtra Exam Gloves, available from Fisher
Scientific and impervious to hydrazine, were employed in all
operations.
(25) (a) Caglioti, L. Tetrahedron 1966, 22, 487. (b) Hutchins, R. O.; Natale,
N. R. J. Org. Chem. 1978, 43, 2299.
(26) (a) Hutchins, R. O.; Maryanoff, B. E.; Milewski, C. A. J. Am. Chem.
Soc. 1971, 93, 1793. (b) Hutchins, R. O.; Milewski, C. A.; Maryanoff,
B. E. J. Am. Chem. Soc. 1973, 95, 3662.
(27) Kabalka, G. W.; Baker, J. D. J. Org. Chem. 1975, 40, 1834.
(28) Kabalka, G. W.; Summers, S. T. J. Org. Chem. 1981, 46, 1217.
(29) Furrow, M. E.; Myers, A. G. J. Am. Chem. Soc. 2004, 126, 5436.
578
•
Vol. 13, No. 3, 2009 / Organic Process Research & Development

mixture, the vapors in the head space of the reaction cannot
become flammable due to the fact that water is more volatile
and will be enriched in the vapor phase. Based on the described
process, if no more than 325 g of water per kilogram of 7/8
are removed during the distillation of water (raising the
temperature to 155 °C), a flammable atmosphere cannot be
generated.32 This was based on the presence of 398 g water/kg
8 from the hydrazine charge, 73 g water/kg of 8 from the KOH
water content (∼15%), and the addition of 290 g of water/liter
hydrazine hydrate charged. In addition, the reaction generates
an additional 39 g water/kg of 7 during the course of the
reaction. Because this reaction was operating above the flash
point (123 °C) of the solvent (diethylene glycol), all ignition
sources must be avoided and inertion maintained.
In order to fully evaluate the safety of the Wolff-Kishner
reduction of 8, reaction samples were evaluated by Differential
Scanning Calorimetry (DSC)33,34 and Accelerating Rate Calo-
rimetry (ARC).35 The starting ketone 8 was found to be
thermally unstable (DSC), decomposing with an energy release
of 172 Kcal/kg, with a detected onset temperature of 185 °C.
The isolated solids and reaction liquors from the end of reaction
slurry by DSC were found to be thermally stable below 200
°C. In addition, the isolated product 7 was also found to be
thermally stable below 200 °C.
The ARC study was carried out to check for potential
exothermic activity in the reaction mixture. The reagents were
added to a titanium cell with a stir bar and taken through a
series of temperatures to look for heat generation. There was
no evidence of hazardous exothermic activity at or below the
operating region of the reaction (<160 °C) in the ARC run of
the reaction mixture. One mole of gas per mole of substrate
(8) was observed to form between 80-160 °C. The ARC test
did identify a secondary decomposition, generating at least 7
moles of gas per mole of 8. The onset of this decomposition
exotherm was measured at 235 °C,36 although gas generation
was seen earlier, at 180 °C. A more detailed analysis of the
decomposition data showed no evidence of autocatalysis. In
addition, the temperature at which the adiabatic time to
maximum rate was 2 weeks at 183 °C, and the temperature at
which the adiabatic time to maximum rate was 24 h at 213 °C.
Based on these three facts, it was recommended to control the
reaction temperature below 160 °C, with a high temperature
alarm to warn of an unexpected temperature increase. To further
increase safety, the use of indirect heating, such as Syltherm
jackets, is preferred. This both enables better avoidance of
overshoot during the heat-up of the reaction mixture and allows
the jacket to provide a means of cooling if the temperature rises
unexpectedly. The use of this type of reaction vessel, while
preferred, is not absolutely required as long as a slow, controlled
heatup from 145 to 155 °C is maintained.
Due to the potential interaction of hydrazine with metal and
metal oxides, only glass and/or Teflon surfaces should be in
contact with the hot reaction mixture. If a stainless steel vessel
is to be employed, it should be carefully checked for rust or
corrosion prior to use. The high reaction temperatures in
conjunction with KOH also results in slight etching of glass
vessels.37
Waste Disposal. Hydrazine is rapidly destroyed by com-
mercial bleach (5%), and aqueous waste streams may be
processed with alkali and bleach prior to disposal.38,39 Alter-
natively, hydrazine is neutralized with hydrogen peroxide.39
According to the International Register of Potentially Toxic
Chemicals (IRPTC), hydrazine can be diluted with water and
neutralized with dilute sulfuric acid prior to disposal. Hydrazine
waste can be diluted with alcohol or another hydrocarbon and
burnt in a chemical incinerator equipped with an after-burner
and effectively scrubbed. It is advised that adequate commercial
bleach be on hand to rapidly destroy any spilled material.40
Conclusion
In conclusion, we have demonstrated a safe and practical
method for the kilogram preparation of the highly functionalized
imidazole 7 which involved the Wolff-Kishner reduction of
ketone 8. The three step sequence provided 7 in 74% overall
from acid 9 and required no chromatographic purifications. Key
to the successful implementation of this reaction was a thorough
understanding of the safety margins within which the reaction
could be conducted. This included temperature, amount of water
removed from the reaction, and amount of water remaining in
the reaction mixture which ensured no flammable atmosphere
would be encountered. While this reaction was safely and
efficiently conducted on kilogram scale, further scale-up of this
reaction, or any Wolff-Kishner reduction, would require further
testing of all the reaction parameters prior to execution.
Experimental Section
Melting points are uncorrected. All solvents and reagents
were used as received from commercial sources. Analytical
samples were obtained by chromatography on silica gel using
an ethyl acetate-hexane mixture as the eluent. The water
content (KF) was determined by Karl Fisher titration. Hydrazine
concentrations were monitored as follows: hydrazine was eluted
at room temperature from an Ion Pac CS 17 (4 × 250 mm)
with 6 mM methane sulfonic acid using an electochemical
detector, injection volume of 100 µL, flow rate 1.0 mL/min.30
Post column, 0.1 N NaOH was mixed with the stream for
electrochemical detection. Standards of hyrazine and unknowns
were freshly prepared for quantitation (9:1 water/MeOH).
Retention time of hydrazine was 3.1 min.
Preparation of 1-Trifluoromethyl-cyclopropanecarboxy-
lic acid methoxy-methyl amide (10). In a 75 L round-bottom
flask37 charged with 40.0 L of CH2Cl2 was added 4.00 kg (26.0
(32) A flammable atmosphere would be one containing >38 mol %
hydrazine.
(33) The DSC runs were conducted on TA Instruments 2920 DSC cells
operating under TA Instruments’ Thermal Advantage software, ver.
1.1 A. A reusable pressure-tight cell made of either Hastelloy B-2 or
tantalum-lined Hastelloy B-2 was used.
(37) A glass vessel was employed. Etching was significant after three
consecutive runs in the same flask.
(38) Armour, M. A. Hazardous Laboratory Chemicals Disposal Guide;
CRC Press: Boca Raton, FL, 1991; pp 170-171.
(39) Speer, S. E.; Pasricha, A.; Quinn, R. F. Ind. Water Treatment 1995,
48.
(34) Tuma, L. Thermochim. Acta 1992, 179.
(35) Townsend, D. I.; Tou, J. C. Thermochim. Acta 1980, 37, 1.
(36) An onset threshold of 0.025 K/min was used, and the adiabatic
correction (φ) factor was 1.27.
(40) For an alternative waste disposal procedure, see ref 10.
Vol. 13, No. 3, 2009 / Organic Process Research & Development
•
579

mol) of 1-(trifluoromethyl)cyclopropane-1-carboxylic acid 9 at
rt. The solution was cooled to 10 °C, and 5.89 kg (34.3 mol)
of 1,1′carbonyldiimidazole (CDI) was added in portions over
1 h, maintaining the reaction temperature below 20 °C. The
addition of CDI was slightly exothermic, and CO2 gas was
evolved. The reaction was stirred at rt for 30 min, and 5.07 L
(36.3 mol) of NEt3 was added in one portion followed by 3.54
kg (36.3 mol) of N,O-dimethylhydroxylamine hydrochloride.
The resulting mixture was stirred at rt for 15 h and then reverse
quenched into a cooled (5 °C) 3 N HCl solution (25.0 L). The
layers were separated, and the organic layer was washed
sequentially with 25 L of saturated aqueous NaHCO3 and 25
L of water. The organic layer was passed through a plug of
anhydrous Na2SO4 (KF < 1000 ppm) and then concentrated
under vacuum to give 4.86 kg (95%) of 10 which was
sufficiently pure for use without further purification. An
analytical sample was prepared by passing over a plug of silica
gel to give 10 as a colorless liquid: 1H NMR (CDCl3, 400 MHz)
reflux condensor was charged 2.49 kg (5.28 mol, 97.7 wt %)
of 8 as a solid and 1.18 kg (21.1 mol) of powdered KOH
(∼85%, charge not based on wt % of KOH). To the flask was
then charged sequentially 25 L of diethylene glycol, 2.10 L of
hydrazine hydrate, and 500 mL of water. The stirring was
started, and the reaction mixture was heated to an internal
temperature of 143 °C over a 2-h period at which point the
reaction mixture was refluxing. The reaction was maintained
at this temp for ∼30 min, and the refluxing mixture was diverted
to an inline Dean-Stark apparatus, and water began to collect
in the apparatus. Caution: hydrazine is likely to be present
and particular care should be taken to avoid contact with
any vapors or liquids. After approximately 300 mL of water
had collected, the internal reaction temperature was increased
to 155 °C over a 45 min period, and the reaction mixture stirred
at this temperature for 3-3.5 h during which time a total amount
of ∼790 mL of water had collected. The reaction mixture was
then allowed to cool to ∼ 70-75 °C, and the sides of the flask
were rinsed with 7.5 L of MeCN into the reaction mixture. The
resulting slurry was further cooled to ∼55 °C, and 17.5 L of
water was added dropwise. The slurry was allowed to slowly
cool to rt and stirred overnight at rt. The slurry was filtered41
and the wet cake was washed with 35 L of 1:1.5 MeCN:water
and then with 25 L of water and dried under vacuum/N2 sweep
to give 2.00 kg (85%) of 7 as an analytically pure white solid:
mp 165-166 °C; 1H NMR (CDCl3, 400 MHz) δ 0.70 (m, 2H),
0.84 (m, 2H), 1.61 (s, 3H), 2.93 (s, 2H), 6.49 (s, 1H), 7.12 (m,
6H), 7.32 (m, 9H); 13C NMR (CDCl3, 100 MHz) δ 7.0, 17.3,
22.1 (q, J ) 31 Hz), 27.5, 74.8, 119.9, 124.1 (q, J ) 271.0
Hz), 126.4, 127.8, 128.1, 133.3, 142.5, 146.5; 19F NMR (CDCl3,
75 MHz) δ -71.4. Anal. Calcd For C28H25F3N2: C, 75.32; H,
5.64; N, 6.27. Found: C, 75.42; H, 5.66; N, 6.23.
δ 1.24 (m, 2H), 1.27 (m, 2H), 3.26 (s, 3H), 3.72 (s, 3H); 13
C
NMR (CDCl3, 100 MHz) δ 9.9, 28.0 (q, J ) 31.0 Hz), 34.1,
53.4, 60.9, 124.1 (q, J ) 271.0 Hz), 165.0; 19F NMR (CDCl3,
75 MHz) δ -66.7.
Preparation of (2-Methyl-1-trityl-1H-imidazol-4-yl)-(1-
trifluoromethyl-cyclopropyl)-methanone (8). Into a 75 L
round-bottom flask charged with 35.0 L of CH2Cl2 was added
5.93 kg (13.2 mol) of 1112 at rt. The solution was cooled to 5
°C, and 4.40 L (13.2 mol) of EtMgBr (3.0 M in Et2O) was
added while maintaining the internal reaction temperature below
25 °C. The reaction was stirred at rt for 5 min postaddition of
the EtMgBr. To the solution was added in one portion 2.36 kg
(12.0 mol) of 10 in 5 L of CH2Cl2. The resulting mixture was
stirred at rt for 15 h and then reverse quenched into a solution
of 40 L of 5% KH2PO4. The layers were separated, and the
organic layer was washed with 40 L of saturated aqueous
NaHCO3. The organic layer was concentrated under reduced
pressure 15-25 °C to remove about 20 L of the solvent.
Isopropanol (25 L) was slowly added while concentration was
continued to a final volume of approximately 25 L. The resulting
slurry was heated to 60 °C for 1 h, and 25 L of water was
added over 1 h. The slurry was aged at 60 °C for 1 h, cooled
to rt, and filtered. The wet cake was washed with 12 L of 1:2
isopropanol/water and dried under vacuum/N2 sweep to afford
5.24 kg (97.7 wt %, 92%) of 8 as a colorless solid: mp 134-135
°C; 1H NMR (CDCl3, 400 MHz) δ 1.44 (m, 2H), 1.62 (s, 3H),
2.25 (m, 2H), 7.12 (m, 6H), 7.14 (s, 1H), 7.35 (m, 9H); 13C
NMR (CDCl3, 100 MHz) δ 13.1, 17.4, 27.4, 33.6 (q, J ) 30.0
Hz), 75.9, 123.9 (q, J ) 271.0 Hz), 128.3, 129.3, 129.9, 130.1,
136.7, 141.4, 147.6, 188.1; 19F NMR (CDCl3, 75 MHz) δ
-68.0. Anal. Calcd For C28H23F3N2O: C, 73.03; H, 5.03; N,
6.08. Found: C, 72.89; H, 4.99; N, 6.01.
Characterization of (2-methyl-1-trityl-1H-imidazol-4-yl)-
(1-trifluoromethyl-cyclopropyl)-methanol (12): Colorless oil:
1H NMR (CDCl3, 400 MHz) δ 0.72 (m, 1H), 0.85 (m, 2H),
0.91 (m, 1H), 1.62 (s, 3H), 4.07 (br s, 1H), 5.08 (s, 1H), 6.32
(s, 1H), 7.12 (m, 6H), 7.32 (m, 9H); 13C NMR (CDCl3, 100
MHz) δ 5.3, 7.4, 17.2, 27.8 (q, J ) 31 Hz), 66.0, 75.1, 118.2,
128.0, 128.3 (q, J ) 271.0 Hz), 128.4, 130.0, 137.7, 142.3,
146.7; 19F NMR (CDCl3, 75 MHz) δ -67.9.
Acknowledgment
We thank Dr. David J. Waterhouse of Merck & Co., Inc.
for analytical assistance.
Received for review February 5, 2009.
OP9000274
Preparation of 2-Methyl-4-(1-trifluoromethyl-cyclopro-
pylmethyl)-1-trityl-1H-imidazole (7). In a 72-L round-bottom
flask equipped with a mechanical stirrer, thermocouple, and
(41) The filtrate containing aqueous hydrazine waste was disposed of by
careful dilution with commercial bleach and sent for disposal to an
outside vendor.
580
•
Vol. 13, No. 3, 2009 / Organic Process Research & Development