Case study of a γ-butyrolactone alkylation with 1,3-dimethyl-2-imidazolidinone as a promoter

Article abstract of DOI:10.1021/op010224h

1,3-Dimethyl 2-imidazolidinone (DMI) is of lower toxicological risk than 1,3-dimethyl-3,4,5,6-tetrahydro-2(1H)-pyrimidinone (DMPU), hexamethyl-phosphorus triamide (HMPT), and hexamethylphosphoramide (HMPA). Formation of dialkylation byproducts is a common

Full text of DOI:10.1021/op010224h

Organic Process Research & Development 2001, 5, 609−611
Case Study of a γ-Butyrolactone Alkylation with 1,3-Dimethyl-2-imidazolidinone
as a Promoter
Bryan Li,* Richard A. Buzon, and Michael J. Castaldi
Chemical Research and DeVelopment, Pfizer Global Research and DeVelopment, Groton Laboratories, Eastern Point
Road, Groton, Connecticut 06340, U.S.A.
Scheme 1
Abstract:
1,3-Dimethyl 2-imidazolidinone (DMI) is of lower toxicological
risk than 1,3-dimethyl-3,4,5,6-tetrahydro-2(1H)-pyrimidinone
(DMPU), hexamethyl-phosphorus triamide (HMPT), and hexa-
methylphosphoramide (HMPA). Formation of dialkylation
byproducts is a common problem in lactone alkylation. DMI,
used in stoichiometric amount, increases the rate of alkylation
of γ-butyrolactone 1 by >30-fold, therefore minimizing the
dialkylation in multi-kilogram preparations. The isolated yield
of the monoalkylated product 2 is >90%. The reaction protocol
is also demonstrated to work on other lactone substrates and
alkylating agents.
reaction variable, had to be kept below -70 °C, as elevated
reaction temperature was found to compromise the diastereo-
selectivity.
Functionalized lactone 2, an intermediate of a drug
candidate, is prepared from γ-butyrolactone 11 (Scheme 1).
R,R-Dialkylation was a major concern in this reaction. Initial
attempts at the reaction without addition of additives, such
as hexamethylphosphoramide (HMPA), led to either a
substantial level of dialkylation or significant amount of
starting material 1.2
Use of HMPA4 or hexamethylphosphorus triamide (HMPT)
as a cosolvent in lactone alkylations is very common in both
academic and industrial laboratories.5 Both HMPA6 and
HMPT7 are known to be potentially carcinogenic. 1,3-
Dimethyl-3,4,5,6-tetrahydro-2(1H)-pyrimidinone (DMPU),
another widely used HMPA substitute, is a possible chemical
mutagen.8 These reagents are typically used in large excess
as a cosolvent in the reactions. 1,3-Dimethyl 2-imidazoli-
dinone (DMI) has been used as a substitute for HMPT in
dehydration and dehalogenation reactions,9 and it has been
used as solvent or cosolvent in acetylene alkylations,10
Ullmann ether synthesis,11 silane substitution,12 and lactone
alkylation.13 More recently, it was shown to exert regiocon-
trol on a diene-diolate alkylation14 and acylations.15 It is
The formation of dialkylated product and the presence
of unreacted lactone were most likely due to proton exchange
between the monoalkylated product 4 and enolate 3 (Figure
1). There are two competing pathways for enolate 3: reaction
with the bromide to give the desired product 2 (k1) via the
formation of 4 or proton exchange with 4 to give the lactone
1 (k2) and enolate 6, which in turn reacts with the bromide
to give rise to the dialkylation product 7. In our initial
experiments, to keep the reaction temperature below -70
°C, the alkyl bromide was added slowly to the enolate. The
addition took several hours on ∼1 mol scale, as the reaction
was extremely exothermic. Factors including the long
reaction time3 and localized exotherm during dimethylallyl
bromide addition helped increase levels of 5 and 6 and hence
the higher levels of 1 and 7, respectively. It was evident that
a more reactive enolate species and a higher concentration
of the bromide would increase the reaction rate r1 and hence
the ratio of r1/r2. (Figure 1). Temperature, another important
(4) Herrmann, J. L.; Schlessinger, R. H. J. Chem. Soc., Chem. Commun. 1973,
711.
(5) (a) Molander, G. A.; Harris, C. R. J. Am. Chem. Soc. 1996, 118, 4059. (b)
Pellissier, H.; Michellys, P.-Y.; Santelli, M. J. Org. Chem. 1997, 62, 5588.
(c) Takacs, J. M.; Weidner, J. J.; Newsome, P. W.; Takacs, B. E.;
Chidambaram, R.; Shoemaker, R. J. Org. Chem. 1995, 60, 3473. (d)
Kigoshi, H.; Ojika, M.; Ishigaki, T.; Suenaga, K.; Mutou, T.; Sakakura,
A.; Ogawa, T.; Yamada, K. J. Am. Chem. Soc. 1994, 116, 7443. (e) Tanaka,
Y.; Grapsas, I.; Dakoji, S.; Cho, Y. J.; Mobashery, S. J. Am. Chem. Soc.
1994, 116, 7475.
(6) Sarrif, A. M.; Krahn, D. F.; Donovan, S. M.; O’Neil, R. M. Mutat. Res.
1997, 380, 167
(7) Ashby, J.; Styles, J. A.; Paton, D. Br. J. Cancer 1978, 38, 418.
(8) Zijlstra, J. A.; Vogel, E. W. Mutat. Res. 1988, 201, 27.
(9) Spangler, C. W.; Kjell, D. P.; Wellander, L. L.; Kinesella, M. A. J. Chem.
Soc., Perkin. Trans. 1 1981, 2287.
(10) Shimizu, N.; Mori, N.; Kuwahara, Y.; Tsuji, T. Biosci. Biotechnol. Biochem.
1999, 63, 1478.
(11) Oi, R.; Shimakawa, C.; Takenaka, S. Chem. Lett. 1988, 5, 899.
(12) Ito, H.; Ishizuka, T.; Okumura, T.; Yamanaka, H.; Tateiwa, J.-I.; Sonoda,
M.; Hosomi, A. J. Organomet. Chem. 1999, 574, 102.
(13) Preparation of 2-benzyl-3-hydroxy-γ-butyrolactone on millimole scale
was reported using DMI as a cosolvent. In this reaction, benzyl bromide
(1.0 equiv) was used as limiting reagent (DMI, 14 equiv, and the
γ-butyrolactone, 5.2 equiv): Higamie, K.; Furukawa, Y.; Katsumura, S.;
Takehira, Y. Jpn. Kokai Tokkyo Koho, JP 11189589, 1999.
* To whom correspondence should be addressed. Fax: (860)715-7305.
E-mail: bryan_li@groton.pfizer.com.
(1) Kath, J. C.; Brown, M. F.; Poss, C. S.. PCT Int. Apl. WO 9940061, 1999;
Chem. Abstr. 1999, 131, 157767.
(2) Use of 2.00-2.05 equiv of base gave 10-15% of dialkylation product,
whereas 1.85 equiv of base gave ∼10% of starting material 1 and 3-5%
dialkylation product.
(3) Goto, M.; Akimoto, K.-I.; Aoki, K.; Shindo, M.; Koga, K. Tetrahedron
Lett. 1999, 40, 8129 wherein it was reported that the ratio of the
monoalkylated product to the dialkylated product increased in a shorter
reaction time.
10.1021/op010224h CCC: $20.00 © 2001 American Chemical Society and The Royal Society of Chemistry
Published on Web 10/13/2001
Vol. 5, No. 6, 2001 / Organic Process Research & Development
•
609

Table 1. Reactions of lactone enolates with alkyl halides20
reaction, the addition of dimethylallyl bromide solution in
THF (precooled to -78 °C) to the enolate solution at -85
°C (cooling to this temperature was needed to pre-empt the
huge exotherm during the bromide addition) was completed
within 2 min while maintaining the reaction temperature
below -72 °C. Immediately after the electrophile addition,
the reaction was determined to be completed by HPLC
analysis of a reaction aliquot. Typically, the desired product
2 was formed in 90-92% yield with ∼1% of diastereomer,
∼2% of dialkylated product, and ∼4% of the unreacted
lactone also present.18,19 The SN2′ product was not detected
in the reaction. The presence of DMI did not seem to have
any impact on the diastereoselectivity which is probably
controlled by the bulky group at the γ-position of the lactone.
Reverse quench of the reaction by addition of the enolate
(18) By HPLC area, monitored at 210 nm wavelength.
(19) In a typical experiment: Under nitrogen atmosphere, 1.0 M LHMDS
solution in THF (6.49 L, 2.10 equiv) was charged to a three-neck flask,
and cooled to -78 °C. Lactone 1 (1000 g, 3.09 mol) in 5.0 L of anhydrous
THF was pre-cooled to -25 °C and then added to the LHMDS solution
slowly while keeping the reaction temperature below -65 °C. After the
addition, the reaction was stirred for 30 min at -65 to -78 °C. Dimethyl
imidazolidinone (423 g, 3.71 mol, 1.2 equiv) was then added neat, and
reaction mixture was then cooled to -85 °C. Dimethyl allyl bromide (497.8
g, 3.34 mol, 1.1 equiv) in 5.0 L of anhydrous THF was pre-cooled to -78
°C and added quickly (in 2 min) to the enolate solution while maintaining
the reaction temperature below -72 °C. After the addition, the reaction
was stirred for 10 min. The reaction was quenched with acetic acid (743
mL, 4.2 equiv)/THF (5.0 L), and ethyl acetate (20 L) is then added. After
water wash (3 × 10 L), the ethyl acetate solution was dried over Na2SO4
and concentrated to an oil (1.14 kg, 94.3% yield). The crude product was
sufficiently pure to be used in the downstream steps. An analytically pure
sample can be obtained by silica gel chromatography: 1H NMR (CDCl3,
400 MHz) δ 1.34 (s, 9H), 1.56 (s, 3H), 1.64 (s, 3H), 1.87-1.94 (m, 1H),
2.20-2.26 (m, 1H), 2.36-2.42 (m, 1H), 2.63-2.71 (m, 1H), 2.82-2.88
(m, 2H), 3.96 (dt, 2H, J ) 8.4, 8.4 Hz), 4.41 (t, 1H, J ) 6.4 Hz), 4.72 (d,
1H, J ) 9.6 Hz), 4.99-5.00 (m, 1H), 6.87-7.00 (m, 3H), 7.20-7.26 (m,
1H); 13C NMR (CDCl3, 400 MHz) δ 17.78, 25.68, 28.08, 29.27, 29.33,
38.73, 38.92, 54.32, 78.41, 79.91, 113.43, 113.64, 115.98, 116.18, 119.42,
124.86, 124.88, 129.92, 130.00, 135.21, 139.70, 139.78, 155.79, 161.53,
163.98, 179.38.
Figure 1.
structurally analogous to DMPU, but more water soluble and
hence readily removed by aqueous wash in the extractive
workup. More importantly, DMI appears to be of lower
toxicological risk than HMPA, HMPT, and DMPU,16 which
makes it more attractive in the production environment.
We found the rate of the alkylation reaction was increased
by >30-fold17 in the presence of DMI (1.2 equiv). To
increase the bromide concentration and lower the potential
localized temperature spike, precooling of the bromide was
needed so that fast transfer of the bromide (highly exother-
mic!) could be achieved while maintaining the reaction
temperature at <-70 °C. In a 9.3 mol (3 kg of 1) scale
(14) Brun, E. M.; Gil, S.; Parra, M. Syn. Lett. 2001, 1, 156; Tetrahedron
Asymmetry, 2001, 12, 915.
(15) Kondo, J.; Inoue, A.; Shinokubo, H.; Oshima, K. Angew. Chem., Int. Ed.
2001, 40, 2085.
(16) Lo, C. C.; Chao, P. M. J. Chem. Ecol. 1990, 16, 3245.
(17) In the presence of DMI, the reaction time needed was less than 2 min, 2
h or longer was needed in the absence of DMI.
(20) Enolates were generated at -78 °C with 1.0 M LHMDS (1.05 equiv)
solution in THF as base. DMI (1.2 equiv) was added prior to the addition
of the alkyl halide (1.10 equiv). The reaction temperature shown was
maintained during the alkyl halide addition. Yields and diastereomer ratios
were based on HPLC analysis at 210 nm.
610
•
Vol. 5, No. 6, 2001 / Organic Process Research & Development

solution to the dimethylallyl bromide solution in THF was
also investigated. At temperatures below -70 °C, the reaction
gave results comparable to that of the normal quench. N-
alkylation on the Boc-NH group was observed at <1% under
the reaction conditions on a 30 mmol scale, while in the
normal quench, N-alkylation was not detected. Nonetheless,
the reverse quench would require two large cryogenic
reaction vessels, and the resource constraint makes it less
practical oftentimes in production settings, whereas the
normal quench only needed one cryogenic reaction tank and
a flask to hold the bromide solution that was cooled in a dry
ice-acetone bath (on a 10 mol scale, a 22 L flask would
suffice).
alkylation with allylbromide at elevated reaction temperature
(entry 3), and the reaction also worked well for δ-valero-
lactone (entry 4).
In summary, we have shown that DMI, an ecologically
friendly reagent that is yet to gain popularity among the
community of synthetic chemists, effectively promotes the
rate of reaction in γ-butyrolactone alkylation when used in
stoichiometric amount. The protocol described has been
successfully used in multikilogram preparation of 2.
Acknowledgment
We thank Dr. Charles K-F Chiu, Dr. Frank J. Urban, and
Mr. V. John Jasys for helpful discussions. We also thank Dr.
John A. Ragan for his comments on the manuscript.
The approach was found to be general in reactions of other
lactone substrates and alkylating agents (Table 1). The
diastereoselectivity appeared to be dependent on reaction
temperature and the size of the alkylating group (entries 1
and 2). Comparable results were obtained for γ-butyrolactone
Received for review August 13, 2001.
OP010224H
Vol. 5, No. 6, 2001 / Organic Process Research & Development
•
611