A highly efficient synthesis of triisopropylsilyldifluorobromopropyne yields a versatilegem-difluoromethylene building block

摘要

A highly efficient synthesis of triisopropylsilyldifluorobromopropyne yields a versatile gem-difluoromethylene building block ZhiGang Wang and Gerald B. Hammond* Department of Chemistry and Biochemistry, The University of Massachusetts Dartmouth, North Dartmouth, Massachusetts 02747, USA. E-mail: ghammond@umassd.edu Received (in Corvallis, OR, USA) 23rd September 1999, Accepted 8th November 1999 Triisopropylsilyldifluorobromopropyne, readily prepared in excellent yield from the reaction of lithium triisopropylsilylacetylide with CF2Br2, provides a convenient entry into a functionalized CF2 synthon.The gem-difluoromethylene unit is a key structural motif in many fluorine containing compounds of biological and pharmaceutical significance. For this reason, it is an important synthetic target.1 Two complementary approaches to such important unit exist. These are (i) substitution of a carbonyl or an active methylene group by fluorine;2 and (ii) use of small gemdifluoromethylene- containing building blocks.3 When it comes to fluorinate complex molecules, the latter approach is preferred because of the reactivity, thermal instability, hazards and cost associated with electrophilic and nucleophilic fluorinating agents.The two most frequently used gem-difluoromethylene synthons, (EtO)2P(O)CF2Br and EtOC(O)CF2Br, were developed in the late 70s and early 80s by Burton4 and Fried,5 respectively. Stimulated by our earlier work in the synthesis of fluorinated phosphonates,6 we sought a new generation of difluorinated building blocks from inexpensive industrial fluorine feedstock such as CF2Br2 (not included in the list of CFCs to be phased out).Our initial target, triisopropylsilyldifluoropropyne 1 (Fig 1), is a highly functionalized threecarbon backbone that contains a propargyl silane moiety. This feature should facilitate multiple synthetic conversions containing the gem-difluoromethylene unit. A literature search revealed that difluoropropargyl substrates, without exception, have been prepared in disappointingly low yields.7 The low yields obtained have probably contributed to the lack of use of the difluoropropargyl building block in the literature.To our satisfaction, 1 was very efficiently assembled in one step (92 GC-MS, 81 isolated) by the reaction of CF2Br2 with lithium triisopropylsilylacetylide. This reaction has been carried out on 5, 16 and 50 g scales and the results have shown excellent reproducibility.8 We attributed the success of this reaction to the presence of the TIPS group, which possesses remarkably different properties compared to other alkylsilyl groups.9 Chiefly, the presence of TIPS enhances the stability of the triisopropylsilyldifluoropropyne anion intermediate.10 Although the reaction mechanism is still unclear, our experimental observation of a typical 5ndash;10 min induction period supports the difluorocarbene based ionic chain path proposed by Wakselman and co-workers.7a With a highly efficient preparation of 1 in hand, we explored the synthesis of various gem-difluoromethylene- containing compounds.Preliminary results, summarized in Scheme 1,11 unveiled a highly versatile building block. Reduction with LiAlH4 gave allene 2, whereas nucleophilic substitution with MP(O)(OEt)2 afforded difluoropropyne 3 and difluoropropargyl phosphonate 4, although the latter was obtained in low yield. Because compound 4 is a potential precursor for isosteric and isoelectronic phosphate mimics of enzyme inhibitors,12 an optimization of this reaction is in progress.Using ultrasound, 1 reacted rapidly with zinc dust yielding the alkynyl organozinc reagent 5 in situ. This organozinc intermediate is a useful synthetic building block as can be seen in the Reformatsky-type reactions shown at the bottom of Scheme 1. Addition of Zn to 1 produced dimer 6 in nearly quantitative yield. Compound 6 is a potential intermediate in the synthesis of CF2CF2-containing bioactive molecules.13 If 5 is quenched with powdered iodine, it produces difluoroiodopropyne 7, another important gem-difluoromethylene synthon.14 Addition of trans-cinnamaldehyde to 5 afforded difluoro alcohol 8 in 70 yield.When needed, the TIPSprotecting group can be easily removed, as demonstrated by the conversion of 8 to 9 under mild conditions, and in excellent yield. The latter result will allow an easy entry to the preparation of propargylic, and possibly allylic, a,a-difluorocarbonyl compounds, after oxidation of the alcohol and hydrogenation of the triple bond.Fig. 1 Building block potential of 1. Scheme 1 Reagents and conditions: i, BuLi, CF2Br2, THF, 220 deg;C to room temp.; ii, LiAlH4 (0.6 equiv.), THF, 280 deg;C; iii, NaN (TMS)2, HP(O)(OEt)2 (1.0 equiv.), THF, 210 deg;C; iv, HP(O)(OEt)2, BuLi (1.5 equiv.), 210 deg;C; v, Zn (1.2 equiv.),THF, room temp., ultrasound; vi, I2 (1 equiv.), 0 deg;C; vii, trans-cinnamaldehyde, 20 h, room temp.; viii, 1 (1.0 equiv.), Zn (2.0 equiv.), room temp.; ix, TBAF (1.0 equiv., 1 M in THF), 280 deg;C.This journal is copy; The Royal Society of Chemistry 1999 Chem. Commun., 1999, 2545ndash;2546 2545Other synthetic modifications of 1 and the preparation of a monofluorinated counterpart to 1 are under investigation. The generous support of the National Science Foundation (CHE-9711062), Petroleum Research Fund (PRF#32595-B1) and the Camille and Henry Dreyfus Foundation (TH-96-012) is gratefully appreciated. Notes and references 1 For a recent leading reference see: K.Uneyama, G. Mizutani, K. Maeda and T. Kato, J. Org. Chem., 1999, 64, 6717. For a recent review see: M. J. Tozer and T. F. Herpin, Tetrahedron, 1996, 52, 8619. 2 Organo-Fluorine Compounds, ed. B. Baasner, H. Hagemann and J. C. Tatlow, Thieme, Stuttgart, 1999, vol. E 10 a. 3 For an updated compilation of references see: J. M. Percy, Top.Curr. Chem., 1997, 193, 131. 4 D. J. Burton and R. M. Flynn, J. Fluorine Chem., 1977, 10, 329. For recent reviews, see: D. J. Burton and L. Lu, Top. Curr. Chem., 1997, 193, 45; D. J. Burton, Z. Y. Yang and W. M. Qiu, Chem. Rev., 1996, 96, 1641. 5 E. A. Hallinan and J. Fried, Tetrahedron Lett., 1984, 25, 2301. Also: T. Taguchi, O. Kitagawa, Y. Suda, S. Ohkawa, A. Hashimoto, Y. Iitaka and Y. Kobayashi, Tetrahedron Lett., 1988, 29, 5291 and references therein; S. Mcharek, S.Sibille, J. Y. Nedelec and J. Perichon, J. Organomet. Chem., 1991, 401, 211; J. M. Altenburger and D. Schirlin, Tetrahedron Lett., 1991, 32, 7255; J. M. Andres, M. A. Martinez, R. Pedrosa and A. Perez-Encabo, Synthesis, 1996, 1070. 6 F. Benayoud, D. J. deMendonca, C. A. Digits, G. A. Moniz, T. C. Sanders and G. B. Hammond, J. Org. Chem., 1996, 61, 5159; F. Benayoud and G. B. Hammond, Chem. Commun., 1996, 1447. 7 (a) I. Rico, D. Cantauzene and C. Wakselman, J. Chem. Soc., Perkin.Trans. 2, 1982, 1063; (b) P. Y. Kwok, F. W. Muellner, C. K. Chen and J. Fried, J. Am. Chem. Soc., 1987, 109, 3684; (c) Y. Hanzawa, K. Inazawa, A. Kon, H. Aoki and Y. Kobayashi, Tetrahedron Lett., 1987, 28, 659. 8 Under Ar atmosphere, a solution of triisopropylsilylacetylene (36 g, 0.2 mol) in THF (300 ml) was cooled to 220 deg;C. BuLi (125 ml, 1.6 M solution in hexane) was added via syringe and the resulting solution was stirred for 30 min before CF2Br2 (62 g, 1.5 equiv.) was added.The solution was allowed to warm to room temperature and stirred for 9 h. After solvent removal, NH4Cl (120 ml) was added and the oily layer was extracted with Et2O (3 3 100 ml). Standard work-up gave the crude product as a viscous orange oil (92 yield by GC2MS) which was purified by distillation (bp 51ndash;52 deg;C/0.1 mmHg) to yield 1 (49.9 g, 81); dH(300 MHz, CDCl3) 1.11 (s, TIPS protons); dC(75 MHz, CDCl3) 100.61 (t, 1JCF 290), 97.16 (t, 2JCF 36.5), 95.1 (t, 3JCF 4.7), 18.01, 11.28; dF 232.60 (s, 2F); m/z (GC-MS) 312 (M+ + 1, 4), 310 (M+ 2 1, 4), 269 (14), 267 (14), 143 (27), 77 (100) (calc.for C12H21SiF2Br: C, 46.33; H, 6.75. Found: C, 46.91; H, 6.83). The triisopropylsilylacetylene starting material, purchased from GFS Co., contained diisopropylpropenylsilylacetylene (11) and diisopropylpropylsilylacetylene (14). Fractional distillation did not remove these impurities. The GCMS analysis of 1 shows all three components were alkylated and their ratios perfectly match that of the starting material.The calculated C and H were based on the formula of triisopropylsilyldifluorobromopropyne. If the two impurities are removed from the calculation, the values found for C and H are 46.84 and 6.67, respectively. 9 For a recent review of the TIPS group in organic chemistry see: C. Ruuml;cker, Chem. Rev., 1995, 95, 1009. 10 Similarly, whereas ynolates lack convenient methods for their generation a lithium silylynolate has been used in ketenylation reactions: H.Kai, K. Iwamoto, N. Chatani and S. Murai, J. Am. Chem. Soc., 1996, 118, 7634. 11 Selected data for 2: dH(300 MHz, CDCl3) 1.06 (s, TIPS protons), 4.81 (t, 1H, 4JHH 7.23), 4.30 (d, 2H, 4JHH 7.23); m/z (GC-MS) 196 (M+), 157, 153, 125, 97, 83, 67. For 3: dH(300 MHz, CDCl3) 1.09 (s, TIPS protons), 6.17 (t, 2JHF 54.8); dF 2105.49 (s); m/z (GC-MS) 232 (M+), 189, 161, 133, 105, 81, 77. For 4: dH(300 MHz, CDCl3) 1.11 (s, 21H, TIPS protons), 1.39 (t, 6H, 3JHH 7.00), 4.32 (m, 4H); dF 296.79 (d, 2JPF 109.01); dP 4.23 (t, 2JPF 109.10); m/z (GC-MS) 325 (M+ 2 43), 297, 269, 153, 109, 81.For 6: dH(300 MHz, CDCl3) 1.10 (s, TIPS protons); dF 299.01 (s, 4F); m/z (GC-MS) 462 (M+), 377, 307, 239, 183, 115, 77. For 7: dH(300 MHz, CDCl3) 1.11 (s, TIPS protons); dF 228.13 (s); m/z (GC-MS) 315 (M+ 2 43), 265, 237, 189, 165, 119, 105, 77. For 8: dH(300 MHz, CDCl3) 7.41ndash;7.26 (m, 5H), 6.82 (d, 1H, 3JHH 15.9), 6.26 (dd, 1H, 3JHH 15.9, 3JHH 6.2), 4.55 (m, 1H, CHOH), 2.05 (s, 1H, OH), 1.06 (s, 21 H, TIPS protons); dF 294.3 (d, 1F, 2JFF 274), 296.7 (d, 1F, 2JFF 274). For 9: dH(300 MHz, CDCl3) 7.45ndash;7.30 (m, 5H), 6.85 (dd, 1H, 4JHH 0.9, 3JHH 15.9), 6.24 (dd, 1H, 3JHH 15.9, 3JHH 6.3), 4.55 (m, 1H, CHOH), 2.85 (t, 1H, 3JHF 5.0), 2.04 (s, 1 H); dF 295.9 (d, 1F, 2JFF 277), 296.9 (d, 1F, 2JFF 277); m/z (GC-MS) 206 (M+, 5),170 (3), 159 (2), 133 (100), 115 (30), 77 (26), 55 (46). 12 The major by-product is 3, an indication that, compared to the bromide, the triisopropylsilyldifluoropropyne anion is a better leaving group. For a leading reference on fluorinated phosphate mimics, see: D. Orsquo;Hagan and H. S. Rzepa, Chem. Commun., 1997, 645. 13 W. C. Sun, C. S. Ng and G. D. Prestwich, J. Org. Chem., 1992, 57, 132. 14 We have also prepared 8 in 73 yield by the reaction of lithium triisopropylsilylacetylide with CF2I2. However, the prohibitive cost of commercial CF2I2 will most likely limit its use in large-scale synthesis. Communication 9/07784G 2546 Chem. Commun., 1999, 2545ndash;2546