TITLE OF THE INVENTION
LEWIS ACID CATALYSIS USING CHIRAL METAL COMPLEXES
This application claims priority to provisional application 60/254,396 filed on December 7, 2000, incorporated herein by reference.
STATEMENT OF GOVERNMENT INTEREST
This invention was made with government support from the National Science Foundation (under grant CHE-9816028) and the National Institutes of Health (under grant GM 46503).
BACKGROUND OF THE INVENTION FIELD OF THE INVENTION
The invention is related to catalysts and their use in the synthesis of organic molecules. The synthesis of ligands for the catalysts is also included in the invention.
DISCUSSION OF THE BACKGROUND Catalysts are widely used in the preparation of organic chemicals. Catalysts provide a means to prepare organic molecules in high enantioselective yields that would otherwise not be possible using traditional techniques of organic synthesis. Catalysts further allow chemical reactions that would normally be too slow to be of economic or practical utility to be significantly accelerated. U.S. Patent 5,296,595 describes the use of dimetal complexes bridged with heteroatom ligands as catalysts in a number carbenoid insertion of reactions.
U.S. Patent 5,302,737 describes catalysts containing a dirhodium(II) metal center coordinated to one or more bridging fluorinated carboxylate ligands. These catalysts were used for inserting a carbon atom into a Si-H, bond in a regioselective and stereoselective manner. The prior art catalysts include metal carboxylate species which have turnover, rates that are too low to be of significant commercial value.
A benchmark for reactivity of dirhodium(II) catalysts is the dirhodium(II) tetracetate complex Rh2(OAc)4 (1) where OAc represents a carboxylate ligand which bridges two Rh atoms.
.0
CH2C.
V
The rhodium atoms of this complex is normally viewed as Rh(II), i.e., rhodium in a +2 oxidation state. The reactivity and selectivity ofthe dirhodium complex can be tailored by changing the bridging ligand. The addition of a chiral ligand can, for example, lead to limited enantioselectivity ofthe catalyst. However, given the unpredictable nature of catalysts, it is difficult to prepare catalysts that offer specific reaction behavior and favored end products (see for example M.P. Doyle in "Catalysis by Di- and Polynuclear Metals Cluster Complexes, R.D. Adams and F.A. Cotton, Ed.s., VCH Publishers, New York, 1998, chapter 17 - incorporated herein by reference).
An important measure of a catalyst's effectiveness is substrate-to-catalyst loading. This ratio is a reflection ofthe number of molecules of substrate that can be transformed per molecule of catalyst. This ratio is preferably much greater than 1, meaning that a given quantity of catalyst material can transform a much greater quantity ofthe substrate material into a desired end product. A large substrate-to-catalyst loading is desirable since the quantity of catalyst required to complete a reaction is minimized. This is an important consideration where the catalyst contains a precious metal or a complex ligand and is expensive to produce. The substrate-to-catalyst loading can be represented as the turn over number (TON) of the catalyst. Each time a catalyst transforms a substrate molecule it is said to have "turned over". A large TON implies that a single catalyst molecule is able to transform many substrate molecules. Turnover rate is also used to judge the usefulness of a catalyst. Turnover rate is a function ofthe number of molecules turned over (transformed into the desired end- product) per unit of time. Methods of determining TON are commonly known in the art, and are described in, for example, Catalytic Asymmetric Synthesis, I. Ojima, Ed., Wiley- VCH, New York, NY, 2000, incorporated herein by reference.
The ligand structural conformation of two types of catalyst complexes, Rh2(OAc)4 and Rh2(carboxamidate)4, are shown below as A and B respectively. While these catalysts individually show regioselectivity or stereoselecivity no single catalyst is able to perform both tasks simultaneously and, importantly, it has not possible to purposefully design a catalyst that performs all ofthe desired functions (M.P. Doyle, D.J. Timmons, J. Organomet. Chem., 2001, 617-618, pp. 94-104 - incorporated herein by reference).
Rh2(OAc)4 Rh2(carboxamidate)4
A B
In U.S. Patent 5,175,311 (incorporated herein by reference), dirhodium(II) catalysts containing carboxylate ligands (e.g., carboxamidate ligands) other than OAc were shown to offer improvements in activity and selectivity over Rh2(OAc)4. Chirality was imparted to the catalyst by adding chiral substituents to the carboxamidate ligands. The catalysts were applied to diazo decompositions and addition and polar addition reactions. Turnover and enantioselectivity were low however..
The application of di- or polymetallic catalysts containing late transition metals' in enantioselective Lewis acid catalyzed reactions is limited (see for example - K.A. Jorgensen, Angew. Chem., Int. Ed., 2000, 39, 3558 - incorporated herein by reference). As has been commonly believed in the art, the steric bulk ofthe bridging ligands together with the relatively electron rich nature of carboxamidates would not be expected to yield effective Lewis acid catalyst. Generally, Lewis acids are electron poor complexes. The early transition metals typically exhibit greater Lewis acidity than the later transition metals.
Diels-Alder and hetero-Diels-Alder reactions are commercially important as applied to the synthesis of, for example, dihydropyrans (see for example Boger, D.L.; Weinreb, S.H., Hetero-Diels-Alder Methodology in Organic Synthesis; Academic Press; New York, 1987,
Nicolaou, K.C.; Sorensen, E.J., Classics in Total Synthesis; Targets, Strategies, Method; VCH; New York, 1996. Waldmann, H. Synthesis 1994, 535. Tietze, L.F.; Kettschau, G., Top. Curr. Chem. 1997, 190, 1. Tietze, L.F., Curr. Org. Chem. 1998, 2, 19 - of each of which are incorporated herein by reference). Although Lewis acids have been shown to be effective in accelerating Diels-Alder reactions and may provide improved enantioselectivity, their use has been limited due to the high substrate-to-catalyst loadings required.
Diazo decompositions are a route to addition reactions, by generating an electrophilic metal carbene as an intermediate. Catalytic transformations of organic diazo compounds can be used in a variety of synthetic methods. Procedures for the formation of carbon-carbon bonds by cyclopropanation, dipolar addition, carbon-hydrogen insertion, aromatic substitution reactions, and ylide generation/rearrangement with allylamines, allyl sulfides, and allyl ethers are known (see U.S. Patent 5,175,311). These methods suffer from lack of stereo- and regio specificity. The catalysts so far used in diazo decompositions have not demonstrated reactivity towards vinyldiazoacetates or diazomalonates (see for example H.M.L. Davies, Curr. Org. Che. 1998, 2, pp. 463-468; H.M.L. Davies & B.D. Doan, J. Org. Chem. 1999, 64, pp. 8501-8508; P. Muller & D. Fernandez, Helv. Chim. Ada. 1995, 78, pp. 947-957).
SUMMARY OF THE INVENTION
The present invention includes the application of certain dirhodium(II) carboxamidates as enantiosselective Lewis acid catalysts. The use of these catalysts species in inter alia Diels-Alder, hetero-Diels-Alder reactions, Mukiyama-aldol reactions, aldol reactions, epoxide ring opening, ylide reactions, diazo decomposition reactions, diene reactions and addition reactions to provide reaction products of high sy anti yield and/or high ee yield is a part of this invention.
Dirhodium(II) catalysts containing multiple chiral sites are another aspect ofthe invention. Dirhodium(II) carboxamidates containing fluorinated ligands are a further aspect ofthe invention.
Other aspects ofthe invention will become clear with a further understanding ofthe invention by reference to the detailed description provided herein.
DETAILED DESCRIPTION OF THE INVENTION
The catalysts o the present invention are dirhodium(II) complexes where the dirhodium(II) metal center is bridged by at least one carboxamidate ligand. The structures shown below are intended to convey a molecule of formula Rh2(L)4 where L may be a carboxamidate ligand that bridges the dimetal center. For clarity the molecules depicted are shown with only a single bridging ligand. The remaining coordination sites ofthe Rh dimer are indicated by solid lines and may be occupied by up to three additional bridging carboxamidate ligands and are indicated by a solid lines. The Rh(II) complex may contain four identical carboxamidate ligands, a mixture of different carboxamidate ligands or a mixture of carboxamidate ligands and other bridging and non-bridging ligands that fill the coordination sphere ofthe two metal centers. The halogenated analogs, c and d below, ofthe oxaazetidine and oxapyrrolidinine complexes (a and b) are preferred and their reaction chemistry, together with the reaction chemistry ofthe non-fluorinated catalysts, is a preferred embodiment.
Each ofthe molecules above contains at least one carboxamidate ligand having a chiral center attached to the nitrogen atom ofthe bridging ligands. The carbon atom bonded to both the nitrogen and the ester group imparts a first center of chirality to the ligand. A second chiral center can be added to the catalyst by further substituting the ligand, for example by adding a chiral substituent (chiral R2) to the ester group. A third center of chirality can be added to the Rh2(carboxamidate)4 complex by incorporating oxazolidinone- carboxylate or oxaimidazolidinone-carboxylate ligands containing a third chiral substituent as R3 in e above.
Because the catalyst has an active catalytic site on both sides ofthe catalyst, i.e. at each ofthe metal atoms, it is preferred that either (1) both sides ofthe catalyst have a chiral center to thereby effect enantioselectivity, or (2) one side have a chiral center, and the other side have a blocking structure which would substantially impair approach to the metal atom on the other side. Otherwise, the enantioselectivity ofthe catalyst would be greatly reduced with the chiral side ofthe catalyst effecting enantioselectivity, while the "free" side ofthe catalyst produces a racemic mixture.
The first of these options is most preferred. In other words, it is preferred to have at least one chiral center on each side ofthe catalyst, i.e., bonded to the complexing atom which is bonded to each ofthe metal atoms. This second chiral center should have the same R S configuration as the first chiral center.
Preferably, a chiral center is oriented on both sides ofthe catalyst by having one ligand oriented with its chiral center on one side ofthe catalyst and another ligand having its chiral center on the other side ofthe catalyst. Alternatively, one ligand can have a chiral center bonded to both of its complexing atoms.
The concept of stereoselectivity and methods of determining enantioselectivity are commonly known in the art (see, for example, J. March, Advanced Organic Chemistry, 3 rd Ed., John Wiley & Sons, New York (1985), pp. 82-109, incorporated herein by reference).
The bridging ligands are preferably carboxamidates that bond to the metals through both a nitrogen and an oxygen atom. The preferred ligands are derived from azetidinones (I), pyrrolidinones (II), oxazolidinones (III), and imidazolidinones (IV) whose general formulas are shown below. The ligands may be prepared as described in M.P. Doyle et al. Adv. Synth. Catal. 2001, 343, p. 112 - incorporated herein by reference.
III IV
R1 is a hydrogen atom in the uncomplexed ligand. In I-IV and a-f above R2 and R3 may be any organic or organometallic group, including, but not limited to, linear, branched or cyclic alkanes, polycyclic groups, aromatic groups, mixed aromatic-aliphatic polycyclic groups, metal-organic groups, etc., constituted ofthe atoms C, O, N, P, S, H, etc., preferably when R2 or R3 contains carbon, it contains 1-35 carbon atoms. In this sense, the identifiers R2 and R3 are used herein to denote a chemical functionality that imparts chirality to the ligand. The R2 and R3 substituents described here may, of course, be unsubstituted or substituted with any and all functionalities. Examples of such functionalities include NO2, halogen, NR3 where R is defined as R2 above, CN, etc. In this sense the invention methods reside in the causation at the various reaction centers and not in the identity ofthe particular appended substituents. Those of ordinary skill in the art are well suited to carry out the invention reactions as described herein with broad variation in regard to the ligand groups surrounding the reaction center in view ofthe guidance provided herein.
Preferred R2 and R3 substituents include methyl, ethyl, i-propyl, neopentyl, octadecyl or benzyl. The incorporation of a second and a third chiral group, such as menthyl, lactate or mandelate as R2 and/or R3 is especially preferred.
In another preferred embodiment ofthe invention R2 and/or R3 and can represent polymeric materials, preferred polymeric materials include polystyrene-polyethyeneglycol and Merrifield's resin.
The substituents X1 and X2 may be hydrogen or a halogen, preferably fluorine.
The dirhodium(II) complexes may contain other uni-dentate, bi-dentate or multi- dentate ligands bonded to the metal center in a dative, covalent, or ionic manner in place of, or in addition to, the at least one carboxamidate ligand. Other ligands may include but are not limited to halogen, XR, XR2, XR3, and their respective salts, where X is any element ofthe main group ofthe periodic table ofthe elements and R is a hydrogen atom any organic or organometallic group, including, but not limited to, linear, branched or cyclic alkanes, ' polycyclic groups, aromatic groups, mixed aromatic-aliphatic polycyclic groups, metal- organic groups, etc., constituted ofthe atoms C, O, N, P, S, H, etc. The R substituent described here may, of course, be unsubstituted or substituted with any and all functionalities.
Examples of such functionalities include NO2, halogen, NR3 where R is defined as above, CN, etc. In this sense the invention methods reside in the causation at the various reaction centers and not in the identity ofthe particular appended substituents. Those of ordinary skill in the art are well suited to carry out the invention reactions as described herein with broad variation in regard to the organic groups surrounding the reaction center in view ofthe guidance provided herein.
Enantiomerically pure 3,3-difluoro-2-oxaazetidine-4-carboxylates were prepared from D-mannitol by the sequence of synthetic steps that is described in Scheme 1. Thus oxidative cleavage ofthe acetone ketal of D-mannitol (see D.Y. Jackson, Synth. Commun. 1988, 18, pp. 337-341), followed by imine formation, afforded 3 which was then treated with zinc dust and ethyl bromodifluoroacetate (see T. Taguchi et al. Tetrahedron Lett. 1988, 29, pp. 5291-5294; J.E. Baldwin et al. J. Chem. Soc, Chem. Commun. 1991, pp. 736-738) to produce 4 as a mixture of diastereoisomers whose ratio was dependent on R1 (Bn, syn/anti = 4.7; p- MeOC6H4CH2, syn/anti = 4.0; Ph2CH, syn/anti - 2.0). Chromatographic separation ofthe syn isomer, alcoholysis, oxidation (see C. Palomo et al. J. Org. Chem. 1997, 62, pp. 2070-2079), and esterification produced the azetidin-2-one product 5 that was deprotected (see J. Podlech & S. Steurer Synthesis, 1999 pp. 650-654) to afford the desired ligand. The key to this' synthesis was the success of a [2+2] cylcoaddition reaction (step d), for which the only deficiency was the stereoisomer ratio with imines whose R1 substituent was removable at a later stage. With these considerations the /r>-methoxybenzyl group was determined to be optimal for the construction of 5 (R1 = H). Although several esters of 5 (R1 = H) were
prepared, only two (R = z'Bu and cHex) were converted by standard procedure (see M.P. Doyle et al. Org. Syn. 1996, 73, pp. 13-24) to dirhodium(II) tetrakis(carboxamidates).
Scheme 1
e, f, g, h
aMe2C(OMe)2, TsOH (46%). δ NaIO4/CH2Cl2-H2O. Ε.'NH^gSO/CH.Cl,. d Zn/THF/TMSCl/BrCF2COOEt (61% from 2).e CH2Cl2/TFA/MeOH/PhMe./Me2CO-H2O (3:1) /NaIO4/KMnO4/HOAc. * tBuOH/pTsOH Dean- Stark trap (57 % from 4). h CAN MeCN- H2O (2:1) with R1 =jo-MeOC6H4CH2 (69%). ' Rh2(OAc)4/PhCl (65% with R2 = /Bu, 73% with R2 = cHex).
The invention catalysts include 2-oxaazetidine-4-carboxylat.es and 2-oxa-pyrrolidine- 5-carboxylates such as those tabulated below.
Table 1 : Representative Invention Complexes
Chiral imidazolidinone ligands were prepared from Ν-CBZ-protected L-asparagine via the Hofinann reaction, esterification, acylation and then deprotection (Scheme 2). Improved procedures for the syntheses of 8 andlOare given in the Examples. Three structurally different imidazolidinones (10) were prepared by procedures that varied with the structure of R.
Scheme 2
Methyl l-[(d and l)-menthoxyacetyl]-2-oxaimidazolidine-4(S)-carboxylates (11) were prepared by the standard acylation-deprotection procedure (equation 1) (see M. P. Doyle et al. Inorg. Chem., 1996, 35, 6064. M. P. Doyle, et al. J. Am. Chem. Soc, 1995, 117, 5763. The analogous methyl l-(l-menthoxyacetoxyacetyl]-2- oxaimidazolidine-4(S)-carboxylate (12) utilized a bromoacetyl intermediate (13)
11
to good effect (Scheme 3). A similar construction to that for the synthesis of 9 was used to prepare the prolinate derivative 11.
Scheme 3
12
The dirhodium catalysts are prepared by refluxing Rh2(OAc)4 in monochlorobenzene with the ligand as previously described (see M.P. Doyle et al., J. Am. Chem. Soc, 1993, 115, p. 9968 - the synthesis ofthe catalysts species described herein incorporated by reference) and isolated as solids.
The dirhodium(II) derivatives of 11, 12 and 14 (c9-cl 1) were prepared from dirhodium(II) acetate by the standard acetate displacement reactions (Scheme 4), and each was characterized spectroscopically. Several transformations were used to test the viability of these catalysts.
Scheme 4
c9 : R = /-menthoxy-CH2 clO : R = d-menthoxy-CH2 el l ,: R = /-menthoxy-CH2COOCH2
The catalysts ofthe invention may be employed in a variety of Lewis acid catalyzed reactions that include but are not limited to Diels-Alder reactions and addition reactions. Preferred reactions include Diels-Alder, hetero-Diels-Alder reactions, Mukiyama-aldol reactions, aldol reactions, epoxide ring opening, diazo decomposition, ylide formation and other reactions. The catalysts are especially useful in hetero-Diels-Alder reactions, especially with the Danishefsky diene.
Lewis acid catalyzed reactions are defined herein as those reactions wherein the reaction rate is accelerated or the stereoselectivity or regioselectivity is improved when the reaction is conducted in the presence of an effective catalyst. The reactions ofthe invention are carried out by bringing the substrate (reactant) material or materials (diene, addition molecule etc) together in the proximity ofthe invention compounds (catalysts). The reactions may be conducted without regard to the order of addition. Therefore the catalyst may be
added to the substrate material(s) or the substrate material(s) may be added to the solid catalyst. Reactions may be carried out in diluting solvents that are polar, non-polar, protic or aprotic. Preferred solvents include hexane, pentane, TiNF, diethylether and water. Alernatively the reactions may be accomplished without the presence of a solvent. The components of a reaction may be heated to initiate interaction ofthe catalyst and substrates materials or cooled to remove excess heat evolved during the course ofthe reaction. Preferred temperatures are from 20-80 °C. The catalyst may be present in any concentration either in solution or heterogeneously. The dirhodium(II) complexes ofthe invention are preferably employed in amounts ranging from about 0.01 to 1 mol% based upon the amount of substrate molecule being reacted. The substrate materials may be present in any range of concentration providing that contact between the catalyst and substrate is sufficient to allow reaction ofthe reaction components. The reaction times may vary depending on the conditions described above. Those of ordinary skill in the art are well suited to carry out the invention reactions as described herein with broad variation in regard to the reaction conditions and substrate materials in view ofthe guidance provided herein and in the Examples.
Enantioselective Lewis acid catalyzed hetero-Diels-Alder reactions can be carried out with the invention complexes. In a typical Diels-Alder reaction a diene is reacted with an olefin to yield a cycloaddition product (eq 4). In this reaction the diene may be substituted with those substituents defined for R, R2 and R3 ofthe invention complexes described above, or unsubstituted. Any conjugated diene is suitable. Any olefin is suitable. Preferred dienes include isoprene, cyclopentadiene, Danishevsky's diene and l-phenyl-l,3-butadiene.
Hetero-Diels Alder reactions involve the reaction of a heterodiene containing a heteroatom functionality such as an aldehyde or ketone. Diels-Alder and hetero-Diels-Alder reactions can be carried out by reacting a diene and an olefin in the presence ofthe invention compounds. The invention complexes are especially useful for providing enantioselective hetero-Diels-Alder reactions products. These reactions may be performed on the same series of dienes as those described above for Diels-Alder reactions. Preferred heteroatom substituted
olefins include benzaldehyde, acrolein, cinnamaldehyde, acrylonitrile and methylmethacrylate.
In a preferred embodiment ofthe invention presented in the reaction below, Danishefsky's diene was reacted with p-nitrobenzaldehyde to yield the cycloaddition products SI -a and Sl-b. Those of ordinary skill in the art can carry out the Diels-Alder and hetero- Diels-Alder reaction, otherwise known as a [4+2] cycloaddition of a diene and olefin, by way ofthe reaction specifics provided in the attendant examples and/or as described in, for example, J. March, Advanced Organic Chemistry, 3rd Ed., John Wiley & Sons, New York (1985), pp. 745-758, incorporated herein by reference, and the references listed therein.
Sl-a Sl-b
Results for various invention catalysts are tabulated below.
Table 3: Enantioselectivity in Catalytic Cycloaddition of -Nitrobenzaldehyde to Danishefsky Diene*
Catalyst yield, %b ee, %c
1 (symanti 2) 1 (config.)
Rh2(OAc)4 67 (100:0) —
Rh2(5R-MEPY)4 53 (100:0) 73 (R)
Rh2(5S-dFMEPY)4 53 (50:50) 78 (S)
Rh2(4S-MEAZ)4 63 (95:5) 56 (S)
Rh2(4S-IBAZ)4 62 (100:0) 66 (S)
Rh2(4R-dFIBAZ)4 68 (60:40) 70 (R)
Rh2(4R-dFIBAZ)4 93d 72 (R)
Rh2(4S-CHAZ)4 54 (100:0) 61 (S)
Rh2(4R-dFCHAZ)4 9&d 76 (R)
Rh2(4S-MACIM)4 76d 74 (S)
Rh2(4S-MPPIM)4 S2d 95 (S) a Unless indicated otherwise, reactions were performed at room temperature in anhydrous CH2C12 using equivalent amounts of reactants and 1.0 moP/o of catalyst with a reaction time of 24 h. b Isolated yield after column chromatography. The symanti ratio was determined by !H NMR after quenching with 5% Et3N in MeOH. c Determined by HPLC using a Chiralpak AD column (hexaneJPrOH = 85:15). d Reactions were performed with a five-fold molar excess of aldehyde.
Surprisingly, the highest degree of enantiocontrol was obtained with the Rh2(4S- MPPIM)4 catalyst which exhibits lower Lewis acidity than complexes containing electron withdrawing fluorine-substituents. The higher degree of enantiocontrol is thought to be due to the addition ofthe second chiral center on the carboxamidate ligand. Varying the
substituent ofthe unsaturated alkene indicated that the electronic properties exert a significant effect on the enantiocontrol ofthe reaction (see table below).
Table 4: Enantioselectivity in Catalytic Cycloaddition of substituted dienes'
The invention catalysts allowed a dramatic and unexpected increase in the substrate- to-catalyst loadings to be realized. Prior art Lewis acid catalysts show substrate-to-catalyst loadings that are normally in the range from 10 to 50 whereas the invention catalysts are able to provide substrate-to-catalyst loadings of as much as 10,000. Results with the Danishefsky
diene, obtained by employing various unsaturated aldehydes and catalysts are presented below. Of particular note is the high enantioselectivity at high substrate-to-catalyst ratios.
Table 5: Substrate-to-catalyst Ratios in hetero-Diels-Alder Reactions of Aldehydes with the Danishefsky Diene°
Diazo decompositions involve the decomposition of a diazo (N2=) moiety of an organic molecule. Diazo decomposition reactions may be carried out by reacting a diazo compund in the presence ofthe invention compounds. Those of ordinary skill in the art can carry out the diazo decomposition reactions by way ofthe reaction specifics provided in the Examples or by way of guidance in "Catalysis Di- and Polynuclear Metal Cluster Compounds" below. The desired reaction product ofthe diazo decomposition may be a cyclopropanation product (U.S. Patent 5,175,311) or other addition of cylcoaddition product, see also M.P. Doyle in "Catalysis by Di- and Polynuclear Metal Cluster Complexes, R.D. Adams and F.A. Cotton, Ed.s., VCH Publishers, New York, 1998, chapter 17.
Dimetallic Rh(II) complexes containing difluorinated carboxamidate ligands show increased reactivity towards the decomposition of vinyl diazo compounds such as of 2- methyl-2-propenyl-l-yl phenyldiazoacetate. Carrying out the decomposition shown in
scheme 5, the difluoro complex provided the cyclopropanation product in approximately the same yield and enantiomeric excess as that ofthe nonfluorinated catalyst although the fluorinated catalyst yielded the R isomer. Surprisingly, however, it was found that when equal amounts of catalyst were used and the reaction conditions were the same, the fluorinated analog was demonstrated to react at least eight times faster than the nonfluorinated catalyst.
Scheme 5
(1R,5S) 68%, ee
This enhanced reactivity was also demonstrated for the pyrrolidinone complexes where reaction ofthe nonfluorinated catalyst with 2-mefhyl-2-propen-l-yl did not occur at room temperature whereas reaction ofthe catalyst containing the fluorinated ligand occurred within minutes.
In the diazo decomposition reaction of ethyldiazoacetate by to yield ylide formation and subsequent [2, 3] sigmatropic rearrangement of and allyl iodide (eq 6) the fluorinated catalysts were able to provide the end product in greater enantiomeric excess.
N
2CHOOOEt +
20 21
Table 6: Enantioselectivity in Catalytic Diazodecomposition.
"Reaction performed in refluxing CH2C12 with 1.0 mol% of catalyst and 10 equiv of allyl iodide. *Weight yield after chromatography.
Those of ordinary skill in the art can carry out the addition reaction in view ofthe specifics provided in the Examples and as described in J. March Advanced Organic Chemistry, 3rd Ed., John Wiley & Sons, New York, 1985 chapter 15. Hydrosilation is an addition reaction across a Si-H bond.
With cyclohexyl diazoacetate, which offers an estimate of diastereoselectivity as well as enantioselectivity equation 2, remarkable differences were observed with catalysts c9-cl6
(Table 2). Here there is a demonstrable difference between clO and c9 in their stereochemical influence on selectivity, with clO showing a significant enhancement in enantiocontrol. Similar results were not seen with cl 1.
15 16 17
With 3-phenylpropyl diazoacetate, (see J. W. Bode et al. J. Org. Chem., 1996, 61, 9146) carbon-hydrogen insertion gave one product equation 3. Enantioselectivity in this case was enhanced with c9 (82% ee) but diminished with clO (27% ee).
18 19 ',
The outcome of these reactions demonstrates that remote chiral attachments on a carboxamidate ligand to dirhodium have a significant influence on enantiocontrol,
With diastereoisomers c9 and clO virtually opposite influences are observed in the outcomes of reactions 2 and 3. In contrast, el l appears to have destroyed any significant enantiocontrol.
Table 2. Diastereoselectivity and enantioselectivity in carbon-hydrogen insertion reactions from diazo decomposition of cyclohexyl diazoacetatea
15+16 %ee %ee
catalyst yield, %b 15 : 16 15 16
Rh2(OAc)4 14 46:54
Rh2(4S-MLMIM)4 (12) 64 81:19 53 16
Rh2(4S-MDMIM)4 (13) 33 91:9 >9 67
Rh2(S,S-MAOIM)4 (14) 53 76:24 0 nd
Rh2(S,S-BOPCI)4 (15) 80 54:46 6 10
Rh2(R,S-BOPCI)4 (16) 83 42:58 14 13
a Reactions were performed in refluxing CH2C12 as previously described (ref. 8 and 10). b Yield after chromatographic separation ofthe catalyst.
Addition reactions are among the most common and widely employed reactions in organic synthesis. In a typical addition reaction an unsaturated carbon-carbon bond is reacted with an insertion molecule XY to effect the addition ofthe XY molecule across the unsaturated carbon-carbon bond (eq 8). The addition typically yields a racemic mixture, see J. March, Advanced Organic Chemistry, 3rd Ed., John Wiley & Sons, New York, 1985, chapter 15 incorporated herein by reference. XY can be any hetero or homo atom moiety where X and Y are atoms ofthe main group or transition metals ofthe periodic table ofthe elements. XY may be singly or multiply bonded and may be substituted with any organic or organometallic combination of atoms.
C
II + X-Y — C— C- (8) .C
X Y
Hydrosilation is achieved by the insertion of an organic moiety into a Si-H bond (U.S. Patent 5,296,595). Hydrosilation reactions can be catalyzed by reacting a diazo compound and a Si-H containing compound in the presence ofthe invention compounds. The dirhodium catalysts containing fluorinated ligands have further established their higher reactivity in the hydrosilation reactions shown in equation 7.
Reaction times are an order of magnitude less with Rh2(5S-dFMEPY)4 than with Rh2(5S-MEPY)4, enantiocontrol is diminished. However, when tri(isobutyl)silane is used in place of phenyldimethylsilane, enantiomeric excesses for the Si-H insertion product are comparable.
EXAMPLES
General Experimental Details: Danishefsky's diene was purchased from Lancaster, and all other chemicals purchased from Aldrich and used without purification. Dichloromethane was distilled from CaH2 prior to use tefrahydrofuran was distilled prior to use from sodium and benzophenone; all other solvents were used without further purification. !H NMR and 13C
NMR spectra were obtained as solutions in CDC13, and chemical shifts are reported in parts per million (ppm, δ) downfield from the internal standard, Me4Si (TMS), using either an AM-250, DRX-400, DRX-500 or DRX-600 NMR spectrometer. Optical rotations were measured using a JASCO DIP- 1000 digital polarimeter. N-Fluorobenzenesulfonimide was recrystallized from ethyl acetate :hexanes prior to use.
Dirhodium(II) tetrakis [methyl l-(/-menthoxyacetyl)-2-oxaimidazolidine-4(S)- carboxylate], Rh2(4S-MLMIM)4. Rhodium(II) acetate (0.265 g, 0.60 mmol), methyl 1-/- menthoxyacetyl-2-oxaimidazolidine-4(S)-carboxylate (2.03 g, 5.96 mmol), and 20 mL of anhydrous chlorobenzene were mixed in a round bottom flask fitted with a Soxhlet extraction apparatus into which was placed a thimble containing an oven-dried mixture of two parts
Na2CO3 and one part sand. The resulting blue-green solution was heated at reflux under nitrogen for about 22 h. The progress of ligand displacement was followed by HPLC (m- Bondapak-CN column, MeOH:CH3CN = 92:2, flow 1.5 niL/min). The initial Rh2(OAc)4 band disappeared and was replaced by several bands with longer retention volumes. After 22 h one principal band (retention time 6.3 min.), in addition to that for the ligand (retention time 2.0 min), was observed. The resulting blue solution was cooled to room temperature, and the solvent was evaporated under reduced pressure. The residue in a minimal volume of MeOH was chromatographed on reverse phase silica (Bakerbond Cyano 40 mm prep LC packing) eluting with MeOH after removing gray decomposition products by filtration. The first slightly brown band containing excess of ligand (0.990 g) was removed. The following two purple bands were collected and concentrated under reduced pressure. The first purple band contained the title compound (1.12 g), which was recrystallized from MeOH/CH3CN. A second recrystallization yielded pure purple crystals. The second purple band contained an unidentified side product (0.121 g) presumed to be the [3,l]-isomer. The yield of recrystallized purple crystals 0.645 g (0.39 mmol, 65%). 1H NMR (300 MHz), δ: 4.55
(comp, 8 H); 4.12-3.88 (comp, 12 H); 3.78 (s, 6 H); 3.62 (s, 6 H); 3.24 (comp, 4 H); 2.38 (s,
6 H); 2.32 (comp, 4 H); 2.17 (comp, 4 H); 1.65 (comp, 8 H); 1.32 (comp, 8 H); 0.93 (comp, 36 H); 0.83 (comp, 12 H); 13C NMR (75 MHz, CDC13), δ: 172.9; 172.4; 168.8; 168.6; 164.6; 115.3; 79.5; 79.4; 67.5; 67.4; 59.9; 59.5; 52.4; 51.8; 48.5; 48.3; 46.5; 46.2; 40.1; 34.3; 31.5; 31.4; 25.3; 23.1; 22.3; 22.2; 21.0; 16.1; 2.22. [α]D 21-7 = -287.9 (c = 0.51, CH3CN).
General experimental procedure for the cycloaddition reaction of 4-nitrobenzaldehyde with Danishefsky's diene. To a solution of Danishefsky's diene (0.057 g, 0.33 mmol) in dry dichloromethane (2 mL) was added 4-nitrobenzaldehyde (0.050 g, 0.33 mmol) and the dirhodium catalyst (1 mol%, 3.3 mol) and the reaction stirred at room temperature for 24h. Afterwhich, a small sample was taken from the reaction mixture and added to a 5% solution of triethylamine in methanol (1 mL) and then evaporated to give a mixture ofthe acetals 2 whose ratio was determined from !H NMR spectroscopy. To the remaining solution trifluoroacetic acid was added and the reaction stirred for 2 min and then quenched with aq sodium hydrogen carbonate solution. The resulting mixture was extracted with dichloromethane dried (MgSO4) and the dichloromethane removed under reduced pressure to yield the crude dihydropyranone 1 which was purified by column chromatography
(Dichloromethane as eluent). The enantiomeric excess ofthe pure dihydropyranone 1 was determined by HPLC on a Chiralpak AD column with peaks at 12.59 and 16.82 min [flow 1.5 mL/min, Hexane: TrOH (85:15), λmax = 254 nm].
Exp. 1:
a) Cycloaddition reaction in the presence of Rh2(OAc)4
The reaction with Rh2(OAc)4 (1.50 mg, 1 mol%) yielded the dihydropyranone 1 (0.046 g, 67% yield).
b) Cycloaddition reaction in the presence of Rh2(5R-MEPY)4
The reaction with Rh2(5R-MEPY)4 (2.83 mg, 1 mol%) yielded the dihydropyranone 1 (0.037 g, 53°,% yield) and was shown to possess an ee of 80%.
c) Cycloaddition reaction in the presence of Rh2(4S-MEAZ)4
The reaction with Rh2(4S-MEAZ)4 (2.60 mg, 1 mol%) yielded the dihydropyranone 1 (0.043 g, 63% yield) and was shown to possess an ee of 59%.
d) Cycloaddition reaction in the presence of Rh2(4S-IBAZ)4
The reaction with Rh2(4S-IBAZ)4 (3.00 mg, 1 mol%) yielded the dihydropyranone 1 (0.043 g, 62% yield) and was shown to possess an ee of 69%.
e) Cycloaddition reaction in the presence of Rh2(4S-CHAZ)4
The reaction with Rh2(4S-CHAZ)4 (3.30 mg, 1 mol%) yielded the dihydropyranone 1 (0.038 g, 54% yield) and was shown to possess an ee of 65%.
f) Cycloaddition reaction in the presence of Rh2(4R-dFIBAZ)4
The reaction with Rh2(4R-dFIBAZ)4 (3.50 mg, 1 mol%) yielded the dihydropyranone
1 (0.047 g, 68% yield) and was shown to possess an ee of 84%.
g) Cycloaddition reaction in the presence of Rh2(5S-dFMEPY)4
The reaction with Rh2(4S-dFMEPY)4 (3.04 mg, 1 mol%) yielded the dihydropyranone 1 (0.037 g, 53% yield) and was shown to possess an ee of 82%.
h) Cycloaddition reaction in the presence of Rh2(2S-DOSP)4
The reaction with Rh2(2S-DOSP)4 (6.00 mg, 1 mol%) yielded the dihydropyranone 1 (0.047 g, 68% yield) and was shown to possess an ee of 20%.
i) Cycloaddition reaction in the presence of Rh2(2S-TBSP)4
The reaction with Rh2(2S-TBSP)4 (4.63 mg, 1 mol%) yielded the dihydropyranone 1 (0.042 g, 61 % yield) and was shown to possess an ee of 16%.
Exp. 2:
a-d) The procedure for Exp. 1 was followed but after 2 and 4 hr a sample was removed the solvent evaporated and the crude Η NMR spectrum recorded to determine the progress of the reaction.
e) The procedure for Exp. 1 was followed but using 0.1 mol% Rh2(4R-dFIBAZ)4 and stirred for 24h. The pyranone 1 was shown to possess an ee of 78%) using the previously described method (see Exp. 1). f) The prodecdure for Exp. 1 was followed but the reaction was carried out at 0°C and quenched after 6h. HPLC Analysis, using the previously described method (see Exp. 1), showed the product to possess an ee of 94%.
Exp.3a)
a) The procedure for Exp. 1 was followed but using S-vitro-2-finfuraldehy.de (4.041 g, 0.290 mmol) and Danishefsky diene (O.OSO g, 0.290 mmol) to yield the dihydropyranone (0.029 g, 48 %) which was purified by column chromatography (Dichloromethane as eluent). The enantiomeric excess ofthe pure dihydropyranone 1 was determined by HPLC on a Chiralpak AD column with peaks at 18.11 sand 27.38 min [flow 1.0 mL/min, Hexane: rOH (8S:1 S), λmax = 254 run].
b) The procedure for Exp. 1 was followed using 4-vitro-traps-cinnamaldehyde (0.050 g, 0.282 mmol), Danishefsky's diene (4.049 g, 0.282 mmol) and Rh2(OAc)4 (1.5 mg, 1 mol%) to yield the dihydropyranone (4.024 g, 35%) which was purified by column chromatography (Dichloromethane as eluent).
c) The procedure for Exp. 1 was followed using 4-vitro-traps-cinnamaldehyde (4.050 g, 0.282 mmol), Danishefsky's diene (0.049 g, 0.282 mmol) and Rh2(4R-dFIBAZ)4 (3.00 mg, 1 mol%)to yield the dihydropyranone (0.024 g, 35%) which was purified by column chromatography (Dichloromethane as eluent). The enantiomeric excess ofthe pure dihydropyranone 1 was determined by HPLC on a Chiralcel OD column with peaks at 35.37 and 44.91 min [flow 1.0 mL/min, Hexane:rPrOH (8S:1 S), λmax = 254 nm].
d) The procedure for Exp. 1 was followed but using ethyl glyoxalate (50% solution in toluene, 0.118 mL, 0.582 mmol), Danishefsky's diene (0.050 g, 0.291 mmol) and
Rh2(4R.dFIBAZ)4 (3.00 mg, 1 mol%) and the reaction stirred for 3 days, to yield the dihydropyranone (0.021 g, 43%). The enantiomeric excess of the pure dihydropyranone 1
was determined by GC on a Chiraidex γ-TA column with peaks at 8.30 and 8.99 min [flow 1.0 mL/min, temp = 160°C isotherm]. The pyranone Z possessed an ee of 54%.
Exp. 4a)
a) The procedure for Exp. 1 was followed but using benzaldehyde (0.050 g, 0.471 mmol), Danishefsky's diene (0.081 g, 0.471 mmol) and Rh2(OAc)4 (2.08 mg, 1 mol%) to yield the dihydropyranone (0.038 g, SO %) which was purified by column chromatography (Dichloromethane as eluent).
b) The procedure for Exp. 1 was followed but using benzaldehyde (0.050 g, 0.471 mmol), Danishefsky's diene (0.081 g, 0.471 mmol) and Rh2(SR-MEPY)4 (4.04 mg, 1 mol%) to yield the dihydropyranone (0.018 g, 23 %) which was purified by column chromatography
(Dichloromethane as eluent). The enantiomeric excess ofthe pure dihydropyranone 1 was determined by GC on a Chiraldex β-BM column with peaks at 36.59 and 42.68 min [flow 1.0 mL/min, temp =150°C isotherm]. The pyranone 1 possessed an ee of 44%.
c) The procedure for Exp. 1 was followed but using benzaldehyde (0.00 g, 0.471 mmol), Danishefsky's diene (0.081 g, 0.471 mmol) and Rh2(4S-MEAZ)4 (3.78 mg, 1 mol%) to yield the dihydropyranone (0.033 g, 43 %) which was purified by column chromatography (Dichloromethane as eluent). The enantiomeric excess ofthe pure dihydropyranone 1 was determined by GC on a Chiraldex β-DM column with peaks at 36.59 and 42.68 min [flow 1.0 mL/min, temp =150°C isotherm]. The pyranone 1 possessed an ee of 27%.
Table 7: Enantioselectivity of Cycloaddition Reactions.
Reagents: 1) 1 Mol% catalyst, CH2C12, 24h, r.t.; ii) TFA; iii) 5% Et3N/MeOH. "Isolated yield after column chromatograph, yields are unoptimised. feEe's were determined by HPLC using a Chiralpak AD column. Tndicated the ratio of syn to anti glycosides after quenching the reaction with 5% tiiethylamine in methanol and was detemined from !H NMR spectroscopy.
*Epimerization ofthe C, center is often facilitated by stronger Lewis acids. eAfter 24h only the pyranone was present in the reaction mixture indicative ofthe electrophilicity ofthe dirhodium core.
Table 8: Enantioselectivity of Cycloaddition Reactions.
Εe's were determined by HPLC using a Chiralpak AD column. ^Indicates the ratio of aldehyde to product in the crude reaction determined by 1H NMR spectroscopy. The pyranone 1 was also present in significant amounts and was incorporated into the % conversion. ^Reaction was carried out at 0°C.
Table 9: Enantioselectivity of Cycloaddition Reactions.
"Isolated yield after column chromatograplhy, yields are unoptimized. έEe's were determined by HPLC using a Chralpak AD column. Εe's were determined by HPLC using a Chralcel OD column. *Ee's were determined by GC using a Chraldex γ-TA column at 160°C isotherm.
Table 10: Enantioselectivity of Cycloaddition Reactions.
Εe's were determined by GC using a Chraldex β-BM column at 150°C isotherm.
l,2:5,6-Diisopropylidene-D-mannitol: A mixture of powdered D-mannitol (28 g, 0.154 mol), ptoluenesulfonic acid (0.5 g) and 2,2-dimethoxypropane (46 mL, 0.375 mol) in dry Me2SO (50 mL) was stirred at room temperature under nitrogen. Within 1 hour the suspended solids had dissolved, and after 16 hours the reaction solution was poured into 150 mL of 3% NaHCO3. The mixture was extracted with ethyl acetate (3 x 200 mL), and the extracts were concentrated under reduced pressure until a solid mass occurred. Hexanes (300 mL) was added, and the mixture was dissolved under refluxing conditions. The solution was allowed to cool slowly overnight. The resulting crystalline material was collected by filtration and then washed with 100 mL of cold etherlhexanes (1 :20) and dried to give the diacetonide
(18 g, 0.07 mol) in 46% yield. Η NMR (250 MHz, CDC13) δ 4.23-4.05 (m, 1-H, 4H), 3.97 (dd, J = 8.1, 5 .6Hz, 2H), 3 .74 (t, J = 6.2 Hz, 2H), 2.63 (br s, 2H), 1.41 (s, 6H), 1.3 5 (s, 6H); ,3C NMR (62.5 MHz, CDC13) δ 109.4, 76.3, 71.2, 66.7, 26.7, 25.2.
(4R)-l-(4,-Methoxylbenzyl)-3,3-difluoro-4-(l»R-l',2'-O-isopropylideneethyI)azetidiπ-2- one: To a stirred solution of 2 (5.0 g, 19.0 mmol) in CH2C12 (50 mL) was added sodium periodate (8.1 g, 38.0 mmol) followed by addition of saturated sodium bicarbonate (2.0 mL). The mixture was stirred at room temperature for 2 hours. Magnesium sulfate (3 g) was added and stirring was continued for 20 min. The slurry was filtered through a plug of celite to give a CH2C12 solution of crude 2,3-O-isopropylidene-D-glyceraldehye (3). [7] To the above solution was added anhydrous magnesium sulfate (10 g) then pmethoxybenzylamine (4.69 g, 34 mmol) during 30 min. After an additional 30 min the resulting slurry was filtered, and the filtrate was washed with CH2C12 (2 x 20 mL). Imine that was recovered after the solvent
evaporated was subjected to the Reformatsky reaction according to the literature procedure, and the resulting product was fully characterized. The syn to anti ratio was determined by !H NMR, and the syn-lactam 4 was isolated in 61% yield (7.6 g, 23.2 mmol). [α]D25 = -93.8 (c 1.5, CH2C12); Η NMR (250 MHz, CDC13) δ 7.26-7.21 (comp, 2H), 6.90-6.71 (comp, 2H), 4.85 (d, J= 14.7 Hz, IH), 4.27-4.16 (comp, 3H), 3.38 (s, 3H), 3.76-3.68 (comp, 2H), 1.38 (s, 3H), 1.35 (s, 3H); 13C NMR (62.5 MHz, CDC13) δ 159.9, 130.1, 126.1, 119.9 (t, JF.C = 292 Hz), 114.1, 110.5, 74.8, 66.3 (t, JF.C = 23.7 Hz), 66.1, 55.3, 44.8 (t, JF.C = 2.8 Hz), 26.5, 24.9.
(4R)-l-(4'-Methoxylbenzyl)-3,3-difluoro-4-(isobutyloxycarbonyl)azetidin-2-one: A solution of γ-lactam 4 (7.6 g, 23.2 mmol) in 166 mL of 6:3:1 CH2Cl2/TFA/CH3OH (100:50: 16 mL) was stirred at room temperature for 1 h (step e, Scheme 1). The solvent was evaporated under vacuum, and to the residue was added toluene (50 mL) which was evaporated to remove residual TFA. The residue was dried thoroughly under vacuum, then a mixture of 320 mL of 3:1 acetone H20 (240:80 mL) was added to dissolve the residue, then NaIO4 (9.93 g, 46.4 mmol) was added (step f, Scheme 1). The mixture was stirred at room temperature for 2 hours. KMnO4 (10.6 g, 93.6 mmol) was added, followed by acetic acid (10 mL). The resulting red mixture was stirred at room temperature for 3 h then filtered, and the filter cake was washed with acetone until the acetone eluent was colorless. The aqueous eluent was extracted with ethyl acetate (2 x 200 mL). The resulting organic layer was washed with brine (50 L), dried over anhydrous MgSO4, and the solvent was evaporated under reduced pressure to give crude acid. To the crude acid was added benzene (150 mL), 2-methyl-l-propanol (20 mL) and ptoluenesulfonic acid (0.43 g). The resulting solution was refluxed (step g, Scheme 1) until water was no longer evolved in a Dean-Stark apparatus. Solvent was evaporated and the residue was subjected to column chromatography on silica gel (hexanes: ethyl acetate = 20:1) to afford product as a colorless viscous oil (4.3 g, 13.2 mmol) in 57% yield. [ ]D29 = +37.4 (c 0.4, CH3CN); !H NMR (500 MHz, CDC13) δ 7.15 (d,
J = 8.5 Hz, 2H), 6.89 (d, J = 8.5 Hz, 2H), 4.93 (d, J = 15.0 Hz, IH), 4.25 (dd, JF.H = 6.8, 2.5 Hz, IH), 4.22 (dd, J= 15.0, 1.8 Hz, IH), 4.05 (dd, J= 10.5, 7.6 Hz, IH), 3.98 (dd, J= 10.5, 7.6 Hz, IH), 3.82 (s, 3H), 2.01-1.91 (m, IH), 0.96 (d, J = 6.6 Hz, 6H); 13C NMR (125 MHz, CDC13) δ 164.8, 159.8, 159.3 (t, JF.C = 30 Hz), 130.0, 126.1, 124.7, 119.5 (t, JF.C = 292 Hz), 114.6, 72.4, 64.0 (t, JF.C = 25.3 Hz), 55.3, 44.7, 27.6, 18.8.
2-MethyI-I-propyl 3,3-Difluoro-2-oxaazetidine-(4R)-carboxylate: To the above N- protected azetidinone (2.7 g, 8.3 mmol) was added 75 mL of 2:1 CH3CN/H2O (50:25 mL). The resulting solution was cooled to 0°C under an ice- water bath. Ceric ammonium nitrate (18.1 g, 33.2 mmol) was added to the mixture which was stirred at the same temperature for 5 1 h.[10] The solution was allowed to warm to room temperature, and stirring was continued for another 3 h. The reaction mixture was extracted with ethyl acetate (2 x 200 mL). The combined organic layer was washed with water (30 mL) and brine (50 mL), dried over anhydrous MgSO4, and the solvent was evaporated under reduced pressure. The residue was subjected to column chromatography on silica gel (hexanes: ethyl acetate = 10:1) to afford 10 product as a colorless oil ( 1.18 g, 5.7 mmol) in 69% yield. [α]D25 = +26.6 (c 5.0, CH3CN); Η NMR (250 MHz, CDC13) δ 7.40 (br s, IH), 4.60 (dd, JF.H = 6.5, 3.5 Hz, IH), 4.09 (dd, J= 10.5, 6.9 Hz, IH), 3.98 (dd, J = 10.5, 6.6 Hz, IH), 2.01-1.91 (m, IH), 0.93 (d, J = 6.8 Hz, 6H); 13C NMR (62.5 MHz, CDC13) δ 165.6, 160.3, 120.9 (t, JF.C = 292 Hz), 72.7, 62.2 (t, JF.C = 24.8 Hz), 27.6, 18.7; MS(FAB), m/z 208 (MH)+.
15 Dirhodium(II) Tetrakis[2-methyl-I-propyl 2-oxa-3,3-difluoroazetidine-4(R)- carboxylate]: Prepared and purified in 65% overall yield by standard methods: [11,13,14] [oc]D25 = +250.8 (c 0.43, MeCN); found: C, 38.3; H, 4.1; N, 6.9. C32H40N4O12F8Rh2«CH3CN requires C, 38.1; H, 4.1; N, 6.5; Η NMR (250 MHz, CDC13) δ 4.91-4.89 (m, 2H), 4.43-4.41 (m, 2H), 4.10-3.92 (m, 8H), 2.02-1.86 (m, 4H), 0.96 (d, J= 6.6 Hz, 12H), 0.92(d, J= 6.8 Hz,
20 12H); 13C NMR (62.5 MHz, CDC13) δ 168.2, 167.2, 72.2, 72.1, 66.4, 66.1, 52.7, 27.7, 27.5,
18.7; MS (FAB) m z 1031 ([MH]+, 100%); HRMS (FAB) for (C32H40N4O12F8Rh2H)+, m/z calcd. 1031.0703, found 1031.0712.
2,2-Dimethyl-7-fiuoro-8-oxo-l-aza-3-oxabicyclo [3.3.0] octane: Prepared by a modification to the method reported by Coward and Konas.[16] To a cooled stirred solution of »5 diisopropylamine (5.4 mL, 41.9 mmol) in dry THF (30 mL) at -78°C, n-butyllithium in hexanes (14.2 mL, 35.5 mmol) was added via syringe pump over a 1 h period and the resulting mixture stirred for 1 h. A solution of (5S)-2,2-dimethyl-8-oxo-l-aza-3-oxabicyclo- [3.3.0]octane[18] (5.00 g, 32.3 mmol) in dry THF (20 mL) was added over a 1 h period to the
LDA mixture during which time an orange color developed. After 2 h, N- fluorobenzenesulfonimide (11.2 g, 35.5 mmol) in dry THF (30 mL) was added dropwise over a 1 h period, and the reaction mixture was stirred for a further 1 h. The reaction mixture was then quenched with aqueous saturated ammonium chloride (20 mL) and allowed to warm to 5 room temperature, then THF was removed under reduced pressure. The resulting aqueous solution was extracted with dichloromethane (3 x 50 mL), the organic layers were combined and dried (MgSO4), and the solvent was removed under vacuum to yield a yellow oil which was subjected to silica gel column chromatography [EtOAc:hexanes (3:2) as eluent]. The title compound was isolated as a yellow oil (5.22 g, 94%) which was a mixture of 10 diastereoisomers: 'H NMR (600 MHz, CDC13) syn-isomer 8 5.31 (ddd, J = 52, 10, 8 Hz, 1
H), 4.19 (dd, J = 16, 6 Hz, 1 H), 3.93-3.98 (m, 1 H), 3.50 (dd, J = 9, 9 Hz, 1 H), 2.79 (ddd, J= 13, 8, 6 Hz, IH), 1.87-1.96 (m, IH), 1.71 (s, 3H), 1.48 (s, 3H); anti-isomer 8 5.08 (dd, J = 52, 6 Hz, 1 H), 4.46 (m, 1 H), 4.15 (dd, J = 8, 6 Hz, 1 H), 3.42 (dd, J = 9, 9 Hz, 1 H), 2.42 (ddd, J = 21, 15, 6 Hz, IH), 1.96-2.06 (m, IH), 1.66 (s, 3H), 1.52 (s, 3H).
15 (5S)-2,2-Dimethyl-7,7-difluoro-8-oxo-l-aza-3-oxabicyclo[3.3.0]octane: Prepared as described for the monofluorination procedure using the 2,2-dimethyl-7-fluoro-8-oxo-l-aza-3- oxabicyclo[3.3.0]octane (4.95 g, 28.6 mmol), column chromatography [chloroform:EtOAc (4:1) as eluent] yielded the title compound (4.37 g, 80%) as a white solid: mp 35-37 °C (from ethyl acetate/light petroleum); [α]D28 = +89.2 (c 0.85, CHC13); 1H NMR (500 MHz, CDC13)
20 δ 4.26 (dd, J = 9, 6 Hz, IH), 4.13 (m, IH), 3.47 (dd, J= 9, 9 Hz, IH), 2.77 (ddd, J= 15, 15, 6
Hz, IH), 2.13 (dddd, J= 27, 15, 13, 8 Hz, IH), 1.68 (s, 3H), 1.54 (s, 3H);13C NMR (100 MHz; CDC13) δ 159.8 ( dd, JC.F = 33.0, 28.4 Hz), 121.2 (dd, JC.F = 256.9, 251.6 Hz), 92.4, 69.8, 54.0, 35.1 (dd, JC.F = 24.7, 21.7 Hz), 26.3, 23.3; 19F NMR (376 MHz; CDC13, C6F6 external standard) δ 59.77 (dd, JF.F = 264.4 Hz, JH.F = 12.9 Hz), 58.44 (ddd, JF.F = 264.0 Hz,
!5 JH.F = 26.8, 14.4 Hz); MS (Cl, NH3) m/z 192 ([MH]+, 100%).
(5S)-3,3-Difluoro-5-(hydroxymethyl)-2-pyrrolidinone: To a solution ofthe (5S)-2,2- dimethyl-7,7difluoro-8-oxo-l-aza-3-oxabicyclo[3.3.0]octane (1.91 g, 10.0 mmol) in a 1:1 mixture of dioxane and water (60 mL), Amberlite IR-120 (H1") resin (1.91 g) was added, and
the reaction mixture was heated at reflux for 4 h. After cooling to room temperature, the Amberlite resin was removed via filtration, washed with water (25 mL) and the solvent was removed under reduced pressure to give the crude product which was purified by silica gel chromatography [dichloromethane:methanol (5:1) as eluent]. The title compound (1.38 g, 91%) was isolated as a white crystalline solid: mp 144-145 °C (from methanol/dichloro ethane); [α]D21 = +30.6 (c 1.01 in MeOH); found: C, 39.5, H; 4.5; N, 9.1; C5H7F2NO2 requires C, 39.7; H, 4.7; N, 9.3%. Η NMR (600 MHz, CD3OD) δ 3.74-3.80 (m, IH), 3.60 (dd, J= 11, 4 Hz, IH), 3.48 (dd, J= 11, 5 Hz, IH), 2.59-2.66 (m, IH), 2.31-2.40 (m, IH); 13C NMR (100 MHz, DMSO-d6) δ 165.1 (dd, JC.F = 30.2, 30.2 Hz), 1 19.4 (dd, JC.F = 247.4, 245.7 Hz), 63.2, 50.2, 32.6 (dd, JC.F = 22.0, 21.9 Hz); 19F NMR (376 MHz, DMSO-d6, C6F6 external standard) δ 60.75-59.96 (br ), 59.23 (ddd, JF-F = 265.5 Hz, JH.F = 16.1, 2.3 Hz); MS (Cl, NH3) m/z 169 ([M+NH4]+, 100%).
Methyl (5S)-3,3-difluoro-2-pyrrolidinone-5-carboxylate: Freshly prepared Jones reagent was added in portions to a solution of (5S)-3,3-difluoro-5-(hydroxymethyl)-2-pyrrolidinone (0.48 g, 3.20 mmol) in wet acetone (25 mL), and the reaction was monitored by TLC. Upon complete consumption of starting material, 2-propanol (10 mL) was added carefully and the reaction mixture was stirred for 1 h. The reaction mixture was diluted with water (40 mL), and the organic solvents were removed in vacuo. The aqueous residue was adjusted to pH 3.0 with NaHCO3 and continuously extracted with ethyl acetate for 2 days. The organic phase was concentrated under vacuum and the crude acid residue taken up in ether (100 mL). To this solution was added a freshly prepared solution of diazomethane in ether in portions, and upon complete consumption of intermediate acid, as judged by TLC, the reaction was quenched by addition of acetic acid (2 mL). The reaction mixture was concentrated in vacuo to give the crude product. Flash chromatography (30:5:65, ethyl acetate/methanol/light petroleum) gave the title compound as an off-white oil (0.47 g, 83%); [ ]D25 = +1.5 (c 1.02 in CHC13); IR (film) cm'1 3282 (NH), 2958, 1751, 1732; 'H NMR (400 MHz, CDC13) δ 7.85 (br s, IH), 4.35 (ddd, J= 7.6, 5.0, 2.1 Hz, IH), 3.50 (s, 3H), 2.94-2.81 (m, IH), 2.71-2.59 (m, IH); 13C NMR (100 MHz, CDC13) δ 170.0, 165.6 (dd, JC.F = 30.7, 30.7 Hz), 116.5 (dd, JC.F = 250.1, 248.6 Hz), 53.2, 50.0, 34.1 (dd, JC.F = 24.7, 24.0 Hz); 19F NMR (376 MHz, CDC13, C6F6 external standard) δ 56.51-55.70 (m), 54.70 (dddd, JC.F = 273.0 Hz, JH.F = 16.7, 14.4, 2.3
Hz); MS (Cl, NH3) m/z 197 ([C6HπN2O3F2+NH4]+, 100%); HRMS (Cl) for [C6HπN2O3F2+NH4]+: m/z calcd 197.0738, found 197.0735.
Dirhodium(II) Tetrakis [methyl 3,3-difIuoro-2-oxapyrroIidine-(5S)-carboxyϊate] :
Prepared and purified in 65% overall yield by standard methods: [11,13,14] nip > 275 °C (from ethyl acetate/hexane); [α]D24 - -326.8 (c 0.12, CH3CN); found: C, 30.3; H, 2.8; N 5.5,
C24H24FgN4O12Rh2-2H2O requires C, 30.2; H, 3.0; N, 5.8%; IR (KBr) cm"1 3448 (OH), 2960, 2924, 2852, 1749, 1653; 'H NMR (500 MHz, CDC13) δ 4.28-4.30 (m, 2 H), 4.14-4.16 (m, 2H), 3.81 (s, 6H), 3.73 (s, 6H), 2.53-2.72, (m, 6H), 2.30 (ddd, 29, 15, 4.5 Hz, 2 H); 13C NMR (100 MHz, DMSO-d6) δ 178.2 (dd, JC.F = 27.8, 27.7 Hz), 177.6 (dd, JC.F = 27.5, 27.5 Hz), 173.2, 172.6, 118.6 (dd, JC.F = 247.9, 245.4 Hz), 118.5 (dd, JC.F = 249.5, 246.6 Hz), 58.5,
58.0, 52.5, 52.0, 36.3 (dd, JC.F = 24.5, 23.9 Hz), 35.5 (dd, JC.F = 23.9, 23.3 Hz); 19F NMR (376 MHz, DMSO-d6, C6F6 external standard) δ 66.14 (dddd, JF.F = 256.3 Hz, JH.F = 21.6, 18.6, 4.1 Hz), 64.47 (ddd, JF.F = 256.6 Hz, JH.F = 19.3, 4.7 Hz), 63.69 (ddd, J = 17.7, 15.2, 2.7 Hz); MS (FAB) m/z 917.7. HRMS (FAB) for [C24H24N4O12F8Rh2]+, m/z calcd 917.9373, found 917.9377.