Carbanion-Accelerated Claisen Rearrangements Asymmetric Induction with Chiral Phosphorus-Stabilized Anions

Abstract

The carbanion-accelerated Claisen rearrangement has been extended to include phosphorus carbanion-stabilizing groups. The appropriately substituted allyl vinyl ethers are synthesized by the nucleophilic addition of allyloxides to phosphorus-substituted allenes, which are obtained in one step from simple starting materials. The phosphorus-stabilized, carbanion-accelerated Claisen rearrangements proceed rapidly at room temperature in high yield, and the rearrangements are highly site and stereoselective. The first examples of asymmetric induction in the Claisen rearrangement with chiral, phosphorus, anion-stabilizing groups are described. The observed asymmetric induction is highly dependent on the structure of the auxiliary and the metal counterion involved. Both internal and relative diastereoselectivity are high. A model for the observed sense of internal diastereoselectivity is proposed that is founded in the current understanding of the structure of phosphorus-stabilized anions.

Introduction

The aliphatic Claisen rearrangement of allyl vinyl ethers enjoys recognition as one of the preeminent members of the important class of [3,3]-sigmatropic rearrangements. The vast literature on all aspects of the reaction continues to grow documenting new variations, synthetic applications, and mechanistic and theoretical details.¹ Synthetically, the most commonly employed variants are the thermally induced processes illustrated in Scheme 1 which produce a number of γ,δ-unsaturated carbonyl derivatives with varying degrees of facility. The highly selective and predictable creation of the new double bond and stereogenic centers is a hallmark of the rearrangement. These familiar issues of relative and internal stereogenesis attend all reactions within or between molecules containing enantio- or diastereotopic faces.² A less common manifestation of relative stereogenesis in the Claisen rearrangement involves the influence of remote stereocenters not contained within the pericyclic array. This is encountered in substrates bearing stereogenic carbons attached to positions 1 and 6 (A and B, Scheme 2). An intriguing subset of this phenomenon is the special case of auxiliary-based, relative stereocontrol wherein the preexisting stereogenic unit is ultimately recoverable unchanged (Scheme 2). A few, but notable examples of this concept have been described for variants of the Claisen rearrangement,^{3, 4} as has the use of stoichiometric amounts of chiral modifiers.⁵

Scheme 1.

Scheme 2.

To further extend the synthetic utility and improve stereoselectivity, catalysis of the Claisen rearrangement constitutes an important objective.⁶ Significant rate and selectivity enhancements are now on record employing catalysis with Lewis acids (both main group⁷ and transition metal⁸) Brønsted acids,⁹ antibodies,¹⁰ and N-heterocyclic carbenes.¹¹These advances have significantly improved the rate and enantioselectivity of the Claisen rearrangement. The continuing efforts in this area are testimony to the fact that a truly general solution to asymmetric Claisen rearrangements is still elusive. In addition, to enable catalysis by Lewis and Brønsted acids, Lewis basic binding sites (generally a carboalkoxy group at C(2)) are required to enable activation and provide rigidity for stereocontrol. This fact is not so much a criticism as a statement of opportunity to develop new concepts for addressing these key issues.

Our study of the Claisen rearrangement stems from a long-standing interest in the electronic modulation of the rearrangement.¹² Early reports from our laboratories documented the accelerating effect of a carbanionic (π-donor) at the 2-position of an allyl vinyl ether, Scheme 3.¹³ The arylsulfonyl stabilizing group (G = ArSO₂) has been extensively investigated and the resulting carbanion-accelerated Claisen rearrangements (CACR)¹⁴ display the following characteristics: (1) >300-fold acceleration, (2) exclusive γ-site selectivity, (3) substitution compatible at all positions, (4) exclusive formation of trans olefins and (5) high diastereoselectivity (95:5 dr) for syn or anti isomers. Although extension of the CACR with other sulfur-based anion stabilizing groups was unsuccessful, a number of phosphorus-based groups have shown considerable potential. As shown in Scheme 3, phosphine oxides, phosphonates and phosphonamides are superior to sulfones in rate and selectivity in the CACR.¹⁵

Scheme 3.

In all of these cases, the anion-stabilizing groups are achiral. Thus, the two limiting chair-like transition-states i and iv are enantiomeric giving rise, necessarily, to racemic products, ii (iii) and v (vi), Scheme 4. If, however, G* were chiral, i and iv would be diastereomeric leading to diastereomers ii and v which, after removal of G* constitute enantiomers iii and vi. Insofar as the anionic charge is responsible for the rate enhancement, effective desymmetrization of that charge by chiral groups G* should lead to significant differences in rates of rearrangement for i and iv. These differences are manifested in the enantiomeric excess of the products.

Scheme 4.

The specific objective of the investigation described below is the development of an auxiliary-based, asymmetric CACR using chiral, anion-stabilizing groups. The success of this enterprise rests on the design of the chiral moiety G* which should satisfy the following criteria: (1) ready construction from available, non-racemic materials, (2) effective acceleration of the CACR, (3) strong diastereofacial bias in the CACR and (4) facile cleavage and recovery of the auxiliary unit. In addition to the stereoselective construction of new carbon-carbon bonds, the results of this study should provide valuable insights into the nature of heteroatom-stabilized allyl anions.¹⁶

Background

The application of chiral phosphorus reagents in organic synthesis is not very common, apart from the ubiquitous use of chiral phosphines as ligands for transition metal catalysts. Chiral organophosphorus reagents have been employed in asymmetric olefination,¹⁷ asymmetric Michael addition,¹⁸ and asymmetric alkylation and amination,¹⁹ and asymmetric Staudinger reactions.²⁰ A small number of common motifs have found good use in several of these transformations such as diazaphospholidines (i) and diazaphosphorinanes (ii) oxazaphospholidines (iii, iv) and oxazaphosphorinanes (v, vi) (Chart 1).

Chart 1.

The de novo design of an auxiliary for a phosphorus-stabilized carbanion was nearly impossible since little was known at the time about the structure of the anion. Early design criteria were guided by our own studies in addition to the NMR and IR spectroscopic studies from Seyden-Penne²¹ on P-stabilized carbanions bearing additional stabilizing groups (-COR, - CO₂R, -CN, Ph, etc.). Two limiting structures were proposed: a parallel, chelated anion vii and an orthogonal, (C-Li contact) anion viii, Chart 2. It was a priori difficult to predict which of these two structures the allyl vinyl ether anion would prefer. Indeed, it embodies features of both being an unstabilized allyl anion, but also containing a potentially coordinating β-oxygen. In either case, we reasoned that a key element in stereocontrol was to desymmetrize the region around the carbanion by maximally differentiating the groups X and Y, Chart 2. This condition was satisfied by the N-substituted 1,3,2-oxazaphosphorinane moiety ix depicted in Chart 2. The phosphorinane (six-ring) was selected in favor of the phospholidine (five-ring) for two reasons: (1) to move the N-ligand closer to the vicinity of the anion and (2) to reduce reactivity at phosphorus.²²

Chart 2.

This structural motif formed the basis for the three types of substrates examined, Chart 3. Preliminary studies, optimization and N-group dependence were carried out with model I in racemic form (derived from an achiral amino alcohol). Models II (cis) and II (trans) (derived from the same enantiomerically pure amino alcohol) were used to establish the absolute stereochemical course of the reaction and to produce enantiomerically enriched products. The synthesis and rearrangements of various allyl vinyl ethers in these families and stereochemical analysis of the keto phosphonamidate products is described in detail below. Some of these studies have been described previously in preliminary form.²³

Chart 3.

Results and Discussion

1. Racemic Oxazaphosphorinanes. 1.1. Synthesis of Amino Alcohol Auxiliaries

The achiral amino alcohols 2a, 2b and 2c required for synthesis of Model I were prepared by methylation of the 3-trifluoroacetamidopropanoates (1a-1c) with excess methylmagnesium bromide, Scheme 5. The use of methyllithium or a simple acetamide protecting group led to considerably lower yields. Compounds 1a-1c were prepared by simple Michael addition to methyl acrylate followed by acetylation with trifluoroacetic anhydride.

Scheme 5.

1.2. Synthesis of Allenes

The amino alcohols were directly converted to the required allenes in a one pot operation via a facile Horner-Mark [2,3]-rearrangement (4 to 5, Table 1). Thus, addition of 3-butyn-2-ol (3a) to a cold solution of PCl3 and 1.0 equiv of N-methyl morpholine (NMM), followed by sequential addition of 2.0 additional equiv of NMM and the amino alcohol 2a, 2b or 2c produced allenes 5a-5c respectively. Allene 6 was prepared under identical conditions by combining 2-methyl-3-butyn-2-ol (3b) with PCl₃ and amino alcohol 2a.

Table 1.

Synthesis of Allenes 5 and 6

entry	R1	R2	Product	yield, %
1	t-Bu	H	5a	58
2	1-Ad	H	5b	70
3	Me	H	5c	79
4	t-Bu	Me	6	53

1.3 Synthesis of Allyl Vinyl Ethers

Previous studies demonstrated that very subtle changes in reaction conditions produce drastic differences in product distribution during allyloxide additions to achiral phosphorus-substituted allenes, and similar behavior with these oxazaphosphorinanes was anticipated. Indeed, four products could be isolated following exposure of potassium or sodium allyloxide to allene 5a (Table 2): the β,γ- and α,β-unsaturated oxazaphosphorinanes 7aa and 8aa resulting from allyloxide addition, and acetylenic oxazaphosphorinanes 9a and 10a resulting from simple tautomerization. Formation of the β, γ- unsaturated oxazaphosphorinane 7aa required the use of sodium allyloxide at room temperature in the presence of t-butyl alcohol (entry 1). tert-Butyl alcohol serves as a proton source to trap the intermediate allyl anion prior to isomerization. The use of sodium hydride without tert-butyl alcohol at lower temperatures (entry 3) afforded a mixture of 7aa and both acetylenic products. Under no circumstances did the addition of potassium allyloxide lead to the exclusive formation of 7aa (entries 2, 4–6). However, the α, β-unsaturated derivative 8aa was cleanly prepared at room temperature in the absence of t-butyl alcohol in high yield by the addition of potassium allyloxide (entry 2).

Table 2.

Optimization of Allyloxide Addition to Allene 5a.

entry	MHa	t-BuOH, equiv	temp, °C	time, h	result
1	NaH	2.4	20	0.5	7aa (64%)
2	KH	0	20	0.5	8aa (71%)
3	NaH	0	−20	3.5	7aa:9a:10a (4:3:3)
4	KH	2.4	0	0.5	7aa:9a:10a (1:1:1)
5	KH	1.8	20	0.25	8aa only
6	KH	solvent	20	1.0	9a and 10a only

^aAll reactions performed with a full equivalent of MH.

The results for all of the optimized allyl- and crotyloxide additions in this series are collected in Table 3. All additions to the monomethylated allenes proceed rapidly at room temperature. Crotyloxide addition to the dimethylated allene 6 proceeded more slowly and was free of contamination from the corresponding acetylenes or the α, β-unsaturated addition product. However, a small amount (< 5%) of inseparable thermal Claisen rearrangement product was invariably formed.

Table 3.

Allyl and Crotyloxide Additions to Allenes 5 and 6.^a

entry	allene	base	t-BuOH, equiv	time, min	product	R1	R2	R3	yield, %
1	5a	NaH	2.4	20	7aa	t-Bu	H	H	64
2	5a	KH	0	30	8aa	t-Bu	H	H	71
3	5a	NaH	2.4	20	7ab	t-Bu	H	CH3	46
4	5a	KH	0	20	8ab	t-Bu	H	CH3	80
5	5b	NaH	2.4	5	7ba	1-Ad	H	H	34
6	5c	NaH	2.4	2	7ca	Me	H	H	40
7	6	KH	0	180	11	t-Bu	CH3	CH3	51

^aAll reactions were performed using 1.2 equiv of base and allylic alcohol.

1.4. Claisen Rearrangements

All anion-accelerated rearrangements in this series were conducted using 2.0–2.5 equiv of lithium dimsylate generated from either equimolar amounts of n-BuLi and DMSO in THF or KH in 3:1 DMSO/THF in the presence of 6 equivalents of LiCl. Anion accelerated rearrangements of all β,γ -unsaturated isomers are complete within 10–15 min at room temperature while the corresponding thermal rearrangements require 2–4 h in refluxing toluene or at 100 °C in THF (sealed tube). As expected, anion-accelerated rearrangements of the α,β-unsaturated isomers required reaction times on the order of 1.5–3 h.

Two types of diastereoselectivity are manifest with chiral phosphorus-stabilized, carbanion-accelerated Claisen rearrangements. The relative asymmetric induction is defined as the ability of the chiral phosphorus subunit to influence the creation of the new stereogenic centers in the rearrangement. The internal asymmetric induction reflects the extent of chair/boat conformational selectivity. The absolute asymmetric induction describes the configuration of newly created stereogenic centers with respect to the configuration of the existing phosphorus atom. In the racemic series (Model I), the particular substitution pattern of each allyl vinyl ether determined whether relative asymmetric induction could be studied separately or coupled with internal asymmetric induction. Therefore, the rearrangements of each allyl vinyl ether will be discussed individually. Absolute asymmetric induction will only be discussed with Model II compounds (vide infra).

Anion-accelereated Claisen rearrangement of allyl vinyl ether 7aa with LiDMSO afforded a mixture of two diastereomers (l- and u-12aa)²⁴ with as high as 90% diastereoselectivity (Table 4). The diastereomeric ratios were determined by integration of either the distinct singlets in the proton-decoupled ³¹P NMR spectra or the C(3) methyl doublets in the ¹H NMR spectra. The thermal rearrangement is non-selective (entry 1) as is the anionaccelerated rearrangement when KDMSO serves as base (entry 2). All rearrangements of anions with lithium as the counterion gave excellent diastereoselectivities (entries 3–8). The rate of reaction decreased with decreasing solvent polarity. Interestingly, use of the more bulky and more strongly basic lithium diisopropyl sulfoxide (entry 6) increased the rate of rearrangement but did not affect stereoselectivity. Thus it appears that the reaction rate is solvent dependent, but stereoselectivity is counterion dependent.

Table 4.

Claisen Rearrangement of Oxazaphosphorinanes 7aa.^a

entry	baseb	solvent	time, h	yield, %	drc
1d	none	THF	4.0	90	58:42
2	KDMSO	DMSO/THF(3:1)	0.25	77	52:48
3	KDMSO (LiCl)e	DMSO/THF(3:1)	0.25	81	91:9
4	LiDMSOf	DMSO/THF(3:1)	0.25	73	95:5
5	LiDMSO	THF	0.50	64	94:6
6	LiDMSO	Et2O	1.0	31	90:10
7	LiDIPSO	THF	0.25	68	94:6
8	n-BuLi	THF	1.75	51	95:5

^aAll rearrangements carried out at room temperature.

^bUsually 2.0 – 2.5 equivalents of base used.

^cDiastereoselectivity determined by ³¹P NMR or ¹H NMR (500 MHz).

^dReaction performed at 100°C in a sealed tube..

^e6.0 equiv of LiCl used.

^fPrepared with n-BuLi. The reagent was insoluble in THF but dissolved upon addition of 7aa.

The diastereomeric ratios observed in the rearrangement of 7aa are a direct reflection of the control that the chiral environment about the phosphorus atom exerts on the transition state during the rearrangement. Any change in the chair/boat transition state selectivity will not affect the observed product ratio.

The importance of the local environment about the phosphorus atom was further demonstrated by studying the rearrangements of the N-adamantyl and N-methyl derivatives 7ba and 7ca (Table 5). Whereas both thermal Claisen rearrangements are unselective (entries 1 and 4), anion-accelerated rearrangements of 7ba (entries 2–3) proceed with the same high stereoselectivity as seen with 7aa. However, the high degree of stereoselectivity disappears under anion-forming conditions when the ligand bound to the oxazaphosphorinane nitrogen is small (R = methyl, entry 5). Thus a sterically bulky N-substituent is crucial for proper stereocontrol.

Table 5.

Claisen Rearrangement of 7ba and 7ca.^a

entry	educt	R	baseb	solvent	time, h	yield, %	drc
1	7ba	1-Ad	none	toluened	2.0	93	60:40
2	7ba	1-Ad	LiDMSO	THF	0.25	58	94:6
3	7ba	1-Ad	LiDMSO	DMSO/THF	0.25	74	94:6
4	7ca	Me	none	toluened	2.5	86	53:47
5	7ca	Me	LiDMSO	THF	0.25	62	52:48

^aAnionic rearrangements carried out at room temperature.

^b2.5 equiv of base were used.

^cDiastereoselectivities determined by ¹H NMR (500 MHz)

^dReactions run at 110 °C.

The α,β-unsaturated oxazaphosphorinane 8aa underwent Claisen rearrangement under anion-forming conditions to afford a mixture of (l) and (u) keto oxazaphosphorinanes 12aa (Scheme 6). However, this rearrangement is selective in the opposite sense when compared to that of compound 7aa. The origin of the reversal of selectivity may be explained by a permutation of the allyl anion geometry. Chart 4 shows the geometries of the favored anions. α-Deprotonation of 7aa gives E allyl anion 7aa− since the configuration about C(1)-C(2) is fixed during allyloxide addition and a high barrier to rotation relative to rearrangement exists. γ - Deprotonation of 8aa then must be selective to form a preponderance of Z allyl anion 8aa− having the opposite configuration at C(1).

Scheme 6.

Chart 4.

Rearrangement of the trimethyl-substituted allyl vinyl ether 11 also leads to a mixture of two diastereomeric products (13, Scheme 7). And although the chair/boat conformational selectivity does affect the product ratios in this rearrangement, it is not possible to quantify this internal selectivity. This result implies that, assuming complete chair selectivity, the minimum relative asymmetric induction is 92:8. Anything less than complete chair selectivity would make the relative asymmetric induction even better.

Scheme 7.

The rearrangements of allyl vinyl ethers 7ab and 8ab (Table 6) result in the formation of two new stereogenic centers and thus, mixtures of four diastereomers. In these cases, it was possible to quantify the degrees of internal and relative asymmetric induction. The four diastereomers are distinguishable in the ³¹P NMR spectra (note ³¹P chemical shifts in Scheme 8).

Table 6.

Claisen Rearrangement of Oxazaphosphorinanes 7ab and 8ab.

							inductionb
entry	educt	basea	solvent	time, h	temp, °C	yield, %	relative	internal
1	7ab	none	THF	4.0	100	84	58:42	77:23
2	7ab	KDMSO/LiClc	DMSO/THF (3:1)	0.25	20	80	92:8	98:2
3	8ab	KDMSO/LiClc	DMSO/THF (3:1)	3.0	20	60	13:87	94:6

^a2.0–2.5 equiv of base were used.

^bSee text for definition.

^c6.0 equivalents of LiCl were used.

Scheme 8.

Assignment of the diastereomers was made according to the following arguments. The major product from the CACR of compound 7ab (δ³¹P 15.74 ppm) was assumed to be a syndimethyl diastereomer by analogy to all other CACRs (i.e. sulfones, phosphonates, etc.) in that similarly substituted allyl vinyl ethers produce syn-dimethyl diastereomers. Thermal rearrangements are also syn selective and thermal rearrangement of 7ab gave two major diastereomers (δ ³¹P: 15.74, 16.25 ppm). The anti dimethyl diastereomers were identified from the anion-accelerated rearrangement of the α,β-unsaturated oxazaphosphorinane 8ab. These assignments were also based on analogy to other CACRs of similarly substituted allyl vinyl ethers which have been shown to be anti selective. To assign numbers to the internal and relative asymmetric induction selectivities, we still needed to determine which transition state was responsible for each product. Making the reasonable assumption that the major diastereomer in each rearrangement resulted from a chair-like transition state, all products can be assigned as having resulted from either a chair- or boat-like transition state from either face of the allyl anion. The internal diastereoselectivity is then the sum of the integrals of the products arising from a chair- versus a boat-like transitions state. Thus, the internal diastereoselectivities are 98:2 for 7ab and 94:6 for 8ab. The relative diastereoselectivities are the sum of the integrals of the products arising from rearrangement on the same face of the allyl anion. Thus, the relative diastereoselectivities are 92:8 for 7ab and 87:13 for 8ab.

2. Enantiomerically Pure Oxazaphosphorinanes. Model II

Because CACR products 12xx and 13 are racemic and non-crystalline it was not possible to determine the configuration of the newly formed stereogenic centers by degradative or X-ray crystallographic methods. Thus, it was necessary to employ a chiral oxazaphosphorinane which was generated from enantiomerically pure amino alcohol 15.

2.1. Synthesis of Amino Alcohol 15

Amino alcohol 15 was chosen for its structural similarity to amino alcohol 2a. Thus, the oxazaphosphorinane ring, and hence the environment about the phosphorus atom would experience minimal perturbation. The preparation of 15 is shown in Scheme 9. Yeast reduction of ethyl acetoacetate afforded (S)-ethyl-3-hydroxybutanoate in 55% yield (er 98:2).²⁵ The enantiomeric purity was determined by optical rotation and GC analysis of the Mosher esters.²⁶ The observed enantiomeric excess is significantly higher than previously reported and can only be explained by the brand of yeast used.²⁷ Amidation without hydroxyl protection by the method of Weinreb²⁸ gave (S)-tert-butyl-3-hydroxybutyramide in 72% yield (er >99:1). This reaction required a careful quenching protocol followed by acidification to pH ~6. At lower pH, β-elimination to the butenamide occurred. The enantiomeric excess was determined by derivatization to a 3,5-dinitrophenyl carbamate and HPLC analysis on a Pirkle L-Naphthylalanine column.²⁹ Diborane reduction then produced amino alcohol 15 in 67% yield.

Scheme 9.

2.2. Synthesis and Assignment of Allenes cis- and trans-16

Allene formation from 15, PCl₃, and propargylic alcohol 3a using standard reaction conditions produced a mixture of four diastereomers (Scheme 10). From this mixture, diastereomeric allene pairs cis-16 (54%) and trans-16 (20%) could be easily separated by chromatography. Assignment of these phosphorus epimers was tentatively based on the downfield chemical shifts of phosphorus in the ³¹P NMR spectrum and the C(6) proton in the ¹H NMR spectrum of cis-16.³⁰ These assignments were later confirmed by X-ray crystallography (vide infra).

Scheme 10.

2.3. Synthesis and Assignment of Allyl Vinyl Ethers

Allyl vinyl ethers cis-17 and trans-17 were prepared by sodium allyloxide addition to allenes cis- and trans-16, respectively, in the presence of tert-butyl alcohol (Scheme 11). The formation of acetylenic tautomers was responsible for the low yield in the preparation of trans-17.

Scheme 11.

The tentative assignments of configuration at the phosphorus center made earlier on the basis of ¹H and ³¹P NMR spectroscopic data could be confirmed by single crystal X-ray crystallographic determinations of both allyl vinyl ethers cis-17 and trans-17 (Figure 1).³¹ These structures verified the S-configuration at phosphorus and a cis relationship between the C(6) methyl group for cis-17 and the R-configuration at phosphorus and a trans relationship between the C(6) methyl group for trans-17. Furthermore, the geometry about the vinyl ether double bond was also unambiguously established in both cases. As expected, nucleophilic attack on a monomethylated allene is governed by steric approach of the allyloxide to the less hindered face of the allene.

Figure 1.

ORTEP images of cis-17 and trans-17.

Other interesting features of these structures include the observation that the oxazaphosphorinane ring in cis-17 is nearly flat with bond angles between the substituents on nitrogen of approximately 120°. The planarity at nitrogen, in accord with other reported 1,3,2-oxazaphosphorinanes,³² stems from the longer P-O and P-N bond lengths that widen the angles between the atoms in the ring. The wider angles also contribute to the flattening of the ring. However, the most important interaction governing the ring conformation is the geminal-P(1)-N(1) substitution pattern which places the phosphorus substituent in a pseudoaxial position. Bentrude has shown that N-unsubstituted 1,3,2-oxazaphosphorinanes exist predominantly in a chair conformation, whereas N-substituted 1,3,2-oxazaphosphorinanes exist exclusively in a twist-chair conformation.³³ Clearly, the geminal P(1)-N(1) interaction in cis-17 strongly influences 1,3,2-oxazaphosphorinane ring conformation. In trans-17 the nitrogen is lightly pyramidalized and the ring takes up a chair-like conformation wherein the phosphorus substituent again takes up an axial position.

2.4. Claisen Rearrangements

Under anion-forming conditions, allyl vinyl ether cis-17 (Table 7) rearranged rapidly at room temperature with high relative diastereoselectivity. As in the racemic series, the thermal rearrangement required four hours at 100 °C and showed poor diastereoselectivity (compare entries 4 and 7). At this point, the effect of added LiCl on the stereoselectivity of the CACR was studied. In the absence of LiCl, there is no observed asymmetric induction in the anion-accelerated rearrangement of cis-17 (entry 1). As the amount of LiCl is increased from 1–6 equivalents (entries 2–4), the diastereomeric ratios improve from ~2:1 to 9:1. No further increase in selectivity is observed with additional equivalents of LiCl (entry 5). Note also that LiDMSO prepared from n-BuLi and DMSO in THF also results in high relative asymmetric induction. Finally, added LiCl had no positive effect on the selectivity of the thermal rearrangement (entry 8). The anion-accelerated rearrangement of trans-17 also proceeded with high selectivity (Scheme 12).

Table 7.

Carbanionic Claisen Rearrangement of cis-17^a

entry	educt	product	baseb	LiCl, equivc	temp, °C	yield, %d	relative inductione
1	cis-17	cis-18	KDMSO	0	20	62	50:50
2	cis-17	cis-18	KDMSO	1	20	77	65:35
3	cis-17	cis-18	KDMSO	2	20	69	80:20
4	cis-17	cis-18	KDMSO	6	20	78	90:10
5	cis-17	cis-18	KDMSO	12	20	65	89:11
6	cis-17	cis-18	LiDMSOf	0	20	65	90:10
7	cis-17	cis-18	none	0	100	93	66:34
8	cis-17	cis-18	none	6	100	90	64:36

^aAll anionic rearrangements (15 min) were done in 3:1 DMSO/THF except entry 6 which was done in 2:1 THF/DMSO, thermal rearrangements (240 min) were done in THF.

^b2–2.5 equiv of freshly prepared base were used.

^cLiCl was added to KDMSO before addition of 17.

^dYield after chromatography.

^eSee text for explanation.

^fPrepared from n-BuLi.

Scheme 12.

2.5. Degradation and Assignment

To establish the relative configuration at the newly created stereogenic centers with that of the existing phosphorus center, CACR products (S)-(l,l)-18 and (S)-(u,u)-18 were degraded to optically active dimethyl methylsuccinates as shown in Scheme 13. Treatment of the keto oxazaphosphorinanes with KHMDS at –78 °C and trapping with tert-butyldimethylsilyl chloride afforded the silyl enol ether. Ozonolysis followed by an oxidative workup provided an enantiomerically enriched sample of methylsuccinic acids which were esterified with diazomethane. The overall yields were 59% from cis-18 and 24% from trans-18.

Scheme 13.

The absolute configuration of the succinate esters could not be established by optical rotation because the absolute optical rotation of dimethyl methylsuccinate is quite small.³⁴ With only limited amounts of material in hand, the use of optical rotation became inappropriate. The enantiomeric succinates were instead distinguished by ¹H NMR spectroscopy in the presence of the chiral shift agent (R)-2,2,2-trifluoro-1-(9-anthryl)ethanol.³⁵ The methyl ester regions of the spectra clearly show four singlets. Assignment of the major enantiomers followed from the ¹H NMR spectrum of a mixture of the racemic succinates enriched with authentic R succinate and the chiral shift agent. As shown in Scheme 13, cis-18 produced the (S)-dimethyl methylsuccinate and trans-18 produced the R enantiomer. These data, coupled with the known configurations of the vinyl ether double bond (E) and the phosphorus stereogenic center, allows the unambiguous assignment of the sense of the chair-like folding in the transition state during the rearrangement. Thus, the rearrangement of cis-17 proceeds by bonding to the Re face of the allyl anion at C(1) (c.f. Chart 4) whereas the Si face is preferred for trans-17.

2.6. Proposed Transition States

It is clear that auxiliary structure, anion structure, counterion effects, and conformational preferences are crucial to phosphorus-induced diastereoselectivity. Before developing any reasonable model for the rearrangement, it would be prudent to review the current understanding of the structure of such phosphorus-stabilized anions.

The combination of multinuclear, variable temperature solution NMR studies, X-ray crystallographic analyses³⁶ and computational studies³⁷ provide a very clear picture of the factors that influence the structure of these species. Anions derived from 1,3,2-diazaphosphorinanes and 1,3,2-dioxaphosphorinanes bearing both P-benzyl and P-isopropyl substituents display the following characteristics: (1) the anionic carbons are fully planarized (i.e. sp² hybridized) and are devoid of contacts to lithium, (2) the atoms of the anion are all aligned in nearly eclipsed conformations with respect to the P=O moiety such that the dihedral angle, O(1)-P(1)-C(1)-C(2) spans a mere 10° across all structures, (3) the lithium atoms are coordinated to the phosphoryl oxygens and to two or more solvent (THF) molecules, (4) the nitrogen atoms in the 1,3,2-diazaphosphorinanes are pyramidalized, and (5) rotational barriers around the P-C bonds are very low. Although phosphorus-stabilized allyl anions have been studied only computationally, our experimental observations are consistent with the results of those studies.

The currently preferred model proposed to explain relative diastereoselectivity in the carbanion-accelerated Claisen rearrangement incorporates the various steric and electronic components that have been established in the foregoing studies. The model, shown in Figure 2, possesses the following characteristics: (1) a planar carbanion in which the atoms are aligned parallel to the P=O bond, (2) chelation of a lithium ion between the phosphoryl and vinyl ether oxygens, and (3) a sterically directed, diastereofacial preference of chair conformers such that the allyloxy side chain folds away from the bulky nitrogen substituent. This model corresponds to the parallel anion model vii in Chart 2.

Figure 2.

CACR transition state models for cis-17 and trans-17.

In this model, the strong preference for the ul transition state in the anion-accelerated Claisen rearrangements of 1,3,2-oxazaphosphorinanes is a result of approach of the allyloxy side chain to the sterically less hindered side, i.e., away from the large N-tert-butyl group. The effect of the size of this group is dramatic as selectivity is completely lost in the N-methyl substrate, 7ca, but slightly enhanced in the N-adamantyl substrate 7ba. This model also comports with the high selectivity observed in the CACR of trans-17 which displayed an even higher level of diastereoselectivity than cis-17. The high diastereoselectivity observed with trans-17 is particularly noteworthy. In a chair-like ring conformation, the allyl vinyl ether would have an axial orientation (Figure 1). Following the same logic, the fixed anion conformation and the biased folding of the allyl ether side chain away from the N-tert-butyl group leads to the correct prediction for the configuration of the rearrangement product. However, the higher selectivity for trans-17 is puzzling and may hint that the ring conformations are not chairs but rather twist chairs with similar orientations of the allyl vinyl ether substituent because of the bulky N-tertbutyl group.

Conclusion

The carbanion-accelerated Claisen rearrangement allyl vinyl ethers bearing chiral 1,3,2-oxazaphosphorinanes takes place under extremely mild conditions (room temperature, 15 min) to afford γ, δ-unsaturated ketones with high levels of internal and relative diastereoselectivity. The internal diastereoselectivity followed the well-established paradigm of chair-like transition structures for these rearrangements. The relative diastereoselectivity with respect to the 1,3,2-oxazaphosphorinane ring was dependent upon the auxiliary structure and the reaction conditions. Lithium salts were shown to greatly enhance the diastereoselectivity which was interpreted to arise from chelation of the phosphoryl oxygen with the allyl vinyl ether oxygen in the stereodetermining transition structure. The size of the nitrogen substituent was shown to be critical for high diastereoselectivity. A model was constructed for the origin of high selectivity that was grounded in structural studies on phosphorus-stabilized anions.

EXPERIMENTAL³⁸

Racemic 1,3,2-Oxazaphosphorinane-2-oxides

Methyl 3-(N-tert-Butyl-N-trifluoroacetyl)aminopropionate (1a)

A 15-mL, 3-necked, flask equipped with a stirring bar, septum, and N₂ inlet was charged with 5.08 g (25.2 mmol) of methyl 3-tert-butylaminopropionate and 25 mL CH₂Cl₂. The mixture was placed in an ice bath and 5.35 mL (37.9 mmol) of trifluoroacetic anhydride was added via syringe over 5 minutes. The reaction mixture was warmed to room temperature and stirred for two hours. The solvent was removed by rotary evaporation and the crude product purified by Kugelrohr distillation to give 5.30 g (84.3%) of amido ester 1a, as a colorless oil. Data for 1a: bp 90 °C (0.3 mm Hg); ¹H NMR (300 MHz) 3.71 (m, 2 H), 3.66 (s, 3 H), 2.60 (m, 2 H), 1.44 (s, 9 H); ¹³C NMR (75.5 MHz) 170.5 (C(1)), 157.0 (q, J = 34.3, CF₃C=O)), 116.4 (q, J = 289.4, CF₃), 59.7 ((CH₃)₃C), 51.8 (OCH₃), 39.9 (q, J = 3.2, C(3)), 36.0 (C(2)), 27.8 ((CH₃)₃C); IR (neat) 2961 (m), 1742 (s), 1694 (s), 1482 (m), 1439 (s), 1422 (s), 1401 (m), 1383 (s), 1372 (s), 1327 (s), 1294 (m), 1266 (s), 1198 (s), 1123 (s), 1048 (s), 1026 (m), 986 (w), 899 (w), 804 (w); MS (70 eV) 240 (M⁺-15, 38), 208 (38), 199 (38), 168 (30), 167 (1), 166 (20), 139 (11), 57 (100), 56 (15), 55 (42). Anal. Calcd for C₁₀H₁₆F₃NO₃ (225.25): C, 47.06; H, 6.32; N, 5.49; F, 22.33. Found: C, 47.04; H, 6.24; N, 5.56; F, 22.03.

Methyl 3-[N-(1-Adamantyl)-N-trifluoroacetyl]aminopropionate (1b)

To a suspension of 1-adamantylamine (12.5 g, 80.2 mmol) in methanol (50 mL) at 0 °C was added methyl acrylate (8.02 mL, 88.2 mmol) over 20 minutes. The resulting mixture was stirred for 2.5 days at room temperature and then the methanol was removed by rotary evaporation. The crude product was used without further purification. The amino ester (19.0 g, 80.2 mmol) was placed in a dry flask and dichloromethane (100 mL), 4-dimethylaminopyridine (0.49 g, 4.0 mmol), and triethylamine (13.4 mL, 92.2 mmol) were added sequentially and the mixture was cooled to 0 °C. A solution of trifluoroacetic anhydride (13.0 mL, 92.2 mmol) in CH₂Cl₂ (25 mL) was added dropwise via addition funnel over 30 min. The solution was warmed to room temperature and stirred for 2 days. The mixture was then poured into Et₂O (250 mL), washed with sat. aq. NaHCO₃ solution (3×50 mL), and brine (50 mL), then dried (MgSO₄), filtered and the solvents removed by rotary evaporation. The crude product was purified by vacuum distillation to give 24.9 g (93% for 2 steps) of 1b as a clear, colorless oil. Data for (1b): bp 160 °C (0.6 mm Hg); ¹H NMR (500 MHz) 3.75–3.72 (m, 2 H), 3.70 (s, 3 H), 2.66–2.63 (m, 2 H), 2.19 (s, 6 H), 2.15 (s, br, 3 H), 1.73–1.66 (m, 6 H); ¹³C NMR (125.8 MHz) 170.6 (C(1)), 156.8 (q, J = 34.3, CF₃C=O), 116.3 (q, J = 289.9, CF₃), 61.3 (adamantyl C), 51.8 (OCH₃), 39.0 (adamantyl CH₂CN), 38.8 (s, C(3)), 36.7 9(s, C(2)), 36.0 (s, adamantyl CH₂CH CH₂CN), 30.0 (s, adamantyl CH); MS (70 eV) 333 (M⁺, 0.77), 136 (10), 135 (100), 134 (15); Anal. Calcd for C₁₆H₂₂F₃NO₃(333.35): C, 57.65; H, 6.65; N, 4.20; F, 17.10. Found: C, 57.43; H, 6.56; N, 4.25; F, 17.46.

Methyl- 3-(N-methyl-N-trifluoroacetyl)aminopropionate (1c)

To a solution of N-methyl trifluoroacetamide (16.0 g, 123.4 mmol) and methyl acrylate (28.6 mL, 308.5 mmol) in t-butanol (30 mL) was added potassium t-butoxide (0.75 g, 6.17 mmol). The resulting solution was stirred at room temperature for 8.5 h. Acetic acid (0.2 mL) was added and the mixture stirred for an additional hour. The reaction mixture was then quenched with acetic acid (1.0 mL), filtered through silica gel (3" column), and the column washed with Et₂O (250 mL). The solvents were removed in vacuo and the crude product purified by vacuum distillation to yield 23.4 g (89%) of 1c as a clear, colorless oil. Data for (1c): bp 153–154 °C (75 mm Hg); ¹H NMR (500 MHz) major rotamer 3.70–3.67 (m, 2 H), 3.69 (s, 3 H), 3.17 (s, 3 H), 2.65-2.63 (m, 2 H), minor rotamer 3.70- 3.67 (m, 2 H), 3.70 (s, 3 H), 3.02 (s, 3 H), 2.65-2.63 (m, 2 H); ¹³C NMR (CDCl₃, 125.8 MHz) major rotamer 171.5 (s, C(1)), 156.8 (q, J = 36.0, CF₃CO), 116.2 (q, J = 287.3, CF₃), 51.7 (OCH₃), 45.7 (C(3)), 35.5 (NCH₃), 31.2 (C(2)), minor rotamer 170.6 (C(1)), 156.6 (q, J = 36.1, CF₃C=O), 116.3 (q, J = 287.4, CF₃), 51.8 (s, OCH₃), 44.9 (C(3)), 34.3 (NCH₃), 32.8 (C(2)); IR (neat) 3378 (w), 2959 (m), 2361 (w), 1738 (s), 1688 (s), 1520 (m), 1497 (w), 1441 (s), 1387 (m), 1319 (s), 1248 (s), 1146 (s), 1096 (s), 1048 (m), 985 (w), 887 (w), 835 (w), 793 (w), 760 (s), 725 (w); MS (70 eV) 213 (M⁺, 9), 182 (18), 181 (13), 153 (25), 140 (100), 116 (24), 102 (15), 84 (13), 69 (32), 59 (15), 55 (42); Anal. Calcd for C₇H₁₀F₃NO₃(213.15): C, 39.44; H, 4.73; N, 6.57; F, 26.74. Found: C, 39.81; H, 4.95; N, 6.60; F, 27.02.

4-(t-Butylamino)-2-methyl-2-butanol (2a)

A 250-mL, 3-necked flask equipped with a mechanical stirrer, addition funnel (with septum and N₂ inlet), and thermometer was charged with 4.00 g (15.7 mmol) of amido ester 1a and 100 mL of Et₂O. The mixture was cooled to 5°C and the addition funnel charged with 36.2 mL (2.7M, 99.7 mmol) of methylmagnesium bromide. The Grignard reagent was added dropwise over 30 minutes at 0–5°C. The ice bath was removed and the mixture was stirred for 3 h. The mixture was then cooled to 0°C and 50 mL of a sat. aq. K₂CO₃ solution was added. The mixture was diluted with 200 mL of a sat. aq. solution of Rochelle's salt. The aqueous layer was extracted with EtOAc (3×250 mL) and the organic layers dried (MgSO₄), filtered, and the solvent removed by rotary evaporation. The crude product was purified by Kugelrohr distillation to yield 1.47 g (58.8%) of 2a, as a clear, colorless oil. Yield range, 59–76%. Data for 2a: bp 80 °C (0.3 mm Hg); ¹H NMR (300 MHz) 2.86 (t, 2 H, J = 5.5), 1.57 (t, 2 H, J = 5.5), 1.21 (s, 6 H), 1.10 (s, 9 H); ¹³C NMR (75.5 MHz) 71.0 (C(2)), 50.4 ((CH₃)₃C), 40.8 (C(3)), 38.7(C(2)), 29.6 (C(1), CH₃C(2)), 28.5 ((CH₃)₃C); IR (neat) 3276 (m), 2971 (s), 2932 (s), 2869 (m), 1717 (w), 1653 (w), 1480 (m), 1426 (m), 1391 (m), 1362 (s), 1267 (w), 1233 (m), 1215 (m), 1154 (m), 1086 (w), 1017 (w), 967 (w), 938 (w), 887 (w); MS (70 eV) 144 (M⁺- 15, 100), 126 (90), 88 (14), 86 (30), 81 (11), 70 (75), 58 (43), 57 (11); Anal. Calcd for C₉H₂₁NO (159.28): C, 67.87; H, 13.29; N, 8.79. Found: C, 67.52; H, 13.28; N, 8.84.

4-N-Methylamino-2-methyl-2-butanol (2c)

To a refluxing ethereal solution of methylmagnesium bromide (3.0 M, 75.1 mL, 225 mmol) under N₂ was added dropwise a solution of amido ester 1c (8.0 g, 37.5 mmol) in Et₂O (40 mL) over 30 min. The resulting thick suspension was refluxed for 1 h and quenched with the dropwise addition of a sat. aq. solution of NH₄Cl (15 mL). Ether and absolute ethanol (100 mL each) were added and the mixture heated to reflux for 30 min. The mixture was then cooled and filtered through a plug of silica gel (5 g). After washing the silica gel with absolute ethanol (150 mL), the combined filtrates were evaporated under reduced pressure. The residue was diluted with methanol, filtered, and purified by column chromatography (5–15% ammonia-saturated methanol in Et₂O) to give 2.30 g (52.3%) of 2c as a clear, colorless oil. Data for (2c): bp 66–68 °C (1.2 mm Hg); ¹H NMR (500 MHz) 2.86 (t, 2 H, J = 5.8), 2.40 (s, 3 H), 1.58 (t, 2 H, J = 5.8), 1.22 (s, 6 H); ¹³C NMR (125.8 MHz) 71.0 (C(2)), 48.4 (C(4)), 39.9 (C((3)), 36.1 (NCH₃), 29.7 (C(1) and H₃CC(2)); IR (neat) 3293 (s, br, OH, NH), 2967 (s), 2930 (s), 2845 (s), 2799 (m), 1643 (w), 1474 (s), 1377 (s), 1362 (s), 1296 (w), 1266 (m), 1198 (m), 1171 (s), 1142 (m), 1111 (m), 1038 (w), 941 (m), 909 (m), 882 (m), 810 (m), 734 (m); MS (10 eV) 117 (M⁺, 12), 102 (10), 84 (13), 59 (20), 58 (10), 44 (100); high resolution MS calcd for C₆H₁₅NO 117.1157, found: 117.1155; TLC R_f 0.25 (Et₂O/MeOH (NH₃sat.), 3:1).

Synthesis of 2-Allenyl 1,3,2-Oxazaphosphorinane-2-oxides. General Procedure 1

A 250-mL, 3-necked, round-bottomed flask equipped with a stirring bar, septum, thermometer, and N₂ inlet was charged with 100 mL of CH₂Cl₂ and cooled to 0°C. Phosphorus trichloride (13.7 mmol) and N-methylmorpholine (13.7 mmol) were added sequentially via syringe. After stirring for 5 minutes, the appropriate propargylic alcohol (13.7 mmol) in 5 mL of CH₂Cl₂ was added via syringe (0–5°C). After stirring for 20 minutes, N-methylmorpholine (27.4 mmol) was added. After stirring for 5 minutes, the appropriate amino alcohol (13.7 mmol) in 5 mL of CH₂Cl₂ was added. The reaction mixture was warmed to room temperature and stirred for 16 hours. The mixture was diluted with 100 mL of water and extracted with CH₂Cl₂ (3×100 mL). The organic layers were washed with water and brine (1×50 mL each). The combined organic layers were dried (MgSO₄), filtered, and the solvent removed by rotary evaporation. Purification is given for each individual compound.

3-tert-Butyl-2-(1',2'-butadienyl)-6,6-dimethyl-2-oxo-1,3,2-oxazaphosphorinane (5a)

Following General Procedure 1, the crude product was purified by column chromatography (hexane/acetone; 1:1) to give 2.05 g (58%) of allene 5a. Data for 5a are reported for a distilled sample: bp 175 °C (0.2 mm Hg); ¹H NMR (300 MHz) 5.46–5.42 (m, 1 H), 5.30-5.20 (m, 1 H), 3.26-3.07 (m, 2 H), 1.94-1.80 (m, 2 H), 1.74-1.66 (m, 3 H), 1.44 (d, 3 H, J = 2.5), 1.38 (s, 12 H); ¹³C NMR (75.5 MHz) 209.6 (C(2')), 209.2 (C(2')), 88.7 (d, J = 6.0, C(3')), 86.3 (d, J = 6.9, C(3')), 79.9 (d, J = 6.3, C(6)), 79.8 (d, J = 7.2, C(6)), 54.5 (d, J = 4.2, (CH3)3C), 28.6 ((CH3)3C), 12.2 (d, J = 7.3 C(4')), 11.7 (d, J = 7.0, C(4')). The following resonances could not be unambiguously assigned: 86.4, 86.3, 85.7, 85.5, 38.7, 38.5, 38.4, 38.3, 28.7, 28.5, 28.4; ³¹P NMR (121.4 MHz) 9.26, 9.24; IR (neat) 2977 (s), 2874 (m), 2357 (w), 1952 (m), 1639 (w), 1466 (m), 1370 (s), 1291 (w), 1254 (s, P=O), 1208 (s), 1152 (s), 1102 (m), 1061 (s), 997 (s), 936 (m), 889 (m), 855 (m); MS (70 eV) 257 (M⁺, 6.9), 243 (13), 242 (100), 204 (11), 201 (47), 186 (94), 174 (40), 148 (91), 146 (59), 145 (58), 84 (16), 70 (61), 69 (48), 58 (37), 57 (78), 56 (13), 55 (19), 53 (28), 43 (15), 42 (20), 41 (82), 39 (13), 31 (19); high resolution MS calcd for C₁₃H₂₄NO₂P 257.1542, found 257.1577; TLC R_f 0.40 (hexane/acetone, 1:1).

3-tert-Butyl-6,6-dimethyl-2-(3'-methyl-1',2'-butadienyl)-2-oxo-1,3,2-oxazaphosphorinane (6)

Following General Procedure 1, the crude product was purified by column chromatography (hexane/acetone, 1:1) to give a 1.36 g (53.1%) of allene 6. Data for 6 are reported for a distilled sample: bp 205 °C (0.15 mm Hg); ¹H NMR (300 MHz) 5.55-5.47 (m, 1 H), 3.20-3.08 (m, 2 H), 2.04-1.94 (m, 2 H), 1.89-1.82 (m, 6 H), 1.57(s, 3 H), 1.48 (s, 12 H); ¹³C NMR (75.5 MHz) 207.3 (C(2')), 95.3 (d, J = 17.2, C(3')), 86.7 (d, J = 182.9, C(1')) 80.1 (d, J = 8.6, C(6)), 54.9 (d, J = 5.2, (CH₃)₃C), 38.9 (d, J = 13.7, C(4)), 38.8 (C(5)), 29.1 (CH₃C(6)), 28.9 ((CH₃)₃C), 28.8 (CH₃C(6)), 19.2 (d, J = 6.6, C(4')), 18.7 (d, J = 7.7, CH₃C(3')); ³¹P NMR (121.4 MHz) 9.6; IR (CCl₄) 2980 (s), 2940 (s), 2872 (m), 1964 (m, C=C=C), 1632 (w), 1559 (w), 1466 (m), 1445 (m), 1395 (m), 1385 (m), 1362 (s), 1291 (s), 1252 (s, P=O), 1235 (s), 1204 (s), 1152 (s), 1103 (m), 1061 (m), 999 (s), 932 (m), 889 (m); MS (70 eV) 271 (M+, 11.5), 257 (1), 256 (76), 215 (73), 201 (11), 200 (94), 198 (15), 188 (28), 160 (56), 159 (75), 148 (99), 132 (15), 126 (12), 84 (24), 70 (73), 69 (50), 67 (29), 58 (49), 57 (80), 56 (11), 55 (19), 43 (38), 42 (20), 41 (100), 39 (19), 31 (27); high resolutions MS calcd for C₁₄H₂₆NO₂P 271.1711, found 271.1706; TLC R_f 0.38 (hexane/acetone, 1:1).

3-Adamantyl-2-(1',2'-butadienyl)-6,6-dimethyl-2-oxo-1,3,2-oxzazphosphorinane (5b)

To a refluxing ethereal solution of methylmagnesium bromide (3.0 M, 84.0 mL, 252 mmol) under N₂ was added a solution of 1b (14.0 g, 42.0 mmol) in Et₂O (50 mL) over 30 min. The solution was heated reflux for 1.5 h further and then quenched by the dropwise addition of a sat. aq. NH₄Cl solution (10 mL). Ether and absolute ethanol (150 mL each) were added, the mixture was stirred for 1 h and filtered through a pad of Celite. After washing the precipitate with absolute ethanol (250 mL), the combined filtrates were evaporated under reduced pressure. The residue was diluted with saturated methanolic ammonia in Et₂O (1:1, 250 mL), filtered, and twice purified by column chromatography (5–15% saturated methanolic ammonia in Et₂O) to give 6.52 g (65.4%) of 4-[N-(1-adamantyl)amino]-2-methyl-2-butanol (2b).

To a solution of phosphorus trichloride (0.84 mL, 9.69 mmol) in CH₂Cl₂ (50 mL) at 0 °C under N₂ was added dropwise N-methylmorpholine (1.12 mL, 10.1 mmol) over 5 min. After stirring for 5 min, (±)-3-butyn-2-ol (0.73 mL, 9.69 mmol) was added dropwise over a 5 min period. After stirring for 30 min, N-methylmorpholine (2.24 mL, 20.2 mmol) was added. After an additional 5 min, a solution of 2b (2.0 g, 8.43 mmol) in CH₂Cl₂ (40 mL) was added by cannula over a 20 min period. The resulting mixture was stirred for 20 h and poured into Et₂O (400 mL). The mixture was washed with water (30 mL), 1N HCl (30 mL), and brine (30 mL). The organic phase was then dried (MgSO₄), filtered, and concentrated in vacuo. The crude product was purified by column chromatography (EtOAc) to give 2.32 g (70.4%) of 5b as a clear, colorless oil. Data for (5b): ¹H NMR (500 MHz) major diastereomer 5.50-5.44 (m, 1 H), 5.31-5.24 (m, 1 H), 3.27-3.20 (m, 1 H), 3.143-3.07 (m, 1 H), 2.10 (d, 3 H, J = 10.8), 2.09 (s, 3 H), 1.95 (d, 3 H, J = 10.8), 1.76-1.71 (m, 3 H), 1.63 (s, 6 H), 1.45 (s, 3 H), 1.35 (s, 3 H), minor diastereomer 5.50-5.44 (m, 1 H), 5.31-5.24 (m, 1 H), 3.27-3.20 (m, 1 H), 3.14-3.07 (m, 1 H), 2.10 (d, 3 H, J = 10.8), 2.09 (s, 3 H), 1.95 (d, 3 H, J = 10.8), 1.76-1.71 (m, 3 H), 1.63 (s, 6 H), 1.46 (s, 3 H), 1.36 (s, 3 H); ¹³C NMR (125.8 MHz) 210.0 (C(2')), 209.6 (C(2')), 80.3-80.2 (m, C(6)), 55.92 (adamantyl C), 55.90 (adamantyl C), 41.05 (adamantyl CH₂CN), 40.99 (adamantyl CH₂CN), 39.45-39.35 (m, C(5)), 37.1 (C(4)), 36.1 (adamantyl CH₂CH CH₂CN), 29.6-29.5 (m, CH₃C(6) and adamantyl CH) 28.91 (CHC(6)), 28.87 (CH₃C(6)), 12.9 (d, J = 7.0, C(4')), 12.3 (d, J = 7.0, C(4')); the following peaks could not be unambiguously assigned: 88.9, 88.8, 87.4, 86.1, 86.0, 85.7, 85.6; IR (neat) 2977 (s), 2909 (2), 2851 (s), 1952 (m, C=C=C), 1453 (m), 1370 (s), 1298 (m), 1242 (s, P=O), 1204 (s), 1150 (s), 1115 (m), 1094 (s), 995 (s), 930 (m), 891 (m), 808 (s), 776 (m), 752 (m), 706 (m), 669 (m); MS (70 eV) 335 (M⁺, 7.4), 280 (10), 279 (24), 136 (11), 135 (100), 106 (10), 93 (16), 79 (19), 55 (11), 53 (13); high resolution MS calcd for C₁₉H₃₀O₂NP 335.2015, found 335.2016; TLC R_f 0.24 (EtOAc).

N-Methyl-2-(1',2'-butadienyl)-6,6-dimethyl-2-oxo-1,3,2-oxazaphosphorinane (5c)

From 2c (5.1g, 32.0 mmol) following General Procedure 1, the crude product was purified by column chromatography (hexane/acetone, 3/2) to give 6.54 g (79%) of 5c as a clear, colorless oil. Data for (5c): ¹H NMR (500 MHz) major diastereomer 5.38-5.35 (m, 1 H), 5.32-5.25 (m, 1 H), 3.21-3.08 (m, 2 H), 2.72 (d, 3 H, J = 9.4), 1.95-1.91 (m, 2 H), 1.76-1.71 (m, 3 H), 1.48 (s, 3 H), 1.41 (s, 3 H), minor diastereomer 5.38-5.35 (m, 1 H), 5.32-5.25 (m, 1 H), 3.21-3.08 (m, 2 H), 2.72 (d, 3 H, J = 9.4), 1.95-1.91 (m, 2 H), 1.76-1.71 (m, 3 H), 1.48 (s, 3 H), 1.40 (s, 3 H); ¹³C NMR (125.8 MHz) major diastereomer 209.5 (C(2')), 85.5-84.5 (C(3')), 82.4 (d, J = 205.8, C(1'), 81.5 (C(6)), 45.7 (C(4)), 35.8 (C(5)), 34.6 (NCH₃), 28.4 (CH₃C(6)), 27.5 (CH₃C(6)), 12.1 (CH₃C(4')), minor diastereomer 209.3 (C(2')), 85.5-84.5 (C(3')), 82.3 (d, J = 190.2, C(1')), 81.5 (C(6)), 45.8 (C(4)), 35.8 (C(5)), 34.6 (NCH₃), 28.4 (CH₃C(6)), 27.5 (CH₃C(6)), 12.3 (CH₃C(4')); ³¹P NMR (121.5 MHz) 11.9, 11.8; IR (neat) 2977 (m), 2490 (w), 1950 (m, C=C=C), 1649 (w), 1470 (m), 1372 (m), 1298 (s), 1248 (s, P=O), 1223 (s), 1157 (m), 1227 (s), 1061 (m), 997 (s), 970 (s), 878 (m), 808 (m), 785 (m), 750 (s); MS (70 eV) 215 (M⁺, 22), 162 (31), 160 (55), 159 (57), 144 (10), 123 (16), 98 (19), 95 (11), 69 (52), 44 (100); high resolution MS calcd for C₁₀H₁₈NO₂P 215.1071, found: 215.1073; TLC R_f 0.18 (4.5% i-PrOH in CH₂Cl₂).

Synthesis of Allyl Vinyl Ethers. General Procedure 2

A 15-mL, 3-necked flask equipped with stirring bar, septa, and N₂ inlet was charged with NaH dispersion (50%, 0.48 mmol). The NaH dispersion was rinsed with hexane (3×1 mL/0.1 g NaH) and suspended in 3.0 mL of THF. t-Butyl alcohol (0.96 mmol) and the appropriate allyl alcohol (0.48 mmol) were added sequentially. After stirring for 10 minutes, a solution of the allene (0.40 mmol) in 1.0 mL of THF was added via syringe, The reaction was monitored by TLC. Upon completion, the reaction was quenched with 5 mL of water and extracted with Et₂O (3×15 mL). The organic layers were washed with brine (15 mL), dried (MgSO₄), filtered, and the solvent removed by rotary evaporation. Purification is given for each individual compound.

(E)-3-tert-Butyl-6,6-dimethyl-2-oxo-[2'-(2-propenyloxy)-2'-butenyl]-1,3,2-oxazaphosphorinane (7aa)

Following General Procedure 2, the crude product was purified by radial chromatography (hexane/acetone, 5:1) to give 155 mg (64%) of 7aa. Data for 7aa: ¹H NMR (300 MHz) 6.00-5.88 (m, 1 H), 5.29 (dd, 1 H, J = 1.3, 17.4), 5.19 (dd, 1 H, J = 1.1, 10.4, 4.54-4.46 (m, 1 H), 4.20-4.08 (m, 2 H), 3.15-2.93 (m, 3 H)), 2.61-2.50 (m, 1 H), 1.89-1.77 (m, 2 H), 1.62-1.51 (m, 3 H), 1.48 (s, 3 H), 1.37 (s, 9 H), 1.26 (s, 3 H); ¹H NMR (500 MHz); ¹³C NMR (75.5 MHz) 149.8 (d, J = 12.2, C(2')), 133.8 (CH=CH₂), 116.9 (CH=CH₂), 93.9 (d, J = 10.2, C(3')), 79.2 (d, J = 8.2, C(6)), 67.8 (OCH₂), 55.1 ((CH₃)₃C), 39.4 (C(4)), 39.3 (C(5)), 35.2 (d, J = 134.6, C(1')), 29.7 (d, J = 8.0, CH₃C(6)), 29.1 ((CH₃)₃C), 28.9(CH₃C(6)), 12.3 (C(4')); ³¹P NMR (121.4 MHz) 20.5; IR (neat) 2980 (s), 2934 (s), 2874 (m), 1665 (s), 1466 (s), 1399 (m), 1387 (m), 1372 (s), 1362 (m), 1348 (m), 1293 (s), 1256 (s), 1231 (s), 1202 (s), 1152 (s), 1102 (s), 1063 (m), 997 (s), 926 (s), 889 (m); MS (70 eV); 315 (M⁺, 1.42), 300 (13), 274 (22), 242 (15), 218 (25), 186 (15), 174 (18), 162 (100), 150 (33), 148 (36), 146 (12), 126 (15), 84 (16), 80 (14), 70 (73), 69 (29), 58 (36), 57 (46), 56 (10), 55 (16), 53 (11) 42 (14), 41 (73), 39 (14); high resolution MS calcd for C₁₆H₃₀NO₃P 315.1975, found 315.1969; TLC R_f 0.45 (hexane/acetone, 1:1). Anal. Calcd for C₁₆H₃₀NO₃P (315.20): C, 60.93; H, 9.59; N, 4.44; P, 9.82. Found: C, 60.72; H, 9.49; N, 4.53; P, 9.88.

(E)-3-tert-Butyl-6,6-dimethyl-2-oxo-[(E)-2'-(2-propenyloxy)-1'-butenyl]-1,3,2-oxazaphosphorinane (8aa)

Following General Procedure 2, the crude product was purified by column chromatography (hexane/acetone, 1.5:1) to give 175 mg (71%) of 8aa. Data for 8aa are reported for a recrystallized sample: mp 60–62 °C; ¹H NMR (300 MHz) 5.99-5.86 (m, 1 H), 5.34-5.20 (m, 2 H), 4.54 (d, 1 H, J = 7.5), 4.23 (d, 2 H, J = 5.1), 3.25-3.05 (m, 2 H), 2.66-2.45 (m, 2 H), 2.09-2.00 (m, 1 H), 1.86-1.79 (m 1 H), 1.51 (s, 3 H), 1.30 (s, 12 H), 1.09 (t, 3 H, J = 7.5); ¹³C NMR (75.5 MHz) 171.7 (d, J = 19.7, C(2')). 132.5 (CH=CH₂), 117.4 (CH=CH₂), 92.5 (d, J = 195.4, C(1')), 78.9 (d, J = 7.5, C(6)), 67.9 (OCH₂), 54.7 ((CH₃)₃C), 40.1 (d, J = 8.1, C(4)), 39.3 (C(5)), 29.9 (CH₃C(6)), 29.4 (d, J = 5.3, CH3C(6)), 29.0 ((CH₃)₃C), 25.5,(C(3')), 11.4 (C(4')); ³¹P NMR (121.4 MHz) 15.8; IR (CCl₄) 2977 (s), 2940 (s), 2874 (m), 1615 (s, C=C), 1464 (m), 1394 (m), 1385 (m), 1370 (s), 1362 (m), 1343 (m), 1287 (s), 1248 (s), 1225 (s, P=O), 1186 (s), 1150 (s), 1094 (s), 1059 (m), 1015 (s), 968 (s), 926 (s), 887 (s); MS (70 eV) 315 (M⁺, 19.2), 301 (15), 300 (89), 260 (11), 258 (11), 244 (32), 242 (19), 204 (46), 203 (16), 174 (11), 148 (15), 134 (12), 126 (76), 84 (13), 70 (100), 69 (11), 58 (31); TLC R_f 0.45 (hexane/acetone, 1:1). Anal. Calcd for C₁₆H₃₀NO₃P (315.40): C, 60.93; H, 9.59; N, 4.44; P, 9.82. Found: C, 60.92; H, 9.65; N, 4.26; P, 9.73.

(E)-3-tert-Butyl-6,6-dimethyl-2-oxo-[2'-((E)-2-butenyloxy)-2'-butenyl]-1,3,2-oxazaphosphorinane (7ab)

Following General Procedure 2, the crude product was purified by column chromatography (hexane/acetone, 1:1) to give 59.0 mg (46%) of 7ab. Data for 7ab: ¹H NMR (300 MHz) 5.77-5.60 (m, 2 H), 4.52-4.48 (m, 1 H), 4.11-4.06 (m, 2 H), 3.13-2.94 (m, 3 H), 2.57 (dd, 1 H, J = 18.1, 15.2), 1.87-1.82 (m, 2 H), 1.71 (d, 3 H, J = 5.8) 1.62 (d, 3 H, J = 6.7, 4.5), 1.50 (s), 1.39 (s, 9 H), 1.28 (d, 3 H, J = 1.1); ¹³C NMR (75.5 MHz) 149.8 (d, J = 15.5, C(2')), 129.4 (CH=CHCH₃), 126.7 (CH=CHCH₃), 93.7 (d, J = 9.9, C(3')), 79.2 (d, J = 9.6, C(6)), 67.5 (OCH₂), 55.1 ((CH₃)₃C), 39.3 (d, J = 7.2, C(4)), 35.1 (d, J = 133.5, C(1')), 29.7 (CH₃C(6)), 29.1 ((CH₃)₃C), 28.8 (CH₃C(6)), 17.7 (CH₃CH=CH), 12.3 (C(4')); ³¹P NMR (121.4 MHz) 19.2; IR (neat) 2977 (s), 2587 (w), 1719 (w), 1665 (s, C=C), 1464 (m), 1397 (m), 1372 (m), 1293 (s), 1256 (s, P=O), 1231 (s), 1200 (s), 1152 (s), 1102 (s), 1065 (m), 997 (s), 926 (m), 901 (m), 847 (m); MS (70 eV) 329 (M⁺, 10.2), 314 (16), 275 (14), 274 (74), 273 (15), 219 (39), 218 (39), 218 (62), 204 (24), 190 (13), 188 (11), 164 (27), 163 (70), 162 (100), 151 (66), 150 (36), 148 (15), 134 (16), 125 (11), 84 (14), 70 (21), 58 (11); high resolution MS calcd for C₁₇H₃₂NO₃P 329.2120, found 329.2121; TLC R_f 0.43 (hexane/acetone, 1:1).

(E)-3-tert-Butyl-6,6-dimethyl-2-oxo-[2'-((E)-2-butenyloxy)-1'-butenyl]-1,3,2-oxazaphosphorinane (8ab)

Following General Procedure 2, the crude product was purified by column chromatography (hexane/acetone, 1:1) to afford 230 mg (69%) of 8ab. Data for 8ab: bp 200 °C (0.5 mm Hg); ¹H NMR (300 MHz) 5.83-5.56 (m, 2 H), 4.55 (d, 1 H, J = 7.6), 4.17 (d, 2 H, J = 5.9), 3.27-3.07 (m, 2 H). 2.66-2.47 (m, 2 H). 2.11-2.03 (m, 1 H), 1.88-1.81 (m, 1 H), 1.72 (d, 3 H, J = 5.7), 1.53 (s, 3 H), 1.33 (s, 12 H), 1.10 (t, 3 H, J = 7.5); ¹³C NMR (75.5 MHz) 171.8 (d, J = 19.5, C(2')), 130.0 (CH=CHCH₃), 125.3 (CH=CHCH₃), 91.7 (d, J = 195.3, C(1')), 78.7 (d, J = 8.6, C(6)), 67.9 (OCH₂), 54.6 (d, J = 4.7, (CH₃)₃C), 40.0 (d, J = 7.5, C(4)), 39.2 (C(5)), 29.8 (CH₃C(6)), 29.3 (d, J = 5.6, CH₃C(6)), 28.9 ((CH₃)₃C), 25.5 (C(3')), 17.6 (CH₃CH=CH), 11.4 (C(4')); ³¹P NMR (121.4 MHz) 16.0; IR (neat) 2975 (s), 2876 (m), 2361 (w), 1709 (w), 1613 (s, C=C), 1464 (m), 1370 (m), 1347 (m), 1289 (m), 1250 (s. P=O), 1223 (s), 1186 (s), 1152 (s), 1094 (s), 1063 (m), 1017 (m), 967 (s), 932 (m), 889 (m); MS (70 eV) 330 (M⁺+1, 11.4), 329 (M⁺, 30.5), 314 (46), 273 (15), 272 (20), 260 (46), 258 (24), 256 (11), 246 (12), 219 (15), 218 (55), 217 (20), 216 (13), 206 (13), 205 (13), 204 (55), 202 (19), 192 (18), 190 (25), 188 (18), 164 (91), 163 (30), 162 (15), 160 (10), 152 (23), 151 (19), 148 (34), 135 (13), 134 (52), 127 (10), 126 (100), 84 (42), 83 (11), 70 (99), 58 (41), 55 (10); high resolution MS calcd for C₁₇H₃₂NO₃P 329.2120, found 329.2124; TLC R_f 0.43 (hexane/acetone, 1:1).

(E)-3-Adamantyl-6,6-dimethyl-2-oxo-[2'-(2-propenyloxy)-2'-butenyl]-1,3,2- oxazaphosphorinane (7ba)

From 5b (3.8 g, 11.3 mmol) Following General Procedure 2, the crude product was purified by column chromatography (petroleum/acetone ether, 3:1) to give 1.52 g (34%) of 7ba as a clear, colorless oil. Data for (7ba): ¹H NMR (500 MHz) 6.02-5.94 (m, 1 H), 5.31 (dd, 1 H, J = 17.1, 1.5), 5.20 (d, 1 H, J = 10.5), 4.53-4.49 (m, 1 H), 4.21-4.11 (m, 2 H), 3.19-3.14 (m, 1 H), 3.03 (dd, 1 H, J = 11.4), 2.07 (s, 3 H), 1.97 (d, 3 H, J = 11.4), 1.86-1.81 (m, 1 H), 1.71 (s, 6 H), 1.66-1.60 (m, 1 H), 1.63 (dd, 3 H, J = 6.5, 4.4), 1.51 (s, 3 H), 1.27 (s, 3 H); ¹³C NMR (125.8 MHz) 149.7 (d, J = 12.9, C(2')), 133.9 (CH=CH₂), 117.0 (CH=CH₂), 94.0 (d, J = 10.3, C(3')), 79.2 (d, J = 9.9, C(6)), 67.8 (OCH₂), 56.0 (d, J = 10.3, C(3')), 41.0 (adamantyl CH₂CN), 40.0 (d, J = 6.6, C(5)), 37.3 (d, J = 2.1, C(4)); 36.3 (adamantyl CH₂CH CH₂CN), 35.5 (d, J = 131.8, C(1')), 29.7 (s, adamantyl CH), 29.6 (CH₃C(6)), 29.5 (CH₃C(6)), 12.3 (s, C(4')); IR (neat) 2977 (m), 2909 (s), 2853 (s), 2681 (w), 2448 (w), 1665 (s), 1456 (m), 1404 (m), 1387 (m), 1362 (m), 1300 (m), 1248 (s, P=O), 1200 (s), 1148 (s), 1100 (s), 992 (s), 926 (s), 891 (m), 849 (s), 776 (m); MS (70 eV) 393 (M⁺, 1.3), 352 (13), 136 (11), 135 (100), 106 (11), 93 (11), 79 (11); Anal. Calcd for C₂₂H₃₆NO₃P (393.50): C, 67.15; H, 9.22; N, 3.56; P, 7.87. Found: C, 67.41; H, 9.28; N, 3.45; P, 7.82; TLC R_f 0.31 (EtOAc).

(E)-N-Methyl-6,6-dimethyl-2-oxo-[2'-(2-propenyloxy)-2'-butenyl]-1,3,2-oxazaphosphorinane (7ca)

From 5c (700 mg, 3.25mmol) Following General Procedure 2, the crude product was purified by column chromatography (CH₂Cl₂/i-PrOH, 19:1) to give 356 mg (40%) of 7ca as a clear, colorless oil. Data for (7ca): ¹H NMR (500 MHz) 6.01-5.90 (m, 1 H), 5.32 (dd, 1 H, J = 17.2, 1.4), 5.21 (dd, 1 H, J = 10.5, 1.4), 4.60-4.51 (m, 1 H), 4.20-4.08 (m, 2 H), 3.20-2.90 (m, 2 H), 2.90-2.75 (m, 2 H), 2.72 (d, 3 H, J = 8.8), 1.94-1.66 (m, 2 H), 1.62 (dd, 3 H, J = 6.7, 4.5), 1.50 (s, 3 H), 1.33 (d, 3 H, J = 1.6); ¹³C NMR (125.8 MHz) 149.1 (d, J = 13.4, C(2')), 133.4 (CH=CH₂), 116.7 (CH=CH₂), 94.1 (d, J = 10.2, C(3')), 80.9 (d, J = 8.7, C(6)), 67.4 (OCH₂), 45.8 (C(4)), 35.9 (C(5)), 34.9 (NCH₃), 33.5 (d, J = 46.0, C(1')), 30.2 (CH₃C(6)), 26.0 (CH₃C(6)), 11.9 (C(4')); ³¹P NMR (121.5 MHz) 22.2; IR (neat) 2980 (m), 2926 (m), 2442 (w), 2357 (w), 2220 (w), 1968 (w), 1667 (m), 1470 (w), 1387 (m), 1348 (w), 1300 (m), 1252 (s, P=O), 1196 (m), 1127 (s), 1100 (m), 1065 (m), 997 (s), 974 (s), 926 (m), 882 (m), 847 (m), 777 (m), 712 (m); MS (70 eV) 273 (M⁺, 1.7), 176 (15), 162 (40), 160 (28), 159 (16), 98 (52), 71 (11), 69 (23), 57 (93), 44 (100), 43 (14); high resolution MS calcd for C₁₃H₂₄NO₃P 273.1494, found 273.1495; TLC R_f 0.21 (CH₂Cl₂/i-PrOH 19:1).

(E)-3-tert-Butyl-6,6-dimethyl-2-oxo-[2'-((E)-2-butenyloxy)-3'-methyl-2'-butenyl]-1,3,2-oxazaphosphorinane (11)

Following General Procedure 2, the crude product was purified by column chromatography (hexane/isopropanol, 10:1) to give 130 mg (51%) of 11. Data for 11: ¹H NMR (300 MHz) 5.72-5.60 (m, 2 H), 4.10-3.96 (m, 2 H), 3.13-3.04 (m, 2 H), 2.89 (dd, 1 H, J = 18.9, 15.6), 2.62 (dd, 1 H, J = 15.9, 15.5), 1.83 (t, 2 H, J = 6.3), 1.68 (d, 3 H, J = 7.5), 1.67-1.62 (m, 6 H), 1.48 (s, 3 H), 1.37(s, 9 H), 1.25 (s, 3 H); ¹³C NMR (75.5MHz) 142.6 (d, J = 12.5, C(2')), 127.7 (CH=CHCH₃), 123.8 (CH=CHCH₃), 117.5 (d, J = 11.0, C(3')), 78.8 (d, J = 8.7, C(6)), 69.2 (OCH₂), 54.7 ((CH₃)₃C), 38.7 (C(4)), 38.6 (C(5)), 32.5 (d, J = 142.7, C(1')), 29.3 (CH₃C(6)), 29.0 ((CH₃)₃C), 28.9 (CH₃C(6)), 19.1 (CH=CHCH₃), 16.8 (CH₃C(3')), 16.7 (CH₃C(3')); ³¹P NMR (121.4 MHz) 20.5; IR (neat) 2973 (s), 2926 (s), 2732 (m), 1676 (w), 1609 (w), 1464 (m), 1387 (m), 1370 (m), 1293 (s), 1254 (s, P=O), 1196 (s), 1152 (s), 1094 (s), 1065 (m), 997 (s), 928 (m), 889 (m); MS (70 eV) 343 (M+, 3.38), 328 (26), 289 (11), 288 (28), 286 (12), 233 (28), 232 (29), 190 (18), 177 (46), 176 (42), 165 (56), 164 (19), 149 (11), 148 (100), 147 (12), 126 (26), 86 (17), 84 (15), 70 (28), 58 (14); high resolution MS calcd for C₁₈H₃₄NO₃P 343.2276, found 343.2278; TLC R_f 0.45 (hexane/acetone, 1:1).

Anion-Accelerated Claisen Rearrangement

A. KH/LiCl/DMSO

General Procedure 3

A 3-necked, 15-mL, flask equipped with a stirring bar, septa, and a vacuum/N₂ inlet was charged with 35% KH dispersion (0.68 mmol). The dispersion was rinsed with hexane (3×0.5 mL/0,1 g KH) and DMSO (3.0 mL) was added. After stirring until H₂ evolution ceased (~10 minutes), LiCl (6 equiv) was added all at once (if necessary), followed by stirring for 10 more minutes. Then a solution of allyl vinyl ether (0.30 mmol) in THF (1.0 mL) was added via syringe. The reaction was monitored by TLC. Upon completion, the reaction mixture was quenched with water (10 mL) and extracted with ether (3×15 mL). The organic layers were washed with water (3×15 mL) and brine (1×15 mL). The combined organic layers were dried over MgSO₄, filtered, and the solvent removed by rotary evaporation. Purification is given for each individual compound.

B. n-BuLi/DMSO/THF

A 15-ml, 3-necked, round-bottomed flask equipped with a septum, stirring bar, thermometer, and N₂ inlet was charged with 2.5 mL of THF and 1.5 mL of DMSO. n-BuLi (1.6 M in hexane, 0.46 mmol) was added dropwise via syringe to produce a clear, colorless solution. After stirring for 5 minutes, a solution of allyl vinyl ether in 1 mL THF was added via syringe. The solution was stirred at room temperature until judged complete by TLC. The reaction was quenched with water (5 mL) and extracted with Et₂O (3×15 mL). The organic layers were washed with water (3×5 ml) and brine (1×15 mL each). The combined organic layers were dried over MgSO₄, filtered, and the solvent removed by rotary evaporation. Purification is given for each individual compound.

C. LiDMSO/THF

Thermal Claisen Rearrangement. General Procedure 4

The allyl vinyl ether (0.10 mmol) THF (1.5 mL/mmol) was placed in a high pressure sealed vial and heated at 100 °C until starting material was consumed as judged by TLC. The contents of the vial were transferred to a pear-shaped flask with ether and the solvent removed in vacuo.

[R,S](Pl,3'l)-3-tert-Butyl-6,6-dimethyl-2-(3'-methyl-2'-oxo-5'-hexenyl)-2-oxo-1,3,2-oxazaphosphorinane (l)-12aa

Anionic (General Procedure 3.A.): radial chromatography (hexane/acetone, 3:1) afforded 74.0 mg (81%) of 12aa; thermal: radial chromatography (hexane/acetone, 3:1) afforded 90.7 mg (90%) of 12aa. Data for (l)-12aa and (u)-12aa: ¹H NMR (500 MHz) Major diastereomer (l)-12aa 5.76-5.69 (m, 1 H), 5.07-5.02 (m, 2 H), 3.39 (dd, 1 H, J_P-H = 21.5, J_H-H = 13.5), 3.18-3.05 (m, 2 H), 2.93 (dd, 1 H, J_P-H = 21.5, J_H-H = 13.5), 2.92-2.87 (m, 1 H), 2.43-2.37 (m, 1 H), 2.09-2.03 (m, 1 H), 2.00-1.95 (m, 1 H), 1.88-1.84 (m, 1 H), 1.51 (s, 3 H), 1.37 (s, 9 H), 1.30 (s, 3 H), 1.09 (d, 3 H, J = 7.2). Minor diastereomer (u)-12aa: 5.76-5.69 (m, 1 H), 5.07-5.02 (m, 2 H), 3.40 (dd, 1 H, J_P-H = 21.5, J_H-H = 13.5), 3.18-3.05 (m, 2 H), 2.93 (dd, 1 H, J_P-H = 21.5, J_H-H = 13.5), 2.92-2.87 (m, 1 H), 2.43-2.37 (m, 1 H), 2.09-2.03 (m, 1 H), 2.00-1.95 (m, 1 H), 1.88-1.84 (m, 1 H), 1.51 (s, 3 H), 1.37 (s, 9 H), 1.30 (s, 3 H), 1.06 (d, 3 H, J = 7.0); ¹³C NMR (125.8 MHz) 206.6 (d, J = 7.0, C(2')), 135 (C(5')), 116.7 (C(6')), 80.5 (d, J = 7.9, C(6)), 55.2 ((CH₃)₃C), 46.7 (d, J = 118.3, C(1')), 45.7 (C(3')), 40.0(d, J = 7.1, C(4)), 39.3 (C(5)), 36.2 (C(4')), 29.9 (CH₃C(6)), 29.2 (d, J = 7.0, CH₃C(6)), 28.8 ((CH₃)₃C), 16.0 (CH₃C(3')); ³¹P NMR (121.4 MHz) 47: 15.9, 48: 15.8; IR (neat) 3077 (w), 2975 (s), 2934 (s), 1705 (s, C=O), 1642 (w), 1460 (m), 1397 (m), 1291 (s), 1254 (s, P=O), 1194 (s), 1102 (m), 1063 (m), 992 (s), 920 (s), 891 (m), 868 (m); MS (70 eV) 315 (M⁺, 5.52), 301 (15), 300 (100), 259 (12), 258 (30), 244 (23), 242 (11), 232 (20), 204 (27), 192 (24), 174 (12) 148 (15), 135 (13), 126 (28), 109 (12), 84 (17) 70 (98), 58 (44); TLC R_f 0.39 (benzene/acetone, 2:1). Anal. Calcd for C₁₆H₃₀NO₃P (315.40): C, 60.93; H, 9.59; N, 4.44; P, 9.82. Found: C, 60.92; H, 9.34; N, 4.61; P, 10.09.

[R,S](Pl,3'l)-3-Adamantyl-6,6-dimethyl-2-(3'-methyl-2'-oxo-5'-hexenyl)-2-oxo-1,3,2-oxazaphosphorinane (l)-12ba

Anionic Rearrangement (n-BuLi). From 7ba (194 mg, 0.490 mmol), the crude product was purified by column chromatography (acetone/hexane, 2:1) to give 164 mg (74.2%) of 12ba as a clear, colorless oil. Thermal Rearrangement. From 7ba (270 mg, 0.69 mmol), the crude product was purified by column chromatography (hexane/acetone, 7/3) to give 251 mg (93.0%) of 12ba. Data for 12ba: ¹H NMR (500 MHz) major 5.76-5.69 (m, 1 H), 5.04 (d, 1 H, J = 15.9), 5.01 (d, 1 H, J = 10.1), 3.41 (dd, 1 H, J = 13.4, 11.1), 3.26-3.18 (m, 1 H), 3.01-2.87 (m, 3 H), 2.43-2.37 (m, 1 H), 2.12-1.83 (m, 3 H), 2.10 (d, 3 H, J = 11.8), 2.09 (s, 3 H), 1.92 (d, 3 H, J = 11.8), 1.64 (s, 6 H); 1.52 (s), 1.28 (s), 1.09 (d, 3 H, J = 7.1), minor 5.76-5.69 (m, 1 H), 5.05 (d, 1 H, J = 15.9), 5.02 (d, 1 H, J = 10.1), 3.42 (dd, 1 H, J = 13.4, 13.4), 3.26-3.18 (m, 1 H), 3.01-2.87 (m, 3 H), 2.43-2.37 (m, 1 H), 2.12-1.83 (m, 3 H), 2.10 (d, 3 H, J = 11.8), 2.09 (s, 3 H), 1.92 (d, 3 H, J = 11.8), 1.64 (s, 6 H), 1.52 (s), 1.28 (s), 1.07 (d, 3 H, J = 6.8); ¹³C NMR (125.8 MHz) 206.6 (d, J = 6.8, C(2')), 135.7 (C(5')), 116.6 (C(6')), 8.04 (d, J = 8.8, C(6)), 56.1 (d, J = 2.5, adamantyl C), 47.0 (d, J = 117.5, C(1')), 45.9 (C(3')), 41.0 (s, adamantyl CH₂CN), 40.6 (d, J = 5.3, C(5)), 37.1 (s, C(4)), 36.2 (s, C(4')), 36.1 (s, adamantyl CH₂CH CH₂CN), 30.2 (CH₃C(6)), 29.6 (adamantyl CH), 29.2 (d, J = 6.9, CH₃C(3')); ³¹P NMR (121.5 MHz) 16.5, 16.4; IR (neat) 2977 (s), 2909 (s), 1707 (s, C=O), 1640 (w), 1456 (m), 1387 (m), 1372 (m), 1298 (m), 1244 (s, P=O), 1196 (s), 1148 (s), 1115 (m), 1096 (s), 992 (s), 928 (s), 891 (m), 868 (m), 818 (m), 768 (w); MS (70 eV) 343 (M⁺, 8.9), 258 (22), 136 (11), 135 (100), 106 (37), 93 (21), 79 (17), 69 (13), 67 (10); high resolution MS calcd for C₂₂H₃₆NO₃P 393.2421, found 393.2427; TLC R_f 0.29 (hexane/acetone, 2:1).

[R,S](P1,3'1)-N-Methyl-6,6-dimethyl-2-(3'-methyl-2'-oxo-S'-hexenyl)-2-oxo-1,3,2-oxazaphosphorinene (12ca)

Anionic rearrangement (Li⁺DMSO₋). From 7ca (125 mg, 0.460 mmol) Following General Procedure 4, the crude product was purified by column chromatography (CH₂Cl₂/i-PrOH, 19:1) to yield 78 mg (62%) of 12ca as a clear, colorless oil. Thermal rearrangement. From 7ca (88 mg, 0.32 mmol) the crude product was purified by column chromatography (CH₂Cl₂/i-PrOH, 19:1) to give 76 mg (86%) of 12ca. Data for (12ca): ¹H NMR (500 MHz) major 5.72-5.63 (m, 1 H), 4.99 (d, 1 H, J = 16.0), 4.97 (d, 1 H, J = 9.2), 3.18-2.97 (m, 4 H), 2.84-2.78 (m, 1 H), 2.66 (d, 3 H, J = 9.6), 2.37-2.32 (m, 1 H), 2.08-2.01 (m, 1 H), 1.90-1.84 (m, 1 H), 1.79-1.74 (m, 1 H), 1.46 (s, 3 H), 1.32 (s, 3 H), 1.05 (d, 3 H, J = 6.4), minor 5.72-5.63 (m, 1 H), 4.99 (d, 1 H, J = 16.0), 4.97 (d, 1 H, J = 9.2), 3.18-2.97 (m, 4 H), 2.84-2.78 (m, 1 H), 2.662 (d, 3 H, J = 9.7), 2.37-2.32 (m, 1 H), 2.08-2.01 (m, 1 H), 1.90-1.84 (m, 1 H), 1.79-1.74 (m, 1 H), 1.46 (s, 3 H), 1.32 (s, 3 H), 1.03 (d, 3 H, J = 6.6); ¹³C NMR (125.8 MHz) major 206.6 (d, J = 6.9, C(2')), 135.3 (C(5')), 116.9 (C(6')), 82.12 (d, J = 5.9, C(6)), 46.4 (C(3')), 46.0 (C(4)), 43.7 (d, J = 18.3, C(1')), 36.7 (C(5)), 36.7 (C(4')), 35.1 (NCH₃), 29.9 (CH₃C(6)), 26.8 (CH₃C(6)), 15.5 (CH₃C(3')), minor 206.5 (d, J = 7.2, C(2')), 135.3 (C(5')), 116.9 (C(6')), 82.06 (d, J = 4.5, C(6)), 46.4 (C(3')), 46.0 (C(4)), 43.7 (d, J = 118.3, C(1')), 36.7 (C(5)), 36.7 (C(4')), 36.5 (CH₃N), 29.9 (CH₃C(6)), 26.8 (CH₃C(6)), 15.2 (CH₃C(3')); ³¹P NMR (121.5 MHz) 17.05, 16.96; IR (neat) 2979 (s), 2932 (s), 2469 (w), 1705 (s, C=O), 1640 (w), 1458 (m), 1389 (m), 1374 (m), 1298 (s), 1252 (s, P=O), 1194 (s), 1125 (s), 1063 (s), 1036 (m), 997 (2), 972 (s), 926 (m), 882 (s), 808 (w), 712 (w); MS (70 eV) 273 (M⁺, 2.4), 174 (11), 162 (63), 160 (34), 159 (32), 98 (40), 69 (50), 44 (100); high resolution MS calcd for C₁₃H₂₄NO₃P 273.1496, found 273.1495; TLC R_f 0.18 (CH₂Cl₂/i-PrOH, 19:1).

[R,S](Pl,3'l,4'u)-3-tert-Butyl-6,6-dimethyl-2-(3',4'dimethyl-2'-oxo-5'-hexenyl)-2-oxo-1,3,2-oxazaphosphorinane ((l)-12ab)

Anionic (General Procedure 3.A.): radial chromatography (hexane/acetone, 1:1) afforded 40.1 mg (80%) of 12ab; thermal: radial chromatography (hexane/acetone, 1:1) afforded 32.5 mg (84%) of 12ab. Data for 12ab are reported for a distilled sample: bp 150 °C (0.05 mm Hg); ¹H NMR (500 MHz) Major diastereomer syn-(l)-12ab: 5.81-5.74 (m, 1 H), 5.01-4.96 (m, 2 H), 3.27 (dd, 1 H, J_P-H = 17.7, J_H-H = 13.7), 3.16-3.11 (m, 2 H), 2.98 (dd, 1 H, J_P-H = 16.9, J_H-H =13.7), 2.81-2.77 (m, 1 H), 2.55- 2.51 (m, 1 H), 1.99-1.93 (m, 1 H), 1.88-1.84 (m, 1 H), 1.50 (s, 3 H), 1.36 (s, 9 H), 1.30 (s, 3 H), 1.04 (d, 3 H, J = 7.1), 0.95 (d, 3 H, J = 6.9). Minor diastereomer syn-(u)-12ab: 5.81-5.74 (m, 1 H), 5.01-4.96(m, 2 H), 3.46 (dd, 1 H, J = 18.7, 13.3), 3.16-3.11 (m, 2 H), 2.90 (dd, 1 H, J = 16.9, 13.7), 2.81-2.77 (m, ¹H), 2.55-2.51 (m, ¹H), 1.99-1.93 (m, ¹H), 1.88- 1.84 (m, 1 H), 1.50 (s, 3 H), 1.37 (s, 9 H), 1.30 (s, 3 H), 0.99 (d, 3 H, J = 6.9), 0.93 (d, 3 H, J = 6.8); ¹³C NMR (125.8 MHz) 206.6 (d, J = 4.9, C(2')), 142 (C(5')), 114 (C(6')), 80.5 (d, J = 66.1, C(6)), 55.3 ((CH₃)₃C), 51.3 (C(3')), 47.0 (d, J = 120.2 C(1')), 40.0 (C(4)), 39.3 (C(5)), 38.7 (C(4')), 30.0 (CH3C(6)), 29.1 ((CH₃)₃C) 15.3 (CH₃C(3')), 12.3 (CH₃C(4')); ³¹P NMR (101.3 MHz) 15.2; IR (neat) 2977 (s), 2878 (m), 1705 (s, C=O), 1638 (w), 1456 (m), 1456 (m), 1397 (m), 1372 (s), 1291 (s), 1254 (s, P=O), 1196 (s), 1150 (s), 1100 (m), 1063 (m), 995 (s), 918 (m), 891 (m), 868 (m); MS (70 eV) 329 (M⁺, 16.5), 315 (26), 314 (100), 273 (25), 272 (43), 258 (20), 256 (10), 246 (16), 218 (18), 206 (27), 205 (10), 190 (16), 189 (13), 188 (11), 148 (12), 126 (33), 124 (12), 109 (10), 86 (22), 84 (21), 70 (65), 58 (25); high resolution MS calcd for C₁₇H₃₂NO₃P 329.2120, found 329.2128; TLC R_f 0.40 (benzene/acetone, 2:1).

[R,S](Pl,3'l,4'l)-3-tert-Butyl-6,6-dimethyl-2-(3',4'dimethyl-2'-oxo-5'-hexenyl)-2-oxo-1,3,2-oxazaphosphorinane ((u)-12ab)

Anionic (General Procedure 3.A.): radial chromatography (hexane/acetone, 3:1) afforded 48.7 mg (60%) of (u)-12ab. Data for (u)-12ab are given for a distilled sample: bp 150 °C (0.05 mm Hg); ¹H NMR (500 MHz) 5.67-5.60 (m, 1 H), 5.02-4.96 (m, 2 H), 3.45 (dd, 1 H, J_P-H = 18.1, J_H-H = 13.5), 3.17-3.04 (m, 2 H), 2.87 (dd, ¹H, J_P-H = 16.7, J_H-H = 13.5), 2.91-2.83 (m, 1 H), 2.52-2.48 (m, 1 H), 2.01-1.95 (m, 1 H), 1.88-1.84 (m, 1 H), 1.52 (s, 3 H), 1.38 (s, 9 H), 1.30 (s, 3 H), 1.04 (d, 3 H, J = 8.6), 1.01 (d, 3 H, J = 8.6); ¹³C NMR (125.8 MHz) 206.9 (d, J = 7.4, C(2')), 140.3 (C(5')), 115.0 (C(6')), 80.5 (d, J = 7.8, C(6)), 55.3 ((CH₃)₃C), 51.3 (C(3')), 48.5 (d, J = 118.4, C(1')), 40.1 (C(4')), 39.4 (C(4)), 39.4 (C(5)), 30.0 (CH₃C(6)), 29.2 (CH₃C(6)), 29.0 ((CH₃)₃C), 18.0 (CH₃C(3')), 12.6 (CH3C(4')); ³¹P NMR (121.4 MHz) 16.1; IR (neat) 2975 (s), 2878 (m), 1703 (s, C=O), 1639 (m), 1456 (m), 1397 (m), 1372 (s), 1291 (s), 1252 (s, P=O), 1190 (s), 1150 (s), 1100 (m), 1063 (s), 995 (s), 918 (s), 891 (s), 868 (m); MS (70 eV) 329 (M⁺, 16.2), 315 (18), 314 (100), 273 (19), 272 (39), 258 (15), 246 (15), 218 (15), 206 (24.0), 190 (12), 126 (24), 84 (15), 70 (46), 58 (18); high resolution MS calcd for C₁₇H₃₂NO₃P 329.2120, found 329.2122; TLC R_f 0.40 (benzene/acetone, 2:1).

[R,S](Pl,4'u)-3-tert-Butyl-6,6-dimethyl-2-(3',3',4'trimethyl-2'-oxo-5'-hexenyl)-2-oxo-1,3,2-oxazaphosphorinane (13)

Anionic (General Procedure 3.A): radial chromatography (hexane/acetone, 1:1) afforded 89.5 mg (94%) of 13; thermal: radial chromatography (hexane/acetone, 1:1) afforded 50.0 mg (65.8%) of 13. Data for 13 are reported for a distilled sample: bp 160 °C (0.05 mm Hg); ¹H NMR (500 MHz) Major diastereomer: 5.68-5.60 (m, 1 H), 5.09-5.00 (m, 2 H), 3.53 (dd, 1 H, J_P-H = 17.5, J_H-H = 15.7), 3.31-3.26 (m, 1 H), 3.16-3.10 (m, 1 H), 2.80 (dd, 1 H, J_P-H = 17.9, J_H-H = 15.7), 2.46-2.43 (m, 1 H), 1.97-1.88 (m, 2 H), 1.51 (s, 3 H), 1.38 (s, 9 H), 1.27 (s, 3 H), 1.11 (s, 3 H), 1.04 (s, 3 H), 0.93 (d, 3 H, J = 6.7). Minor diastereomer: 5.68-5.60 (m, 1 H), 5.09-5.00 (m, 2 H), 3.54 (dd, 1 H, J_P-H = 17.5, J_H-H = 15.7), 3.31-3.26 (m, 1 H), 3.16-3.10 (m, 1 H), 2.79 (dd, 1 H, J_P-H = 17.9, J_H-H = 15.7), 2.46-2.43 (m, 1 H), 1.97-1.88 (m, 2 H), 1.51 (s, 3 H), 1.36 (s, 9 H), 1.27 (s, 3 H), 1.11 (s, 3 H), 1.04 (s, 3 H), 0.89 (d, 3 H, J = 6.8); ¹³C NMR (75.5 MHz) 208.4 (d, J = 8.2, C(2')), 140.6 (C(5')), 115.6 (C(6')), 80.3 (d, J = 9.3, C(6)), 55.0 (d, J = 4.8, (CH₃)₃C), 51.2 (C(3')), 44.0, (C(4')), 42.7 (d, J = 127.5, C(1')), 40.1 (d, J = 7.0, C(4)), 39.1 (C(5)), 30.1 (CH₃C(6)), 29.4 (d, J = 7.0, CH₃C(6)), 29.0 ((CH₃)₃C); 21.4 (CH₃C(3')), 20.4 (CH₃C(3')), 14.9 (CH₃C(4')); ³¹P NMR (121.4 MHz) 16.1; IR (neat) 2975 (s), 1701 (s, C=O), 1636 (w), 1466 (m), 1389 (m), 1372 (m), 1291 (s), 1254 (s, P=O), 1233 (s), 1200 (s), 1103 (w), 1065 (m), 997 (s), 920 (s), 891 (m), 853 (w); MS (70 eV) 343 (M⁺, 7.91), 329 (13), 328 (68), 300 (11), 287 (10), 286 (46), 232 (10), 220 (13), 190 (22), 148 (16), 126 (22), 110 (49), 88 (10), 86 (65), 84 (100), 75 (14), 70 (22), 58 (13); high resolution MS calcd for C₁₈H₃₄NO₃P 343.2276, found 343.2288; TLC R_f 0.40 (benzene/acetone, 2:1).

Enantiomerically Enriched 1,3,2-Oxazaphosphorinane-2-oxides

(S)-3-tert-Butyl-3-hydroxybutyramide ((S)-14)

The enantiomeric composition of ethyl (S)-3-hydroxybutanoate was measured by two methods. Optical rotation gave an enantiomeric excess of 97% (${[\alpha ]}_{589}^{28}=+42.5$ (c 1.4; CHCl₃)). Analytical gas chromatographic separation of the Mosher esters²⁶ gave an enantiomeric excess of 96% (cOV-17, 165 °C, t_R = 14.5 (minor diastereomer), t_R = 14.9 (major diastereomer)). A 100-mL, 3-necked, flask equipped with septum, stirring bar, thermometer, and N₂ inlet was charged with 7.95 mL (75.7 mmol) of t-butylamine and cooled to 0°C. Trimethylaluminum (2 M in toluene, 37.8 mL, 75.7 mmol) was added slowly via syringe (0 °C – 10 °C). The mixture was warmed to room temperature for 30 minutes. After cooling to 0°C, 5.00 g (37.8 mmol) of ethyl(S)-3-hydroxybutanoate was added dropwise via syringe (0 °C – 10 °C). The mixture was warmed to room temperature and stirred for 20 hours. The mixture was cooled to 0 °C and water (50 mL) was added dropwise. CAUTION: Upon the initial addition of water (< 1 mL), there is an induction period of several minutes followed by rapid, exothermic evolution of methane. The mixture was acidified to pH~ 6 with 2 N HCl, filtered through a Buchner funnel and continuously extracted with Et₂O for 20 h. The organic layer was dried (K₂CO₃), filtered, and the solvent removed by rotary evaporation. The crude product was purified by column chromatography to afford 4.33 g (71.9%) of (S)-14 as white solid. Data for (S)-14 are given for a recrystallized sample: mp 89–90 °C ${[\alpha ]}_{589}^{28}=+14.2$ (c 1.4; CHCl₃); ¹H NMR (300 MHz), 5.53-5.37 (m, 1 H), 4.21-4.10 (m, 1 H), 3.92 (d, J = 2.9, 1 H), 2.28-2.14 (m, 2 H), 1.35 (s, 9 H), 1.20 (d, 3 H, J = 6.2); ¹³C NMR (75.5 MHz) 171.9 ((1)), 64.7 (C(3)), 51.2 ((CH₃)₃C), 44.4 (C(2)), 28.7 ((CH₃)₃C), 22.6 (C(4)); IR (CCl₄) 3443 (m) , 2971 (s), 1744 (w), 1698 (m), 1671 (s, C=O), 1511 (m), 1455 (m), 1420 (m), 1393 (w), 1366 (m), 1250 (w), 1198 (s), 1167 (s), 1142 (s), 928 (w); MS (70 eV) 159 (M⁺, 11.0), 144 (24), 86 (12), 59 (18), 58 (100), 57 (17); Anal. Calcd for C₈H₁₇NO₂ (159.23): C, 60.35; H, 10.76; N, 8.80. Found: C, 60.65; H, 10.99; N, 8.79.

Preparation of (S)-4-N-tert-Butylamino-2-butanol ((S)-15)

A 500-mL, 3-necked flask equipped with an addition funnel, thermometer, stirring bar, and an N₂ inlet was charged with 120 mL (1.0 M, 120 mmol) of BH₃ THF and was cooled to 3–5 °C. A solution of the amide (S)-52 (6.81 g, 42.8 mmol) dissolved in 50 mL of THF was added dropwise via addition funnel (3–7 °C). The mixture was stirred at 3–10 °C for 2.5 h and then warmed to room temperature and stirred for an additional 5 h. The mixture was cooled to 5 °C and 150 mL of 6 N HCl was cautiously added (N.B.: The first 5 mL was added 3–5 drops at a time maintaining the internal temperature <7 °C. The remainder was added dropwise (<20 °C) being careful to avoid a violent exothermic evolution of gas). The mixture was warmed to room temperature and stirred for 30 min and then cooled to 3–5°C. KOH pellets were added cautiously until the pH=11 maintaining a temperature below 40 °C. The aqueous layer, which contained some solid KOH, was extracted with 250 mL Et₂O. Water was then added to dissolve the solid KOH and the homogeneous aqueous layer was extracted with Et₂O (3 × 250 mL). The combined organic fractions were dried (K₂CO₃), filtered and concentrated by rotary evaporation. The crude product was purified by Kugelrohr distillation to afford 5.85 g (94.3%) of (S)-53 as a clear, colorless oil. Data for (S)-15: bp 55 °C (0.15 mm Hg); ¹H NMR (300 MHz) 3.99-3.92 (m, 1 H), 2.99 (dt, 1 H, J = 11.7, 4.0), 2.67 (td, 1 H, J = 11.4, 2.8), 1.67-1.59 (m, 1 H), 1.44-1.30 (m, 1 H), 1.15 (d, 3 H, J = 7.3), 1.09 (s, 9 H)); ¹³C NMR (75.5 MHz) 69.2 (C(2)), 50.1 ((CH₃)₃C), 41.0 (C(4)), 37.4 (C(3)), 28.4 ((CH₃)₃C), 23.3 (C(1)); IR (neat) 3278 (s), 2965 (s), 2928 (s), 2867 (s), 1655 (w), 1480 (s), 1443 (s), 1391 (m), 1364 (s), 1335 (m), 1231 (s), 1215 (s), 1138 (s), 1121 (s), 1094 (m), 1030 (m), 976 (m), 911 (m); MS ((70 eV) 145 (M⁺, 0.69), 130 (100), 112 (22), 72 (17), 70 (17), 58 (16); Anal. Calcd for C₈H₁₉NO (145.25): C, 66.15; H, 13.18; N, 9.65. Found: C, 66.12; H, 13.23; N, 9.71.

(S)-3-tert-Butyl-3-hydroxybutryamido-3',5'-dinitrophenyl carbamate ((S)-15’)

Compound (S)-15’ was prepared by the method of Pirkle.²⁹ Data for (S)-15’: ¹H NMR (300 MHz) 9.59-9.54 (m, 1 H), 8.84-8.81 (m, 2 H) 8.56-8.52 (m, 1 H) 6.84-6.79 (m, 1 H) 5.36-5.21 (m, 1 H) 2.53 (dd, 1 H, J = 7.2, 19.8), 2.38 (dd, 1 H, J = 7.2, 19.8),1.33 (d, 3 H, J = 7.8 Hz), 1.29 (s, 9 H); HPLC (column B; hexane/EtOAc, 4:1; 2 mL/min) t_R (S)-54, 10.50 min; >99.5%.

[S](Pl,6l)-3-tert-Butyl-2-(1',2'-butadienyl)-6-methyl-2-oxo-1,3,2-oxazaphosphorinane (cis-16) and [R](P l,6 u)-3-tert-Butyl-2-(1',2'-butadienyl)-6-methyl-2-oxo-1,3,2-oxazaphos-phorinane (trans-16)

A 250-mL, 3-necked flask equipped with stirring bar N₂ inlet, thermometer, and septum was charged with 100 mL of dry CH₂Cl₂ and cooled to −5 °C. Phosphorus trichloride (1.20 mL, 13.8 mmol) and N-methylmorpholine (1.51 mL, 13.8 mmol) were added sequentially via syringe. After stirring for 5 min, a solution of (±)-3-butyn-2-ol (1.08 mL, 13.8 mmol) in 5 mL of CH₂Cl₂ was added via syringe (0–5 °C). After stirring for 20 min, N-methylmorpholine (3.01 mL, 27.6 mmol) was added. After for an additional 5 min, a solution of (S)-N-tert-butyl-4-amino-2-butanol (1.08 mL, 13.8 mmol) in 5 mL of CH₂Cl₂ was added. The reaction mixture was warmed to room temperature and stirred for 6 h. The mixture was diluted with 100 mL of water and extracted with CH₂Cl₂ (3×100 mL). The organic layers were washed with water and brine (1×50 mL each). The combined organic layers were dried (MgSO₄), filtered, and the solvent removed by rotary evaporation. The crude product was twice purified by column chromatography (hexane/acetone, 3:1) to give 650 mg (19.5%) of trans-16 and 1.82 g (54.3%) of cis-16 as clear colorless oils. Data for cis-16 are reported for a distilled sample: bp 160 °C (0.03 mm Hg); ¹H NMR (300 MHz) 5.47-5.39 (m, 1 H), 5.38-5.23 (m, 1 H), 4.62-4.49 (m, 1 H), 3.31-3.03 (m, 2 H), 2.09-1.93 (m, 1 H), 1.76-1.61 (m, 4 H), 1.35 (s, 9 H), 1.32 (d, 3 H, J = 6.3); ¹³C NMR (75.5 MHz) 210.6 (C(2')), 210.3 (C2')), 54.7((CH₃)₃C), 40.3 (C(4)), 34.2 (C(5)), 29.1 ((CH₃)₃C), 29.0 ((CH₃)₃C), 12.5 (d, J = 6.9, CH₃C(6)), 12.1 (d, J = 7.0, CH₃C(6)). The following resonances could not be unambiguously assigned: 87.6, 87.5, 86.3, 86.1, 85.9, 85.7, 85.2, 71.9, 71.8, 71.7, 22.0, 21.9; ³¹P NMR (121.4 MHz), 13.1, 12.9; IR (neat) 2975 (s), 1952 (s, C=C=C), 1653 (w), 1445 (m), 1366 (s), 1279 (s), 1254 (s, P=O), 1204 (s), 1129 (s), 1044 (s), 1013 (s), 990 (s), 949 (s), 895 (m), 828 (m), 808 (m); MS (70 eV) 243 (M⁺, 3.20), 229 (12), 228 (100), 187 (49), 186 (26), 174 (16), 164 (11), 148 (10), 146 (14), 145 (14), 134 (12), 70 (20), 58 (29), 57 (12), 55 (10); high resolution MS calcd for C₁₂H₂₂NO₂P 243.1388, found 243.1385; TLC R_f 0.30 (hexane /acetone, 3:1). Data for trans-16 are reported for a distilled sample: bp 160 °C (0.03 mm Hg); ¹H NMR (500 MHz) 5.44-5.39 (m, 1 H), 5.33-5.26 (m, 1 H), 4.35-4.30 (m, 1 H), 3.32-3.23 (m, 1 H), 3.08-3.00 (m, 1 H), 1.80-1.73 (m, 5 H), 1.35 (s, 9 H), 1.34-1.32 (m, 3 H); ¹³C NMR (125.8 MHz), 210.6 (C(2')), 210.3 (C(2')), 55.7 ((CH₃)₃C), 42.4 (C(4)), 34.9 (C(5)), 29.4 ((CH₃)₃C). The following resonances could not be unambiguously assigned: 85.9, 85.8, 85.7, 84.7, 84.5, 84.2, 83.1, 82.9, 75.1, 75.0, 74.9, 22.5, 22.4, 13.2, 13.1, 13.0; ³¹P NMR (121.4 MHz), 9.1, 8.9; IR (neat) 2975 (s), 2724 (w), 1949 (s, C=C=C), 1638 (m), 1557 (w), 1474 (w), 1366 (s), 1321 (m), 1281 (s), 1256 (s, P=O), 1204 (s), 1152 (s), 1111 (s), 1046 (s), 1016 (s), 989 (s), 893 (s), 853 (s), 810 (s); MS (70 eV) 243 (M⁺, 4.59), 229 (13), 228 (100), 187 (41), 186 (14), 145 (13), 134 (13); high resolution calcd for C₁₂H₂₂N₂P 243.1388, found 243.1387; TLC R_f 0.39 (hexane /acetone, 3:1).

[S](Pl,6l)(E)-3-tert-Butyl-6-methyl-2-oxo-[2'-(2-propenyloxy)-2'-butenyl]-1,3,2-oxazaphosphorinane (cis-17)

A 100 mL, 3-necked flask equipped with stirring bar, septa, and N₂ inlet was charged with 50% NaH dispersion (221 mg, 4.60 mmol). The NaH dispersion was rinsed with hexane (3 × 2 mL) and suspended in 40 mL of dry THF. After 5 min, t-butanol (930 µL, 9.86 mmol) and allyl alcohol (307 µL, 4.52 mmol) were added sequentially via syringe. After 10 min, a solution of cis-16 (1.00 g, 4.11 mmol) in 2 mL of THF was added via syringe. After an additional 15 min, the reaction was quenched with 10 mL of water and extracted with EtOAc (50 mL). The water layer was separated and further extracted with CH₂Cl₂ (3 × 50 mL). The organic layers were washed with brine (1 × 10 mL), combined, dried (MgSO₄), filtered and the solvent removed by rotary evaporation. The crude product was purified by column chromatography (hexane/acetone, 2:1) to give 670 mg (54.0%) of cis-17 as a white solid. Data for cis-17: mp 42–44 °C; ¹H NMR (300 MHz) 6.02-5.89 (m, 1 H), 5.30 (dd, 1 H, J = 17.5, 1.2), 5.18 (dd, 1 H, J = 10.4, 1.0); 4.57-4.45 (m, 2 H), 4.21-4.09 (m, 2 H), 3.15-2.95(m, 3 H), 2.63 (dd, 1 H, J = 17.1, 15.4), 2.01-1.91 (m, 1 H), 1.65-1.51 (m, 4 H), 1.37 (s, 9 H), 1.27 (d, 3 H, J = 6.4); ¹³C NMR (75.5 MHz) 149.2 (d, J = 6.4, C(2')), 133.7 (CH=CH₂), 116.9 (CH=CH₂), 94.3 (d, J = 10.6, C(3')), 70.3 (d, J = 7.9, C(6)), 67.8 (OCH₂), 55.1 (d, J = 4.9, (CH₃)₃C), 40.5 (C(4)), 34.4 (C(5)), 34.1 (d, J = 131.7, C(1')), 29.4 ((CH₃)₃C), 22.3 (d, J = 9.0, CH₃C(6)), 12.2 (C(4')); ³¹P NMR (121.4 MHz) 24.4; IR (CCl₄) 2977 (m), 2932 (m), 1667 (m), 1458 (w), 1401 (w), 1364 (w), 1347 (w), 1277 (m), 1258 (s, P=O), 1233 (s), 1202 (s), 1146 (m), 1127 (m), 1102 (m), 1044 (m), 1009 (m), 992 (m), 947 (m), 893 (w); MS (70 eV) 301 (M⁺, 2.9), 286 (14), 260 (52), 204 (100), 162 (36), 110 (24), 70 (11); high resolution MS calcd for C₁₅H₂₈NO₃P 301.1807, found 301.1799; TLC R_f 0.35 (hexane/acetone, 1:1).

[S](Pu,6l)(E)-3-tert-Butyl-6-methyl-2-oxo-[2'-(2-propenyloxy)-2'-butenyl]-1,3,2- oxazaphosphorinane (trans-17)

A 15 mL, 3-necked flask equipped with stirring bar, septa, and N₂ inlet was charged with 50% NaH dispersion (97.1 mg, 2.02 mmol). The NaH dispersion was rinsed with hexane (3×1 mL) and suspended in 16 mL dry THF. After 5 min, t-butanol (380 µL, 4.03 mmol) and allyl alcohol (137 µL, 2.01 mmol) were added sequentially via syringe. After an additional 5 min, a solution of allene trans-16 (407 mg, 1.67 mmol) in 1 mL of THF was added via syringe. After stirring at room temperature for 15 min, the reaction was quenched with 10 mL of water and extracted with EtOAc (75 mL). The aqueous layer was separated and further extracted with CH₂Cl₂ (3×50 mL). The organic layers were washed with brine (1×10 mL), combined, dried (MgSO₄), filtered, and the solvent removed by rotary evaporation. The crude product was purified by column chromatography (hexane/acetone, 2:1) to give 220 mg (30.0%) of trans-17 as a white solid. Data for trans-17: mp 72–75 °C; ¹H NMR (500 MHz) 5.99-5.93 (m, 1 H), 5.31 (d, 1 H, J = 17.4); 5.20 (d, 1 H, J = 10.3); 4.58-4.55 (m, 1 H), 4.44-4.40 (m, 1 H), 4.21-4.15 (m, 2 H), 3.28-3.20 (m, 1 H), 3.05-2.95 (m, 1 H), 2.99 (dd, 1 H, J = 17.9, 15.1), 2.69 (dd, 1 H, J = 16.5, 15.7), 1.80-1.68 (m, 5 H), 1.35 (s, 9 H), 1.31 (d, 3 H, J = 6.2); ¹³C NMR (125.8 MHz) 163.0 (C(2')), 133.6 (CH=CH₂), 117.2 (CH=CH₂), 94.6 (d, J = 11.4, C(3')), 76.0 (d, J = 6.0, C(6)), 67.9 (OCH₂), 55.6 ((CH₃)₃C), 42.1 (C(4)), 35.1 (C(5)), 32.5 (d, J = 120.5, C(1')), 29.5 ((CH₃)₃C), 22.8 (d, J = 7.6, CH₃C(6)), 12.2(C(4'); ³¹P NMR (121.4 MHz) 19.9; IR (CCl₄) 2977 (s), 2936 (m), 2870 (m), 1711 (w), 1667 (m), 1553 (m), 1458 (m), 1399 (m), 1385 (m), 1366 (m), 1346 (w), 1281 (s), 1260 (s, P=O), 1202 (s), 1102 (s), 1048 (s), 1017 (s), 992 (s), 930 (m), 893 (m); MS (70 eV) 301 (M⁺, 6.16), 286 (54), 260 (55), 245 (12), 204 (100), 176 (15), 174 (12), 162 (36), 148 (10), 135 (36), 134 (15), 112 (20), 70 (27); high resolution MS calcd for C₁₅H₂₈NO₃P 301.1807, found 301.1816; TLC R_f 0.44 (hexane /acetone, 1:1).

[S](Pl,6l,3'l)-3-tert-Butyl-6-methyl-2-(3'-methyl-2'-oxo-5'-hexenyl)-2-oxo-1,3,2-oxazaphosphorinane (S)-(l,l)-18

Anionic (General Procedure 3.A.): radial chromatography (hexane/acetone, 1:1) afforded 470 mg (78.3%) of cis-57. Anionic (General Procedure 3.B.): radial chromatography (hexane/acetone, 1:1) afforded 28.5 mg (65.2%) of (S)-(l,l)-18. Data for (S)-(l,l)-18 are reported for a distilled sample: bp 160 °C (0.5 mm Hg); ¹H NMR (300 MHz) Major diastereomer: 5.80-5.67 (m, 1 H), 5.09-5.00 (m, 2 H), 4.60-4.54 (m, 1 H), 3.38 (dd, 1 H, J = 17.8, 13.3) 3.21-2.90 (m, 4 H), 2.46-2.35 (m, 1 H), 2.16-2.01 (m, 2 H), 1.70-1.59 (m, 1 H), 1.38 (s, 9 H), 1.30 (d, 3 H, J = 6.0), 1.10 (d, 3 H, J = 6.8). Minor diastereomer: 5.80-5.67 (m, 1 H), 5.09-5.00 (m, 2 H), 4.60-4.54 (m, 1 H), 3.39 (dd, 1 H, J = 17.8, 13.3) 3.21-2.90 (m, 4 H), 2.46-2.35 (m, 1 H), 2.16-2.01 (m, 2 H), 1.70-1.59 (m, 1 H), 1.38 (s, 9 H), 1.30 (d, 3 H, J = 6.0), 1.08 (d, 3 H, J = 6.6); ¹³C NMR (75.5 MHz) 206.5 (d, J = 6.8, C(2')), 135.7 (CH=CH₂), 116.7 (CH=CH₂), 70.7 (d, J = 7.1, C(6)), 55.3 (d, J = 3.2, (CH₃)₃C), 46.3 (C(3')), 45.5 (d, J = 116.1, C(1')), 40.1 (C(4')), 36.4 (C(5)), 34.5 (d, J = 3.5, C(4)), 29.3 ((CH₃)₃C), 22.2 (d, J = 2.2, CH₃C(6)), 16.0 (CH₃C(3')); ³¹P NMR (121.4 MHz) 19.2; IR (neat) 2975 (s), 2934 (m), 1707 (s, C=O), 1642 (w), 1458 (m), 1399 (m), 1366 (m), 1277 (s), 1254 (s, P=O), 1198 (s), 1125 (s), 1044 (s), 990 (s), 947 (m), 893 (m), 870 (m), 810 (m); MS (70 eV) 301 (M⁺, 5.1), 286 (76), 244 (33), 228 (12), 192 (39), 174 (19),148 (12), 124 (21), 112 (48), 93 (12), 70 (100), 55 (26), 41 (44); high resolution MS calcd for C₁₅H₂₈NO₃P 301.1807, found 301.1810; TLC R_f 0.28 (benzene/acetone, 2:1).

[S](Pu,6l,3'u)-3-tert-Butyl-6-methyl-2-(3'-methyl-2'-oxo-5'-hexenyl)-2-oxo-1,3,2-oxaza-phosphorinane (S)-(u,u)-18

Anionic (General Procedure 3.A.): radial chromatography (hexane/acetone, 1:1) afforded 35.4 mg (70.8%) of (S)-(u,u)-18 as a colorless oil. Data for (S)- (u,u)-18 are reported for a distilled sample: bp 160 °C (0.5 mm Hg); ¹H NMR (500 MHz) Major diastereomer: 5.78-5.70 (m, 1 H), 5.07-5.00 (m, 2 H), 4.44-4.37 (m, 1 H), 3.35-3.25 (m, 1 H), 3.30 (dd, 1 H, J = 17.0, 13.0), 3.11-3.00 (m, 3 H), 2.47-2.42 (m, 1 H), 2.13-2.07 (m, 1 H), 1.86- 1.73 (m, 2 H), 1.34 (s, 1 H), 1.34-1.31 (m, 3 H), 1.12 (d, 3 H, J = 7.1). Minor diastereomer: 5.78- 5.70 (m, 1 H), 5.07-5.00 (m, 2 H), 4.44-4.37 (m, 1 H), 3.35-3.25 (m, 1 H), 3.30 (dd, 1 H, J = 17.0, 13.0), 3.11-3.00 (m, 3 H), 2.47-2.42 (m, 1 H), 2.13-2.07 (m, 1 H), 1.86-1.73 (m, 2 H), 1.34 (s, 1 H), 1.34-1.31 (m, 3 H), 1.11 (d, 3 H, J = 7.1); ¹³C NMR (125.8 MHz) 207.2 (C(2')), 135.7 (C(5')), 116.8 (C(6')), 67.1 (d, J = 9.8, C(6)), 56.0 ((CH₃)₃C), 43.7 (d, J = 105.4, C(1')), 36.5 (C(4')), 35.2 (C(5')), 29.4 ((CH₃)₃C), 22.6 (d, J = 7.3, CH₃C(6)), 16.1(CH₃C(3')); ³¹P NMR (121.4 MHz) 14.7; IR (neat) 2977 (s), 1703 (s, C=O), 1642 (w), 1458 (w), 1368 (m), 1283 (s), 1256 (s, P=O), 1196 (m), 1152 (w), 1109 (m), 1043 (s), 1017 (s), 992 (s), 972 (m), 893 (m), 862 (w); MS (70 eV); 301 (M⁺, 5.25), 287 (16), 286 (100), 244(17), 192(16), 112 (24), 70 (34); high resolution MS calcd for C₁₅H₂₈NO₃P 301.1807, found 301.1811; TLC R_f 0.35 (benzene/acetone, 2:1).

Degradation and Assignment of Rearrangement Products. Silyl Enol Ethers. General Procedure 5

A 15-mL, 3-necked flask equipped with septa, stirring bar and N₂ inlet was charged with a solution of the Claisen rearrangement product (20.2 mg, 0.0613 mmol) in 0.5 mL THF. The solution was cooled to −78 °C. A solution of potassium hexamethyldisilazide in THF (0.815 M, 165 µL, 0.134 mmol) was added via syringe. After two minutes, a solution t-butyldimethylsilyl chloride in THF (22.8 mg, 0.151 mmol) was added via syringe. After stirring for 30 min at −78 °C, the solution was warmed to room temperature, transferred to a centrifuge tube, and centrifuged for 5 min at 5000 rpm. The supernatant was transferred to a round-bottomed flask and the solvent removed. The resulting silyl enol ether was carried on without further purification.

Preparation of Dimethyl Succinates from Degradation

A-15 mL, 3-necked flask equipped with stirring bar, thermometer, and ozone inlet and outlet tubes (outlet tube connected to a bubbler containing a 0.1 N KI solution) was charged with 3 mL of methyl acetate and cooled to −10 °C. A solution of the silyl enol ether (0.0613 mmol) in 1 mL methyl acetate was added via syringe. Ozone was bubbled through the solution until the blue color persisted. Hydrogen peroxide (30%, 0.5 mL) and formic acid (88%, 1 mL) were then added sequentially. The solution was stirred at room temperature for 16 h and then the solvents were removed in vacuo. The residue was taken up in 3% NaOH and extracted with Et₂O (2×15 mL). The water layer was acidified with 2 N H₂SO₄ and continuously extracted with Et₂O for 16 h. The Et₂O layer was dried (Na₂SO₄) filtered, and the solvents removed. The resulting 2,3-dimethylsuccinic acids were esterified by treatment with an ethereal solution of diazomethane. The dimethyl 2,3-dimethylsuccinates were compared to authentic meso-and d,1-succinates by capillary GC analysis. The syn diastereomer was the major diastereomer present. GC (COV-17; 80°C isothermal) t_R (d,l), 9.95 min; t_R (meso), 11.14 min.

(S)[Pl,6l,3'l)-3-tert-Butyl-6-methyl-2-(3'-methyl-2'-tert-butyldimethylsiloxy-1,5-hexadienyl)-1,3,2-oxazaphosphorinane (cis-19)

General Procedure 5. Data for cis-19: ¹H NMR (300 MHz) 5.80-5.71 (m, 1 H), 5.08-5.01 (m, 2 H), 4.61 (d, 1 H, J = 15.4), 4.15-4.10 (m, 1 H), 3.29-3.20 (m, 1 H), 2.83-2.91 (m, 1 H), 2.42-2.38 (m, 1 H), 2.23-2.18 (m, 1 H), 2.15-2.07 (m, 1 H), 1.34 (s, 9 H) 1.29 (d, 3 H, J = 6.0), 1.08 (d, 3 H, J = 8.1), 0.97 (s, 9 H), 0.30 (s, 3 H), 0.23 (s, 3 H). Ozonolysis/esterification of cis-19 afforded 17.3 mg (59.3%) of dimethyl (S)-(+)-methylsuccinate as assigned above. The ¹H NMR spectrum was identical to that from an authentic sample of the racemate.

(R)[Pl,6u,3'l)-3-tert-Butyl-6-methyl-2-(3'-methyl-2'-tert-butyldimethylsiloxy-1,5-hexadienyl)-1,3,2-oxazaphosphorinane (trans-19)

General Procedure 5. Data for trans-19: ¹H NMR (300 MHz) 5.82-5.70 (m, 1 H), 5.10-5.01 (m, 2 H), 4.66 (d, 1 H, J = 15.4), 4.62-4.53 (m, 1 H), 3.19-3.04 (m, 1 H), 2.45-2.35 (m, 1 H), 2.42-2.38 (m, 1 H), 2.23-2.18 (m, 1 H), 2.15-2.07 (m, 1 H), 1.34 (s, 9 H) 1.33 (d, 3 H, J = 6.1), 1.08 (d, 3 H, J = 6.1), 0.99 (s, 9 H), 0.27 (s, 6 H). Ozonolysis/esterification of trans-19 afforded 3.7 mg (24.1%) of dimethyl (R)-(+)-methylsuccinate as assigned above. The ¹H NMR spectrum was identical to that from an authentic sample of the racemate.

Chiral shift agent (CSA) study of Dimethyl (R)- and (S)-Methylsuccinates

To a solution of the appropriate dimethyl methylsuccinate (0.25 M) in CDCl₃, was added (R)-2,2,2- trifluoro-1-(9-anthryl)ethanol in increments of 1.0 equivalents until separation was observed by ¹H NMR spectroscopy (500 MHz). The methoxy groups are diagnostic: for dimethyl (R)- methylsuccinate 3.649 and 3.625 (2 s); for dimethyl (S)-methylsuccinate 3.646 and 3.620 (2 s). These assignments were verified by spiking with an authentic sample of dimethyl (R)- methylsuccinate.