Rh(I)-Catalyzed Cyclocarbonylation of Allenol Esters to Prepare Acetoxy 4-Alkylidene Cyclopent-3-en-2-ones

Abstract

A Rh(I)-catalyzed cyclocarbonylation reaction of allenol esters has been examined and its synthetic viability established for the conversion of trisubstituted allenes to bicyclo-[4.3.0], and -[5.3.0] skeletons possessing an α-acetoxy cyclopentadienone. Tetrasubstituted allenol acetates gave elimination products, providing examples of a cyclocarbonylation reaction between an alkyne and a latent cumulene or cumulene equivalent. Cleavage of the acetate affords a free hydroxyl group illustrating the utility of this method for accessing α-hydroxy carbonyls from allenol esters.

Introduction

α-Hydroxy carbonyls are important building blocks in organic synthesis and are present in a number of biologically active molecules.¹ A variety of methods exist for forming α-hydroxyl carbonyls, the most commonly used protocol being the oxidation of enol ethers.^1c In view of incompatibilities of some functional groups to these oxidation conditions,^1b the regioselectivity requirements for the enolization step,² and redox economy considerations,³ synthetic alternatives to this late stage enolization/oxidation strategy would be useful.

The Pauson-Khand reaction provides a powerful method for the preparation of highly functionalized cyclopentenone-containing compounds.⁴ However, missing from the cyclocarbonylation reaction arsenal is efficient ways of accessing an α-hydroxycarbonyl moiety via an enol ether or vinyl ester precursor. Shore demonstrated that alkoxycyclopentenones could be prepared from enol ether precursors⁵ using a Pauson-Khand reaction and subsequently a number of groups have rendered this reaction asymmetric.⁶ However, relatively harsh conditions are required for the conversion of the alkoxycyclopentenone to a hydroxycyclopentenone, thus the alkoxy group serves as a control element in the reaction and then is typically removed.⁶ The Co₂(CO)₈-mediated Pauson–Khand reaction of vinyl esters to form an α-acetoxy- or α-benzyloxy-cyclopentenones, results in concurrent loss of the ester moiety through a proposed single electron reduction.⁷

The Rh(I)-catalyzed cyclocarbonylation reaction⁸ of allene-ynes is an efficient method for synthesizing a variety of alkylidene cyclopentenones including bicyclo[5.3.0]undecadienones, a long sought after ring system previously inaccessible to cyclocarbonylation methodologies.⁹ It was hypothesized that a Rh(I)-catalyzed cyclocarbonylation reaction of allenol acetates affording α-acetoxy cyclopentadienones may be possible due to the mildness of the reaction conditions and the unlikely prospect of the rhodium(I) catalyst participating in a single electron transfer process under these reaction conditions.¹⁰ Moreover, allenol acetates are well-known and are prepared via a formal [3,3]-sigmatropic rearrangement of a propargyl acetates using a variety of transition metal catalysts such as Ag, Au, Cu, Pt, and Rh.¹¹ Thus, readily available allenol acetates afford an opportunity to directly access α-acetoxy cyclopentadienones, which in turn can be used as precursors to α-hydroxycarbonyl-containing compounds. Herein, a Rh(I)-catalyzed cyclocarbonylation reaction to form α-acetoxycyclopentadienones is reported.

Results and Discussion

Substrate Design

Once the feasibility of the cyclocarbonylation reaction of allenol acetates to produce α-acetoxy carbonyls has been established, our research objective is to examine several cycloaddition substrates. Guided by a plethora of natural product substructures that would benefit from this reaction: 1) the chain length of the tether between the allene and alkyne will be varied (n = 1-4); 2) substitution on the allene, alkyne and tether altered; and 3) stereochemical consequences of the [3,3]-sigmatropic rearrangement of 1 to give 2 and the cyclocarbonylation reaction to give 3 will be examined (Scheme 1). Furthermore, imbedding the allene into a conformationally anchored cyclohexane ring will serve as an effective method for examining diastereoselective issues.

Scheme 1.

Substrate Design

Preparation of Propargyl Acetates

Propargyl acetates were prepared using two general procedures. Preparation of cyclohexane-based substrates began by alkylating the lithium enolate of dimethyl hydrazone 4, with the corresponding halides or triflates 5a-c (Scheme 2).¹² Acidic hydrolysis with oxalic acid gave ketones 6a and 6c (n = 1, 3) in 44% and 83% yield.¹³ Ketone 6b was obtained in only 5% yield, possibly due to a competing E2 elimination of the triflate to form an enyne. The diastereomers of 6a-c were separated via column chromatography and the major diastereomer was carried forward. Addition of ethynyl or 1-propynylmagnesium bromide to 6a-c gave the corresponding propargyl alcohols in 63-90% yields with diastereoselectivities ranging from 1:1 to 3:1. The major diastereomers were assigned based upon the predisposition of small nucleophiles to add axially to substituted cyclohexanones.¹⁴ Separation of the two diastereomers was readily accomplished via column chromatography. The major diastereomers were acetylated using triethylamine, DMAP, and acetic anhydride yielding a single diastereomer of propargyl acetates 7a-d in 39-47% yield from ketones 6a-c. Two substrates were prepared to examine electronic and steric effects of the carboxy group. A bulky pivaloyl group was appended to the corresponding propargyl alcohol from 6c using trimethylacetic anhydride and catalytic Sc(OTf)₃ to give 7e in 81% yield.¹⁵ An electron withdrawing p-nitrobenzoate was attached to the same propargyl alcohol using 4-nitrobenzoyl chloride and DMAP to give 7f in 74% yield.¹⁶

Scheme 2.

Preparation of Propargyl Acetates 7a-f

Linear propargyl acetates were prepared using two different methods (Scheme 3). Propargyl acetates 9a-d (n = 1, 2) were prepared by addition of ethynylmagnesium bromide or 1-propynylmagnesium bromide to aldehyde 8a or ketones 8b-c followed by in situ acetylation with acetyl chloride furnishing the desired products in 55-83% yields.¹⁷ For propargyl acetate 9e (R² = n-Pr) addition of the lithium acetylide of 1-pentyne to ketone 8c gave the propargyl alcohol, which was then acetylated using triethylamine, DMAP, and acetic anhydride to obtain propargyl acetate 9e in 49% yield over two steps. Propargyl acetates 9f-g were prepared by reacting a slight excess of the sodium salt of methyl acetoacetate with iodides 5c and 5d.¹⁸ After an aqueous work up, the crude material was subjected to Krapcho decarboxylation by heating the ketoester to 130 °C with lithium chloride in wet DMSO to obtain 8-(trimethylsilyl)oct-7-yn-2-one or 9-(trimethylsilyl)non-8-yn-2-one.¹⁹ Purification of the ketones proved impractical and as a result the crude material was subjected to alkynylation with ethynylmagnesium bromide. After chromatographic purification, the alcohols were acetylated using triethylamine, DMAP, and acetic anhydride to obtain propargyl acetates 9f and 9g in 38 and 22% yields over 4 steps.

Scheme 3.

Preparation of Propargyl Acetates 9a-g

Formation of Allenol Acetates From Propargyl Acetates

Previously reported conditions to form allenol acetates were screened for efficiency and diastereoselectivity. Reacting a single diastereomer of 7c with AuCl₃ at 60 °C (entry 1, Table 1) gave complete conversion to the allene-yne to 10c in 30 min in a 1:1 diastereomeric ratio (dr).^11k Performing this same reaction at room temperature with AuCl₃ gave 10c as a 1:1 mixture of diastereomers in 30 min (entry 2). Upon lowering the temperature to −30 °C, the reaction took 5 h to go to 55% completion and afforded 10c a 1:1 dr (entry 3). Increasing the temperature to 90 °C decreased the reaction time to give full conversion to allenol acetate 10c in 12 min in a 1:1 dr (entry 4). Two Ag(I) catalysts, AgBF₄ and AgSbF₆, were examined (entries 5-8) with no significant changes in diastereoselectivity, but significant loss of the TMS group was observed.^11f,i Using [Rh(OCOCF₃)₂]₂ gave only a trace amount of product after 14 h (entry 9).^11j Using the PtCl₂ conditions reported by Malacria,^11h clean conversion to 10c was observed with a 1:1 dr (entry 10). Among the conditions examined, AuCl₃ afforded the allenol acetates in the highest yield, and shortest reaction times. The reaction temperature of the AuCl₃-catalyzed process had no effect on the diastereoselectivity.

Table 1. Catalyst Screening for Rearrangement.

entry	catalyst	mol %	T (°C)	time (h)	conversion% (isolated %)a	drb
1	AuCl3	10	60	0.5	100 (71)	1:1
2	AuCl3	10	rt	0.5	100 (74)	1:1
3	AuCl3	20	−30	5	55	1:1
4	AuCl3	10	90	0.2	100 (62)	1:1
5	AgBF4	50	rt	4	84 c	1:1
6	AgSbF6	10	60	3	83	2:1
7	AgSbF6	10	rt	19	90c	1:1
8	AgSbF6/PPh3	20	rt	17	73	1:1
9	[Rh(OCOCF3)2]2	2	60	14	2	1:1
10	PtCl2	10	40	96	100	1:1

^aConversion determined by ¹H NMR

^bDiastereomeric ratio determined comparing allenyl protons by ¹H NMR

^cSignificant loss of TMS observed by ¹H NMR

Interestingly, reacting both diastereomers of 7c independently with AuCl₃ at room temperature showed little difference in reaction time, yield, or dr (entries 1 and 2, Table 2). However, when using [Rh(OCOCF₃)₂]₂ significant differences in reactivity for the two diastereomers were observed (compare entries 3 and 4). With the alkyne cis to the t-butyl group only trace amounts of product were observed after 72 h (entry 3). Conversely, the trans diastereomer showed complete conversion to 10c in 4 h in 94% yield (entry 4).

Table 2. Au(III)- and Rh(II)-Catalyzed Rearrangement.

entry	R1	R2	catalyst (mol %)	time (h)	yield (%)	dr
1	CCH	OAc	AuCl3 (10)	0.5	74	1:1
2	OAc	CCH	AuCl3 (10)	0.5	81	1:1
3	CCH	OAc	[Rh(OCOCF3)2]2 (2)	72	2a	1:1
4	OAc	CCH	[Rh(OCOCF3)2]2 (2)	4	94	1:1

^aApproximate conversion by ¹H NMR

Similar results were seen when 11 (7:1 dr) was reacted with [Rh(OCOCF₃)₂]₂ (Scheme 4). After 5 h at 60 °C, the minor diastereomer possessing a trans relationship between the alkyne and the t-butyl group was completely converted to allene 12, and no change was observed for the major diastereomer of 11, based upon ¹H NMR. Thus the stereochemistry of the propargyl acetate significantly impacts the reaction time and efficiency of the rearrangement when using [Rh(OCOCF₃)₂]₂, with the axially oriented alkyne being the slowest and least efficient. It is postulated that developing 1,3-diaxial interactions of the alkyne coordinated to [Rh(OCOCF₃)₂]₂ slow this reaction.

Scheme 4.

[Rh(OCOCF₃)₂]₂ Catalyzed Rearrangement of 11

With optimized rearrangement conditions in hand, propargyl acetates 7a-c were transformed to the trisubstituted allenol acetates 10a-c using AuCl₃ in 74–84% yield as 1:1 mixtures of diastereomers (entries 1-3, Table 3). Tetrasubstituted allenol acetate 10d was isolated in 54% yield due to the incomplete consumption of starting material and the relative instability of 10d evidenced by decomposition within 4 h of standing in CDCl₃ at rt (entry 4). Reacting pivaloate ester 7e to AuCl₃ readily formed allene-yne 10e in 1 h (91%, 2:1 dr, entry 5). Subjecting p-nitrobenzoate ester 7f to AuCl₃-catalyzed conditions gave allene-yne 10f in 87% yield in a 1:1 dr (entry 6). Reacting the acyclic secondary propargyl acetate 9a to the AuCl₃-catalyzed reaction conditions yielded only trace amounts of 1,3-disubstituted allene even if the reaction temperature was increased to 60 °C; [Rh(OCOCF₃)₂]₂ and PtCl₂ conditions were also tried, but to no avail (entry 7). Tertiary acyclic propargyl acetates 9b and 9c (entries 8 and 9) rearranged to give the desired allenol acetates 13b and 13c in 46% and 67% yields. Longer reaction times were required when compared to the analogous cyclohexane based propargyl acetates (compare entries 1 and 2 to entries 8 and 9). Substituting an alkyl group for the proton on the terminus of the alkyne had little effect on the yield for the rearrangement of the linear system compared to the analogous cyclohexane based system (compare entries 9 and 10 to 14 and 15). Propargyl acetates 9f and 9g (n = 3, 4) reacted in significantly shorter reaction times and gave 13f and 13g in 78 and 76% yields, respectively (entries 12 and 13). The reaction time was dependent upon the tether length; for example, the more removed the appended alkyne from the propargyl acetate, the faster the reaction. It is postulated that increasing the distance of the appended alkyne from the ring minimizes coordination of the gold catalyst allowing for more rapid catalyst turnover. Reacting propargyl acetate 14a (R = H)^9a to AuCl₃ gave allene-yne 15a in near quantitative yield as a 1.3:1 dr (entry 14). However, treating propargyl acetate 14b (R = Me)^9a to AuCl₃ gave allene-yne 15b in a 43% yield (entry 15). The low yield is attributed to slow reaction times, because after 19 h propargyl acetate 14b was still observed in the crude ¹H NMR along with the appearance of a byproduct containing alkene resonances by ¹H NMR.

Table 3. Preparation of Allenol Acetates^a.

entry	propargyl acetate	time (h)	isolated yield (%)	drc
1b	7a n = 1 R1 = H, R2 = Me	3	84	1:1
2 b	7b n = 2 R1 = H, R2 = Me	0.5	80	1.1:1
3 b	7c n = 3 R1 = H, R2 = Me	0.5	74	1:1
4 b	7d n = 3 R1 = Me, R2 = Me	5	53	1.2:1
5	7e n = 3 R1 = H, R2 = t-Bu	1	91	1.4:1
6	7f n = 3 R1 = H, R2 = p-NO2-C6H4	16	87	1.4:1
7	9a n = 1, R1 = H, R2 = H	20	tracec	NA
8	9b n = 1, R1 = Me, R2 = H	19	46	NA
9	9c n = 2, R1 = Me, R2 = H	4	67d	NA
10	9d n = 2, R1 = R2 = Me	1.5	64	NA
11	9e n = 2, R1 =Me, R2 =n-Pr	1.5	58	NA
12	9f n = 3, R1 = Me, R2 = H	0.5	79	NA
13	9g, n = 4, R1 = Me, R2 = H	0.5	77	NA
14 e	14a R = H	4	quant	1.3:1
15f	14b R = Me	19	43	1.5:1

^aConditions: AuCl₃ (10 mol %), toluene, rt

^bReacted as a single diastereomer

^cObserved by ¹H NMR

^dContaminated with unknown impurity

^eReacted as a 4:1 dr

^fReacted as a 5:1 dr

Rh(I)-Catalyzed Cyclocarbonylation Reaction of Allenol Acetates to Form α-Acetoxy Cyclopentadienones

Next, the feasibility, scope and limitation studies of the Rh(I)-catalyzed cyclocarbonylation reaction of allenol acetates for the formation of bi- and tricyclic ring systems were explored (Table 4). Previously developed and optimized conditions were used to effect the cyclocarbonylation reaction of allenol acetates.⁹ Reaction of allenol acetate 10a to the standard Rh(I) cyclocarbonylation conditions gave only a 19% yield of 16a after 8 h (entry 1). Allene-ynes 10b and 10c underwent cyclocarbonylation to produce 16b and 16c in 67% and 76% yields, respectively (entries 2 and 3). Formation of the [6-7-5] ring system took significantly longer than the analogous [6-6-5] ring system (17 h vs 1 h). Subjecting 10e to cyclocarbonylation conditions gave a 2.3:1 dr of α-acetoxy 4-alkylidene cyclopentadienone 16e in 51% yield (entry 4). Cyclocarbonylation of 10f gave the cyclized product 16f as a 1.8:1 mixture of diastereomers in 34% yield (entry 5). Thus, it appears that neither steric nor electronic changes in the carboxy group significantly impact the diastereomeric ratio of the products. Reaction of linear allene-yne 13b (entry 6) gave 17b in 28% yield along with significant decomposition products, indicating that either the starting material or the bicyclic [5-5] ring systems are unstable to the reaction conditions. Cyclocarbonylation of allene-yne 13c and 13f are readily accomplished giving 17c and 17f in 53% and 61% yields, respectively (entries 7 and 8). The reaction of 13g to produce an [8-5] ring system resulted in only trace amounts of the product with the majority of allenol ester rearranging to the conjugated dienol acetate (entry 9).²⁰ Cyclocarbonylation of allenol-acetate 15a gave the α-acetoxy 4-alkylidene cyclopentenone 19a in 19 h and 74% yield.

Table 4. [Rh(CO)₂Cl]₂ Catalyzed Cyclocarbonylation Reactions^a.

entry	allene-yne	time (h)	yield (%)	drb
1	10a n = 1, R1 = Me 1:1 dr	8	19	1.3:1
2	10b n = 2, R1 = Me 1.1:1 dr	1	67	1:1.2
3	10c n = 3, R1 = Me 1:1 dr	17	76	1.9:1
4	10e n = 3, R1 = t-Bu 2:1 dr	18	51	2.3:1
5	10f n = 3, R1 = p-NO2-C6H4 1.4:1 dr	18	35	1.8:1
6	13b n = 1	22	28	NA
7	13c n = 2	2	53	NA
8	13f n = 3	46	62	NA
9	13g n = 4	120	tracec	NA
10	15a 1.3:1 dr	19	74	2.3:1

^aConditions: [Rh(CO)₂Cl]₂ (10 mol %), CO(g) (1 atm), toluene, 90 °C.

^bDiastereomeric ratio determined comparing proton resonances by ¹H NMR.

^cObserved by ¹H NMR.

Cyclocarbonylation reactions of tetrasubstituted allenes of 10d and 15b afforded a 26% yield of 20 in 69 h and a 45% yield of 21 in 47 h, respectively (Scheme 5). Linear allenol acetate 13d gave a 49% yield of trienone 22d, along with a trace amount of the desilylated product observed in the crude ¹H NMR. Similarly, allenol acetate 13e afforded triene 22e in 42% yield along with desilyated 22e′ in 13% yield. Compounds 20, 21, and 22d,e proved relatively unstable with observable decomposition by ¹H NMR after 24 h in a freezer. These are examples of a cyclocarbonylation reaction between an alkyne and a latent cumulene or cumulene equivalent. Two possible mechanisms for the formation of these compounds have been considered. One involves elimination of the acetoxy group prior to cyclocarbonylation and the other mechanism involves elimination of the acetoxy group after the cyclocarbonylation process. Studies are underway to elucidate the reaction pathway and synthetic utility of this transformation.

Scheme 5.

The [Rh(CO)₂Cl]₂ Catalyzed Cyclocarbonylation Reaction of Tetrasubstituted Allenol Esters

The cyclocarbonylation of allene-ynes 24a and 24c gave cycloadducts 25a and 25c in 32% and 70% yields (Scheme 6), paralleling the yields obtained for analogous [6-5-5] and [6-7-5] ring systems 16a and 16c (Table 4, entries 1 and 3) suggesting that the acetoxy group has little influence on the yields of the cyclocarbonylation. Conversely, the reaction times for the cyclocarbonylation of 24a and 24c were 30 and 40 min, compared to 8 and 17 h for 16a and 16c showing that the acetoxy group slowed the cyclocarbonylation reaction considerably.

Scheme 6.

Preparation and Cyclocarbonylation of Allene-ynes 24a and 24c

Diastereoselective Considerations

To probe the origin of the slight increase in diastereomeric ratios (1:1 to 2:1) in the transformation of 10c to 16c, the cyclocarbonylation of diastereomerically enriched allene-yne 10c was performed in d₈-toluene and monitored via ¹H NMR (Table 5). Starting with 3:1 dr (or 1:3) of 10c, rapid isomerization of the allenol acetate was observed giving a 1:1 mixtures of allenol acetates in 40 min with no evidence of 16c (entries 1 and 2). Performing the reaction at room temperature slowed the rate of isomerization but still resulted in a 1:1 dr of 10c after 7 h with no evidence of 16c (entries 3 and 4). Heating allene 10c in the absence of rhodium catalyst for 24 h at 90 °C resulted in no change in dr. Thus, under the Rh(I)-catalyzed cyclocarbonylation reaction conditions, rapid isomerization of the allene of 10c is occurring. Moreover, subjecting a 5 : 1 dr of 16c to the cyclocarbonylation reaction conditions afforded 16c with no change in the dr after 24 h. Therefore, the diastereoselectivity occurs by a selective transformation of one allenol acetate to the cyclocarbonylation product over the other.

Table 5. Isomerization of Allenol Acetate 10c.

entry	starting dr (major/minor)	catalyst (mol %)	T (°C)	timea
1	3:1	10	90	40 minb
2	1:3	10	90	40 minb
3	3:1	10	rt	7 hb
4	1:3	10	rt	7 hb
5	3:1	0	90	–c
6	1:3	0	90	–c

^aTime indicating a 1:1 dr

^bPerformed under 1 atm CO atm

^cNo epimerization observed after 24 h under Ar.

Finally, reacting acetate 16c with K₂CO₃ in MeOH/H₂O gave alcohol 26 in 60% yield (Scheme 7), thus demonstrating the synthetic utility of the Rh(I)-catalyzed allenic cyclocarbonylation reaction for accessing α-hydroxy containing cyclopentadienones.

Scheme 7.

Deprotection of 16c

Conclusions

We have demonstrated the first Rh(I)-catalyzed cyclocarbonylation reaction for the formation of α-acetoxy 4-alkylidene cyclopentenones from both cyclohexane derived and linear allene-ynes. Cyclohexane and linear allene-ynes [6-5] and [7-5] ring systems were prepared in good yields, however [5-5] ring systems proved less successful. Control experiments confirmed that yields for the cyclocarbonylation reactions for acetoxy-containing allenes parallel that of hydrocarbon-only allenes, but that the former reactions are significantly decelerated. Liberation of the acetate to the free alcohol was also readily accomplished yielding an α-hydroxy ketone. Studies are underway to expand the scope of this reaction and to render the cyclocarbonylation reaction stereoselective.

Experimental Section

General Procedure for the AuCl₃ Catalyzed Allenol Ester Formation

A flame-dried, 5 mL round-bottomed flask equipped with a Teflon-coated stir-bar was charged with AuCl₃ (0.1 equiv) in a nitrogen filled glove box. The flask was removed from the glove box, wrapped in aluminum foil, and placed under N₂. A solution of propargyl acetate in toluene (0.2 M, toluene degassed by bubbling with nitrogen for ∼5 min) was added rapidly via cannula. The reaction was stirred at rt in a darkened hood. When the reaction was complete as observed by TLC, the mixture was passed through a plug of silica gel using hexanes/EtOAc and concentrated in vacuo.

3-Methyl-9-(trimethylsilyl)nona-1,2-dien-8-ynyl acetate (13f)

Following the General Procedure for the AuCl₃ Catalyzed Allenol Ester Formation, AuCl₃ (4 mg, 0.01 mmol) and propargyl acetate 9f (32 mg, 0.12 mmol) were reacted in toluene (0.61 mL) for 30 min. Purification via flash chromatography (hexanes/EtOAc, 9:1, v/v) afforded the title compound (25 mg, 79%) as a colorless oil. ¹H NMR (300 MHz, CDCl₃) δ 7.30 (sext, J = 2.1Hz, 1H), 2.26-2.20 (m, 2H), 2.16-2.02 (m, 2H), 2.13 (s, 3H), 1.83 (d, J = 2.1 Hz, 3H), 1.61-1.49 (m, 4H), 0.14 (s, 9H); ¹³C NMR (75 MHz, CDCl₃) δ 189.4, 168.8, 115.7, 109.7, 107.2, 84.6, 34.6, 28.0, 26.3, 20.9, 20.4, 19.6, 0.1; IR: 3065, 2943, 2862, 2174, 1976, 1750, 1456, 1369, 1249, 1215, 1066, 1039, 843 cm⁻¹; MS m/z (relative intensity): 249 (36%, M-CH₃), 222 (26%), 117 (100%), 84 (88%), 75 (79%), 73 (95%); HRMS-EI (m/z): [M – CH₃]⁺ calcd for C₁₄H₂₁O₂Si, 249.1311; found, 249.1308.

General Procedure for the [Rh(CO)₂Cl]₂ Catalyzed Cyclocarbonylation Reaction

A flame-dried test tube (10 × 100 mm) equipped with a Teflon-coated stir-bar was charged with allene-yne and toluene (0.1 M). The tube was evacuated for 3-5 s and refilled with CO(g) (3×). To the allene-yne solution was added [Rh(CO)₂Cl]₂ (0.10 equiv) in one portion and the test tube was evacuated and refilled with CO(g) (3×). The test tube was placed in a preheated 90 °C oil bath and stirred under CO(g). After the reaction was complete by TLC, the mixture was cooled to rt, passed through a short plug of silica gel using hexanes/EtOAc and concentrated in vacuo. The crude material was purified by flash chromatography.

(8E)-1,2,4,5,6,7-Hexahydro-8-methyl-3-(trimethylsilyl)-2-oxoazulen-1-yl acetate (17f)

Following the General Procedure for the [Rh(CO)₂Cl]₂ Catalyzed Cyclocarbonylation Reaction, allene-yne 13f (27 mg, 0.10 mmol) and [Rh(CO)₂Cl]₂ (3 mg, 0.008 mmol) were reacted in toluene (1.0 mL) for 46 h. Purification via flash chromatography (hexanes/EtOAc, 95:5, v/v) afforded the title compound (18 mg, 62%) as a slightly yellow oil. ¹H NMR (300 MHz, CDCl₃) δ: 5.73 (s, 1H), 2.87 (dt, J = 14.7, 5.4 Hz, 1H), 2.81-2.70 (m, 1H), 2.51-2.32 (m, 2H), 2.15 (s, 3H), 1.94-1.72 (m, 4H), 1.85 (s, 3H), 0.25 (s, 9H); ¹³C NMR (75 MHz, CDCl₃) δ: 204.4, 181.8, 170.0, 143.5, 136.8, 134.5, 73.5, 34.1, 30.6, 25.9, 24.1, 23.7, 20.7, −0.2; IR: 2937, 2865, 1746, 1697, 1528, 1369, 1246, 1224, 1049, 842.1 cm⁻¹; MS m/z (relative intensity): 292 (14%, M⁺), 266 (24%), 249 (21%), 232 (71%), 217 (28%), 117 (51%), 75 (88%), 73 (100%); HRMS-EI (m/z): [M]⁺ calcd for C₁₆H₂₄O₃Si, 292.1495; found, 292.1485.