Et₂Zn‐Mediated Gem‐Dicarboxylation of Cyclopropanols with CO₂

Abstract

An unprecedented Et₂Zn‐mediated gem‐dicarboxylation of C─C/C─H single bond of cyclopropanols with CO₂ is disclosed, which provides a straightforward and efficient methodology for the synthesis of a variety of structurally diverse and useful malonic acids in moderate to excellent yields. The protocol features mild reaction conditions, excellent functional group compatibility, broad substrate scope, and facile derivatization of the products. DFT calculations confirm that the transition‐metal‐free transformation proceeds through a novel ring‐opening/α‐functionalization/ring‐closing/ring‐opening/β‐functionalization (ROFCOF) process, and 1,8‐diazabicyclo[5.4.0]undec‐7‐ene (DBU) plays dual important roles in the transformation.

A zinc‐mediated gem‐dicarboxylation of cyclopropanols with CO₂ is reported. The reaction proceeds through a novel ring‐opening/α‐functionalization/ring‐closing/ring‐opening/β‐functionalization (ROFCOF) process, providing a direct and efficient method for the synthesis of a variety of useful malonic acid derivatives. Mild conditions, excellent functional group compatibility and broad scope are the features of the protocol.

1. Introduction

Carbon dioxide (CO₂), a main greenhouse gas, has been regarded as an ideal C1 feedstock in organic synthesis because it is nontoxic, abundant, inexpensive, and renewable.^{[ 1 ]} In this context, the synthesis of carboxylic acids via carboxylation reaction with CO₂ as the carbonyl source has attracted great interest from chemists,^{[ 2 ]} because carboxylic acids are widely used as important synthetic intermediates and key precursors for polymers, and are also privileged motifs found in numerous bioactive natural products and pharmaceuticals.^{[ 3 ]} Notably, in the past decades, with great effort from many groups, significant progress has been made in dicarboxylation to generate dicarboxylic acids via transition metal catalysis,^{[ 4 ]} photocatalysis,^{[ 5 ]} or electrochemistry.^{[ 6 ]} However, most of the methods involve dicarboxylation of C─C multiple bonds of unsaturated hydrocarbons, and in sharp contrast, the example of direct dicarboxylation of C─C or C─H single bonds is rare. In 2022, Yu and co‐workers elegantly developed an electrochemical ring‐opening dicarboxylation of C─C single bonds in strained rings with CO₂, providing efficient routes to a variety of glutaric acid and adipic acid derivatives from cyclopropanes and cyclobutanes.^{[ 6e ]} To the best of our knowledge, direct dicarboxylation of C─C or C─H single bond to construct useful malonic acids has not yet been reported.^{[ 4} , ^{5 ]} Thus, the development of novel strategies for direct and efficient synthesis of diacids including malonic acids and their derivatives via dicarboxylation of C─C or C─H bond under mild conditions is still highly desirable due to their widespread application in organic synthesis.

Recently, cyclopropanols have been extensively investigated as versatile synthons in C─C or C‐heteratom bond‐forming reactions for the construction of an array of value‐added organic compounds.^{[ 7 ]} One of the most frequently used reaction mode of cyclopropanols is metal‐catalyzed or mediated ring‐opening/functionalization (ROF) via a metal homoenolate intermediate (Scheme  1A). The ROF reaction mode has become a powerful platform for the synthesis of a range of β‐substituted ketones.^{[ 7} , ^{8 ]} Interestingly, a new ring‐opening/functionalization/ ring‐closing (ROFC) reaction mode has recently been developed as a potential strategy for the synthesis of other valuable cyclopropane derivatives.^{[ 9 ]} The pioneering work was reported by Rousseaux and co‐workers, in which they developed an elegant method for converting cyclopropanols into valuable trans‐cyclopropylamines (Scheme 1B, upper).^{[ 9a ]} This ROFC reaction involved in situ generation of zinc homoenolates, condensation with amines, and subsequent ring closure processes. Very recently, Yoshikai et al. reported a zinc‐catalyzed β‐allylation of cyclopropanols with Morita‐Baylis‐Hillmancarbonates, affording bicyclic cyclopropane derivatives in a diastereoselective fashion. This transformation also underwent a ROFC process, but involved a bis‐nucleophilic zinc enolate‐homoenolate species (Scheme 1B, bottom).^{[ 9b ]}

Scheme 1.

Reaction modes of cyclopropanols for the construction of valuable molecules.

Inspired by the above significant achievements and as our continued interest in the fixation of CO₂,^{[ 6} , ^{10 ]} we wondered whether we could develop an efficient dicarboxylation of cyclopropanols with CO₂ via a ring‐opening/functionalization/ring‐closing/ring‐opening/ functionalization (ROFCOF) process. We envisioned that cyclopropanols might undergo ring‐opening in the presence of zinc salts to yield a zinc enolate‐homoenolate species I, which would react with CO₂ to give carboxylate II. The ring‐closing of II then results in the formation of intermediate III. Subsequent ring‐opening of intermediate III would generate alkylzinc species IV. The insertion of another CO₂ to the C─Zn bond of IV would deliver dicarboxylic acid salt V, which would give final diacid products after hydrolysis (Scheme 1C). If successful, this strategy would provide an attractive route to a variety of malonic acid derivatives through dicarboxylation of C─C/C─H single bond of cyclopropanols. However, such a scenario faces several challenges. First, in the presence of znic salts, cyclopropanols might generate zinc homoenolate species,^{[ 7} , ⁸ , ^{9 ]} which would undergo protonation or monocarboxylation with CO₂ to give ketone or monocarboxylic acid products, rendering the chemoselective dicarboxylation particularly problematic. Second, the site‐selective insertion of multiple CO₂ units might be another challenge. For example, the nucleophilicity of the carbon─zinc bond of the intermediate II would lead to the formation of succinic acid salts via direct insertion of another CO₂ instead of ring‐closing to give intermediate III.^{[ 11 ]}

2. Results and Discussion

With this idea in mind, we initiated our investigations by using 1‐phenylcyclopropanol (1a) as the model substrate to react with CO₂ under different conditions (Table  1 ). When the reaction was conducted in the presence of two equivalents of Et₂Zn without any base in dry tetrahydrofuran (THF) at 40 °C, no desired malonic acid derivative 2 was formed and only 14% of propiophenone was obtained as the main product (entry 1). To our delight, the reaction could proceed smoothly to give the product 2 in 81% yield by adding 1 equivalent of DBU (entry 2). A control experiment showed that both zinc salt and base are essential (entry 3). Other organic or inorganic bases, including Et₃N, 1,4‐diazabicyclo[2.2.2]octane (DABCO), DBU, 1,1,3,3‐tetramethylguanidine (TMG), and K₂CO₃, were also investigated, but no better results were observed (entries 4–7). Replacing Et₂Zn with other zinc salts such as Me₂Zn, Zn(CN)₂, and ZnCl₂ led to no detectable yield of 2 (entries 8–10). Pleasingly, product 2 was obtained in 96% yield upon isolation by increasing the amount of DBU to 1.5 equivalents and extending reaction time to 14 h (entry 11). It was also showed that two equivalents of Et₂Zn were necessary to obtain quantitative yield since decreasing the amount of Et₂Zn to 1.5 equivalent resulted in a low yield (entry 12). The reaction temperature and solvent also have great impact on the transformation, and 40 °C and THF were proved to be optimal (Tables S1–S4, Supporting Information).

Table 1.

Optimization of the reaction conditions.

Entry a)	Zn source (equivalent)	Base (equivalent)	t [h]	Yield [%] b)
1	Et2Zn (2)	–	12	n.d.
2	Et2Zn (2)	DBU (1)	12	81
3	–	DBU (1)	12	n.d.
4	Et2Zn (2)	Et3N (1)	12	26
5	Et2Zn (2)	DABCO (1)	12	33
6	Et2Zn (2)	TMG (1)	12	64
7	Et2Zn (2)	K2CO3 (1)	12	trace
8	Me2Zn (2)	DBU (1)	12	n.d.
9	Zn (CN)2 (2)	DBU (1)	12	n.d.
10	ZnCl2 (2)	DBU (1)	12	n.d.
11 c)	Et2Zn (2)	DBU (1.5)	14	99 (96) d)
12	Et2Zn (1.5)	DBU (1.5)	14	64

^a) Reaction conditions: 1 (0.3 mmol), Et₂Zn (2.0 equivalents), CO₂ (1 atm), dry THF (2 mL), 40 °C, 12 h; then acidification with HCl (2 m);

^b) Yields were determined by 1H‐NMR with CH2Br2 as internal standard;

^c) 14 h;

^d) Isolated yield.

Having determined the optimal reaction conditions, we then investigate the generality and limitation of the dicarboxylation, and the results are summarized in Scheme  2 . To our delight, a variety of 1‐aryl‐substituted cyclopropanols could undergo the reaction to give the corresponding malonic acids 2−16 in moderate to excellent yields. Both electron‐donating groups (Me, t‐Bu, Cy, and OMe) and electron‐withdrawing groups (Ph, Cl, Br, and CF₃) at the para‐, meta‐, and ortho‐position of the benzene ring were well tolerated under the dicarboxylation reaction. However, the ortho‐substituted substrates gave the products in lower yields than their para‐ and meta‐substituent analogues (15 vs 3 and 11), which might be due to the steric effect. The disubstituted substrate 1‐(3,4‐difluorophenyl)cyclopropan‐1‐ol worked well to give the corresponding product 17 in 85% yield. Cyclopropanols containing fused or heteroaryl rings were applicable to the reaction, giving the desired products 18−23 in satisfactory to excellent yields. Pleasingly, challenging alkenyl‐substituted substrates such as (E)−1‐styrylcyclopropan‐1‐ol and 1‐(cyclohex‐1‐en‐1‐yl)cyclopropan‐1‐ol could also undergo the reaction smoothly, giving the dicarboxylic acids 24 and 25 in 40% and 56% yield, respectively. Notably, 1‐alkyl‐substituted cyclopropanols uneventfully took part in the double carboxylation to give rise to the corresponding products 28−31 in moderate to near‐quantitative yields. Encouraged by the above results, we then investigated unsymmetrical 1,2‐disubstituted cyclopropanols. A broad range aryl and alkyl groups in cyclopropanols were found to be compatible with the zinc‐mediated dicarboxylation reaction to afford the desired malonic acid derivatives 32–42 although higher reaction temperature and longer reaction time were required in order to obtain satisfactory yields of the products in these cases. It is noteworthy that the reaction is highly regioselective, as the dicarboxylation occurred exclusively at the secondary rather than the tertiary carbon of the cyclopropanols. The structure of 37 was unambiguously confirmed by X‐ray diffraction analysis.^{[ 12 ]} Interestingly, the bicyclic substrate 1,1a,2,3‐tetrahydro‐7bH‐cyclopropa[a]naphthalen‐7b‐ol was also able to participate in the reaction, and the seven‐membered cyclic product 43 was predominantly formed in a high yield.^{[ 13 ]}

Scheme 2.

Substrate scope for gem‐dicarboxylation of cyclopropanols. a) Reaction conditions: cyclopropanol (0.3 mmol), Et₂Zn (2 equivalents), CO₂ (1 atm), DBU (1.5 equivalents), dry THF (2 mL), 40 °C, 14 h; then acidification with HCl (2 m). Isolated yields were reported. b) DBU (1.5 equivalents), 60 °C, 20 h; then esterification with TMSCHN₂ before separation.

The present method can be employed for late‐stage functionalization of complex molecules derived from natural products and drugs. As can be seen from Scheme  3 , cyclopropanols derived from perfume molecules such as celestolide, fixolide, and β‐ionone could undergo the transformation smoothly, yielding the desired products 44–46 in moderate to high yields. Moreover, cyclopropanols derived from the precursor of selective PDE4 inhibitor roflumilast, anti‐gout drug probenecid, and retinoic acid compound adapalin were compatible with this system, affording the expected products 47, 48, and 49 in 62%, 46%, and 41% yield, respectively.

Scheme 3.

Applications of natural products and drug molecules: a) Reaction conditions: cyclopropanol (0.3 mmol), Et₂Zn (2 equivalents), CO₂ (1 atm), DBU (1.5 equivalents), dry THF (2 mL), 40 °C, 14 h; then acidification with HCl (2 m). isolated yields. b) 45 °C, 24 h.

To showcase the practicability of the protocol, a gram‐scale synthesis of product 2 was performed under the standard conditions, which readily gave access to the product in 89% yield (Scheme  4a). The derivatizations of product 2 were also investigated to demonstrate the utilities of this method (Scheme 2b). First, in the presence of acetic anhydride, 2 could undergo decarboxylative cyclization to give the unsaturated γ‐lactone 50 in 82% yield.^{[ 14 ]} The reaction of 2 and 1,3‐dibutylurea under acidic conditions gave barbituric acid derivative 51.^{[ 14 ]} Moreover, the esterification of 2 with TMSCH₂N₂ furnished ester 52 in quantitative yield,^{[ 15 ]} which could be reduced to triol 53 by Lithium aluminohydride.^{[ 16 ]} Treatment of 52 with propargyl bromide under basic conditions, propargylated product 54 was obtained in almost quantitative yield.^{[ 17 ]} Furthermore, compound 52 could be used for the construction of a variety of heterocycles. For instance, condensation of 52 with hydrazine hydrate afforded the corresponding tetrahydropyridazine derivative 55,^{[ 18 ]} while the reduction cyclization of 52 using NaBH₄ as the reductant would yield γ‐lactone 56.^{[ 19 ]}

Scheme 4.

Gram‐scale synthesis and derivatizations of 2.

To gain some insight into the reaction pathway, we first performed the model reaction in the presence of 2 equivalents of TEMPO as the radical scavengers, and 2 could be obtained in 56% yield (Scheme  5a). Second, to probe the role of Et₂Zn and DBU in this transformation, phenylcyclopropanol 1a was treated in the absence of CO₂ under otherwise standard conditions, and ethyl phenyl ketone 2′ was formed in 90% yield. However, the reaction without DBU or Et₂Zn only gave 2′ in very low yields, indicating the synergistic effect between DBU and Et₂Zn in the ring opening of 1a (Scheme 5b). Third, neither 2′ nor 4‐oxo‐4‐phenylbutanoic acid (2″) could undergo carboxylation with CO₂ under standard conditions to give the desired product 2, excluding the possibility of 2′ and 2″ as the key intermediates for the transformation (Scheme 5c).

Scheme 5.

Mechanistic studies.

To further elucidate the mechanism of the reaction, density functional theory (DFT) calculations were conducted on the model reaction of 1‐phenylcyclopropanol (1a) with CO₂. The most favorable pathway is shown in Figure  1 for mechanism discussion (see Supporting Information for more details). First, the metalation of 1a by ZnEt₂ can take place with the aid of a molecule of DBU as a ligand to give an intermediate B. This process has an energy barrier of 26.1 kcal mol⁻¹ (via TS1) and is significantly exothermic by 20.4 kcal mol⁻¹. The ring‐opening of cyclopropane moiety in B can subsequently occur through TS2 (∆G ^‡ = 22.9 kcal mol⁻¹) to form intermediate C. It is found that the direct insertion of CO₂ into the Zn─C bond in C needs to overcome high energy barrier (∆G ^‡ = 34.7 kcal mol⁻¹, via TS3ʹʹ in Figure S1‐1, Supporting Information), which is kinetically inaccessible under the experimental conditions. Alternatively, the intermediate C could be easily converted to an ion‐pair species D via the deprotonation of the methylene (TS3) by DBU base. Then, the proton transfer from DBU‐H⁺ to a new incoming ZnEt₂ through TS4 could result in a dizinc enolate‐homoenolate species E, which is significantly more stable than D by 38.2 kcal mol⁻¹. To be noted, CO₂ insertion into E through six‐membered ring transition state TS5 can take place with a moderate energy (∆G ^‡ = 19.8 kcal mol⁻¹), leading to mono‐carboxylation species F. Next, the sequential ring‐closing and ring‐opening events through TS6, G, and TS7 could give a slightly more stable species H. The isomerization via EtZn‐transferred to adjacent oxygen forms the intermediate I with a newly formed C═C moiety. The subsequent CO₂ insertion could take place via six‐membered ring TS9, similar to TS5, giving the dicarboxylation species J. It is worth noting that during this transformation, DBU plays two important roles. One is acting as a ligand to stabilize the Zn center(s) in some key intermediates and transition states and the other is serving as a Brønsted base to deprotonate the methylene (via TS3). Remarkably, DFT calculations reveal a new mechanism in difunctionalization of cyclopropanols, which involves ring‐opening/α‐functionalization/ring‐closing/ring‐opening/β‐functionalization (ROFCOF) process.

Figure 1.

Energy profiles of Et₂Zn‐mediated gem‐dicarboxylation of 1‐phenylcyclopropanol a) with CO₂ calculated at the level of M06‐2X(SMD,THF)/6‐311+G(d,p)//M06‐2X/6‐31G(d).

3. Conclusion

In summary, the Et₂Zn‐mediated germinal dicarboxylation of cyclopropanols with CO₂ has been reported for the first time, providing a novel, straightforward and efficient strategy for the synthesis of a variety of malonic acid derivatives. This reaction features mild reaction conditions, excellent functional group compatibility, broad substrate scope, and easy derivatization of the products. DFT calculations revealed that the transformation might proceed through an unprecedented ring‐opening/ functionalization/ring‐closing/ring‐opening/ functionalization (ROFCOF) process, in which DBU plays two crucial roles. This work represents a new reaction mode for the currently widely studied cyclopropanols and opens a new window for broadening the utility of this type of substrate. Further investigations on application of this protocol are ongoing in our laboratories.

Conflict of Interest

The authors declare no conflict of interest.

Supporting information

Supporting Information

Liu H., Shi L., Tan X., Kang B., Luo G., Jiang H., Qi C., Et₂Zn‐Mediated Gem‐Dicarboxylation of Cyclopropanols with CO₂ . Adv. Sci. 2024, 11, 2307633. 10.1002/advs.202307633

Data Availability Statement

The data that support the findings of this study are available in the supplementary material of this article.