Rapid and Safe Continuous-Flow Simmons-Smith Cyclopropanation using a Zn/Cu Couple Column

: Flow chemistry has recently opened new chemistry windows thanks to the safer use of hazardous and sensitive reagents. Furthermore, flow procedures usually outperform their batch counter-parts due to improved mass and heat transfer, offering a good opportunity for industrial application. Herein, a rapid Simmons-Smith cyclopropanation flow process is presented. Taking advantage of the in situ production of the zinc carbenoid species, several olefins bearing aromatic rings of different electronic nature, aliphatic chains or heterocycles were smoothly cyclopropanated with a residence time of just 15 minutes. In addition, industrial applicability of the protocol is assured thanks to a successful 12-mmol scale experiment, which represents a 3.59 grams per hour production of a selected example, and to the satisfactory synthesis of pharmaceutical drugs.

Flow chemistry protocols have been gaining relevance and importance over the last decades, as evidenced by the increasing investigations that have been recently published regarding this field. [1]Many well-known transformations have been adapted from batch to flow, [2] as the use of (micro)reactors provide different advantages such as a better heat and mass transfer or the improved handling of sensitive and very reactive species. [3]Regarding the latter one, the use of continuous-flow setups has resulted in a very useful manner of scaling up processes that relay on sensitive, reactive or short-time living species.The straightforward in-line use of this kind of compounds has allowed new transformations that were not possible using batch methods, while reducing both hazards and waste production. [4]n this sense, the use of cartridges filled with heterogeneous catalysts and reagents has resulted in a more efficient manner to handle hazardous or reactive substances.As flow setups usually rely on (micro)reactors with very small internal volumes, only small amounts of reagents are mixed at a time, avoiding the hazards of handling with big amounts of reactive, corrosive, or explosive species. [5]Moreover, cartridges filled with different metals have been presented over the last years for their use in several organic transformations where the in situ production of sensitive organometallic species was accomplished. [6]n the other hand, Simmons-Smith reaction is one of the most useful organic transformations to obtain a cyclopropane from the corresponding olefin (Scheme 1a). [7]Traditionally, this powerful process relies on the synthesis of a zinc carbenoid using dihalomethanes in the presence of a zinc-copper couple. [8]Over the years, several modifications have taken place, substituting either the metal alloy by diethylzinc, [9] or the dihalogenated species by a diazo derivative. [10]lthough the transformation presents very potent advantages, such as high functional group tolerance, high yields, and great control of stereochemistry, [7] the main drawback of all its versions is the use of a sensitive and reactive species.The use of hazardous reagents such as diethylzinc or diazo derivatives could represent a great disadvantage when trying to use this cyclopropanation protocols in industrial scale.In this sense, a flow chemistry approach can provide useful solutions by the implementation of a procedure that in situ generates the desired compounds.In 2015, Ley and coworkers presented a method utilizing an activated manganese dioxide-filled packed-bed reactor for the on-demand synthesis of various diazo species (Scheme 1b). [11]This approach involved the conversion of different hydrazones into their corresponding diazo derivatives, which were subsequently employed for the cyclopropanation of multiple alkenes.However, this process was conducted in a batch manner.In subsequent years, Charette [12] and Lathrop [13] devised a scalable and safer approach for the continuous-flow production of diazo compounds, eliminating the need for a metal cartridge.Despite their efforts, no significant advancements were made in the cyclopropanation reaction, and the requirement for batch conditions remained unchanged.Moreover, and to the best of our knowledge, no flow adaptation has been studied for traditional Simmons-Smith cyclopropanation reaction.With all these and based on our own experience in flow chemistry [14] and the use of metal-filled packedbed reactors, [15] we envisioned a continuous-flow protocol that relies on the use of a novel zinc-copper couple column reactor (Scheme 1c).The corresponding dihalomethane will in situ generate the zinc carbenoid species without compromising stability and reactivity, and will then undergo the stereospecific cyclopropanation of different olefins.Desirably, short reaction times, operational simplicity of the process and safer handling of sensitive reagents as the zinc carbenoid species, will make this protocol highly attractive for the industrial adaptation of the powerful Simmons-Smith cyclopropanation reaction.
Initially, we started our investigation by setting a model flow system up.As depicted in Table 1, a Scheme 1. a. Classical Simmons-Smith reaction conditions and its variations through time.b.Semi-flow cyclopropanation using diazo derivatives.c.This work.

Table 1. Optimization of continuous-flow Simmons-Smith cyclopropanation conditions [a]
[a] Reaction conditions: pump A: 1 a (0.25 mmol), CH 2 I 2 (2 equiv.) in dry DCE (1 M) under nitrogen atmosphere; pump B: saturated aqueous NH 4 Cl solution.c] Isolated yield in brackets.See Supporting Information for the complete evaluation of reaction parameters (solvent, stoichiometry, pressure, temperature, concentration, and reactor residence time).
These are not the final page numbers!�� solution of both benzyl-protected alkene 1 a and diiodomethane in dry 1,2-dichloroethane (DCE) was pumped through the zinc-copper couple column reactor.Zinc carbenoid species will then be formed in situ, which will be immediately involved in the cyclopropanation process.Evaluation and optimization of different reaction parameters such as solvent, temperature or concentration was then performed in order to obtain the best conditions for the cyclopropanation reaction (see Table S2, Supporting Information).Optimization process resulted in the use of a 1 M solution of dry DCE with a molar ratio alkene:diiodomethane of 1:2 as starting reaction mixture, and the flow protocol was run at 40 °C and 75 psi pressure for 15 minutes reactor residence time.Under these adjusted conditions, cyclopropane 2 a was obtained in an excellent 94% isolated yield (entry 1, Table 1).However, inferior outcomes were observed in cases where the system lacked a back pressure regulator (BPR) or when it was operated under an air atmosphere (entries 2 and 3, Table 1).Moreover, no conversion was detected when the reaction was conducted at room temperature (entry 4, Table 1), or when either diiodomethane or the metallic couple were removed from the reaction medium (entries 5 and 6, Table 1).Besides, following a well-stablished procedure described in the literature, [16] same result was obtained when cyclopropanation was performed under batch conditions for fifteen minutes (entry 7, Table 1), showing how flow processes are able to outperform their batch counterparts once again.
Having determined the optimal conditions, we proceeded to assess the applicability for the cyclopropanation reaction (Table 2).As zinc carbenoid species obtained by Simmons-Smith conditions is known to be electrophilic, [7] this will rule the reactivity of the in situ generated carbenoid solution against the different alkenes.As expected, when strongly electrondonating groups (i.e., methoxy group) were assembled in the aromatic region of the cinnamyl alcohol, excellent yields of the corresponding cyclopropanes were achieved (2 a-2 e).Moreover, by introducing a methoxy group in the para position (2 a, 2 c, and 2 d), the alkenes were strongly activated, enabling us to reduce the reaction residence time to just 10 minutes without a notable decrease in yields.This improvement led to enhanced reaction outcomes.When electronically neutral cinnamyl alcohol derivative 1 e was subjected to the optimized flow cyclopropanation conditions, the respective cyclopropane (2 e) was obtained in a 57% isolated yield.Then, different electronic nature of the phenyl ring was evaluated by means of the incorporation of electron-withdrawing groups.The incorporation of halogens into organic compounds is a highly desirable process, as it provides potential sites for future post-functionalization, often through metal-catalyzed cross-coupling reactions.
Moreover, fluorine and chlorine atoms are known to cause huge improvements in the biological and pharmacokinetical properties of bioactive compounds. [17]In this sense, a whole bunch of cinnamyl derivatives bearing a halogen in the aromatic ring were studied, obtaining the subsequent cyclopropanated products in good yields (2 g-2 i).However, the presence of the strongly electron-withdrawing para-trifluoromethyl group decreased the reaction yield to a 16% (2 j).This result, although expected due to the high electrophilicity of the starting alkene, encouraged us to find a modification in the protocol that allows the use of electron-deficient olefins.Delightfully, the addition of just a 10 mol % of diethylzinc (1 M solution in hexanes) to the starting reaction mixture improved the cyclopropane yield up to a valuable 68%.This adjustment in the protocol makes it now a feasible option for the rapid cyclopropanation of electron-poor alkenes.
We next evaluated various different chemical moieties as plausible protecting groups at the alcohol.In this sense, starting cinnamyl alcohol was then protected as both a trimethylsilyl ether and a different benzyl derivative bearing two methyl groups in the phenyl ring.Both cases underwent cyclopropanation reaction smoothly, obtaining the corresponding products in good yields (2 k, 2 l).Nevertheless, trimethylsilyl ether group is known for its lability and low chemical resistance.In this way, stronger silyl ethers were studied, protecting starting para-methoxy alkene as both a tert-butyldimethylsilyl ether and a triisopropylsilyl ether.As expected, cyclopropanation took place efficiently, obtaining the resultant cyclopropanes in remarkable yields (2 m, 2 n).Despite these good results, the use of protected alcohols can still represent a drawback for the industrial application of the method.Encouraged to find a solution, para-methoxy alkene bearing a free alcohol was subjected to optimized conditions.7a] Nonetheless, treating the alcohol with sodium hydride before adding CH 2 I 2 to the reaction media and subjecting the mixture to the cyclopropanation conditions proved to be an efficient strategy to solve this problem, as the subsequent anionic species underwent cyclopropanation in significant yield (78%).Next, the effect of the steric hindrance due to multiple substitution present in the alkene region of the starting material was evaluated.In this respect, a 1,1-diphenylsubstituted olefin was tested, obtaining the corresponding cyclopropane in moderate yield (2 p).Trying to find a solution for low-yielding examples, we subjected 1 p to the improved conditions used for 1 j (using Et 2 Zn).Satisfyingly, this methodology emerged as a valuable alternative to solve these drawbacks, as These are not the final page numbers!�� compound 2 p could now be obtained in 65% yield.Furthermore, 1,2-disubstitution at the alkene was also studied, obtaining the desired product in 57% yield (2 q) as a single diastereoisomer.d] Ar stands for 3,5-dimethylbenzene. [e] Alcohol was treated with NaH (1 equiv.)for 20 minutes and filtered before adding CH 2 I 2 .

COMMUNICATIONS asc.wiley-vch.de
These are not the final page numbers!�� compounds, [18] several cores were studied in order to make the protocol synthetically attractive for structureactivity relationship (SAR) studies.In this regard, aliphatic and aromatic residues were tested under the optimized conditions.Both cyclic and acyclic aliphatic compounds underwent Simmons-Smith cyclopropanation smoothly, giving rise to the respective products in good yields (2 r, 2 s).In the case of the aromatic rings, furan, thiophene and benzothiophene were chosen with the aim of improving the nucleophilic character of the resultant olefin.All of them gave good results when they were subjected to flow cyclopropanation conditions, obtaining the corresponding products in good yields (2 t-2 v).However, we also encountered some limitations, such as the use of an aromatic residue with a dimethylamino group or the utilization of an amino acid precursor.
One of the best advantages of continuous-flow processing is the possibility to easily scale chemical transformations up, as larger experiments can be run just by feeding the reactor with the starting reaction mixture for longer time.Taking this into account, we envisioned the intensification of our protocol by the use of a bigger zinc-copper couple column reactor (see Supporting Information for more details).In this sense, we were able to perform a 12-mmol cyclopropanation run over starting alkene 1 a (Scheme 2a).Pleasingly, the transformation between the olefin and diiodomethane took place smoothly, giving rise to cyclopropane 2 a in a nice 90% yield, which can be translated into a 3.59 grams per hour production.As aforementioned, cyclopropanes are nowadays of great importance in medicinal chemistry, being regularly included in SAR studies due to the plausible modulation of some properties of the resulting drug candidates. [19]In fact, several already approved drugs rely on Simmons-Smith reaction or one of its modifications for the synthesis of the corresponding cyclopropane. [20]To show the applicability of this protocol, we tested the developed flow Simmons-Smith conditions in the formal synthesis of an existing drug, the antidepressant Ropanicant (Scheme 2b).As expected, the drug precursor 3 underwent the cyclopropanation efficiently, obtaining the desired compound 4 in 86% yield, synthetizing 333 milligrams per hour under non-intensified conditions.
Additionally, we have carried out the assessment of the durability of the zinc-copper column (Figure 1).To achieve this, the packed-bed reactor was continuously fed with a 1 M DCE solution of compound 1 a and diiodomethane for an eight-hour operational time (see S. I. for further details).Reaction outcome was monitored by taking crude samples every 30 minutes, showing complete conversion over the first two hours.However, cyclopropane production showed a decay in the following hours, achieving a constant 30% conversion in the second half of the experiment (see S. I.).Despite of this, 50 mmol of product could be successfully obtained after this long-run experiment (Figure 1), which can be translated into a 2.5 mmol of cyclopropane 2 a per gram of zinc-copper column under non-intensified conditions.
In conclusion, we have introduced a continuousflow approach for the gentle and rapid Simmons-Smith cyclopropanation of olefins.Through the utilization of a unique zinc-copper column reactor, zinc carbenoid species are generated in situ within the reaction medium, enabling direct cyclopropanation of the target alkene in a mere 15 minutes.Numerous examples utilizing various frameworks were successfully executed, allowing for the scaling up of the reaction to produce 3.59 grams per hour under intensified conditions.Moreover, the described methodology facilitated the smooth synthesis of actual drug candidates such as a Ropanicant precursor.Given these attributes, this protocol holds promise as a viable option for Scheme 2. a. Scale-up experiment under intensified conditions (see Supporting Information for more details).b.Real drug synthesis application of the protocol.These are not the final page numbers!�� industrial applications, and it serves as an illustration of the benefits offered by continuous-flow methodologies in the synthesis of delicate and short-living chemical entities.

Experimental Section Zn/Cu Couple Synthesis
Copper acetate (0.5 g, 2.75 mmol) was added to an oven-dried flask and dissolved in 50 mL of acetic acid.The mixture was warmed up to 110 °C (solution might be nearly refluxing), and 35 g of Zn granules (536.25 mmol, 195 equiv.)were added and kept at that temperature for 5 minutes.After that, acetic acid was decanted, and the solid was washed with 50 mL of acetic acid and 3×50 mL of Et 2 O.The resulting reddish-grey solid was dried under vacuum and kept under nitrogen atmosphere.

Flow Reactor Setup
All continuous-flow experiments were carried out using a commercially available Vapourtec E-series equipment.The system consisted of two pumps, a temperature-controlled adjustable Omnifit ® column filled with the zinc-copper couple (6.6 mm bore×15 cm length, approximately 8 g Zn/Cu, 1.5 mL), a variable back pressure regulator (BPR), and a Zaiput SEP-10 liquid-liquid separator.See Supporting Information for full details.

General Procedure for the Cyclopropanation of Olefins under Continuous-Flow Conditions
Solution A: the corresponding alkene 1 or 3 (0.25 mmol, 1 equiv.)was added to a Schlenk tube under nitrogen atmosphere.Dry DCE (250 μL, 1 M) and diiodomethane (40.3 μL, 0.5 mmol, 2 equiv.)were added, and the solution was finally homogenized by sonication in an ultrasound bath.Solution B: saturated aqueous NH 4 Cl solution (solubility: 383.0 g L À 1 at 25 °C).
Both solvents were pumped through at a flow rate of 1 mL•min À 1 in order to purge the whole flow system (approximately 10 minutes).
The reaction was performed, if not otherwise stated, by setting column temperature at 40 °C and system pressure at 75 psi (5.2 bar approx.).A second 20 psi BPR was assembled in the organic outcome of the liquid-liquid separator to secure the efficient separation of the phases, as the equipment diaphragm could become "softer" over time when chlorinated solvents are used.Then, solutions A and B were pumped through at 100 μL•min À 1 (15 minutes column residence time).Crude mixture was finally collected after liquid-liquid separation, solvent was removed under reduced pressure and the residue was purified by flash column chromatography to obtain cyclopropanated compounds 2 or 4.

Figure 1 .
Figure 1.Cumulative production of 2 a during the durability test.

Table 2 .
Substrate scope for the continuous-flow Simmons-Smith cyclopropanation.