Accelerated and Concerted Aza‐Michael Addition and SuFEx Reaction in Microdroplets in Unitary and High‐Throughput Formats

Abstract

The sulfur fluoride exchange (SuFEx) reaction is significant in drug discovery, materials science, and chemical biology. Conventionally, it involves installation of SO₂F followed by fluoride exchange by a catalyst. We report catalyst‐free Aza‐Michael addition to install SO₂F and then SuFEx reaction with amines, both occurring in concert, in microdroplets under ambient conditions. The microdroplet reaction is accelerated by a factor of ∼10⁴ relative to the corresponding bulk reaction. We suggest that the superacidic microdroplet surface assists SuFEx reaction by protonating fluorine to create a good leaving group. The reaction scope was established by performing individual reactions in microdroplets of 18 amines in four solvents and confirmed using high‐throughput desorption electrospray ionization experiments. The study demonstrates the value of microdroplet‐assisted accelerated reactions in combination with high‐throughput experimentation for characterization of reaction scope.

Accelerated, catalyst‐free, coupled Aza‐Michael addition and SuFEx reactions between amines and ethenesulfonyl fluoride under ambient conditions in microdroplets are analyzed by electrosonic spray mass spectrometry (ESSI‐MS) while high‐throughput desorption electrospray ionization mass spectrometry (DESI‐MS) is used to follow reaction scope.

Chemical reactions occurring at the interface between water and an organic solvent (on‐water reactions) ^[1] are accelerated as are reactions occurring at the air–water interface.[ ² , ³ , ⁴ ] Studies of reactions in microdroplets[ ⁵ , ⁶ , ⁷ , ⁸ ] have gained significant attention due to their unique reaction environment.[ ⁹ , ¹⁰ , ¹¹ , ¹² ] Microdroplet reactions show significant reaction acceleration[ ³ , ¹³ , ¹⁴ , ¹⁵ ] with enhanced reaction rates.[ ² , ¹⁶ , ¹⁷ , ¹⁸ ] Reaction acceleration at the solution/air interface of microdroplets is attributed to partial solvation[ ² , ³ ] and the existence of a high interfacial electric field.[ ¹⁹ , ²⁰ ] The extreme pH gradient at the interface produced by the high electric field provides highly active superacid or superbase reagents.[ ²¹ , ²² ] Highly reactive species (OH⋅, H₂O₂, and H₂O ⁺ ⋅/H₂O⁻⋅) are suggested to be present at the interface (provided the solvent contains at least some water).[ ⁶ , ⁷ , ⁸ , ⁹ , ¹² , ¹⁹ , ²⁰ ] An associated advantage of microdroplets is that their fast reactions can avoid the need for heterogeneous catalysts but still allow for small‐scale synthesis.[ ¹⁵ , ²³ , ²⁴ , ²⁵ ] Furthermore, the small amounts of reagents and low solvent and energy consumption of microdroplet synthesis make it a green and sustainable synthetic method.[ ²⁵ , ²⁶ ] Scaling up of microdroplet synthesis to obtain gram‐scale products has been achieved using multiplexed electrosprays,[ ¹⁴ , ¹⁵ ] heated ultrasonic nebulized sprays, ^[27] or by recycling the electrosprayed solution. ^[28]

These unique features of microdroplet synthesis have been applied to several reactions[ ² , ³ , ¹⁰ , ²⁹ ] amongst them the Aza‐Michael addition which is known to be accelerated in microdroplets without requiring Lewis acids or metal salt catalysts.[ ¹⁰ , ²⁹ ] The rate of Michael addition is enhanced by using ethenesulfonyl fluoride (ESF), reportedly one of the strongest Michael acceptors.[ ³⁰ , ³¹ , ³² ] ESF is also used as a connecting hub for the sulfur(VI) fluoride exchange (SuFEx) reaction.[ ³⁰ , ³¹ ] The classical SuFEx reaction, developed by Sharpless and co‐workers in 2014, ^[31] involves the installation of a ‐SO₂F group as the first step and then the exchange of F in an S−F bond with nucleophiles (Scheme 1a).[ ³³ , ³⁴ ] SuFEx click chemistry has seen significant applications in materials science,[ ³⁵ , ³⁶ , ³⁷ , ³⁸ , ³⁹ ] chemical biology,[ ⁴⁰ , ⁴¹ ] late‐stage functionalization, ^[42] and drug discovery.[ ³³ , ⁴³ , ⁴⁴ ] Moreover, sulfonyl fluoride (−SO₂F) containing molecules and their derivatives are considered covalent inhibitors for many enzymatic reactions and privileged drug candidates.[ ⁴⁰ , ⁴⁵ ] Representative examples of such compounds are shown in Figure S2. The SuFEx reactivity depends on converting the F atom of the covalent S−F bond into a good leaving group, a process which is assisted by H⁺, R₃Si⁺ and nucleophilic catalysts such as triethylamine, 1,8‐diazabicyclo[5.4.0]undec‐7‐ene (DBU), 2‐tert‐butylimino‐2‐diethylamino‐1,3‐dimethylperhydro‐1,3,2‐diazaphosphorine (BEMP), and bifluoride salts.[ ³⁵ , ³⁷ , ³⁸ , ⁴⁶ , ⁴⁷ , ⁴⁸ ] However, the required multiple steps, prolonged reaction time, and generous amount of catalyst (>30 mol %)[ ³¹ , ³⁸ ] are somewhat disadvantageous in practical applications. ^[48] Given the importance of the SuFEx reaction, microdroplet synthesis is of interest as a means of combining the two steps in an accelerated format while avoiding hazardous and costly catalysts and bases.

Scheme 1.

a) Conventional multistep SuFEx click reaction which requires a long reaction time, catalyst, and high temperature.[ ³² , ⁴⁷ ] b) Accelerated single‐step SuFEx click reaction in microdroplets without using base or catalyst.

We report the installation of SO₂F (an SuFEx hub) on amines by Aza‐Michael addition to ethenesulfonyl fluoride (ESF) with the addition product subsequently undergoing the SuFEx reaction with nucleophiles in a single‐step overall synthesis (Scheme 1b). Both reactions (addition and SuFEx) are performed in microdroplets and show an overall rate acceleration of ∼10⁴ in comparison to the bulk reaction. Note that microdroplets favor the overall reaction even in the absence of catalyst. We suggest that the super acidic nature of the aqueous droplet surface[ ²¹ , ²² ] favors the SuFEx reaction by protonating fluorine.

We studied the concerted addition and SuFEx reaction between amines and ESF in microdroplets using three different solvents: MeOH, EtOH, and water, each under ambient conditions (Scheme 1b). Figure 1 shows the positive mode electrosonic spray ionization (ESSI) mass spectra of the reaction mixture of 4‐aminopyridine (10 mM) and ESF (10 mM) in a 1 : 1 ratio in MeOH, EtOH, and water. The protonated Aza‐Michael addition product, 2 a, and the simultaneously formed methyl ester, 3 a, were observed at m/z 205 and 217, respectively, in methanol. We suggest that the formation of 3 a is due to the concerted Aza‐Michael addition and SuFEx reaction, in which the F atom from 2 a is exchanged for OMe originating from the solvent (MeOH). The corresponding phenomenon was observed for EtOH and water. The reaction in EtOH shows product peaks at m/z 205 and 231, again due to the corresponding addition and SuFEx reactions, respectively. Note that the formation of ethyl ester 4 a is due to the exchange of F by OEt from the solvent. It is reported that the S−F exchange process in the case of aliphatic alcohol nucleophiles is quite challenging[ ⁴⁹ , ⁵⁰ ] and is achieved using high catalyst loadings. ^[48] Here, the SuFEx reaction with a primary alcohol as nucleophile proceeds smoothly in microdroplets. In the case of water as a solvent, the positive mode spectrum (Figure 1c) shows only the addition product, 2 a (m/z 205). The corresponding sulfonic acid (SuFEX product) likely exists in the zwitterionic form so is not observed. However, in the negative mode mass spectrum (Figure S3) the deprotonated sulfonic acid (SuFEX product) was observed at m/z 201, with F being replaced by OH (originating from water). By contrast, the mass spectra in MeOH and EtOH do not show the −SO₃H product (Figure S3) making it obvious that the formation of 3 a and 4 a occurs by the SuFEx reaction with the solvents (MeOH and EtOH) acting as nucleophiles. These esters are not formed by the esterification of −SO₃H. The structures of the synthesized products were characterized by tandem mass spectrometry (MS/MS) using collision‐induced dissociation (CID), as shown in Figure 1d–1f and as discussed in the Supporting Information (MS/MS analysis shows that the amino group of 4‐aminopyridine is responsible for Aza‐Michael addition). In addition, the structural assignments of all the microdroplet synthesized products were confirmed by comparing the isotopic distributions between theoretical and experimental mass spectra as shown in Figure S4–S7.

Figure 1.

Positive mode mass spectra (left column) of concerted Aza‐Michael addition and SuFEX reaction for an equimolar mixture of 4‐aminopyridine and ethenesulfonyl fluoride (ESF) in a) MeOH, b) EtOH, and c) water microdroplets. Labeled peaks, [1 a+H]⁺ , [1 a+H+1 a]⁺ , [2 a], [3 a], and [4 a] represent the protonated 4‐aminopyridine, proton‐bound dimer of 4‐aminopyridine, protonated Aza‐Michael addition product, protonated SuFEX product with MeOH, and protonated SuFEX product with EtOH, respectively. The right‐hand column shows MS/MS spectra of the mass‐selected reaction products: d) [3 a] (m/z 217), e) [4 a] (m/z 231), and f) [2 a] (m/z 205), recorded using collision‐induced dissociation (CID).

Although the installation of SuFEX is relatively simple, conventionally the exchange of F from a strong S(VI)‐F poses a challenge. ^[31] Previous reports suggest that the microdroplet interface is highly acidic, for example, it catalyzes benzimidazole synthesis from o‐diamine and formic acid ^[22] as well as carbonamide formation by reaction of amines and CO₂. ^[21] Here, the interfacial superacidic environment[ ²¹ , ²² ] in microdroplets protonates the alkene allowing SuFEX formation and then assists in converting the stable S(VI)−F bond to a very good leaving group, and this in turn results in a concerted addition and SuFEx reaction. Moreover, previous reports[ ³¹ , ³³ , ⁴⁸ ] suggest that acidic environments (H⁺ and R₃Si⁺) assist in SuFEx reactivity by a similar mechanism.

The overall reaction is accelerated in microdroplets by an acceleration factor of ∼10⁴. The acceleration factor was obtained by taking the ratio of the rate constants of the reaction in microdroplet and bulk, as shown in Figure S8. The enhanced surface‐to‐volume ratio of microdroplets, along with the superacidic environment at their surfaces is suggested to contribute to the rate acceleration. An increase in the product conversion was observed with the increase in nebulization gas pressure, a condition that results in smaller droplets (Figure S9). The unique chemistry and reaction environment of microdroplets is the driving factor for acceleration even without added catalyst. A recent study ^[48] reported SuFEx reaction acceleration using as catalyst Barton's hindered guanidine base (2‐tert‐butyl‐1,1,3,3‐tetramethylguanidine; BTMG) combined with hexamethyldisilazane (HMDS). This reaction occurs on the minutes timescale and the extent of acceleration was determined by comparison with other catalysts. ^[48] In comparison, the reaction in our study occurs on the millisecond‐flight‐time of microdroplets (acceleration factor of ≈10⁴) and it does so without catalyst.

So far, the discussion of the SuFEx reaction has been based on the inherent nucleophilic nature of the solvents, which could be seen as limiting the versatility of the reaction. However, the amines selected for Aza‐Michael addition can also participate in the SuFEx reaction. Figure 2 shows this by reference to the mass spectra for concerted Aza‐Michael addition and SuFEX reaction, in the case of ethylamine and n‐butylamine reactions with 4‐ethoxyaniline and ESF in MeOH microdroplets. The observed peaks at m/z 138, 156, 168, 181, 248, and 273 are due to protonated 4‐ethoxyaniline starting material 1 b, protonated addition product with ethylamine 2 b, SuFEX product 3 b (methyl ester), SuFEX product with ethylamine 4 b, protonated addition product with 4‐ethoxyaniline 5 b, and SuFEX product with ethylamine and 4‐ethoxyaniline 6 b, respectively in Figure 2a. In Figure 2b, the observed peaks at m/z 138, 184, 196, 237, 248, and 301 are due to protonated 4‐ethoxyaniline starting material 1 b, protonated addition product with n‐butylamine 2 c, SuFEX product 3 c (methyl ester), SuFEX product with n‐butylamine 4 c, protonated addition product with 4‐ethoxyaniline 5 c, and SuFEX product with n‐butylamine and 4‐ethoxyaniline 6 c, respectively. Reaction schemes for 4‐ethoxyaniline with ethylamine and n‐ butylamine are shown along with all the product structures in Figures 2c and 2d, respectively. Structures of all these product ions (m/z 156, 168, 181, 184, 196, and 273) were characterized by MS/MS as shown in Figures S10–S12.

Figure 2.

Mass spectra after concerted Aza‐Michael addition and SuFEX reaction of ethenesulfonyl fluoride (ESF) with an equimolar mixture of a) 4‐ethoxyaniline and ethyl amine, and b) 4‐ethoxyaniline and ethyl amine in MeOH microdroplets with the corresponding reaction schemes and the structures of the products being shown in (c) and (d), respectively.

Ethylamine, as an aliphatic amine, is more nucleophilic than aromatic amines, so the in situ SuFEX product 4 b is formed with ethylamine itself. Importantly for product control in the first step of the reaction, the analogous aromatic amine product was not observed. The same favorable phenomenon occurred in the formation of 4 c using n‐butylamine. However, reaction with a mixture of two aliphatic amines (e.g. ethylamine and n‐butylamine) showed the formation of all the possible addition and SuFEx products and their structures are confirmed by conducting MS/MS (Figure S13). Although the aromatic amine (4‐ethoxyaniline) did not act as a nucleophile to produce the SuFEx product in MeOH microdroplets, it showed better nucleophilicity in water microdroplets and resulted in the corresponding SuFEx product as well as mono and bis‐addition product (Figure S14; structures confirmed by MS/MS as shown in Figures S15–19). Overall, these results demonstrate that the SuFEX reaction is not limited to the solvent nucleophiles, rather a range of nucleophiles can participate, establishing the versatility of the reaction in microdroplets but highlighting the need to consider competitive reactivity.

Tandem mass spectrometry (MS/MS) is routinely utilized to characterize product ions and predict fragmentation mechanisms. Here, the CID‐based MS/MS spectra of Aza‐Michael addition products using representative examples of an aromatic and aliphatic amine, were investigated at CID collision energies of 10, 20, and 30 arb. units (Figure 3). In both cases, the fragmentation onset is at relatively high energy (20 arb. unit), suggesting that the new bond formed during the addition reaction is covalent and relatively strong. However, the fragmentation patterns for the aromatic and aliphatic amines are markedly different. The product of the reaction with aniline (Figure 3a) shows fragments due to the neutral loss of HSO₂F and CH₃SO₂F, respectively. The analogous fragmentations were observed previously for the MS/MS spectra of ions 2 a and 3 a derived from aromatic amines (Figure 1d and 1f). The MS/MS spectra of the product with the aliphatic amine (propyl amine, Figure 3b) show neutral loss of HSO₂F and CH₃SO₂F, respectively, and do not show the corresponding radical cation of propyl amine. This contrasting observation is easily explained by the relative stability of the produced aryl cation radical species, which is stabilized by resonance.

Figure 3.

Tandem mass (MS/MS) spectra of the mass‐selected Aza‐Michael addition product of a) aniline and ESF with m/z 204 and b) propyl amine and ESF with m/z 170, using different CID collision energies (10, 20, and 30 arb. units). The labeled peak at m/z 93 corresponds to the aniline cation radical upon fragmentation and is only observed for aromatic amines. MS/MS fragmentation of the product with propyl amine does not show such fragmentation.

The substrate scope of the coupled reaction was explored with a selection of amines and ESF in different solvents from polar protic to polar aprotic (MeOH, EtOH, water, and ACN). The complete list of 18 different amines is shown in Figure 4. An interactive heatmap was generated by using the conversion ratio (CR) for each combination, obtained by calculating the ion intensity of the product over the sum of the product and reactant intensities from mass spectra in positive mode (Figure 4a). Note that the ion intensities for competitive fragmentation effects were not corrected, as adjustments are expected to be relatively small. The heatmap data suggests that the reaction is most favored in MeOH, and least favored in ACN. The SuFEx reaction is effective in a wide range of solvents, however, it was also reported that the nucleophilic displacement of fluoride is assisted by those solvents which can participate in H‐bonding such as polar protic solvents. ^[31] The electron‐rich aromatic amines give a higher conversion to the SuFEx product, suggesting that the Aza‐Michael addition, which requires higher electron density on the N atom of the amine, is the rate‐determining step. Moreover, when the reaction was performed in AcOH solvent (data not shown) the observed slower reaction is ascribed to the decreased nucleophilicity of the amines. Note that in microdroplets, although the first step (Aza‐Michael addition) is hindered by amine protonation, the subsequent SuFEx reaction is greatly favored by the same acidic environment. The electron‐deficient aromatic amines (A7 & A8) barely show any product. We observed a similar trend for the products with 4‐haloanilines (A9, A10, A11) in the heatmap obtained by the negative mode MS (Figure S20). This heatmap suggests that the negative inductive effect of halide groups determines the reaction outcome. Notably, the reaction is also greatly influenced by steric hindrance. Sterically crowded amines (A3 & A6) show no conversion in comparison to the non‐crowded ones (A4). This suggests that the approach of amines towards the α, β‐unsaturated double bond of ESF is crucial, further indicating that the addition reaction is the slow step. All the aliphatic amines show good conversion of the products except those that are sterically crowded (A16 & A17).

Figure 4.

Heatmap for concerted Aza‐Michael addition and SuFEX reaction using different amines (A1–A18) reacting with ESF in different solvents, analyzed by a) ESSI MS by manual reaction mixture preparation and b) automated HT‐based DESI‐MS. Note that the conversion ratio (CR) was used to generate the heatmap using uncorrected ion intensities.

High‐throughput experimentation (HTE) is a fast and systematic method for reaction monitoring using small quantities of reaction mixtures in an array format.[ ⁵¹ , ⁵² , ⁵³ , ⁵⁴ ] In this study we couple HTE methods with accelerated reactions using desorption electrospray ionization mass spectrometric (DESI‐MS) analysis; this combination constitutes a recent very rapid and label‐free method of surveying reactions (up to 1 reaction/s). It can be utilized to monitor product yields as a function of reactants, pH, catalyst, ligand, enzyme, stoichiometric ratio, solvent, and substrate.[ ¹¹ , ⁵⁵ ] In the recent past, this approach has been successfully utilized for different organic reactions such as N‐alkylation,[ ⁵⁶ , ⁵⁷ ] nucleophilic aromatic substitution, ^[58] some enzymatic reactions,[ ⁵⁵ , ⁵⁹ , ⁶⁰ ] and also for late‐stage functionalization of a targeted drug molecule. ^[26] Here, we have studied the concerted Aza‐Michael addition and SuFEx reaction using the HT‐based DESI‐MS in order to compare the performance of the high‐throughput system with single reactions. All reaction conditions were kept identical for the microdroplet reactions except that DESI was used to generate the microdroplets in the HTE experiment. The concerted Aza‐Michael addition and SuFEx reaction were screened using 18 amines in three different solvents (MeOH, EtOH, and ACN) with the HT‐based DESI‐MS. The reaction was prepared at room temperature and each of the reaction mixtures was transferred to a 384 well‐plate subsequently, 16 replicates of each were pinned on a DESI plate for HT‐based DESI‐MS analysis. Figure 4b shows the corresponding heatmap of the conversion ratios of overall product ions in the positive mode. In the heatmap, the reaction in water is not shown since the reaction mixture in water is difficult to pin on the PTFE‐coated DESI plate. Detailed information on HT‐based DESI‐MS and other aspects of the experiments are provided in Supporting Information. The HT‐produced heatmap shows identical trends to those observed in Figure 4a, as expected because both experiments are based on accelerated reactions in microdroplets and the acceleration factors appear to be similar. Another heatmap, displayed in Figure S21), to shows individual product ion intensities (Aza‐Michael addition, methyl ester, and ethyl ester). Note that the DESI array experiment used on the order of 5 ng of the sample, so the treatment gives YES/NO answers to the question of reactivity rather than the graded scale achieved by ESSI examination of individual reaction mixtures. The successful use of HT‐based DESI‐MS analysis for the aza‐Michael addition and SuFEx reaction products encourages extension of the experiments to DESI high‐throughput bio‐assays of these and other reaction products for applications in drug discovery. It is noted that both ESSI and DESI generated microdroplets are utilized in this study. Although, both are ambient ionization methods, there are differences which can lead to the choice of one or the other in particular experiments. In DESI, secondary microdroplets are utilized for analysis, whereas in ESSI, a primary microdroplet stream is subjected to analysis. DESI coupled with automated high‐throughput is utilized for reaction monitoring with minimal requirements regarding sample complexity, whereas ESSI cannot be used for samples with high salt concentrations because of concerns for capillary blockage. ESSI provides more control over the size of the microdroplets and the associated degree of reaction acceleration. In general, both these methods provide useful data and either can be chosen depending on experimental constraints.

In summary, the accelerated Aza‐Michael addition and SuFEx reaction between amines and ESF occur in a concerted fashion in microdroplets. This method shows good product conversion without using any base or catalyst. The intrinsic superacidic nature of the microdroplet surface[ ²¹ , ²² ] favors reaction acceleration in other systems where the bulk reaction calls for acid or base catalysis. The versatility and chemical scope of this reaction is also demonstrated in terms of different solvents and a range of diverse substrates.

The microdroplet coupled Aza‐Michael addition and SuFEx reaction in microdroplets has analogies with the orthogonal SuFEx and CuAAC coupled click reaction of Yang et al. ^[39] It is simple, catalyst‐free, and occurs under ambient conditions. These features make it readily amenable to high‐throughput reaction screening. This example of the important SuFEx chemistry encourages interest in the possible use of accelerated reactions in drug discovery, not just in high‐throughput analysis but also, as shown in other recent work,[ ²⁶ , ⁵⁵ , ⁵⁶ , ⁵⁸ , ⁵⁹ ] as a small‐scale synthetic method with bioassays being performed on the collected products.

Conflict of interest

The authors declare no conflict of interest.

Supporting information

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

Supporting Information

J. Ghosh, J. Mendoza, R. G. Cooks, Angew. Chem. Int. Ed. 2022, 61, e202214090; Angew. Chem. 2022, 134, e202214090.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.