Open Access Article
Marie Caldiero,
Jean-Philippe Bouillon
* and
Thomas Castanheiro
*
Univ. Rouen Normandie, INSA Rouen Normandie, Univ. Caen Normandie, ENSICAEN, CNRS, Institute CARMeN UMR 6064, F-76000 Rouen, France. E-mail: thomas.castanheiro-matias@univ-rouen.fr
First published on 14th April 2026
A mechanosynthesis of α-bromoacrylates was described from commercially available triethyl phosphonoacetate by using a ball-milling-assisted bromination/HWE one-pot process. Initially, the bromination and the HWE protocols were independently optimized under mechanochemical conditions using a weak base and short reaction time. The solvent-free one-pot procedure required neither sequential addition nor an excess of reagents, thereby minimizing synthetic waste. α-Bromoacrylates were obtained with a general high efficiency, either through the solvent-free HWE (36 examples, from 26 to 98% yields) or via the one-pot procedure (9 examples, from 27 to 91% yields), with moderate Z-selectivity. This practical synthetic methodology was further extended to the synthesis of α-iodo- and α-chloroacrylates using NIS and NCS as iodinating and chlorinating agents, broadening the scope of halogenated alkenes formation through our one-pot approach. The α-bromoacrylates were then reacted in a ball-milling-assisted Ullman coupling, demonstrating the synthetic utility of the solvent-free synthesized products.
Wittig and Horner–Wadsworth–Emmons (HWE) reactions are powerful and appealing strategies for the direct construction of polysubstituted alkenes from readily accessible carbonyl substrates.6 α-Bromoacrylates synthesis benefited from these powerful methods. Initially, brominated stabilized ylides have been used to synthesize α-bromoacrylates through the powerful Wittig strategy (Scheme 1A).7 More recently, Karama et al. developed a one-pot procedure for the formation of the desired (Z)-α-bromoacrylates starting from commercially available (carboethoxymethylene)triphenylphosphorane. The in situ formation of the brominated stabilized ylides was enabled by reacting N-bromosuccinimide as a bromine source.8 Huang and co-workers reported a strategy employing α-hypervalent iodine-functionalized phosphonium or arsonium ylides (Scheme 1A).9 In the presence of nBu4NBr as nucleophile, the phosphonium or arsonium ylides acted as Umpolung ylides, generating the corresponding brominated ylides in situ, which subsequently underwent a Wittig-type olefination. This approach predominantly afforded α-bromo-α,β-unsaturated esters with high Z-selectivity. Over the years, HWE strategies employing different brominated phosphonates have also been developed to access α-bromoacrylates with high yields and excellent stereoselectivities. In 2000, Kogen and coll. designed and prepared a phosphonate, bis(2,2,2-trifluoroethoxy)-bromophosphonate (Scheme 1A), which reacted efficiently with aldehydes to afford the desired products with a high E-selectivity.10 In 2004, Brückner and co-workers adopted an Ando's variant of the HWE for the selective synthesis of (E)-α-bromoacrylates.11a A 64:36 mixture of mono- and dibrominated (phenoxy)-phosphonates was synthesized and subsequently reacted with aldehydes in the presence of NaH at 0 °C, affording the desired α-bromoacrylates with a high E-selectivity (Scheme 1A). This strategy was also extended to the synthesis of α-chloroacrylates with a 94:6 E:Z selectivity. However, the corresponding chlorinated (phenoxy)phosphonate was obtained in low yield. Very recently, Ando et al. reported the preparation of t-butyl(diphenylphosphono)-bromoacetate using 1.1 equivalent of NaH and 1.0 equivalent of Br(CCl2)2Br at 0 °C in a two-step, one-pot sequence.11b The corresponding mono-brominated phosphonate was obtained with good selectivity. The α-bromoacrylates were subsequently synthesized in good yields and with high E-selectivity through a two-step sequence conducted at low temperature. In addition, (E)-iodoacrylates were synthesized by reacting t-butyl(diphenylphosphono)bromoacetate with NaI and NaH at 0 °C in a first step, followed by reaction with the corresponding aldehyde at −78 °C in the second step. Ethyl(diethylphosphono)bromoacetate was also employed to produce α-bromoacrylates (Scheme 1A); however, this approach resulted in lower selectivity, favoring the Z-isomer.12
Although these strategies efficiently afforded highly desirable α-bromoacrylates either with Z- or E-selectivity, they require the use of strong bases, strict temperature control (both high or low temperatures), sequential addition of reagents, and harmful organic solvents to achieve satisfactory outcomes. Therefore, the development of more practical and efficient methods for α-bromoacrylates synthesis, with particular attention paid to sustainability, is somewhat needed.
Mechanochemistry, which is known for centuries,13 is re-emerging as a powerful synthetic tool for developing more practical and sustainable synthetic methodologies by minimizing the use of harmful solvents. Removing solvent from the reaction mixture by using ball-milling in synthesis not only enhances sustainability metrics of reactions but also often improves their overall performance. Indeed, ball-milling frequently leads to enhanced or unprecedent reactivities and selectivities, eliminates the need for external heating, reduces reaction times, and enables efficient one-pot procedures, all key advantages of solvent-free approaches.14 In this context, Lamaty and coll. developed in 2010 a HWE procedure for the synthesis of Boc-protected unsaturated amino esters under ball-milling conditions (Scheme 1B).15 This methodology enabled a one-pot process, avoiding the need for a preliminary generation of the deprotonated phosphonate. Furthermore, in contrast to the strong base commonly used in such transformations (NaH, tBuLi), K2CO3 or Cs2CO3 were successfully employed as a milder alternative. However, the reaction still required seven hours to reach completion and was limited to aldehydes. Later, Mack and coll. reported a ball-milling-assisted synthesis of α,β-unsaturated esters starting from commercially available triethyl phosphonoacetate via a HWE process (Scheme 1B).16 Although K2CO3 was used as a base, an additional 12 hours ball-milling oxidation step was necessary to consume the excess aldehyde that remained unreacted after the initial 6 hours solvent-free HWE step. In both cases, the reactions proceeded with good to excellent stereoselectivity. This mechanochemical-assisted HWE process was then applied for the synthesis of α,β-unsaturated amides,17 phosphonocinamic esters,18 and valuable monofluoroalkenes.19 However, none of this strategy was applied for the formation of α-bromoacrylates.
Inspired by the pioneering work of Lamaty's group15 and subsequent ball-milling-assisted HWE strategies that have been developed,16–19 we sought to develop an innovative solvent-free synthesis of valuable α-bromoacrylates starting from commercially available triethyl phosphonoacetate. To establish a practical protocol, we designed a one-pot procedure involving the ball-milling-assisted bromination of triethyl phosphonoacetate, followed by a ball-milling-assisted HWE reaction carried out in the same milling vessel (Scheme 1C).
However, since no solvent-free synthesis of triethyl bromophosphonoacetate 3 had been reported to date, at the outset of our investigation, we investigated the preparation of 3 under ball-milling conditions (Table 1). Fortunately, when 1 was subjected to ball-milling in the presence of 1.06 equivalent of N-bromosuccinimide and NaH, the common base used for the solution-phase bromination of 1, the desired brominated phosphonoacetate 3 was obtained in a good yield (Table 1, entry 1). A minor amount of the dibrominated product 4 was also detected, with 6% 31P NMR yield.
| Entry | Variation from standard conditions | Yield 3a (%) | Yield 4a (%) |
|---|---|---|---|
| a 31P NMR yield calculated with diethyl methylphosphonate as internal standard.b Isolated yield. | |||
| 1 | None | 61 | 6 |
| 2 | 1.06 equiv. of tBuOK | 68 | 9 |
| 3 | 1.06 equiv. of KOH | 80 | 10 |
| 4 | 1.06 equiv. of K2CO3 | 85 | 11 |
| 5 | 1.06 equiv. of Cs2CO3 | 82 (63)b | 8 |
| 6 | 1.06 equiv. of NEt3 | 0 | 0 |
| 7 | 1.06 equiv. of Cs2CO3, 15 min | 84 | 14 |
| 8 | 1.06 equiv. of Cs2CO3, 10 min | 88 (56)b | 11 |
| 9 | 1.06 equiv. of Cs2CO3, 10 min, no ball-milling | 6 | 13 |
We then explored the use of milder inorganic bases (entries 2 to 5). The reaction proceeded efficiently with either K2CO3 or Cs2CO3, both of which have been identified as optimal bases in previously reported ball-milling-assisted HWE process.15–19 However, due to purification difficulties arising from the simultaneous formation of 4, the desired triethyl bromophosphonoacetate 3 was isolated in a 63% yield (entry 5). The use of an organic base such as triethylamine failed to promote the bromination of 1 (entry 6). With the goal of developing a one-pot ball-milling bromination/HWE protocol, the bromination step was evaluated at shorter milling times (15 and 10 minutes, entries 7 and 8). Remarkably, formation of triethyl bromophosphonoacetate 3 was observed with 88% 31P NMR yield after only 10 minutes of milling at 35 Hz (entry 8). Due to the close polarity between 3 and 4, the desired monobrominated product has been isolated with only 56% yield. It is noteworthy that in the absence of ball-milling, the desired product 3 was formed in only a 6% 31P NMR yield (entry 9). Finally, the addition of liquid additives (LAG), and variation of the equivalence of base and number of stainless-steel balls were also tested without any significant impact on the reaction outcome. With the optimized ball-milling-assisted bromination in hand (Table 1, entry 8), we next focused on developing the ball-milling HWE process to access α-bromoacrylates 6 from 3. Building on the pioneering work of Lamaty's group15 and previous reports,16,19 1.1 equivalent of 3 was reacted with benzaldehyde under 35 Hz of milling for 30 min using stainless-steel milling materials (jar and balls). Cs2CO3 was employed as a base to aim for the one-pot bromination/HWE procedure. Pleasingly, the desired product 6a was obtained in a good 77% yield and 1:2.9 E:Z ratio (Scheme 2). The reaction proved to be efficient for the synthesis of α-bromoacrylate on a 3.0 mmol scale with an increased isolated yield (92%). A broad range of substituted aromatic aldehydes were compatible under our solvent-free conditions without any impact of the physical state of the starting aldehyde (Scheme 2), including aliphatic, electron-donating, and electron-withdrawing substituents at the para-, meta- or ortho-position. Notably, steric hindrance had no significant effect on the reaction outcome. The presence of a Me substituent, regardless of its position on the aromatic ring, had little influence on the yield or stereoselectivity (6b to 6d). Polyaromatic aldehydes also reacted efficiently, delivering the desired products 6e and 6f in high yields. Both electron-donating (Ph 5g, NMe2 5h or SMe 5i) and electron-withdrawing (CF3 5j, OCF3 5k, NO2 5l, CN 5m and CO2Me 5n) groups at the para-position enabled the formation of the expected products in high efficiency and moderate E:Z ratios. In addition, shifting the electron-withdrawing substituents enabled the formation of the desired product with similar results, with even better Z-selectivity (6o). It is worth mentioning that valuable post-functionalizable moieties were well tolerated, including an amine (6h), a thioether (6i), a nitro (6l), a cyano (6m), and an ester motif (6n). Halogen atoms, such as iodine (5p), bromine (5q and 5t), chlorine (5r, 5u and 5w), and fluorine (5s and 5v), were also compatible with our solvent-free reaction conditions, affording the desired products with high efficiency. These products open significant opportunities for further transformations. Once again, the presence of a halogen atom at the ortho-position enabled an increase in stereoselectivity in favor of the Z-isomer (6u and 6v). Additionally, a boronic ester-substituted aldehyde (5x) efficiently produced the corresponding product, highlighting potential for subsequent Suzuki–Miyaura-type cross-coupling reactions. Complete regioselectivity towards the aldehyde moiety was observed when using 3-acetylbenzaldehyde (5y). Heteroaryl aldehydes were also suitable substrates. While indole derivative aldehyde (5z) led to lower yields, pyridine-, thiophene- and furan-derived aldehydes (5aa, 5ab, and 5ac) afforded the bromoalkene product in good to excellent yields. Interestingly, improved Z-selectivity was observed when using furan (E:Z = 1:3) and indole derivatives (E:Z > 1:20). Furthermore, a synthetically valuable brominated diene (6ad) was synthesized in 71% yield with moderated Z-selectivity. Aliphatic aldehydes also proved to be competent substrates, giving moderate to good yields and moderate to good stereoselectivities. Phenylpropane- and cyclohexane-derived aldehydes (5ae and 5af) showcased a high efficiency, whereas tBu-derived aldehyde (5ag) provided a lower yield, albeit with improved E:Z ratio. Ethyl oxoacetate afforded product 6ah with a decent yield and reversed stereoselectivity. To further demonstrate the robustness of our solvent-free strategy, more complex aldehydes were evaluated. Citronellal (5ai), responsible for the characteristic lemon fragrance of citronella oil, and retinal (5aj), a polyene chromophore considered as one of the forms of vitamin A, were successfully transformed in moderate to good yields and stereoselectivities. In contrast, ketones failed to react under our optimal solvent-free conditions. Having established the optimal conditions for the ball-milling-assisted formation of triethyl bromophosphonoacetate and delineated the scope and limitations of our solvent-free HWE process for the synthesis of α-bromoacrylates, we then investigated the targeted bromination/HWE one-pot sequence (Scheme 3). Initially, 4-methylbenzaldehyde was reacted with 1.1 equivalent of triethyl phosphonoacetate in the presence of 3.06 equivalents of Cs2CO3 and 1.06 equivalent of NBS. Pleasingly, after 40 minutes of grinding, the combination of both previously optimized reaction conditions enabled the successful formation of the corresponding α-bromoacrylate 6b with a good 64% and moderate stereoselectivity.
![]() | ||
| Scheme 2 Ball-milling synthesis of α-bromoacrylates through solvent-free HWE. a Reaction performed on 3 mmol scale. | ||
These milling conditions were then applied to a variety of aldehydes and proved to be tolerant to a broad range of functional groups. Aldehydes containing either an electron-donating (5i) or an electron-withdrawing substituent (5l) on the aromatic ring were efficiently converted with results comparable to those obtained when performing both reactions separately. When a thiophene-derived aldehyde (5ab) was subjected to the one-pot reaction, the desired product was obtained in decreased, though still with decent efficiency. The reaction with cinnamaldehyde afforded the corresponding product 6ad in a modest 35% yield; however, the good E:Z ratio of 1:4 (previously observed in the stepwise process) was maintained. An aliphatic aldehyde also proved to be a suitable substrate, producing the targeted product in even higher yield (6ae, 83%). Interestingly, the E:Z ratio improved significantly under one-pot conditions, reaching up to 1:9 compared to a 1:4 ratio when performing both steps separately. To further expand the scope of accessible brominated alkenes, tetraethyl methylenebis(phosphonate) and diethyl (cyanomethyl)-phosphonate were also tested under our ball-milling-assisted one-pot bromination/HWE protocol. Gratifyingly, 6ak was obtained in a decent yield, while α-bromoacrylonitrile 6al was isolated in a lower yield, albeit with a good Z-stereoselectivity compared to 6ak. These results underscore the efficiency and versatility of our mechanochemical synthetic strategy for accessing synthetically valuable α-bromoalkene derivatives, and open new avenues for the synthesis of alkenes from functionalized bromo-phosphonoacetate derivatives.
We then wondered whether the corresponding iodinated and chlorinated analogues could be synthesized under our ball-milling procedure by replacing NBS with either NIS and NCS. We initially focused on the formation of triethyl iodophosphonoacetate (7) and triethyl chlorophosphonoacetate (9) under our mechanochemical conditions (Scheme 4A). 7 was formed with a high 95% 31P NMR yield after only 10 minutes of milling, with a minimal formation of the diiodinated by-product (8). In contrast, the mechanosynthesis of the chlorinated analogue required extended milling time (up to 90 minutes) to afford a satisfactory 31P NMR yield, with the dichlorinated by-product 10 also detected in 19% 31P NMR yield. Unfortunately, due to the inherent instability of both desired products 7 and 9, they were isolated in moderate 56% and 48% yields, respectively. Nevertheless, having demonstrated that triethyl iodophosphonoacetate 7 and triethyl chlorophosphonoacetate 9 could be generated under mechanochemical conditions, we next sought to develop one-pot iodination/HWE and chlorination/HWE procedures, which could circumvent the need for isolating 7 and 9 (Scheme 4B). When all reagents were introduced simultaneously at the start of the milling process, the desired α-iodo- and α-chloroacrylates 11 and 12 were obtained in a mixture with non-iodinated and non-chlorinated acrylates, hampering isolation of the desired products. Therefore, triethyl phosphonoacetate (1) was first milled with 1.06 equivalent of Cs2CO3 and NXS (X = I or Cl) for either 10 or 90 minutes at 35 Hz, respectively. The milling jar was then opened to add 1.0 equivalent of the corresponding aldehyde and an additional 2.0 equivalents of Cs2CO3. With this sequential one-pot approach, a range of aldehydes was successfully converted into the corresponding α-iodoacrylates (11a–11d) and α-chloroacrylates (12a–12d) in moderate to good yields. The iodination/HWE sequence afforded the desired products in generally moderate yields (11a–c), regardless of the electronic nature of the substituents on the aromatic ring.
In contrast, the chlorination/HWE sequence provided higher yields for the corresponding products (12a–c). However, the introduction of an electron-withdrawing motif on the starting aldehyde resulted in a decreased yield (see 12c). Heteroaryl and aliphatic aldehydes reacted with similar reaction efficiency within the iodination/HWE sequence (11d–e), whereas lower yields were obtained for the corresponding chlorinated products (12d–e). Overall, moderate E:Z ratios were obtained, except in the case of aliphatic aldehydes.
To demonstrate the synthetic utility of our synthesized products, we sought to exploit the C(sp2)−Br bond to develop a C–N bond formation under mechanochemical conditions. Initially, we developed a solvent-free Ullman-type coupling (Scheme 5). Under ball-milling at 35 Hz in stainless-steel milling equipment, 6a reacted with oxazolidine-2-one 13 in the presence of CuI as catalyst, affording the expected coupling product 14a in a good 65% yield and with high Z-selectivity. These milling conditions were subsequently applied to 6s, leading to the formation of 14b in a decent 56% yield with excellent stereoselectivity (E:Z > 1:20). In contrast, the use of 2-pyrolidinone as a coupling partner resulted in lower reaction efficiency, although a high E:Z ratio was still observed. An electron-withdrawing group-substituted α-bromoacrylate also proved to be a competent substrate in this solid-state Ulmann-type reaction, delivering the desired product 14d in a good 67% yield, with same high Z-selectivity.
| This journal is © The Royal Society of Chemistry 2026 |