Abera Tadie Derese*a,
Mengistu Gemech Menkir
b,
Mihret Kendie Woliec and
Desalegn Addis Yemamd
aDepartment of Chemistry, Debre Tabor University, Ethiopia. E-mail: aberatadie@gmail.com
bDepartment of Chemistry, Bahir Dar University, Ethiopia
cDepartment of Chemistry, Debark University, Ethiopia
dDepartment of Chemistry, Debre Tabor University, Ethiopia
First published on 28th January 2025
DFT calculations were performed to investigate the possible reaction mechanisms underlying catalyst-free chloroboration reactions of carbonyl compounds with BCl3. The interaction between BCl3 and the CO moiety of carbonyl compounds is a two-step reaction. In the first step, B of BCl3 forms a bond with the O of the C
O moiety, followed by the 1,3-Cl migration process from BCl3 to the C of the carbonyl group. To indicate the versatility of our synthetic methodology, a catalyst-free chloroboration of a variety of aldehydes and ketones with a broad range of electron-donating and electron-withdrawing groups with BCl3 was checked. According to DFT results, BCl3-induced chloroboration of aldehydes and ketones progressed under a kinetically favorable condition with <20 kcal mol−1 of activation free energy.
The discovery of transition metal-catalyzed hydroboration, which uses pinacolborane (HBpin) or catecholborane (HBcat) to selectively reduce unsaturated carbonyl compounds, has been a significant advancement in the past years.17–22 Classically, different methods with the involvement of metals, either as catalysts or in stoichiometric proportions, have been discovered to synthesize organoboron compounds.23–26 However, the involvement of metals may render these preparative techniques less economical and environmentally benign, and it may result in heavy metal contamination of the final boron products.27
To eliminate the above-listed drawbacks, transition-metal-free catalytic methods were widely adopted over the traditional approach.28 Among these, transition-metal-free catalytic diboration of unsaturated hydrocarbons,29,30 transition-metal-free catalytic β-boration of α,β-unsaturated compounds,31 transition-metal-free catalytic borylation of allylic and propargylic alcohols,32 and transition-metal-free catalytic hydroboration of unsaturated hydrocarbons33–36 have been reported.
The use of protocols in organic transformations that do not require solvents or catalysts has gained considerable interest lately as a result of the growing need for low-cost, atom-economical, and environmentally safe synthetic processes.37,38 Recently, catalyst-free and solvent-free hydroboration of carbonyl compounds using pinacol borane has been reported. For example, in 2018, Stachowiak and coworkers studied catalyst-free and solvent-free hydroboration of aldehydes.39 In addition, Wang and colleagues experimentally and computationally investigated catalyst-free and solvent-free hydroboration of ketones40 and carboxylic acids.41 However, for the hydroboration of ketones and carboxylic acids, computational results reveal that the reactions are hard to proceed at room temperature because of the high energy barrier.
Furthermore, borylative cyclization of unsaturated compounds without any catalyst, such as intramolecular amination of alkenes and alkynes, metal-free borylative cyclization of alkynes, intramolecular aminoboration of allenes, and catalyst-free annulative thioboration of unfunctionalized olefins was demonstrated using BCl3 as a boron source.42–45 However, the action of BCl3 on catalyst-free chloroboration of carbonyl compounds has not yet studied. Inspired by the work described above, we decided to carry computational work on BCl3-induced chloroboration of carbonyl compounds in a catalyst-free manner. In this study, we performed a DFT calculation for catalyst-free chloroboration of aldehydes and ketones using BCl3 as a boron source.
![]() | ||
Scheme 1 Proposed mechanism for the BCl3-mediated chloroboration of aldehydes and ketones (R![]() ![]() |
We started the study by reacting benzaldehyde and BCl3 without the use of a catalyst. Based on the DFT calculation results shown in Fig. 1, the production of the coordinated complex A-Intr0 between the CO group of the model reactant (benzaldehyde) and BCl3 is found to be 4.9 kcal mol−1. Through A-Ts1, A-Intr0 is subsequently transformed into the zwitterion intermediate A-Intr1 at a lower activation energy of 2.1 kcal mol−1. The generation of the borylated intermediate A-product via the four-membered ring transition state A-Ts2 requires 11.9 kcal mol−1, which corresponds to the formation of C–Cl. The breaking B–Cl distance is 2.12 Å, whereas the forming C–Cl distance is 2.18 Å.
Following the identification of the most advantageous route for the chloroboration of the aldehyde model reactant, we perform analogous DFT computations on ketones using acetophenone as a substrate. The complexation process between BCl3 and the CO moiety of the model reactant, acetophenone, initiates the reaction. Fig. 1 illustrates that the former transition state, K-Ts1, has a lower energy barrier of 1.8 kcal mol−1, making it more kinetically beneficial to afford K-Intr1. Following the formation of K-Intr1, the K-product is formed by the easy transfer of the chloride anion to the carbonyl carbon via K-Ts2, which demands an energy of 13.5 kcal mol−1 (1.6 kcal mol−1 less stable than A-Ts2). Note that 2.09 Å is the breaking B–Cl distance, while 2.17 Å is the forming C–Cl distance.
According to DFT results displayed in Fig. 1, aldehydes exhibited a higher reactivity than ketones owing to the steric hindrance effect and ketones' reduced electrophilicity. Fig. 2 shows the structure of the transition states and optimized intermediates during the chloroboration of aldehydes and ketones.
After we performed DFT calculation for the model substrate, we looked at the extent of the reaction when specific substituted benzaldehydes bearing methyl, methoxy, halogens, and heterocyclic derivatives were functionalized with either electron-withdrawing or electron-donating groups Table 1. Fig. 3 presents DFT results for the BCl3-induced chloroboration reaction of substituted benzaldehydes bearing m-CH3, o-CH3, p-CF3, and p-F. In contrast to the parent model substrate benzaldehyde (Fig. 1), the chloroboration of aldehydes substituted with o-CH3 and p-CF3 proceeded with lower activation energy, yielding more stable boronate ester products (Fig. 3). The reaction proceeded with somewhat greater activation energy to afford the predicted boronate esters when the substituent groups are m-CH3 and p-F when compared with the parent substrate benzaldehyde. Fig. 5 shows the DFT results for the BCl3-induced chloroboration reaction of substituted benzaldehydes bearing p-Br and p-OMe as well as picolinbenzaldehyde and thiophen-2-carbaldehyde as a substrate. p-Br and p-OMe substituted benzaldehyde and thiophen-2-carbaldehyde demands a higher energy barrier than the model reactant. Picolinbenzaldehyde consumes energy almost equal to that consumed by benzaldehyde. Fig. 4 and 6 show the geometric structures for intermediates and transition states in the chloroboration of substituted benzaldehyde.
![]() | ||
Fig. 3 DFT-computed free energies for chloroboration reaction of substituted benzaldehyde bearing m-CH3, o-CH3, p-CF3, and p-F. |
![]() | ||
Fig. 4 Geometric structures of intermediates and transition states in the chloroboration of substituted benzaldehyde bearing m-CH3, o-CH3, p-CF3 and p-F (distances are given in Å). |
![]() | ||
Fig. 5 DFT-computed free energies for the chloroboration reaction of substituted benzaldehyde bearing p-Br and p-OMe, picolinbenzaldehyde and thiophen-2-carbaldehyde. |
The production of the coordinated complex A-Intr2 between the CO group of the reactant (3-methylbenzaldehyde) and BCl3 is found to be 2.4 kcal mol−1 based on the calculation results given in Fig. 3. A-Intr2 is then converted via A-Ts3 into the zwitterion intermediate A-Intr3, which has a lower activation energy of 2.7 kcal mol−1. Notably, 13.3 kcal mol−1 (1.4 kcal mol−1 greater than that of the model reactant) is needed for the four-membered ring transition state A-Ts4 to form the borylated intermediate A-Intr4. The coordinated complex A-Intr5 is determined to be 3.3 kcal mol−1 when CH3 is in ortho-position (2-methylbenzaldehyde), and A-Ts5 forms A-Intr6 (requiring 2.3 kcal mol−1). A-Ts6 converts A-Intr6 into A-Intr7 at 10.3 kcal mol−1, which is 1.6 kcal mol−1 more stable than the model reactant and 3.0 kcal mol−1 from 3-methylbenzaldehyde. By increasing the electron density in the benzene ring, the CH3 group, an activating group, promotes the reaction progress.
The formation of the coordinated complex A-Intr8 between the CO group of 4-(trifluoromethyl)benzaldehyde and BCl3 is reported to be 4.0 kcal mol−1 based on the computation results displayed in Fig. 3. Next, by converting A-Intr8 via A-Ts7 (3.5 kcal mol−1), the zwitterion intermediate A-Intr9 is produced; 11.3 kcal mol−1 is required for the four-membered ring transition state A-Ts8 to produce the borylated intermediate A-Intr10. The coordinated complex A-Intr11 is determined to be 3.6 kcal mol−1 when the substituent is p-F. This is followed by the production of A-Intr12 via A-Ts9, which requires 2.2 kcal mol−1. A-Ts10 converts A-Intr12 into A-Intr13 at a cost of 12.2 kcal mol−1. This indicates that compared to 4-(fluoro)benzaldehyde, 4-(trifluoromethyl)benzaldehyde is more reactive because compared to F atoms, CF3 is a more potent electron-withdrawing group. The structure of the optimal intermediates and transition states during the chloroboration of replacement benzaldehyde is shown in Fig. 4.
ased on the computation findings shown in Fig. 5, the synthesis of the coordinated complex A-Intr14 between the CO group of the reactant (4-bromobenzaldehyde) and BCl3 is found to be 3.6 kcal mol−1. The zwitterion intermediate A-Intr15, which has lower activation energy of 2.9 kcal mol−1, is subsequently created by converting A-Intr14 via A-Ts11. The borylated intermediate A-Intr16 is obtained by converting the four-membered ring transition state A-Ts12 to 12.8 kcal mol−1. When p-OMe is the substituent, A-Intr17 is formed with 3.5 kcal mol−1 of energy, while A-Intr18 through A-Ts13 is formed with 1.8 kcal mol−1. As a result, A-Ts14 forms A-Intr19, requiring 14.8 kcal mol−1 of energy, which makes it 2.9 kcal mol−1 less stable than the model substrate.
The synthesis of the coordinated complex A-Intr20 between CO group picolinbenzaldehyde and BCl3 is found to be 3.5 kcal mol−1, based on the computation results displayed in Fig. 5. Next, using A-Ts15 with 2.7 kcal mol−1, A-Intr20 is converted into the zwitterion intermediate A-Intr21. The four-membered ring transition state A-Ts16 requires 11.8 kcal mol−1 free energy to yield the borylated intermediate A-Intr22. A-Intr23 (3.4 kcal mol−1) is produced when thiophen-2-carbaldehyde is used as the substrate in a reaction with BCl3. A-Ts17 (2.3 kcal mol−1) then converts this product into A-Intr24. A-Intr24 is converted into A-Intr25 via A-Ts18 (15.5 kcal mol−1, 3.6 kcal mol−1 less stable than the model reactant).
Following a computational study on the chloroboration of substituted benzaldehyde, we performed DFT calculations to better understand the scope of BCl3-promoted chloroboration of ketones in a catalyst-free manner when specific substituted acetophenones containing methyl, methoxy, halogens, nitro, and heterocyclic derivatives were functionalized with either electron-donating or electron-withdrawing groups Table 2. Fig. 7 and 9 show DFT-computed free energies for the chloroboration reactions of substituted acetophenone. Fig. 8 and 10 show the geometric structures for intermediates and transition states in the chloroboration of substituted acetophenone.
![]() | ||
Fig. 7 DFT-computed free energies for the chloroboration reaction of substituted acetophenone bearing p-CH3, m-CH3, p-NO2 and p-Br. |
![]() | ||
Fig. 8 Geometric structures of intermediates and transition states in the chloroboration of substituted acetophenone bearing p-CH3, o-CH3, p-NO2 and p-Br (distances are given in Å). |
![]() | ||
Fig. 9 DFT-computed free energies for the chloroboration reaction of substituted acetophenone bearing p-CF3, p-OMe, 1-(pyridin-2-yl)ethanone and 1-(furan-2-yl)ethanone. |
Complexation of BCl3 with 1-(p-tolyl)ethanone, which occurs kinetically at 3.5 kcal mol−1 to produce K-Intr2, is the first step. K-Intr3 is then generated through the transition state K-Ts3, which has a lower energy barrier of 1.8 kcal mol−1. After K-Intr3 (2.4 kcal mol−1) is created, K-Ts4 (15.7 kcal mol−1) facilitates the easy transfer of the chloride anion to the carbonyl carbon to generate K-Intr4. It is less stable by 2.2 kcal mol−1 compared to the model substrate. The coordinated complex A-Intr5 forms with a reduced activation energy of 2.8 kcal mol−1 when the substrate is 1-(m-tolyl)ethanone. With a lower energy barrier of 2.6 kcal mol−1, the previous transition state, K-Ts5, is more kinetically advantageous to produce K-Intr6. After K-Intr6 is produced, K-Ts6, which is 13.4 kcal mol−1, may be used to transfer the chloride anion to the carbonyl carbon with ease, yielding the desired product, K-Intr7. In comparison with K-Ts4 of 1-(p-tolyl)ethanone, K-Ts6 of 1-(m-tolyl)ethanone is stable by 2.3 kcal mol−1.
The formation of the complex intermediate K-Intr8 (3.4 kcal mol−1) occurs when 1-(4-nitrophenyl)ethanone is the substrate. K-Intr9 is generated through the transition state K-Ts7, which has a lower energy barrier of 3.3 kcal mol−1. After K-Intr9 (2.2 kcal mol−1) is created, K-Ts8, which is 12.2 kcal mol−1 (1.3 kcal mol−1 stable than the model substrate), facilitates the easy transfer of the chloride anion to the carbonyl carbon, generating K-Intr10. The coordinated complex A-Intr11 is generated with an activation energy of 3.4 kcal mol−1 when the substrate is 1-(4-bromophenyl)ethanone. With a lower energy barrier of 2.0 kcal mol−1, the previous transition state, K-Ts9, is more kinetically advantageous to produce K-Intr12. K-Ts10, which is 13.2 kcal mol−1, makes it simple to move the chloride anion to the carbonyl carbon once K-Intr12 is formed. K-Intr13 is the final product of this step.
As shown in Fig. 9, the CO moiety of the 1-(4-(trifluoromethyl)phenyl)ethanone reactant molecule combines with BCl3. K-Intr14 is kinetically advantageous since K-Ts11 has a lower energy barrier of 2.9 kcal mol−1. K-Intr16 is produced by the rapid transfer of the chloride anion to the carbonyl carbon via K-Ts12 (12.6 kcal mol−1), which occurs after K-Intr15 (1.0 kcal mol−1) is formed. When OMe is a substituent of acetophenone at para-position, 1-(4-methoxyphenyl)ethanone forms the complex K-Intr17 with BCl3 by demanding 3.0 kcal mol−1. This is followed by the synthesis of K-Intr18 through K-Ts13, which requires 1.5 kcal mol−1. Finally, using K-Ts14, K-Intro19 is generated with an activation energy of 15.3 kcal mol−1 (1.8 kcal mol−1 higher than that of the model reactant) of energy.
The CO moiety of 1-(pyridin-2-yl)ethanone reacts with BCl3 to generate K-intr20 (3.0 kcal mol−1) as the reactant. As can be observed in Fig. 9, K-Ts15, the preceding transition state, has a lower energy barrier of 1.5 kcal mol−1, making K-Intr21 kinetically favorable. When the chloride anion is shifted from K-Intr21 to the carbonyl carbon via K-Ts16 (11.6 kcal mol−1, 1.9 kcal mol−1 stable than the model reactant), K-Intr22 is formed. While 1-(furan-2-yl)ethanone serves as the substrate, K-Intr23 is created at a rate of 2.7 kcal mol−1. For K-Ts17 to convert K-Intr23 into K-Intr24, 1.1 kcal mol−1 is required. Finally, K-Ts18, which requires 17.0 kcal mol−1, transfers K-Intr24 to K-Intr24. K-Ts18 makes 1-(furan-2-yl)ethanone less stable than the model substrate by 3.5 kcal mol−1.
Footnote |
† Electronic supplementary information (ESI) available: Cartesian coordinates (Å). See DOI: https://doi.org/10.1039/d4ra06893a |
This journal is © The Royal Society of Chemistry 2025 |