Toluene (99.5%), ethylbenzene (99.5%), m-xylene (99%), mesitylene (97%), benzene (99.5%), CO₂ (99.99%), and anhydrous AlCl₃ were used for carboxylation, which were commercially purchased and used as received. The anhydrous AlCl₃ was handled under inert atmosphere to avoid decomposing by water. Organosilanes R₄Si such as Me₃SiCl, Me₂PhSiCl, MePh₂SiCl, Ph₃SiCl were also used in carboxylation. HCl (36% - 38%), anhydrous diethyl ether (99.5%) and NaOH (96%) were used for separation of product carboxylic acids.

The instruments used in this work were Nuclear Magnetic Resonance (NMR, Bruker 500), and Infrared Imaging Spectrometer (Nicolet 6700). The carboxylation was carried out in a 100 ml 316 L stainless steel autoclave coated with a heating jacket and a magnetic stirrer.

All materials including Al₂Cl₆ (2.25 mmol, 0.6 g), toluene (20 ml), and Me₃SiCl (0.25 g, 2.25 mmol) were added to the autoclave. After pressurization of CO₂, the temperature was raised to 80˚C, where the carboxylation was conducted for 12 h until the reaction reached equilibrium. Afterwards, the reaction was stopped by cooling down to room temperature. Water was added to the reacted mixture and the products were extracted with diethyl ether and 10% NaOH. The extracted aqueous phase was adjusted to pH = 1 by adding hydrochloric acid. Precipitated white crystals were obtained after the ice bath cooling.

The toluic acid obtained was qualitatively identified by IR. The ratio of the acid isomers was evaluated by relative integration ratio of¹H NMR signals, δH (500MHz, CDCl₃, Me₄Si) 2.56 (H, s, o-Me), 2.39 (H, s, m-Me), 2.31 (H, s, p-Me). The ratio of ortho, meta, and para was found to be 9:4:87.

Other experiments with different Si/Al Based FLPs were conducted in a similar manner as detailed above, and the contrast experiment employing only AlCl₃as catalyst was also conducted.

Several combinations of Si/Al Based FLPs were selected to carry out the carboxylation of toluene by CO₂, to generate corresponding p-toluic acid (p-TA). The results were listed in Table 1. Since the aluminimium chloride tends to dimerize in most organic solvents like toluene, benzene, xylene, etc., here we calculate the isolated yield based on dimer Al₂Cl₆.

a: Molar ratio R₃SiCl/AlCl₃ = 1:1. b: Isolated yield is calculated based on Al₂Cl₆ : p-TA = 1:1. Products obtained by extraction with diethyl ether and NaOH, followed by precipitation in ice bath after adjusted to pH = 1.

According to Table 1, all the Si/Al Pairs were capable to catalyze the carboxylation of toluene by CO₂, giving satisfying yields of 62% - 95%. When Ph₃SiCl was added accompanied with AlCl₃, the p-TA yield reached 95% (entry 4), which exceeded a lot compared to that of only AlCl₃ used (78%, entry 5). However, the yield dropped to 62% when employed Me₃SiCl/AlCl₃ pairs instead (entry 1), suggesting that the catalytic effect is influenced by the steric hindrance of organosilane molecules. This supposition was further demonstrated by the phenomenon of introducing Me₂PhSiCl and MePh₂SiCl that the p-TA yield is ranked as Ph₃SiCl > MePh₂SiCl > Me₂PhSiCl > Me₃SiCl, just corresponds to their size of substituent groups, i.e., the steric hindrance.

As advocate above, the key procedure in carboxylation of aromatics is the activation of CO₂, which means the catalytic performance is depended on the activation effect. According to the formation mechanism of FLPs, it is combined by the weak electronic attractions owing to the bulky structure, thus the organosilane which possessed larger substituent group is favorable to form the FLPs with AlCl₃, leading to a high efficiency in activation of CO₂. The activation process is shown in Figure 3.

Other alkylbenzenes were carboxylated with the acid of different Si/Al Based FLPs, and we proposed a feasible mechanism of carboxylation under Si/Al Based FLPs catalytic systems, involving the activation of CO₂ mainly contributed by AlCl ₃ in cooperation with organosilane, as well as the subsequent electrophilic attack to aromatics, which is illustrated as Figure 4.

The reaction is believed to be an electrophilic aromatic substitution (S_EAR) by an attack to the aromatics nucleus of the activated “CO₂-AlCl₃-R₄Si”. It is known that a trialkysilyl group on the aromatics nucleus promotes the S_EAR reaction, as the silyl group stabilizes the transition state, leading to the (p-σ)_π conjugation between the Si-C bond and the formation of a benzenonium intermediate. Suspension of AlBr₃ and Ph₃SiCl in cyclohexane to CO₂reveals a C=O stretching vibration, assigned to a species consisting of CO₂, AlBr₃, and Ph₃SiCl [6]. This suggests that organosilane could activate CO ₂ in cooperation with AlX₃, which also indicated that the mechanism proposed in Figure 4 is convinced.

a: Molar ratio R₃SiCl/AlCl₃ = 1:1. b: Isolated yield is calculated based on Al₂Cl₆ : p-TA = 1:1. Products obtained by extraction with diethyl ether and NaOH, followed by precipitation in ice bath after adjusted to pH = 1.