Effects of alkyl size of AlR₃ on its reaction with thiophene-2-carbonyl chloride

2.1 Reaction of triethylaluminum (TEA) and TPCC

A TEA–TPCC reaction run was conducted (conditions: [Al] = 300 mmol·L⁻¹, [TPCC] = 600 mmol·L⁻¹, n-heptane as a solvent, V = 20 mL, temperature = 40°C). Six aliquots of the reaction system were taken at reaction time (t) ranging from 126 to 725 s respectively, which were immediately quenched with CD₃OD. The samples were then analyzed by ¹H NMR. The ¹H NMR spectrum of a typical sample taken at t = 260 s is shown in Figure S1, and the selected part of ¹H NMR spectra of the six samples are shown in Figure 1. Compared with the ¹H NMR spectra of blank TPCC solution in n-heptane and TEA in n-heptane that were decomposed in CD₃OD (Figures S2 and S3), the reaction aliquots presented a new quadruplet at 2.98 ppm, which is assigned to methylene protons of ethyl 2-thienyl ketone. This product is formed through the alkylation of TPCC by alkylaluminum as shown in Scheme 4 [36].

Scheme 4

Formation of 2-thienyl ketone in the reaction of TPCC with TEA.

As seen in Figures 1 and 2, and Table S1, the yield of ethyl 2-thienyl ketone reached 25 mol% (based on Al) after about 2 min but gradually decreased by further prolonging the reaction. It means that alkylation of thiophene-2-carbonyl of TPCC by TEA proceeded very fast in the initial stage of reaction, and quickly ceased after a rather short time.

Figure 1

¹H NMR spectra of the products of TEA–TPCC reaction conducted for different durations ((1) 126 s, (2) 260 s, (3) 305 s, (4) 504 s, (5) 600 s, (6) 725 s). The peak marked with Δ is from residual methyl proton of CD₃OD solvent.

Figure 2

Change in the conversion of the ketone product of TEA–TPCC and TOA–TPCC reactions with reaction time (the conversion rate is defined as ketone/Al molar ratio × 100%).

In the full ¹H NMR spectrum of the reaction product, there are several other signals that were not found in the spectra of both TPCC and TEA, including a quadruplet peak at 3.61 ppm, a triplet at 4.31 ppm, and a weak quadruplet at 2.50 ppm (Figure S1). The peak at 3.61 ppm was also present in the spectra of TPCC and TEA (Figures S2 and S3), meaning that it may represent a certain impurity of n-heptane. The intensity of these signals was kept nearly constant when the reaction time was extended from about 2 min to 12 min (Table S1), meaning that they were not directly related to the reaction between TPCC and TEA.

To find the reason for the unique ketone yield−time profile of the TPCC/TEA reaction, a reaction run under typical conditions was recorded using a video camera, and photo shoots of the reaction process were picked from the video record (Figure S8). It is seen that the part of the TEA solution contacting the first drop of TPCC turned red immediately, and the colored zone expanded to the whole solution after about 0.5 s of injection. Further injection of TPCC made the whole solution deep red, and the color became the deepest after 1.8 s when about 1/3 of TPCC ([TPCC] = 200 mmol·L⁻¹) was injected. The color then gradually faded during injecting in the rest 2/3 of TPCC (1.8–5.5 s), and the solution showed an orange color during the later reaction stage (5.5–17 s). This color was kept unchanged when the reaction was extended to over 10 min.

Judging by the nearly constant ketone yield–time profile in Figure 2, the reaction leading to ketone formation must be a fast one. It is likely that ketone formation mainly takes place in the first 3 s of TPCC injection ([TPCC] = 0–300 mmol·L⁻¹) when the color is the deepest. In other words, the reaction proceeds well when TPCC/Al molar ratio is lower than 1:1. According to early literature reports, the yield of ketone based on Al reached 60% or higher in the reaction of R′COCl and Me₂AlCl when R′COCl/Al molar ratio was slightly lower than 1 [36]. It means that the R′COCl/Al molar ratio higher than 1 results in a lower ketone yield. In this work, the solution color evidently faded when the TPCC/Al ratio approached 1:1 (injection continued for 0–3 s). A reaction mechanism was proposed to explain the observed phenomena (Scheme 4). According to this mechanism, the overloaded TPCC can depress the formation of [R′CO]⁺ [AlR₃Cl]⁻ ion pair that is intermediate to ketone production. The deep color seen in the first 3 s of TPCC injection could come from the [R′CO]⁺ cation, as similar ion pairs with benzoylium cation show an orange-red color at room temperature [44]. Fading of this deep color implies a reduction in the ion pair concentration. The ketone product itself can also coordinate with AlEt₂Cl or TEA in the system, leading to a further reduction in ion pair concentration and retarded ketone production.

2.2 Effects of other conditions on the reaction of TEA and TPCC

Effects of other reaction conditions (temperature, speed of injecting TPCC into solution of TEA, concentration of either TEA or TPCC, and order of adding the reactants) on ketone production from TPCC/TEA were investigated. The effect of changing the speed of injecting TPCC on ketone yield was studied first. As shown in Figure 3, the ketone yield increased with a reduction in the duration of TPCC injection when the other conditions were fixed. The yield of ketone was evidently lowered when the reaction temperature was raised from 40°C to 60°C, but the same trend of increasing ketone production by speeding up TPCC injection was present.

Figure 3

Change in the conversion of the ketone product with duration of TPCC injection (TPCC was injected into TEA solution at constant temperature, [Al] = 300 mmol·L⁻¹, [TPCC] = 600 mmol·L⁻¹ after TPCC injection was finished).

When the concentrations of both TEA and TPCC were reduced ([Al] = 100 mmol·L⁻¹, [TPCC] = 200 mmol·L⁻¹ after TPCC injection was finished), the ketone yield was markedly increased, especially when the injection speed was low (Table S2). However, only reducing the TPCC concentration from 600 to 300 mmol·L⁻¹ meanwhile keeping [Al] at 300 mmol·L⁻¹ (TPCC/Al molar ratio was reduced to 1:1) caused a slight decrease in ketone yield. When the order of adding reactants was reversed, namely, TEA was injected into a solution of TPCC, the ketone yield was also enhanced (Table S2).

According to these phenomena, some factors were found to promote ketone formation: (1) quick mixing of reactants in the initial stage of the reaction; (2) low reaction temperature; and (3) low concentration of both reactants. To rationalize these reaction behaviors, it should be assumed that the formation of [R′CO]⁺ [AlR₃Cl]⁻ and the subsequent reaction to ketone are a kinetically fast process, while the formation of R′COCl/AlR₃ donor–acceptor complex is slower (Scheme 4). In this mechanistic model, the ketone is produced only by the ion pair (or acylium salt), meanwhile, the donor–acceptor complex (or oxonium complex) is too stable to form new products. A similar difference in reactivity has been found between acylium salt and the corresponding donor–acceptor complex in R′COCl/Lewis acid combinations, where the former shows the high activity of Friedel–Crafts acylation, but the latter shows no activity [45,46]. In Friedel–Crafts acylation, a slight molar excess of the Lewis acid AlCl₃ is needed to compensate for the complexation of the catalyst with the produced ketone [45].

In the TPCC/AlR₃ system studied in this work, during injecting TPCC into the TEA solution, [(C₄H₃S)CO]⁺ [AlEt₃Cl]⁻ ion pairs and the ketone are quickly formed when TPCC/TEA molar ratio is lower than 1:1, but the ion pairs are gradually converted to donor–acceptor complexes (including TPCC·TEA, TPCC·AlEt₂Cl, ketone·TEA, ketone·AlEt₂Cl) with further addition of TPCC and accumulation of the ketone. This leads to a cease of ketone formation in the later stage of the reaction. Since the formation of the complexes is slower than that of the ion pairs, ketone yield can be enhanced by faster addition of TPCC. Similarly, a lower temperature is beneficial to ion pair formation that has lower activation energy than complexing. When TEA solution was injected into TPCC solution, in the initial seconds of injection, the part of the reaction zone with a high TEA/TPCC ratio became larger than in the case of injecting TPCC into TEA, thus producing more ketone. The increase in ketone yield with a decrease in reactant concentrations could be attributed to the relatively hindered formation of the complexes at low concentrations as compared with the formation of the ion pairs.

According to this mechanistic model, production of the ketone proceeds quickly in the first few seconds of TPCC injection but stops soon when a certain high TPCC/TEA molar ratio (about 1:1) is reached. The slight decrease of ketone yield with t (reduced by about 10% after 10 min, see Figure 2) could be attributed to further reaction of the ketone with “free” TEA in the reaction system (Scheme 3). For the low concentration of such “free” TEA that is not stabilized in the complexes, the conversion rate of the ketone by alkylaluminum was rather low. For low concentrations of the alcohol products formed in such a reaction, they were not identified by ¹H NMR analysis on the reaction products.

2.3 Effects of alkyl size of alkylaluminum on its reaction with TPCC

To investigate the effects of the alkyl size of alkylaluminum on its reaction with an acyl chloride, a reaction of Al(n-Oct)₃ (TOA) with TPCC was conducted under the same conditions as that of TEA and TPCC ([Al] = 300 mmol·L⁻¹, [TPCC] = 600 mmol·L⁻¹, n-heptane as a solvent, V = 20 mL, temperature = 40°C). Five aliquots of the reaction system were taken at t ranging from 90 to 652 s respectively, which were then quenched with CD₃OD and analyzed by ¹H NMR. A typical ¹H NMR spectrum is shown in Figure S4, and partial ¹H NMR spectra of the five samples and the related reagents are shown in Figure 4.

Figure 4

¹H NMR spectra of the products of TOA–TPCC reaction conducted for different durations, t = 90 (1), 180 (2), 310 (3), 450 (4), and 652 s (5). The peak marked with Δ is from residual methyl proton of CD₃OD solvent.

Figure 4 shows two groups of signals that can be assigned to methylene protons of ethyl 2-thienyl ketone (quadruplet at 2.98 ppm) and n-octyl 2-thienyl ketone (triplet at 2.93 ppm), respectively. Assignment of the latter is supported by comparing with the ¹H NMR spectrum of a model compound n-propyl 2-thienyl ketone synthesized in our laboratory (Figure S5). The signal of ethyl 2-thienyl ketone is evidently stronger than that of n-octyl 2-thienyl ketone. The presence of ethyl 2-thienyl ketone in the product of the TPCC–TOA reaction implies that the TOA used in this experiment contains a small amount of Al−Et bonds. For higher reactivity of the Al−Et bonds than the Al−Oct bonds, a larger amount of ethyl 2-thienyl ketone can still be formed from a small amount of Al−Et bonds. The amount of the n-octyl 2-thienyl ketone was determined by deconvoluting the two signals at 2.9–3.2 ppm, and the change in the yield of this ketone with reaction time is shown in Figure 2. Compared with the yield of ethyl 2-thienyl ketone in the TPCC–TEA reaction, the yield of n-octyl 2-thienyl ketone was markedly lower, meaning that the reactivity of Al−Oct bonds was much lower than that of Al−Et bonds. This can be attributed to steric hindrance of the alkyl of AlR₃ toward its approaching acyl chloride.

As shown in Figure S4 and Table S1, there are several ¹H NMR signals in the spectra of the TPCC–TOA reaction product that cannot be clearly identified. By comparing with the time-dependent intensities of the ketone products, they were attributed to unknown impurities introduced by the TOA reagent.

To imitate the reaction between acyl chloride and Al−Pol bonds existing in catalytic olefin polymerization, a mixture of TEA and TOA with a 2:1 molar ratio was prepared, and its reaction with TPCC was studied under the same conditions as the TPCC–TEA system. Since alkyl exchanges take place very fast between alkylaluminum molecules in solution [47], most of alkylaluminum molecules in the TEA/TOA 2:1 mixture will exist as AlEt₂Oct. Therefore, it can be taken as a model molecule of AlEt₂Pol produced by chain transfer with alkylaluminum in catalytic olefin polymerization (Scheme 1). ¹H NMR spectra of the reaction product show two signals representing ethyl 2-thienyl ketone and n-octyl 2-thienyl ketone, respectively, meaning that both Al−Et and Al−Oct bonds of AlEt₂Oct took part in the reaction with TPCC. As seen in Figure 5 and Table S1, the yield of ethyl 2-thienyl ketone (product of Al−Et) was more than four times higher than that of n-octyl 2-thienyl ketone produced from Al−Oct, also showing a strong effect of alkyl size on the reactivity of aluminium-alkyl bonds. It is worth noting that the yield of product from Al−Et in this reaction was only about 1/5 of the ketone yield in the TEA–TPCC reaction, though Al−Et concentration in this reaction was 2/3 of the TEA–TPCC system. It means that the reactivity of Al−Et bonds in AlEt₂Oct was markedly lowered by the bulky n-octyl, as it hinders the rapid approaching of the alkylaluminum molecule to acyl chloride. This strong effect of alkyl size on the reaction forming ketone also implies that the reaction is highly controlled by kinetic factors, not by thermodynamic ones. Factors in favor of fast reaction (e.g., relatively lower temperature and lower reactant concentration, quick mixing of reactants) are found to raise the ketone yield, and factors that weaken its reactivity will reduce the dominance of ion pair and ketone formation over complexation in the initial stage. The increasing size of R in AlR₃ retards the approaching of AlR₃ to acyl chloride, thus resulting in a larger extent of reactivity loss in ion pair/ketone formation than in complex formation.

Figure 5

Change of conversion of the ketone products with reaction time in TPCC–TEA/TOA reaction (ethyl: ethyl 2-thienyl ketone, octyl: n-octyl 2-thienyl ketone).

Judging by the strong deactivation effect of bulky alkyl on the formation of ketone from alkylaluminum and acyl chloride, it is expected that Al–Pol bond (e.g., Al–Pol with polyethylene chain of more than 100 carbons) in AlR₂Pol (R = Me, Et, or i-Bu) will show even lower reactivity in its reaction with an acyl chloride, as most of the propagation chains have more than 50 monomer units and are much bulkier than n-octyl. It is estimated that only less than 0.2 mol% of the Al–Pol bonds will take part in the reaction with acyl chloride under typical conditions of olefin polymerization and quenching. When the polyolefin chain forms crystalline phases in the polymerization system, the aluminum atom of AlR₂Pol will be tightly surrounded by the crystalline lamella of the polymer, making its reaction with acyl chloride even more difficult. In previous work, we have found that the number of Al–Pol bonds formed by chain transfer with TEA was about 5 times the number of active centers after 30 s 1-hexene polymerization with MgCl₂-supported Z-N catalysts [18]. If 0.2% of these Al–Pol bonds were transformed to acyl-capped polymer chains in the subsequent quench-labeling process, it will cause about only a 1% overestimation of the active center number. In more conventional experimental conditions, polymerization lasts for a longer time (e.g., 2–30 min), which may evidently increase Al–Pol concentration in the reaction system. However, the active center concentration in such systems is usually 10–100 times higher than the above-mentioned 1-hexene polymerization [24,25,28,29,30] in the cases of Z–N catalyzed ethylene or propylene polymerization. Based on the results of the previous works, Al–Pol/Ti–Pol ratio of such systems is unlikely to exceed 20 when the polymerization time is shorter than 30 min. Therefore, the overestimation of the active center number caused by the reaction of Al–Pol moieties with RCOCl is estimated to be lower than 5% in counting active centers by acyl chloride quenching. The extent of such overestimation is not larger than the experimental error of the quench-labeling experiment. This will ensure reliable and relatively precise counting of the active center number in olefin polymerization with Z-N catalysts by using acyl chloride as a quenching agent.

Materials

TEA was purchased from Albemarle Co., TOA (25 wt% solutions in n-hexane) was purchased from Sigma-Aldrich, and TPCC (98%) was purchased from J&K Scientific, China. n-Heptane (Shanghai Titan Scientific Co., Ltd., China) was purified by passing through columns of molecular sieve and deoxygenate agent, refluxed over sodium-benzophenone, and distilled under nitrogen prior to use. TEA solution (2 M in n-heptane) was prepared by mixing a calculated amount of neat TEA and n-heptane under a nitrogen atmosphere. TPCC was distilled and diluted with n-heptane to 2 M before use. CH₃OD was purchased from Admas Co. and dried with anhydrous MgSO₄ before use. Other unstated reagents were used as received.

Reaction of alkylaluminum and TPCC

Typical reaction runs of AlR₃ (TEA or TOA) and TPCC were conducted in a Schlenk flask at 40°C under a nitrogen atmosphere for different durations (t). The calculated amount of reactants/solvent was injected into the flask in the order of TEA (or TOA), n-heptane, and TPCC to form a reaction solution with a total volume of 20 mL. At the end of the reactants’ addition, their concentrations in the solution were: [Al] = 300 mmol·L⁻¹ and [TPCC] = 600 mmol·L⁻¹. Injection of TPCC into the reaction solution took about 10 s if not noted. The counting of t started when the final drop of TPCC was injected into the flask. At a series of designed t values, 0.5 mL reaction solution was withdrawn from the flask and immediately injected into 2.5 mL D₃OD under a nitrogen atmosphere. After standing at room temperature for 12 h, 0.5 mL of the CD₃OD solution was transferred into an NMR tube and then accurately weighed CHCl₃ was added as an internal standard for quantitative analysis. Other TEA–TPCC reaction runs in n-heptane were also conducted at 40°C under nitrogen, but the reaction duration was fixed at 120 s, and the speed of injecting TPCC, the reaction temperature, the sequence of adding TEA and TPCC, or the concentration of TEA was respectively changed. The reaction products were sampled and analyzed by ¹H NMR with the same method.

In another set of reaction runs, a mixture of TEA and TOA (TEA:TOA = 2:1, mixed in n-heptane before injecting into the reactor) instead of a single AlR₃ was used as the alkylaluminum reactant and the same method and conditions as the typical reaction runs were adopted to study its reaction with TPCC.

Characterization

¹H NMR spectra were recorded on a Bruker AVANCE III 400 spectrometer operating at 400 MHz in CD₃OD, and 32 scans were accumulated for each sample at 25°C. Accurately weighed CHCl₃ was added as an internal standard, and the peak of CHCl₃ (7.89 ppm) was taken as an internal reference.