Mechanism of cyclic carbonate synthesis from epoxides and CO2.

Three interconnected catalytic cycles account for the title reaction catalyzed by a bimetallic aluminum(salen) complex and Bu(4)NBr. In the first, Bu(4)NBr acts as a nucleophile to activate the epoxide. In the second, Bu(3)N generated in situ serves to activate CO(2). In the third, the aluminum(salen) complex brings the two activated species together so that the key bonds can be formed intramolecularly.

o C/min to 270 o C; hold at final temperature for 5 minutes; T R 7.33 minutes (styrene oxide), T R 12.09 minutes (styrene carbonate). For the first 3.50 minutes, the eluent was routed away from the mass detector. Subsequently, the detector was operated in full EI scan mode.
All kinetics experiments were repeated at least twice and were mutually consistent. The GC response was calibrated with known mixtures of styrene oxide, styrene carbonate, propylene carbonate and tributylamine. The in situ FTIR was calibrated to monitor the disappearance of the styrene oxide peak at 873 cm -1 . A calibration table was constructed with four different concentrations of styrene oxide in propylene carbonate. Experimental procedure for styrene carbonate synthesis monitored by in situ FTIR Styrene oxide (0.83-5.48 mmol), complex 1 (0-0.09 mmol), tetrabutylammonium bromide (0-0.37 mmol) and tributylamine (0-0.04 mmol) were vigorously stirred in propylene carbonate (1.43 mL) until complete dissolution occurred. The mixture was placed in a 200 mL Schlenk flask fitted with the FTIR probe and equipped with a small magnetic stir-bar and placed in a thermostatted bath at 26 o C. An FTIR spectrum was recorded to calibrate the initial concentration of styrene oxide. The reaction flask was thoroughly flushed with CO 2 provided from a balloon inflated with several dry-ice pellets and was then mantained under a CO 2 atmosphere using the inflated balloon. A FTIR spectrum was recorded every 5.5 minutes. When the reaction had reached completion, the FTIR peak areas were converted into concentrations of styrene oxide using the calibration table and plotted against time to obtain the reaction profile.

Experimental procedure for styrene carbonate synthesis monitored by GC
Styrene oxide (0.83-2.74 mmol), complex 1 (0-0.04 mmol) and tetrabutylammonium bromide (0-0.02 mmol) were vigorously stirred in propylene carbonate (1.43 mL) until complete dissolution occurred. The mixture was placed in a 200 mL Schlenk flask equipped with a small magnetic stir-bar and placed in a thermostatted bath at 26 o C. A sample was withdrawn and analysed by GC to calibrate the initial concentration of styrene oxide. The reaction flask was thoroughly flushed with CO 2 provided from a balloon inflated with several dry-ice pellets and was then mantained under a CO 2 atmosphere using the inflated balloon. A sample was withdrawn for GC analysis every 30 minutes. When the reaction had reached completion, the GC peak areas were converted into concentrations of styrene oxide using the calibration table and plotted against time to obtain the reaction profile.
Experimental procedure to allow the order with respect to carbon dioxide to be determined Styrene oxide (3.33 mmol, 0.400 g), complex 1 (0.09 mmol, 0.096 g) and tetrabutylammonium bromide (0.09 mmol, 0.028 g) were vigorously stirred in propylene carbonate (1.43 mL) until complete dissolution occurred. The mixture was placed in a 200 mL Schlenk flask fitted with the FTIR probe and equipped with a small magnetic stir-bar and placed in a thermostatted bath at 26 o C. An FTIR spectrum was recorded to calibrate the initial concentration of styrene oxide. Cylinders of CO 2 and N 2 were fitted with Cole-Parmer (0-250 mL/min) mass-flow control units and connected to a T-joint. The outlet from the T-joint was used to provide a stream of a known % of CO 2 in N 2 which passed through the reaction flask at atmospheric pressure and a FTIR spectrum was recorded every 5.5 minutes. When the reaction had reached completion, the FTIR peak areas were converted into concentrations of styrene oxide using the calibration table and plotted against time to obtain the reaction profile.

Experimental procedure for the repeated use of catalyst 1 with propylene oxide
Catalyst reusability experiments were carried out at 0 o C using propylene oxide as substrate. Complex 1 (2.5 mol%, 0.414 g) and tetrabutylammonium bromide (2.5 mol%, 0.115 g) were placed in a 25 mL round bottomed flask equipped with a magnetic stir bar and a SubaSeal. The mixture was placed in an ice-bath and left for 10 minutes until the system conditioned to 0 o C. At the same time, CO 2 was passed through the flask from a balloon inflated with several dry-ice pellets. Propylene oxide (14 mmol, 0.830 g) was then injected into the reaction and left to vigorously stir for 3 hours. The flask was then opened to air and left to warm to room temperature. The product was isolated directly from the reaction flask by micro-distillation under reduced pressure (0.2 Torr) using a Büchi Kugelrohr system at 140 o C and was obtained as a transparent liquid (yield 30-50 %) which was analysed by 1 HNMR and GC/MS. The catalyst left in the reaction flask was reused in the next cycle following the procedure reported above starting with cooling to 0 o C. After the completion of 16 cycles, a sample of the solid residue in the 25 mL flask was dissolved in methanol and analysed by high resolution electrospray mass spectrometry (positive ion mode) to show that the structure of catalyst 1 was unchanged.

Control experiment on the use of tributylamine in the absence of tetrabutylammonium bromide
Styrene oxide (0.40 g, 3.32 mmol), complex 1 (2.5 mol%, 0.10 g) and tributylamine (5 mol%, 0.20 g) were vigorously stirred in propylene carbonate (1.43 mL) until complete dissolution occurred. The mixture was placed in a 200 mL Schlenk flask equipped with a small magnetic stir-bar and placed in a thermostatted bath at 26 o C. A sample was withdrawn and analysed by GCMS to calibrate the initial concentration of styrene oxide. The reaction flask was thoroughly flushed with CO 2 provided from a balloon inflated with several dry-ice pellets and was then mantained under a CO 2 atmosphere using the inflated balloon. A sample was withdrawn for GCMS analysis after 21 hours which showed just 1.4% conversion of styrene oxide to styrene carbonate.

Kinetic analysis showing the reaction is first order in styrene oxide
First order kinetics plot at four different initial concentrations of styrene oxide.   Black column is % yield of propylene carbonate read against the left y-axis. Grey column is % tributylamine detected read against the right y-axis. A second batch of tributylamine (2.5 mol%) was added after the seventh experiment.

Control experiments on the generation of tributylamine with detection by GCMS
Tributylamine can be generated from tetrabutylammonium bromide under high temperature conditions. Since the injector temperature for the GCMS is 250 o C, tetrabutylammonium bromide could form tributylamine during GCMS analysis rather than during a reaction. To investigate this, the following control experiments were carried out: Tetrabutylammonium bromide (0.005 g, 0.01 mmol) was dissolved in 10 mL of ethyl acetate (10 mL). The mixture was left to stir for 10 minutes, then the solution was analysed by GCMS and tributylamine was detected, showing that if tetrabutylammonium bromide is present in the material injected into the GCMS then tributylamine will be detected.
The peak at 8.7 minutes corresponds to tributylamine (authentic GC and m/z data on pages 11 and 12) Tetrabutylammonium bromide (0.11 g, 0.36 mmol) was dissolved in propylene carbonate (0.48 g, 3.28 mmol) and then stirred until complete dissolution occurred. The mixture was then transferred into a 5 mL flask and purified by microdistillation (yield 93%, 0.45 g). A sample of the distillate was analysed by GCMS, tributylamine was not detected. Thus tetrabutylammonium bromide does not decompose to tributylamine during the distillation process and there is no splash over of tetrabutylammonium bromide during the distillation. The same result was obtained if complex 1 was also added to the propylene carbonate solution of tetrabutylammonium bromide.