(180d) High-P,T Micro-Reactor Processing for the Aqueous Kolbe-Schmitt Synthesis of Hydroquinone and Phloroglucinol

Recently, the high-p,T (high pressure, high temperature) concept was introduced as one original, tailored micro-reactor operation, among some others which were specified as well [1]. The first validation was done for the Kolbe-Schmitt synthesis using resorcinol with water as solvent to give 2,4-dihydroxy benzoic acid [2,3]. Later, a detailed study on the Michael addition with alkylic amines and alpha, beta unsaturated compounds followed [4]. Further, some brief investigations on the radical side-chain bromination of toluene and m-nitro toluene were given [5].

By the high-p,T concept, liquid-phase reactions can be processed under so far mainly unknown conditions much above the boiling point of the process solvent, typically >200°C or more [1]. In this way, even formerly sluggish reactions can be speeded up so that the correspondingly increased heat and mass transfer issues demand for the use of microstructured reactors. The operational range of the Kolbe-Schmitt synthesis using resorcinol was so extended by about 120°C towards 220°C, as compared to the standard batch protocol under reflux conditions (100°C) [2,3]. Among other kinds of process intensification, a 500-times increase in reaction rate compared to the batch processing at comparable yield was found.

In this paper, the high-p,T Kolbe-Schmitt synthesis of phenolic substrates with position isomerism relative to and more hydroxy groups than resorcinol was investigated.

The aqueous Kolbe-Schmitt synthesis of the isomer 1,4-dihydroxy benzene (hydroquinone) gave only very low yields <1% of the 2,5-dihydroxy-benzoic acid. This is explained by a favoured oxidative route to benzoquinone. This is line with the missing aqueous-based batch process counterpart which is only known under non-aqueous high-pressure conditions. Thus, the high-p,T process function here is only to speed up routes, but it cannot enable routes which are not practicable by conventional chemistry.

Detailed investigations were made for the Kolbe-Schmitt synthesis of 1,3,5-trihydroxy benzene (phloroglucinol) to the 2,4,6-trihydroxy benzoic acid which turned out to be successful as continuous micro-reactor variant.

The synthesis needed to be done at lower phenolic content (27 g/l; 61.7 mmol) as compared to the original resorcinol protocol (89 g/l; 203 mmol), but using the same carbonate-to-phenolic excess ratio of about 4. This was done to avoid the solids' deposition during the process and the analytical problems encountered, due to the experiences stemming from the resorcinol synthesis. Indeed, reproducibility and repeatibility of the experiments were much better now, and more consisted results could be obtained.

Yields were determined for a wide temperature range at 80, 100, 110, 120, 140, 160, 180, 200, 210, 220, and 240°C. These experiments showed two main trends. The reaction time to reach maximum yield decreases considerably with temperature. It takes 750 s at 110°C, 140 s at 160°C, and 50 s at 200°C. Up to 200°C, the yields remain fairly constant for reaction times still longer than needed for maximum yield within the range investigated (typically up to 20 min). Decomposition and product generation seem to be in balance here. From there on, a decrease of the yield is observed, which is most pronounced at 240°C.

The second trend observed was that the degree of the maximum yield is strongly dependent on temperature. While at 80°C only 15% yield is achieved, the best trade-off between T and  is given at 120°C with 55% yield. At 220°C, the maximum yield falls down to again only 16%. Thus, the phloroglucinol synthesis can be undergone at much lower temperatures (120°C; 465 s; 55% yield) as compared to the resorcinol one (180°C; 65 s; 43% yield) which is owing to the higher substrate reactivity, i.e. the phloroglucinol core is more electron-rich. However, longer residence times by more than one order of magnitude are needed, which reduces the increase in space-time yield compared to laboratory-flask operation from 440 (resorcinol) to about 20 for phloroglucinol. The decrease in reaction time is as high as by a factor of 144 (at 180°C) as compared to batch synthesis (resorcinol: ~500). The highest phlorglucinol productivity was 12 g/h at (200°C; 1620 ml/h; 20 s) as compared to the 68 g/h manufacture of resorcinol at (200°C; 2000 ml/h; 16 s). This is mainly due to the lower concentration used and the steeper decrease in yield at the high temperatures.

Several side and follow-up products of the phloroglucinol synthesis such as the diverse i,j-dihydroxy benzoic acids and phenolic substrates were identified by GC analysis with standards. This allows for mechanistic speculations on the diverse reaction pathways which will be presented. It is planned to use calorimetric analysis of such selected species (e.g. DSC, TGA) for further evidence on the different kinds of decomposition.

[1] Hessel, V., Löb, P., Löwe, H.; "Development of reactors to enable chemistry rather than subduing chemistry around the reactor - potentials of microstructured reactors for organic synthesis", Curr. Org. Chem. 9, 8 (2005) 765-787.

[2] Hessel, V., Hofmann, C., Löb, P., Löhndorf, J., Löwe, H., Ziogas, A.; "Aqueous Kolbe-Schmitt synthesis using resorcinol in a microreactor laboratory rig under high-p,T conditions", Org. Proc. Res. Dev. 9, 4 (2005) 479-489.

[3] Hessel, V.; Hofmann, C.; Löb, P.; Löhndorf, J.; Löwe, H.; "Minimizing reaction times for the Kolbe-Schmitt synthesis with resorcinol using high p,T-processing in a micro reactor setup", Chem. Eng. Trans. 6 (2005) 49-55. [4] Löwe, H., Hessel, V., Löb, P., Hubbard, S.; "Addition of secondary amines to alpha, beta unsaturated carbonyl compounds and nitriles by using microstructured reactors", Org. Proc. Res. Dev. (2005) submitted.

[5] Löb, P., Löwe, H., Hessel, V.; "Fluorinations, chlorinations and brominations of organic compounds in micro structured reactors", Journal of Fluorine Chemistry 125, 11 (2004) 1677-1694.