58
M. Haumann
Fig. 3.9 Selected ligands used in SILP-catalyzed hydrogenation reactions
The precursor [bis(2-methylallyl)(1,5-cyclooctadiene)ruthenium and the ligand
3-(2,5-(2R,5R)-dimethylphospholanyl-1)-4-di-o-tolylphosphino-2,5-dimethylthiophene (see Fig. 3.9, property of EVONIK, now SOLVIAS) were dissolved in
[C 2 C 1 im][NTf 2 ], and this solution was dispersed onto porous silica. In order to
keep the high-boiling compounds in the gas phase, a large helium flow was adjusted
through a gradient-free recycle reactor. No activity was observed when the Ru complex was dispersed onto the support without ionic liquid. The activity of the SILP
increased linearly with higher ionic liquid loading to reach its maximum at ionic
liquid loading of 100% of the support’s volume. Rinker’s group did a detailed study
for SILP systems in 1973 and explained such behavior by a slow chemical reaction
that is not limited by diffusion effects [26]. Therefore, the more liquid volume
containing catalyst complexes are present, the higher becomes the conversion.
Supercritical CO 2 as the mobile phase has been used for the asymmetric hydrogenation of dimethylitaconate by Hintermair et al. [27]. The chiral bidentate phosphine–phosphoramidite ligand quinaphos-modified (qp, see Fig. 3.9) Rh-qp-SILP
catalyst was immobilized in imidazolium-based ionic liquids, [C 2 C 1 im][NTf 2 ],
[C 4 C 1 im][BF 4 ], and [C 6 C 1 im][OTf], dispersed on silica gel. A reactor setup (see
Fig. 3.10) was established to allow continuous flow reaction with scCO 2 flow [28].
The SILP catalysts achieved full conversion shortly after the start of the reaction and
did not show a decrease in activity during 65 h time on stream. The enantioselectivity
was extremely high (>99%) within the first 10 h of the run, followed by a steady
decline (see Fig. 3.11).
The loss in enantioselectivity was ascribed to partial decomposition of the active
species, resulting in unselective rhodium hydrogenation catalysts. The total turnover
number was remarkable with 115,000 moles of substrate per mole of rhodium. The
STY was calculated to be reasonably high with 0.3 kg m
−3 h
−1 .
In a follow-up study, the same authors reported a more detailed investigation
of scCO 2 /Rh-SILP asymmetric hydrogenation [29]. The water content in the feed
was found to be a crucial factor for catalyst stability. Drying of incoming feed is
well established in industry (e.g., by using redundant guard beds) and would allow
implementation of the combined scCO 2 -SILP catalysis for industrial asymmetric
hydrogenations.
M. Haumann
Fig. 3.9 Selected ligands used in SILP-catalyzed hydrogenation reactions
The precursor [bis(2-methylallyl)(1,5-cyclooctadiene)ruthenium and the ligand
3-(2,5-(2R,5R)-dimethylphospholanyl-1)-4-di-o-tolylphosphino-2,5-dimethylthiophene (see Fig. 3.9, property of EVONIK, now SOLVIAS) were dissolved in
[C 2 C 1 im][NTf 2 ], and this solution was dispersed onto porous silica. In order to
keep the high-boiling compounds in the gas phase, a large helium flow was adjusted
through a gradient-free recycle reactor. No activity was observed when the Ru complex was dispersed onto the support without ionic liquid. The activity of the SILP
increased linearly with higher ionic liquid loading to reach its maximum at ionic
liquid loading of 100% of the support’s volume. Rinker’s group did a detailed study
for SILP systems in 1973 and explained such behavior by a slow chemical reaction
that is not limited by diffusion effects [26]. Therefore, the more liquid volume
containing catalyst complexes are present, the higher becomes the conversion.
Supercritical CO 2 as the mobile phase has been used for the asymmetric hydrogenation of dimethylitaconate by Hintermair et al. [27]. The chiral bidentate phosphine–phosphoramidite ligand quinaphos-modified (qp, see Fig. 3.9) Rh-qp-SILP
catalyst was immobilized in imidazolium-based ionic liquids, [C 2 C 1 im][NTf 2 ],
[C 4 C 1 im][BF 4 ], and [C 6 C 1 im][OTf], dispersed on silica gel. A reactor setup (see
Fig. 3.10) was established to allow continuous flow reaction with scCO 2 flow [28].
The SILP catalysts achieved full conversion shortly after the start of the reaction and
did not show a decrease in activity during 65 h time on stream. The enantioselectivity
was extremely high (>99%) within the first 10 h of the run, followed by a steady
decline (see Fig. 3.11).
The loss in enantioselectivity was ascribed to partial decomposition of the active
species, resulting in unselective rhodium hydrogenation catalysts. The total turnover
number was remarkable with 115,000 moles of substrate per mole of rhodium. The
STY was calculated to be reasonably high with 0.3 kg m
−3 h
−1 .
In a follow-up study, the same authors reported a more detailed investigation
of scCO 2 /Rh-SILP asymmetric hydrogenation [29]. The water content in the feed
was found to be a crucial factor for catalyst stability. Drying of incoming feed is
well established in industry (e.g., by using redundant guard beds) and would allow
implementation of the combined scCO 2 -SILP catalysis for industrial asymmetric
hydrogenations.