Enantioselective hydrogenation of prochiral substrates in catalytic membrane reactors

Abstract

Batch reactors containing catalytic membranes based on molecular asymmetric hydrogenation catalysts immobilized onto hybrid PVA/inorganic polymers were engineered and tested in the enantioselective hydrogenation of prochiral substrates. All reactions were efficiently carried out in mild reaction conditions with ee’s comparable to those of the corresponding homogeneous catalysts as well as minimal metal leaching upon reuse.

---

The development of sustainable, highly selective processes for the large scale production of fine chemicals is a current major industrial issue.¹ Among the solutions proposed to achieve such a goal, the immobilization of chemical catalysts onto solid insoluble support materials provides significant benefits in terms of ease of recovery and reuse of the expensive catalysts, clean catalyst separation from the reaction products as well as integration in existing reactor equipments.² When preformed enantioselective catalysts are immobilized, this strategy enables the preparation of heterogeneous catalysts featuring predictable asymmetric induction, often comparable with that of the homogeneous counterparts.³ In fact, the chemical industry has a strong preference for heterogeneous catalysts, but limited availability of heterogenized catalysts, together with activity loss and leaching problems, have confined their application to a few industrial processes so far.⁴ There is therefore a clear need to foster the research in the interdisciplinary domain integrating homogeneous and heterogeneous catalysis and to develop suitable catalytic equipments.

“Catalytic Membrane Reactors” (CMR) embedding a catalyst in a thin layer are receiving increasing attention as efficient and environmentally friendly production devices.⁵ In particular, compared to metallic or ceramic materials, polymeric-based membranes provide several advantages: (i) the sorption and permeability of reagents and products can be driven so as to improve the catalyst performance, (ii) the catalyst efficiency can be increased by co-incorporation of additives, (iii) the manufacture technology of polymeric membranes is much developed, thus allowing for a wide choice of sizes, morphologies, mechanical, chemical and thermal stability, affinity for reagents and catalysts to be available.⁶

Nonetheless, very few examples of polymeric catalytic membranes used in enantioselective processes have been reported. The (R,R)-MeDuPHOS–Rh, (S)-BINAP–Ru and (S,S)-SALEN–Mn complexes were supported onto PDMS films and used in the asymmetric hydrogenation and epoxidation reactions of olefins.⁷‡ These membranes showed good hydrogenation activity upon incorporation of silica or sulfonic acid derivatives, but insufficient stability in terms of metal leaching. Similar results were obtained by occlusion of ((R,R)-MeDuPHOS)–Rh catalysts into PVA membranes.⁸

During our studies on solid electrolytes for electrochemical applications,⁹ we recently discovered the possibility to immobilize molecular complexes onto inorganic-modified polymeric membranes that do not act as separation medium between two phases.^10,11

Herein we report on the preparation and use of enantioselective catalytic hydrogenation membranes based on asymmetric hydrogenation molecular catalysts immobilized onto hybrid PVA/inorganic polymers and their assembly into CMRs. Preformed hybrid membranes with different PVA, SiO₂, WO₃, Na₃PO₄, polystyrenesulfonic acid and PET content,^9,10 and rhodium(I) complexes of the general formula [(PP*)Rh(NBD)]PF₆ were scrutinized for this purpose (PP* = chiral diphosphino ligand, Scheme 1) (see ESI†, Table S1).

Scheme 1 Sketch of the chiral ligands used.

The immobilization of the rhodium compounds was easily and effectively accomplished by stirring a methanol solution of the Rh complex in the presence of the preformed hybrid membrane, followed by washings. Typical Rh loadings in the range 1.6–2.8% (w/w) were obtained in the dry material (ICP-AES and EDS), corresponding to 63–75% of the starting rhodium used. An electron microscope image of a typical catalytic membrane section is reported in Fig. 1. Although the presence of either WO₃ or SO₃⁻groups in the membrane is required to achieve an effective anchorage, no direct relation of catalyst loading with membrane composition was apparent. This suggests that the sorption of the rhodium specie is based on a complex combination of interactions whose exact nature is difficult to assess. One can hypothesize that both WO₃ (through donor–acceptor interactions as in the case of the Augustine’s method using heteropolyacids)¹² and the sulfonate group (through electrostatic bonds as in the case of ion-exchange resins)¹³ are responsible for the immobilization of the cationic complex.

Fig. 1 ESEM image of a section of the catalytic membrane NK1/[((−)-BINAP)Rh(NBD)]PF₆ (secondary electrons, 2650 magnifications).

The catalytic membranes obtained were used in the enantioselective hydrogenation of MAA (Scheme 2) in a one-pot, two-step procedure, either in a “fixed-frame” or in a rotating membrane device (Fig. 2). In the latter case, a methanol solution of the substrate was introduced into a Teflon®-covered autoclave with the catalytic membrane prepared in situ. This was stirred for the desired time under 5 bar H₂ at room temperature. For comparative purposes, different molecular catalysts were immobilized onto the same membrane type and the hydrogenation reactions were carried out under identical conditions (Table 1 and Table S2, ESI†). The results showed that:

Scheme 2 Hydrogenation reaction of MAA.

Fig. 2 Rotating catalytic membrane assembly.

Table 1 Yields and ee’s (brackets)% for the asymmetric hydrogenation of MAA using catalytic membranes bearing different chiral ligands^a

Membrane type	BINAP	DIOP	Monophos	TMBTP
a Experimental conditions: CH3OH, rt, 5 bar H2, 120 min, MAA ∶ catalyst = 200 ∶ 1, rotating membrane 150 rpm. b 1 bar, 10 min. c 60 min.
Homogeneous	87.0 (21.7)^b	99.0 (66.0)^b	95.2 (98.0)^c	87 (99.0)^b
NK-1	22.3 (15.0)	34.8 (17.3)	20.5 (90.5)	26.4 (98.5)
CSNKW-1	93.0 (11.0)	51.3 (17.6)	24.4 (90.5)	91.8 (98.3)

---

(1) all catalytic membranes are effective under very undemanding conditions,

(2) the activity and enantioselectivity of the heterogenized catalysts may be compared with those of the corresponding homogeneous counterparts,

(3) the activities are strongly dependent on the membrane type, whereas the ee’s are not significantly affected,¹⁴

(4) rhodium leaching in solution is negligible (<1 ppm, GF-AAS) for most catalytic membranes but not acceptable (>1 ppm) for others (e.g.DIOP and TMBTP on CSNKW-1),

(5) none of the hybrid membranes is catalytically active in the absence of anchored rhodium complex,

(6) all reactions are truly heterogeneous, as proved by the absence of catalytic activity of the solutions recovered after catalysis,

(7) except for a minor membrane folding, all catalytic membranes remain intact after catalysis.

A deeper investigation on the reuse of the catalytic membranes was carried out using the immobilized [((−)-Monophos)₂Rh(NBD)]PF₆ complex as probe catalyst. A sequence of three cycles of 2 hours, followed by one 17 h cycle and one 2 h cycle was adopted as a standard experiment. Selected results are reported in graphical format in Fig. 3. All catalytic membranes showed a slight activity decrease upon recycle, although they were still efficient after 25 h use. Complete conversions were achieved after the appropriate reaction time. The ee’s were fairly constant upon reuse, being highest (>90%) for the NK-1 and CSNKW-1 membranes. In most cases, the rhodium leaching was acceptable and generally proportional to the residence time of the catalytic membrane into the methanol solution (i.e. higher for the overnight cycle). We selected the decay of catalyst activity (expressed as variation of the TOF after 25 h, ESI†) as a reliable parameter to evaluate the overall catalytic membrane performance upon recycle. This parameter did not show any clear relationship with the membrane composition, yet there was a correlation with the amount of rhodium leached in solution. The identification of the factors affecting the Rh leaching is not obvious, however. One evidence is that high rhodium losses are accompanied by a detectable tungsten leaching in solution. Accordingly, it is possible that the leaching of WO₃—and thus of the rhodium interacting with it—may be responsible for the deactivation of the catalytic membranes through the release of non-catalytic rhodium species in solution. This fact, coupled with the lower activity decay of the PET-containing membranes, suggests that the membrane mechanical resistance may play a role on their reusability.¹⁵

Fig. 3 Recycle of catalytic membranes.

In light of these results, we designed a “fixed-frame” CMR in which the flow of the stirred methanol solution is tangent to the membrane and we used this device to scale-up the process by a factor of about ten (ESI†). Under these conditions, the (Monophos)Rh/NK-1 membrane gave a 50% higher TOF in each cycle, as compared to the rotating membrane reactor, with analogous enantioselectivity and Rh leaching.

The possibility to hydrogenate different substrates in diverse solvent systems also testify for the versatility of the technology. Table 2 shows representative results for the reduction of prochiral C=C and C=O bonds (DDF) in the compounds sketched in Scheme 3. All reactions proceeded with conversions and ee’s comparable with those obtained in the homogeneous phase and with metal leaching below 1 ppm. It is worth mentioning that hydrogenations could be carried out in water (in which solubility of the catalyst hampers the homogenous phase reaction) although with somewhat lower yield and ee’s compared to methanol, but with lower Rh leaching (<0.05 ppm). Lower conversions were observed in other solvents (e.g.ethanol) due to the lower membrane permeation.⁹

Table 2 Asymmetric hydrogenations using catalytic membranes based on immobilized Rh complexes (values in %)^a

Substrate	Solvent	Membrane	Ligand	Yield, ee	Yield, ee
				CMR	Homogeneous
a Experimental conditions: rt, 5 bar H2, 2 h, substrate ∶ catalyst = 200 ∶ 1. b 17 h. c 24 h. d 24 h, 40 bar H2.
MAA	CH3OH	NK-1	Monophos	93.9, 94.0^b	99.9, 98.0^b
MAA	H2O	NK-1	Monophos	55.2, 70.0^b	—
DMI	CH3OH	CSNKW-1	TMBTP	24.6, 12.4^c	19.6, 14.0
EMC	CH3OH	NK-1	DIOP	12.8, 17.0^d	13.0, 17.0^d
ANL	H2O	NK-1	Monophos	15.0, 24.1	—
DDF	CH3OH : H2O 1 ∶ 1	NK-1	Monophos	33.9, 3.0^d	—

---

Scheme 3 Sketch of the substrates tested.

In conclusion, we communicated here a novel class of enantioselective catalytic membranes prepared by an easy and low-cost procedure devoid of any chemical reaction. Preliminary data show that CMRs based on these membranes result in good reusability, versatility, high ee’s and lower rhodium leaching compared to known polymeric membrane systems. To the best of our knowledge, this is the first Chiral Catalytic Membrane Reactor engineered showing effective enantioselection and remarkable durability, even in polar solvents (i.e. in methanol). Studies aimed at elucidating the interactions between the complexes and the membranes are ongoing in our laboratories and will be reported in due course.

Thanks are due to Prof. S. Recchia and Dr V. Dal Santo (ISTM-CNR) for GF-AAS and ICP-AES analysis.

Notes and references

1. Asymmetric Catalysis on Industrial Scale, ed. H. U. Blaser and E. Schmidt, Wiley-VCH, Weinheim, 2004; N. B. Johnson, I. C. Lennon, P. H. Moran and J. A. Ramsden, Acc. Chem. Res., 2007, 40, 1291; L. A. Saudan, Acc. Chem. Res., 2007, 40, 1309.
2. Catalyst Separation, Recovery and Recycling; Chemistry and Process Design, ed. D. J. Cole-Hamilton and R. P. Tooze, Springer, Dordrecht, 2006; Recoverable and Recyclable Catalysts, ed. M. Benaglia, Wiley-VCH, Weinheim, 2009; Recoverable Catalysts and Reagents, ed. J. A. Gladysz, Chem. Rev., special issue, 2002, vol. 102, issue 10; Chiral Catalyst Immobilisation and Recycling, ed. D. E. De Vos, I. F. J. Vankelecom and P. A. Jacobs, Wiley-VCH, Weinheim, 2000; Fine Chemicals Through Heterogeneous Catalysis, ed. R. A. Sheldon and H. van Bekkum, Wiley-VCH, Weinheim, 2001; N. End and K. U. Schöning, in Topics in Current Chemistry, Springer, Heidelberg, 2004, vol. 242; M. D. Jones, R. Raja, J. M. Thomas and B. F. G. Johnson, Top. Catal., 2003, 25, 71; C. Li, H. Zhang, D. Jiang and Q. Yang, Chem. Commun., 2007, 547; C. E. Song, Annu. Rep. Prog. Chem., Sect. C, 2005, 101, 143; J. M. Thomas and R. Raja, Acc. Chem. Res., 2008, 41, 708; R. T. Baker, S. Kobayashi and W. Leitner, Adv. Synth. Catal., 2006, 348, 1337.
3. Heterogenized Homogeneous Catalysts for Fine Chemicals Production, Catalysis by Metal Complexes Series, vol. 33, ed. P. Barbaro and F. Liguori, Springer, London, 2010; J. M. Fraile, J. I. Garcia and J. A. Mayoral, Chem. Rev., 2009, 109, 360; P. Barbaro, Chem.–Eur. J., 2006, 12, 5666 for a review on asymmetric homogeneous catalysts, see: Catalytic Asymmetric Synthesis, ed. I. Ojima, Wiley-VCH, Weinheim, 3rd edn, 2010.
4. N. Yoneda, T. Minami, K. Hamato, Y. Shiroto and Y. Hosono, J. Jpn. Petrol. Inst., 2003, 46, 229; P. Pugin and H. U. Blaser, in Heterogenized Homogeneous Catalysts for Fine Chemicals Production, Catalysis by Metal Complexes Series, vol. 33, ed. P. Barbaro and F. Liguori, Springer, London, 2010, ch. 7.
5. E. Drioli and E. Fontanova, in Heterogenized Homogeneous Catalysts for Fine Chemicals Production, Catalysis by Metal Complexes Series, vol. 33, ed. P. Barbaro and F. Liguori, Springer, London, 2010, ch. 6; S. Miachon and J. A. Dalmon, Top. Catal., 2004, 29, 59; R. Dittmeyer, K. Svajda and M. Reif, Top. Catal., 2004, 29, 3; J. G. Sanchez Marcano and T. T. Tsotsis, in Catalytic Membranes and Catalytic Membrane Reactors, Wiley-VCH, Weinheim, 2010; Polymer Membranes/Biomembranes, Advances in Polymer Science Series, vol. 224, ed. W. P. Meier and W. Knoll, Springer, London, 2010.
6. I. F. J. Vankelecom and P. A. Jacobs, Catal. Today, 2000, 56, 147; I. F. J. Vankelecom, Chem. Rev., 2002, 102, 3779; B. M. L. Dioos, I. F. J. Vankelecom and P. A. Jacobs, Adv. Synth. Catal., 2006, 348, 1413; S. S. Ozdemir, M. G. Buonomenna and E. Drioli, Appl. Catal., A: General, 2006, 307, 167.
7. A. Wolfson, S. Janssens, I. Vankelecom, S. Geresh, M. Gottlieb and M. Herskowitz, Chem. Commun., 2002, 388; I. F. J. Vankelecom, D. Tas, R. F. Parton, V. Van de Vyver and P. A. Jacobs, Angew. Chem., Int. Ed. Engl., 1996, 35, 1346; I. Vankelecom, A. Wolfson, S. Geresh, M. Landau, M. Gottlieb and M. Hershkovitz, Chem. Commun., 1999, 2407; K. B. M. Janssen, I. Laquiere, W. Dehaen, R. E. Parton, I. E. J. Vankelecom and P. A. Jacobs, Tetrahedron: Asymmetry, 1997, 8, 3481; D. Tas, C. Thoelen, I. F. J. Vankelecom and P. A. Jacobs, Chem. Commun., 1997, 2323.
8. A. Wolfson, S. Geresh, M. Gottlieb and M. Herskowitz, Tetrahedron: Asymmetry, 2002, 13, 465.
9. H. Sawa, US Pat. 7396616, 2008; H. Sawa, US Pat. 7101638, 2006; H. Sawa, JP Pat. 3889605, 2007; H. Sawa, JP Pat. 4041422, 2008.
10. P. Barbaro, C. Bianchini, F. Liguori, H. Sawa and F. Vizza, PCT/JP, 056288, 2010.
11. M. Mulder, Basic Principles of Membrane Technology, Kluwer, Dordrecht, 1991.
12. R. L. Augustine, S. K. Tanielyan, S. Anderson and H. Yang, Chem. Commun., 1999, 1257; R. L. Augustine and S. K. Tanielyan, US Pat. 6025295, 2000; R. L. Augustine, S. K. Tanielyan, N. Mahata, Y. Gao, A. Zsigmond and H. Yang, Appl. Catal., A, 2003, 256, 69.
13. P. Barbaro and F. Liguori, Chem. Rev., 2009, 109, 515.
14. A significant drop in ee’s was observed only for DIOP-based catalysts. Although we have no definitive explanation for this behaviour, a surface effect on the enantioselectivities can be invoked, see: J. M. Fraile, J. I. García, C. I. Herrerías, J. A. Mayoral and E. Pires, Chem. Soc. Rev., 2009, 38, 695; J. I. García, B. López-Sánchez, J. A. Mayoral, E. Pires and I. Villalba, J. Catal., 2008, 258, 378.
15. Polyethylene glycol-containing membranes showed worse reusability likely due to their greater swellability in methanol, see ref. 9.

---

Footnotes

† Electronic supplementary information (ESI) available: Details of experimental data, procedures and membranes composition. See DOI: 10.1039/c0cy00030b

‡ DuPHOS = 1,2-bis-(2R,5R)-dimethyl(phosphacyclopentyl)-benzene; BINAP = 2,2′-bis(diphenylphospino)-1,1′-binapthyle; SALEN = N,N′-bis(3,5-di-tert-butylsalicylidene)-1,2-cyclohexanediamine; PDMS = polydimethylsiloxane; PVA = polyvinyl alcohol; NBD = bicyclo[2.2.1]hepta-2,5-diene; PET = polyethylene terephthalate TMBTP = 4,4′-bis(diphenylphosphino)-2,2′,5,5′-tetramethyl-3,3′-bithiophene; DIOP = 2,3-O-isopropylidene-2,3-dihydroxy-1,4-bis(diphenylphosphino) butane; Monophos = (3,5-dioxa-4-phospha-cyclohepta[2,1-a:3,4-a′]di-naphthalen-4-yl)dimethylamine; MAA = methyl 2-acetamidoacrylate; DMI = dimethyl itaconate; ANL = α-angelica lactone; EMC = ethyltrans-β-methylcinnamate; DDF = dihydro-4,4′-dimethyl-2,3-furandione.

---