HETEROGENEOUS ORGANOCATALYSTS FOR THE GREEN SYNTHESIS OF CHIRAL GLYCINATE INTERMEDIATES

Project Title: HETEROGENEOUS ORGANOCATALYSTS FOR THE GREEN SYNTHESIS OF CHIRAL GLYCIDATE INTERMEDIATES

Project code PNII-ID-PCE-2011-3-0041, No. 321/2011

Contracting Authority: UEFISCDI

Total funding: 1.350.000,00 lei

Project Host Institution:

Host Institution: University of Bucharest

Address: Bdul Mihail Kogalniceanu, 36-46, Bucharest 050107, ROMANIA

City: Bucharest

Institutional Code in the Register of Potential Contractors (http://rpc.ancs.ro):1716

Research team:

Nr. crt.	Last name	First name	Date of birth	Scientific title	Doctor	Position
1	COMAN	Simona Margareta	26/07/1969	Professor	Yes	Project leader
2	SANDULESCU	Madalina	28/02/1974	Lecturer	Yes	Member
3	FLOREA	Mihaela	21/10/1974	Lecturer	Yes	Member
4	CANDU	Natalia	13/08/1983	Research assistant	Yes	Member
5	DOBRINESCU	Claudiu	21/02/1984	Research assistant	No	Member

Project Summary

The use of catalysis in asymmetric synthesis is an efficient way for the production of sophisticated molecules following the atom economy concept. In this context, organocatalysis starts to be more and more important, bringing advantages both in respect of its synthetic range but also for economic reasons. Thus, one might expect that in the near future an increasing number of organocatalytic reactions will make the jump from academic synthesis to industrial application. On the other hand, making organocatalysts insoluble and consequently easily recoverable and reusable is a stately way to answer the principles of “Green chemistry”. In this context, the main objective of the present project is the preparation of efficient heterogeneous organocatalysts (e.g., inorganic carriers grafted chiral ketones, chiral ketones based magnetically recoverable nanocatalysts, and chiral-MOFs) for the asymmetric synthesis of chiral glycidates through the epoxidation of cinnamates derivatives structures. The project is expected to lead to the development of new organocatalysts, to new strategies in organocatalysis, to new insight into reaction mechanism of organocatalyzed reactions, to new methods in synthetic organic chemistry, and arousal of the interest of chemical and pharmaceutical industry in organocatalysis.

Project objectives:

O1. The synthesis of homogeneous chiral ketones.
Relevance: The preparation of structurally modified chiral ketones as building blocks for the synthesis of enantioselective heterogeneous organocatalysts.

O2. The synthesis of heterogeneous grafted organocatalysts
Relevance: The design and preparation of novel enantioselective heterogeneous organocatalysts involving mesoporous organic-inorganic hybrid materials (MCM-41) and LDHs as inorganic carriers and grafted chiral ketone as active site.

O3. The synthesis of Chiral-Metal-Organic-Frameworks (CMOFs)
Relevance: The design and preparation of new chiral porous coordination networks (CMOFs) involving the chiral ketone as pillars. Although still at its infancy stage, chiral porous MOFs have already shown very high catalytic activity and enantioselectivity in several organic transformations. Many more CMOFs will emerge in the future, and they will have a bright future in the asymmetric catalytic synthesis of important optically pure organic molecules.

O4. The synthesis of Magnetically Recoverable Nanocatalysts (MNs)
Relevance: The unique combination of high enantioselectivity and enhanced reactivity combined with its recyclables and ease of separation makes this chiral nanotechnology one of the most promising strategies for the formation of enantiomerical enriched compounds on an industrial scale.

O5. The characterization of the homogeneous and heterogeneous organocatalysts
Relevance: Physico-chemical characterization of the heterogeneous organocatalysts for a rational design and optimization of the final structure.

O6. The catalytic investigation of the prepared homogeneous and heterogeneous organocatalysts (enantioselective epoxidation)
Relevance: The asymmetric heterogeneous synthesis of chiral glycinates through the epoxidation of the commercially available and inexpensive (E)-trans-cinnamates structures. The use of this isomer as raw material is highly desirable since the difficulties in the preparation of (Z)-cis-cinnamic esters are well known.

Synthetic results:

O1, O5 and O6. The synthesis of homogeneous chiral ketones; The characterization of the homogeneous organocatalysts; The catalytic investigation of the prepared homogeneous organocatalysts (enantioselective epoxidation)

The difficulties encountered in the preparation of cis-cinnamic acid esters [L. Deng et al., J. Org. Chem. 57 (1992) 4320], lead to the necessity to find effective methods for the synthesis of (2R, 3S)-phenyl-glycidates by catalytic asymmetric epoxidation of the trans-methyl-cinnamate - much cheaper and commercially available. Due to the disadvantages associated with the use of the organometallic chiral complexes in such transformations, their replacement remains a challenging in the area. In this context, previous reports [JACS 1996, 118, 9806] recommend ketone derived from D-fructose (also known as ketone Shi) as a suitable candidate for the enantioselective epoxidation of a large number of di- and tri-substituted (E) –alkenes, (Z)- and terminal olefins in the presence of Oxone (potassium peroxomonosulfate). Reactions occur with enantioselectivities from high to moderate [J.Org. Chem., 2005, 70, 2904]. However, subsequent reports show that chiral ketone Shi derived from fructose is not effective for the epoxidation of α, β-unsaturated esters due to its decomposition tendency under required basic conditions. Moreover, Oxone undergoes a gradual decay in basic conditions. Therefore, creating and optimizing an organocatalytic system, able to transform the trans-methyl-cinnamate to optically active phenyl-glycidates, represented a challenge of the project.
To achieve the project objectives an important amount of the research was dedicated to the optimization of the homogeneous organocatalytic system able to efficiently transform the trans-methyl-cinnamate to chiral phenyl glycidate through the dioxirane epoxidation (Scheme 1). For this, several ketone structures were synthesized (Scheme 2) and the reaction conditions were optimized for each, by using Oxone as epoxidation agent. The synthesis methodologies of A and B structures are described in the literature [Y. Tu et al., J. Am. Chem. Soc. 118 (1996) 9806; Catal. Sci. & Tech., 2015, 5, 729] while the procedure for the synthesis of C structure is detailed in the scientifically report of phase 1 of the project.

Scheme 1. Catalytic dioxirane-mediated epoxidation in the presence of Oxone

Scheme 2. Chiral ketones with different structures used for the dioxirane-mediated epoxidation of trans-methylcinnamate, in the presence of Oxone

The ¹H-NMR and IR spectroscopy confirmed the successful synthesis of the A, B and C structures. The dioxirane-mediated epoxidation was carried out in buffers, in bi-phase systems, and in the presence of a phase-transfer catalyst. The nature of the phase-transfer catalyst highly influences both the reaction rate and the selectivity to epoxides (C = 21-77%, S_epoxide = 41-56 %). Moreover, if in the presence of the quaternary ammonium salts the (2R, 3S)-phenyl-glycidate configuration is favored (e.e. = 7.6-29%), the presence of the crown ether lead to a stereo-inversion from (2R, 3S)- to (2S, 3R)-phenyl-glycidate configuration (e.e. = 38%), indicating some additional steric hindrance introduced by the phase-transfer catalysts which disfavor some potential transition states. Two mechanistic pathways able to predict the stereochemical path of the dioxirane-mediated epoxidation were already proposed [Tetrahedron Lett., 1987, 28, 3311] involving a spiro or a planar transition state (Scheme 3A). The epoxidation of trans-methylcinnamate takes place through the spiro transition state due to the steric effects, while the discrimination between (2R,3S)- and (2S,3R)-enantiomers it seems to be made by favoring the “spiro 1” state transition (quaternary ammonium salts) or “spiro 2” state transition (crown ether) (Scheme 3B), in agree with [Catal. Sci. & Tech., 2015, 5, 729].

Scheme 3. (A) Planar and spiro transition states for the olefine dioxirane-mediated epoxidation [Tetrahedron Lett., 1987, 28, 3311]. (B) Spiro transition states for dioxirane-mediated epoxidation of trans-methylcinnamate, in the presence of chiral ketone A (Scheme 2)

    Another important factor dramatically influencing the epoxidation reaction is the pH. The addition of K₂CO₃ leads to an increased rate of dioxirane generation due to the increased nucleophilicity of Oxone accompanied by its stability decreases. In addition, a high pH reaction disfavors the secondary Bayer-Villiger (BV) oxidation of the chiral ketone with the concomitant increases of its life time. In the light of these two problems, initial experiments were carried out at pH = 7-8. At this pH, chiral ketone decomposed rapidly, the BV oxidation being the main way of its decomposition. However, the BV reaction could be suppressed by increasing the pH and favoring, in this way, the balance movement towards the efficient formation of dioxirane B (Scheme 1). Therefore, if the chiral ketone catalyst is reactive enough to compete in the two processes, the problems associated with working at high pH may be removed. Among the three structures, the diacetat-ketone seems to be the most adequate structure for a very efficient and highly enantioselective epoxidation of α, β-unsaturated esters, according with literature [Catal. Sci. & Tech., 2015, 5, 729]. The higher reactivity is due to the presence of the two acetate groups with a more electron-withdrawing character.

    However, in spite of their renewable origin (all structures from Scheme 2 were synthesized from D-fructose) and the high reactivity in the synthesis of the chiral epoxides, the preparation involves several reagents which are hardly accepted by green chemistry community. Therefore, to find and explore organocatalysts of which synthesis is more environmental friendly was an important alternative taken into consideration during the project development. In this context, very recently sodium levulinate ([Na][LEV]) was classified as a readily biodegradable and low toxicity ionic liquid [RSC Adv. 6 (2016) 87325-87331]. His preparation is as simple as possible, by treating levulinic acid with sodium bicarbonate, and do not involve dangerous reagents or harsh reaction conditions. On the other hand, the raw material - levulinic acid (LA) - is classified as of the most important platform molecules synthesized through the hydrolysis of cellulose or starch. Obviously, the dioxirane-mediated epoxidation with this organocatalyst does not generate stereoselectivity in the reaction products but the low costs of the organocatalytic system and the lack of harmful elements may competes the costs associated with the enzymatic separation of the obtained racemic phenyl-glycidate into enantiomers. The use of a levulinic acid-based structure as organocatalyst is, at the same time, a novelty in the area. The epoxidation of trans-methylcinnamate take place with moderate conversions (C = 22.4%) but total selectivity to phenyl-glycidate (S_epoxide = 100%), in the presence of [Na][LEV]. The total selectivity to epoxide is another element which indicates the system as a highly green one. The moderate conversion is, most probably, due to the presence of the carboxylate group in γ-position to the –C=O group, with a greatly diminished electron-withdrawing effect compared to the strong electron-withdrawing effect of acetate substitutes in α- position to the carbonyl group from the diacetate-ketone (Structure C, Scheme 2) structure. As such, the -C=O (active catalytic center) from diacetate-ketone structure is much more reactive than the -C=O from [Na][LEV] structure. On the other hand, trans-methyl cinnamate is an unsubstituted electron deficient alkene. As such, it is expected a lowered reactivity toward the electrophilic dioxirane generated in-situ from the -C=O group of the organocatalyst. An increased catalytic efficiency from the conversion point of view should be expected by applying this benign organocatalyst in the diaxirane-mediated epoxidation of cinnamates with donating groups into their structure. For the proposed objective of the project, the most important issue was demonstrated: the viability of the concept which can open a novel area of research in organocatalysis.

    Once established the optimal reaction conditions for the homogeneous dioxirane-mediated epoxidation the next step of the project was to synthesize equivalent heterogeneous organocatalysts, with high efficiency in the process.

O2-O6. The synthesis of heterogeneous grafted organocatalysts. The characterization of the heterogeneous organocatalysts. The catalytic investigation of the prepared heterogeneous organocatalysts.

To be used for practical purposes, the heterogeneous organocatalysts must meet certain requirements: (i) the preparation must be simple, efficient and more generally applicable; (ii) their performances must be comparable or better than those of homogeneous organocatalysts; (iii) their separation from the reaction mixture must be possible by a simple filtration and more than 95% of the catalyst to be recovered; (iv) dissolving the active species to be minimal; (v) reusing the catalyst must be possible without loss of activity; (vi) should be mechanically, thermally and chemically stable. They must be compatible with the solvent and commercially available; (vii) for commercially purposes selectivity is sometimes more important than the activity and life of the catalyst.

Obviously, the structure of the heterogeneous organocatalysts is more complicated than of the homogeneous counterpart and, an inappropriate choice of the inorganic support may lead to solids with low catalytic efficiency also due to the possible mass transfer limitation of the reactants and products through the porous structures. This undesirable effect is even more probable for heterogeneous organocatalysts which already comprises voluminous organic molecules in the carrier pores. Therefore, the best choice of the carriers involves knowledge of the kinetic diameters of all organic molecules involved in the dioxirane-mediated epoxidation (ie, organocatalysts, reactants and products). Therefore, their kinetic diameters have been estimated from the molecular weight correlation, using the equation:

for aromatic hydrocarbons (M_w = molecular weight in g mol^-1) [Combust. Flame 96 (1994) 163-170]. This approximation method for the kinetic diameters of oxygenated molecules of which the critical properties do not have been reported is reasonable, as Huber et al. [J. Catal. 279 (2011) 257–268] demonstrated not long ago.

The kinetic diameters of the organocatalysts the reactant and products of the dioxirane-mediated epoxidation are listed in Table 1.

Table 1. Kinetic diameters of the organocatalysts, reactants and products, estimated from the molecular weight correlation [Combust. Flame 96 (1994) 163-170]

Organic molecule	Organocatalyst	Reactant	Product	Kinetic diameter (σ, Å)
A (Scheme 2)	+			7.8
B (Scheme 2)	+			9.4
C (Scheme 2)	+			8.3
Levulinic acid	+			6.0
[Na][LEV]	+			6.4
Trans-methylcinnamate		+		6.7
Phenyl-glycidate			+	6.9

    Obviously, such large molecules need carriers with larger mezopores in their texture, such as: (i) mesoporous silica (MCM-41); (ii) hydroxylated inorganic fluorides (e.g., AlF_3-x(OH)_x); (iii) LDH materials and (iv) MOF materials, or nanocatalysts, which display a high external surface area, such as: (v) core-shell magnetic nanoparticles/silica (or aluminium fluoride).

    MCM-41, nanoscopic hydroxylated inorganic fluorides (ie, AlF₃) and MgAl-LDH materials were prepared following the corresponding procedures described by Coman et al. [J. Mol. Catal.,146 (1999) 247; Pure and Appl. Chem., 84 (3) (2012) 427; ChemSusChem, 5 (2012) 1708; Top. Catal., 55 (2012) 680; Catal. Today, accepted, under corrections, CATTOD-D-16-00475, (2016)]. For the synthesis of the magnetic nanoparticles (MNP) supports a method in three steps, in which the first step was the common MNP synthesis by co-precipitation, followed by the MNP coating with a silica (MNP/SiO₂) or hydroxylated aluminum fluoride (MNP/AlF₃) layer. Finally, both materials were functionalized with aminopropyl groups necessary for the grafting of the structural modified organocatalysts (Scheme 4). MOF’s structures (ie, MIL-101(Al), MIL-101(Al)-NH₂ and UIO-66(Zr)-NH₂) were prepared following a procedure described in literature [Chem. Mater. 26 (2014) 6722].

    Once the inorganic carriers prepared and typical structures confirmed through characterization techniques (adsorption-desorption of liquid nitrogen at 77 K, transmission electron microscopy (TEM), X-ray diffraction (XRD), TGA/DTA, Mössbauer spectroscopy, magnetic measurements, DLS, and DRIFT spectroscopy), the next stage was the synthesis of heterogeneous organocatalysts. Different preparation methodologies (e.g., grafting, encapsulation, ion-exchange, co-precipitation) were applied as a function of the carrier and organocatalyst nature. The asymmetric epoxidation of trans-methyl-cinnamate in the presence of the heterogeneous organocatalysts was conducted in accordance with the methodology used the homogeneous conditions. Moreover, with the aim to improve as much as possible the green degree of the organocatalytic system, parallel tests in which the Oxone was replaced with the much more benign H₂O₂ epoxidation agent were also made. The obtained products were analyzed by HPLC using a CHIRALPAK IA column and the product identification was done by comparing the retention times with those of standard commercial compounds.

Scheme 4. Schematic synthesis procedure of the core-shell magnetic nanoparticles stabilized with silica or hydroxylated aluminium fluoride

In the following will be summarized the most important results obtained in the dioxirane-mediated epoxidation of trans-mehtylcinnamate in the presence of heterogeneous organocatalysts.

As specified, the immobilization of the organocatalysts generates heterogeneous catalysts with more complicated structures than homogeneous counterparts with unexpected positive or negative effects upon the catalytic efficiency. An important positive effect induced by the grafted ketone B (see Scheme 2) on aminopropyl-based MCM-41 carrier (characterized by a bi-modal porosity with narrow mezoporos of 2.6 and 3.6 nm, respectively), for instance, was the increases of the e.e. in the (2R, 3S)-phenyl-glycidate enantiomer. The catalytic activity was lower than that of the homogeneous ketone Shi (homogeneous Shi ketone (structure A, Scheme 2): C = 76.4%, S_epoxid = 54%, ee = 17.6% (2R, 3S); ketone B/MCM-41: C = 20.8%, S_epoxid = 35%, ee = 38.3% (2R, 3S)), but this is in line with the general features of heterogeneous versus homogeneous catalysis (ie, often the catalytic activity of solid catalysts is lower than of the homogeneous counterparts). The enhanced enantioselectivity arises, most probably, from steric congestion encountered by the prochiral reactant, namely trans-methylcinnamate (Scheme 5).

By using H₂O₂/CH₃CN mixture and in the presence of the ketone B/MNP-SiO₂ core-shell systems the conversion of the trans-methyl-cinnamate was quite high taking into account the reaction temperature: T = 0°C: C = 36.1%, S_epoxid = 100%, ee = 73.7% (2S, 3R). The total selectivity to epoxide indicates the system as a green one (Scheme 6). The stereoselection is very high in the presence of this catalyst but the discrimination is in the favor of the “wrong” enantiomer - (2S, 3R), indicating a reaction pathway through the “spiro 2” transition state (see Scheme 2). In the presence of the ketone B/MNP-AlF₃ sample, the trans-methyl-cinnamate conversions are lowered with 10% comparing with those obtained in the presence of the ketone B/MNP-SiO₂ core-shell system, while e.e. is 100% in the (2S,3R)-isomer regardless of the reaction conditions. These differences can have different reasons: i) the different chemical nature of the MNP shell (ie, SiO₂ versus AlF₃); ii) the lowered amount of the grafted chiral ketone; and/or iii) the higher particles agglomeration in the reaction medium of the ketone B/MNP-AlF₃ sample (DLS measurements). However, in both cases a great advantage of the system refers to the simple separation and recycling of the heterogeneous organocatalyst by applying an external magnetic force.

Scheme 5. Schematic representation of the trans-methylcinnamate epoxidation in the presence of ketone B/MCM-41 catalyst and Oxone

Scheme 6. Schematic representation of the trans-methylcinnamate epoxidation in the presence of ketone B/MNP/SiO₂ catalyst and H₂O₂/acetonitrile

A considerable effort was invested in applying the concepts of green chemistry for developing a synthesis of heterogeneous organocatalysts with a high acceptability in the environment. For this, in the last part of the project have been developed new materials made from renewable organocatalysts intercalated into layered double hydroxides structure (LDH). In literature there is no information on the levulinate intercalation in LDH structures and their use in epoxidation reactions in liquid phase. On the basis of the materials design and of the catalytic system were major green elements such as simple preparation methodology from renewable raw materials, the heterogeneous character of the organocatalyst, the use of H₂O₂ as oxidizing agent and the lack of soluble inorganic base in the reaction medium. Moreover, the layered double hydroxides (LDH) are often preferred in chemical processes to other types of catalysts due to their versatility, simplicity, easy to modify their properties and low price. Such materials are an important basis for the development of new materials with controlled structure, controlled accessibility to the active sites, adjustable pore size and high surface area [Mat. Chem. Phys. 2007, 104, 133]. Not last, their ability to retain and change the inorganic with organic anions makes these materials unique.

Four types of heterogeneous organocatalysts were prepared by different methodologies, as listed in Table 2.

Table 2. Preparation methods for the levulinate-intercalated LDH

Sample

LEV@LDH-air

LEV@LDH-N₂

LEV@LDH-pp

LEV@LDH-mem

Precursors nature

MgAl-LDH
[Na][LEV]

MgAl-LDH
[Na][LEV]

Metal nitrates salts
Levulinic acid

LDH-derived mixed oxide (calcined at 450°C)
[Na][LEV]

Intercalation method

Ion-exchange

(four steps, air)

Ion-exchange (one-step, nitrogen)

Co-precipitation

Reconstruction

The prepared materials were characterized by techniques such as adsorption-desorption isotherms of liquid nitrogen at -196 ° C, X-ray diffraction (XRD), thermogravimetric analysis and differential thermal analysis (TG-DTA), and infrared spectroscopy.

Anion exchange process occurs with the introduction levulinate anions in LDH structure in different concentrations and structural orientations, depending on the method of preparation (Table 1).

IR spectra of LEV@LDH-N₂ sample highlight the characteristic absorption bands of the LDH carrier and a novel band with a maximum at 1567 cm^-1 associated with the asymmetrical vibration of ionized carboxyl groups of the levulinate molecules. This band, which is missing in the IR spectrum of the LEV@LDH-air sample is the main evidence that [LEV]^δ^- species were inserted by ion exchange only in nitrogen atmosphere. The intercalation is also confirmed by TG-DTA analysis. Levulinic acid content of the sample was measured from TG as 14.5wt%. The XRD pattern of the sample does not indicate changes in structural parameters of the pristine LDH, while adsorption-desorption isotherms of liquid nitrogen indicate changes its texture properties: surface BET slightly decreases from 42 m²/g (MgAl-LDH) to 39 m²/g (LEV@LDH-N₂), while the porosity has been change from a bimodal with pores of 3.5 and 14.6 nm (MgAl-LDH) to a monomodal one with pores of 10.6 nm (LEV@LDH-N₂). Summarizing these data it can be supposed the [LEV]^δ^- species are not intercalated between the LDH sheets but, most probably, are anchored on the sheets corners of the LDH (Figure 1). This assumption is also based on the fact that due to high load density of carbonate anions between layers they are in a strong electrostatic interaction with LDH layers, making difficult any ion exchange with species outside the network [Chem. Mater. 13 (2001) 3507-3515].