Enantioselective epoxidation of styrene by manganese chiral schiff base complexes immobilized on MCM-41

2

ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ÝÝ These are not the final page numbers!

Full Papers

1

.993–2.001 ꢁ. In both complexes the MnꢀO distances are

found to be slightly longer than the usual due to Jahn–Teller

[

29]

distortion.

Powder X-ray diffraction study (PXRD)

The PXRD patterns of MCM-41, pyridine carboxaldehyde modi-

fied MCM-41, and MCM-41 with the immobilized metal com-

plex are shown in Figure 3. MCM-41 shows one characteristic

Figure 3. XRD patterns of a) MCM-41 b) pyridine carboxaldehyde modified

Figure 4. Top : N

2

-adsorption isotherm of a) MCM-41, b) pyridine carboxalde-

MCM-41 c) Metal complexes immobilized on MCM-41 as stacked plots.

hyde modified MCM-41, c) complex 1/MCM-41, d) complex 2/MCM-41.

Bottom: Pore size distribution in p) MCM-41, q) pyridine carboxaldehyde

modified MCM-41, r) complex 1/MCM-41, s) complex 2/MCM-41.

intense peak due to reflection at the (100) plane and two low-

intensity peaks due to reflections at the (110) and (200)

[

27]

planes.

On complexation, the intensities of all peaks de-

MCM-41-immobilized catalyst shows a type III isotherm, indi-

cating incorporation of the metal complex into the pores of

MCM-41 through a multistep grafting method (Figure 4top,

crease and the intensities of the 110 and 200 reflections de-

crease more on immobilization of metal complex on MCM-41,

with only a minor shift in the 2q values (Figure 3b,c). This can

be attributed to shrinkage effects due to the presence of the

metal complex in the pore channels of MCM-41, causing a de-

crease in the pore size and an increase in the d-spacing. Never-

theless, the retention of the characteristic peaks indicates that

the long-range order of mesoporous hexagonal channels is

preserved upon modification. Similar shifts in the 2q values

upon immobilization of metal complexes on MCM-41 has been

[24]

c,d).

This is in agreement with the decrease in the pore

III

volume of MCM-41 after incorporation of chiral Mn Schiff

base complexes. BET surface area, pore diameter, and pore

volume data are presented in Table 2. The BET surface areas in

functionalized MCM-41 and complex 1/MCM-41 and com-

plex 2/MCM-41 are less than that of pure MCM-41. Similarly,

pore diameters as well as pore volumes are found to be less in

the case of the modified MCM-41. For example, on immobiliza-

tion of complex 1, the BET surface area decreases from 1015 to

[

27,28]

reported previously

and supports the successful attach-

2

ꢀ1

ment of chiral metal catalysts to modified MCM-41.

623 m g , the pore size from 37 to 25 ꢁ, and the pore

3

ꢀ1

volume from 0.576 to 0.415 cm g , Figure 4bottom, p, q, r,

and s).

Surface area and CHN analysis

The amount of Mn loading in the two immobilized com-

plexes was determined by inductively coupled plasma (ICP)

spectrometer analysis. Carbon, hydrogen and nitrogen con-

tents were determined by CHN analysis. The C/N ratio is close

to that expected (Table 2). This further indicates that metal

complexes are immobilized on the surface of MCM-41.

MCM-41 and pyridine carboxaldehyde modified MCM-41 show

typical type IV isotherms having three well-distinguished re-

[

30]

gions: mono- and multilayer adsorption on the pore walls

region A), capillary condensation (region B), and multilayer ad-

sorption on the outer surface (region C) (Figure 4top, a,b). The

(

ChemPlusChem 0000, 00, 0 – 0

3

ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

These are not the final page numbers! ÞÞ

Full Papers

Table 2. BET and CHN analysis of MCM-41, pyridine carboxaldehyde modified MCM-41, complex 1/MCM-41, complex 2/MCM-41. Mn loading is calculated

from ICP analysis.

Elemental analysis [wt%]

Compound

Mn loading

mg/100 mg]

BJH pore diameter

[ꢁ]

Total pore vol.

BET surface area

C

H

N

C/N

3

ꢀ1

2

ꢀ1

[

[cm g

]

[m g

]

MCM-41

–

37

30

26

25

0.944

0.576

0.423

0.415

1015

745

623

638

–

pyridine carboxaldehyde modified MCM-41

complex 1/MCM-41

complex 2/MCM-41

10.62

69.46

54.90

2.13

3.67

2.44

2.32

7.17

7.35

4.57

9.68

7.46

10.82

11.35

2

9

13

FTIR and solid-state Si and C NMR analysis

sponding to C=N, C=C (aromatic), and CꢀN stretching vibra-

tions and it does not possesses any peaks due to NꢀH vibra-

tions. The presence of the C=N vibrational mode and absence

FTIR spectra of the two Mn complexes are shown in Figure S1

and Figure S2, respectively. Both complexes show characteristic

stretching vibrational bands for C=N, C=C, CꢀO, ꢀCH, and =CH

of NꢀH vibrational bands confirms the coupling of the ꢀNH

2

group of the siloxyl moiety with the ꢀCHO group of pyridine

carboxaldehyde through formation of a C=N bond.

ꢀ

1

ꢀ1

[

n(ꢀC=N, 1600–1617 cm ), n(ꢀC=C, 1526–1540 cm ), n(CꢀO,

ꢀ

1

ꢀ1

1

304–1307 cm ), n(CꢀH, 2932–2948 cm ), and n(=CH, 762–

77 cm )]. The FTIR spectrum of calcined MCM-41(Fig-

The formation of amino-functionalized MCM-41 and pyridine

ꢀ

1

[23]

29

7

carboxaldehyde modified MCM-41 is further confirmed by Si

ꢀ

1

13

29

ure 5a) shows characteristic bands at 1064 and 3464 cm for

and C NMR spectroscopy. The Si NMR spectrum of MCM-41

shows a broad peak at ꢀ97.3 to ꢀ99.3 ppm (Figure 6a). How-

ever, on modification with APTS this chemical shift value shift-

ed to ꢀ110.5 ppm (Figure 6b) and upon further treatment

[

25]

SiꢀO<C-<Si and SiꢀOH bonds, respectively. Treatment with

-aminopropyltriethoxysilane (APTS) yields the amine-function-

alized MCM-41, which shows additional bands at 3274, 2940,

3

1

567, and 1492 due to n(NꢀH stretching), n(CꢀH, stretching),

with pyridine carboxaldehyde a downfield shift occurs

29

n(NꢀH bending), and n(CH bending), respectively (Figure 5b).

(ꢀ93.3 ppm; Figure 7a). These shifts in the Si NMR spectrum

further indicates that the chemical environment near Si in

MCM-41 does not remain the same upon modification. The

2

The presence of NꢀH stretching and ꢀCH bending modes in-

2

dicates the functionalization of MCM-41 with amine via APTS.

After MCM-41 had been functionalized with amino groups, it

was further treated with pyridine carboxaldedyde (Scheme 2).

FTIR spectrum of pyridine carboxaldehyde modified MCM-41

13

C NMR spectrum of pyridine carboxaldehyde modified MCM-

41 shows a signal at 160 ppm due to the aromatic moiety, two

peaks at 23.7 and 11.9 ppm due to the aliphatic carbon atom

of the aminopropyl linker, and peaks at 45.1 and 48.1 ppm due

to the carbon atom in the C=N

ꢀ

1

(

Figure 5c) shows peaks at 1647, 1548, and 1393 cm corre-

bond (Figure 7b). These signals

clearly provide the evidence for

the incorporation of the linker

group on the silica matrix.

The formation of linker

groups on the silica matrix is

thus confirmed. In the next

step, the chiral Mn Schiff base

complexes are anchored to

MCM-41 throug the formation

of an MnꢀN bond with the

linker group (Scheme 2). The

FTIR spectrum of the MCM-41-

immobilized complexes shows

bands similar to those of neat

complexes in the region of

ꢀ

1

600–1200 cm . The presence

of the most characteristics

ꢀ1

bands above 1600 cm due to

the C=N bond confirms the im-

mobilization of the metal com-

plexes over MCM-41 (Fig-

ure 5d).

Figure 5. FTIR spectra of a) MCM-41, b) APTS-MCM-41, c) pyridine carboxaldehyde modified MCM-41, d) com-

plex 1/MCM-41, e) complex 2/MCM-41

&

ChemPlusChem 0000, 00, 0 – 0

4

ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ÝÝ These are not the final page numbers!

Full Papers

istics peaks due to p!p* and n!p* transitions. The bands

appearing below 300 nm are due to p!p* transitions and

above 300 nm due to n!p* transitions. Complex 1 shows

bands at 247, 297, 334, 392, and 479 nm. The bands at 247

and 297 nm are both due to n!p* transitions. The bands at

334 and 392 nm are because the metal-to-ligand charge trans-

fer transition (MLCT) and the band at 479 arises from a d–d

transition. The band at 479 nm is mainly a combination of

a charge transfer and d–d transition band. Complex 2 shows

bands at 239, 281, 307, 344, and 398 nm. The lower-wave-

length bands are due to p!p* transitions and the higher-

wavelength bands are due to n!p* transitions while bands

above 340 nm are due to charge transfer transitions.

The diffuse reflectance spectra (DRS) of the MCM-41-immo-

bilized complexes also show bands approximately in the same

region (Table 3). Small changes in the wavelengths can be at-

tributed to the effect of the MCM-41 matrix. MCM-41-anchored

complex 1 shows transitions due to p!p* and dp!pp. The

most significant changes observed in the case of complex 1

and complex 2 supported on MCM-41 is the appearance of

a band above 600 nm. In the case of complex 1, the band at

653 nm is due to a transition from a dp orbital (HOMO) to

a pp orbital of the ligand (LUMO) mostly concentrated on the

pyridine linkage. Complex 2/MCM-41 shows two additional

bands at 491 nm and 613 nm besides the p!p* transition.

These two transitions originate from the excitation of an elec-

tron from a dp orbital (HOMO) to a pp orbital (LUMO) mostly

concentrated at the linker atoms of the pyridine carboxalde-

hyde moiety. Thus the appearance of charge transfer bands

truly indicates the binding of the metal complex through pyri-

dine carboxaldehyde with amino-functionalized MCM-41.

III

Scheme 2. Synthesis of MCM-41-supported Mn chiral Schiff base complex

3.

2

9

Figure 6. Si NMR spectra of a) MCM-41, b) APTS-MCM-41.

UV/Vis and diffuse reflectance spectra

III

The UV/Vis spectrum of the ligands, Mn metal complexes, and

MCM-41-supported metal complexes are shown in Figure 8.

The transition assignment for each absorption is listed in

Table 3. The UV/Vis spectra of the two ligands show character-

29

Figure 7. a) Si NMR spectrum of pyridine carboxaldehyde modified MCM-

1

3

41, b) C NMR spectrum of pyridine carboxaldehyde modified MCM-41.

ChemPlusChem 0000, 00, 0 – 0

5

ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

These are not the final page numbers! ÞÞ

Full Papers

potential range 1.0 V to ꢀ1.0 V,

no anodic or cathodic peak po-

tentials were observed. Howev-

er, starting from the negative

potential range, we could ob-

serve a strong reduction peak

at ꢀ0.415 V for complex 1 and

at ꢀ0.365 V for complex 2 due

III

II

to the reduction of Mn to Mn .

[31]

Bonadies et al.

has also ob-

served a single irreversible re-

duction peak for similar com-

plexes in acetonitrile and has at-

tributed it to the formation of

III

+

1

:1 electrolyte of [Mn (salen)]

ꢀ

and CH COO . Hence the pres-

3

ence of a single reduction peak

potential confirms the initial ox-

idation state of the metal com-

plexes to be +III; however, in

solution the axial ligand

ꢀ

(

[

CH COO )

dissociates

and

3

III

+

Mn (salen)] forms. A plot of

current peak vs. the square root

of the scan rate is found to be

linear, indicating that this

Figure 8. UV/Vis spectra of a) ligand 1, b) ligand 2, c) complex 1, d) complex 1/MCM-41, e) complex 2, f) com-

plex 2/MCM-41 (and inset).

a

normal diffusion-controlled

process (Figure 10a,b). The correlation is found to be much

more prominent for the peak current values corresponding to

the lower scan rate (Figure 10b).

Table 3. Assignment of UV/Vis transitions in synthesized complexes.

Compound

l [nm]

Transition

For recording the cyclic voltammograms of the MCM-41-sup-

ported chiral Mn Schiff base complexes, the working electrode

was prepared by taking a 1:1 weight ratio of metal complexes

ligand 1

213, 253

p!p*

n!p*

318, 398

ligand 2

211, 229, 279 p!p*

305, 382

n!p*

[32]

in 1 mL of acetonitrile. This suspension was ultrasonicated

complex 1

247, 297

n!p*

for 15 min. Then 10 mL of this dispersion was coated on

a glassy carbon electrode and 5 mL of 5% styrene (as binder

from Aldrich) was applied to this coating and dried. The glassy

carbon electrode was used as the working electrode and Ag/

AgCl/KCl (saturated) was used as reference electrode. Cyclic

voltammograms corresponding to the MCM-41-immobilized

complexes also show a single electron reduction peak due to

3

4

34, 392

79

n!p*

dp!pp*

complex 1/MCM-41 236, 285

p!p*

348

424

653

n!p*

dp!pp*

dp!pp*(pyridine carboxaldehyde)

complex 2

239, 281

p!p*

3

07, 344

98

p!p*

III

II

the Mn !Mn process (Figure 9c,d). However, the reduction

potential values (ꢀ0.565 V and ꢀ0. 599 V) differ from those of

the homogeneous systems. The change in the peak potential

values upon immobilization on the solid support may be due

to the effect of the MCM-41 matrix. In the case of the immobi-

lized system we the plot of the scan rate versus the peak cur-

rent was linear, indicating a surface-controlled process (Fig-

ure 10c).

complex 2/MCM-41 292, 336

p!p*

3

4

6

55, 398

91

13

n!p*

dp!pp*

dp!pp*(pyridine carboxaldehyde)

Cyclic voltammetry study

The voltammograms of the unsupported complexes were re-

SEM analysis

corded with 0.01m solutions of the complexes in acetonitrile

using 0.1m tetrabutyl ammonium bromide (NBu Br) as a sup-

Scanning electron microscopic images of MCM-41 and MCM-

41-anchored salen complexes are shown in Figure 11. SEM mi-

crographs of modified MCM-41 show no significant change in

comparison to its precursor except some formation of agglom-

erate particles due to complex formation. This indicates that

4

porting electrolyte. Cyclic voltammograms of the two metal

complexes were recorded at different scan rates and are

shown in Figure 9a,b. When the voltammograms for com-

plex 1 and complex 2 were recorded starting from the positive

&

ChemPlusChem 0000, 00, 0 – 0

6

ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ÝÝ These are not the final page numbers!

Full Papers

chiral modification of the MCM-

41 surface does not lead to sig-

nificant changes in the surface

morphology. However, the pres-

ence of agglomerate particles

indicates the presence of metal

complexes which block the

pores of the mesoporous

system.

TGA analysis

Thermogravimetric patterns of

the parent Mn Schiff base com-

plexes and of the immobilized

complexes are displayed in

Figure 12. Comparison of the

thermogravimetric analysis indi-

cates that complex 1 has four

weight-loss steps at about 46,

160, 288, and 3948C while com-

plex 2 shows four weight-loss

steps at about 45, 190, 280, and

4028C (Figure 12top (a,b)). On

the basis of the weight changes,

the first weight-loss step in both

complexes corresponds to the

loss of a water molecule as an

endothermic phenomenon. The

second weight loss at 160–

Figure 9. Cyclic voltammograms of a) complex 1, b) complex 2, c) complex 1/MCM-41, d) complex 2/MCM-41 re-

corded at different scan rates (0.1 V to 0.025 V).

1908C may be related to the

loss of an acetate group. There

is a sharp weight change at

3948C (complex 1) and at 4028C

(

complex 2), which is attributed

to the loss of Schiff base ligands.

However, for the corresponding

immobilized complexes the

weight loss extends up to

4748C

(Figure 12top

(c,d)),

which indicates that the thermal

stability is greatly enhanced. A

comparison of the dTGA curve

for the parent complex 1 and

immobilized complex 1 clearly

that immobilization enhances

the thermal stability of the com-

plexes (Figure 12bottom (X and

Y curves)). On the basis of the

thermal analysis data, we may

conclude that MCM-41-immobi-

lized complexes may be treated

thermally without any signifi-

cant decomposition.

Figure 10. a) Plot of square root of scan rate (0.1 V to 0.075 V) vs. peak current (in mA). b) Plot of square root of

scan rate (0.1 V to 0.005 V) vs. peak current (in mA). c) Plot of of scan rate (0.1 V to 0.075 V) vs. peak current (inmA).

Plots determined from the values from cyclic voltammograms of complex 1 (a and b) and complex 1/MCM-41 (c).

ChemPlusChem 0000, 00, 0 – 0

7

ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

These are not the final page numbers! ÞÞ

Full Papers

Table 4. Oxidation of styrene using complex 1 as the catalyst and H

as the oxidant in different solvents.

2

O

2

[

a]

[

c]

Prod. sel. [%]

BA

[

b]

[d]

Solvent

CH OH

t [h]

Conv. [%]

SO

others

% ee

3

36

16.5

13.2

66.2

30.2

31.5

14.3

34.2

–

9.2

8.9

76

52

68

62

2.2

–

89.6

89.4

22.2

48

32

38

85.6

–

1.2

1.7

1.8

–

15

18

32

22

27

29

16

–

C

CH

2

H

5

OH

CN

Cl

3

Figure 11. SEM images of a) MCM-41, b) complex 1/MCM-41, c) complex 2/

MCM-41.

2

DMF

CHCl

3

H

2

O

toluene

hexane

–

[

3

[

a] Reaction conditions: catalyst 10 mg, styrene 5 mmol, solvent 10 mL,

0% H 2 mL, 08C. [b] Determined by GC analysis; conversion (%)=

moles of reactant converted]ꢂ100]/[moles of reactant in feed]. [c] Prod-

2 2

O

uct selectivity (%)=[moles of product formed]ꢂ100]/[moles of reactant

converted]. [d] Selectivity for the R product; determined by HPLC analysis

on a chiral stationary phase.

oxidation of styrene was slow with less conversion and poor

[33]

selectivity towards epoxidation. Liu et al. also observed simi-

lar solvent effects for epoxidation in the presence of a Au–

silica catalyst. In the presence of less polar solvents the oxida-

tion of styrene is better than in protic solvents and also the se-

lectivity towards epoxide formation is improved. Interestingly,

the reaction does not proceed in nonpolar solvents such as

toluene and hexane, indicating the effect of solvent polarity on

the catalytic conversion of styrene to styrene oxide. Of the var-

ious solvents considered, the best results were obtained with

acetonitrile (CH CN): 66% conversion and 76% selectivity

3

(

Table 4). Since CH CN is found to be a suitable solvent for the

3

Figure 12. Top: TGA analysis of a) complex 1, b) complex 2, c) complex 1/

MCM-41, d) complex 2/MCM-41. Bottom: X and Y are corresponding dTGA

curves for complex 1 and complex 1/MCM-41, respectively.

epoxidation process, we carried out the oxidation of styrene

with all of the synthesized catalysts in CH CN.

3

In order to observe the effect of temperature, we again per-

formed the oxidation of styrene in CH CN with complex 1 as

3

Catalytic activity

the test catalyst over a range of temperatures (Table 5). At

room temperature (308C) the oxidation was very slow and we

obtained benzaldehyde as the major product and styrene

oxide with only 5% ee. When the reaction was conducted in

an ice bath (08C) both the conversion and the selectivity to-

wards epoxidation improved and the enantioselectivity in-

creased to 32% ee. However, on lowering the temperature to

ꢀ208C we observed low conversion, low selectivity, and

a slight decrease in enantioselectivity. We attribute the steady

increase in enantioselectivity at 08C to the formation of

a more persistent and a better organized activated complex at

III

As mentioned in the introduction, Mn chiral Schiff base com-

plexes are well-known catalysts for the epoxidation of styrene.

Herein we have also investigated the oxidation of styrene to

styrene oxide (SO) in the presence of the synthesized homoge-

neous and also the MCM-41-supported metal complexes. Hy-

drogen peroxide (30%, H O ) is used as the oxidant in the cat-

2

alytic process. The catalytic oxidation of styrene is dependent

on various parameters such solvent, temperature, amount of

catalyst, and oxidant. Here we discuss the effect of these pa-

rameters on the oxidation of styrene catalyzed by the chiral

[34]

lower thermal energy. Hence, we performed the catalytic ox-

idation of styrene using all other catalysts at 08C. Hutching

and co-workers have also found a decrease in enantioselectivi-

ty at ꢀ108C for the epoxidaton of stillbene using Mn salen

III

Mn complexes.

Effect of solvent and temperature

[21]

complexes.

In order to observe the effect of the solvent on the oxidation

of styrene we employed complex 1 as a test catalyst. We chose

a range of solvents (both protic and aprotic) having different

polarities (i.e. different dielectric constants) and the results are

shown in Table 4. In the presence of highly polar solvents the

These results show a clear trend where the sense of induc-

tion is reversed in accordance with the Arrhenius law and also

improves with decreasing temperature from room temperature

to 08C. Generally, it is believed that the log rate constant in-

creases with absolute temperature, according to the Arrhenius

&

ChemPlusChem 0000, 00, 0 – 0

8

ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ÝÝ These are not the final page numbers!

Full Papers

Effect of oxidant concentration and catalyst amount

2 2

Table 5. Oxidation of styrene using complex 1 as the catalyst and H O

as the oxidant at different reaction temperatures.

[

a]

The effect of the oxidant concentration on the styrene epoxi-

dation was studied by GC analysis and cyclic voltammetry, and

again we used complex 1 as the test catalyst. In the GC analy-

sis, in the absence of oxidant we did not observe any signal

for the formation of epoxide. However, on sequential addition

[

c]

Prod. sel. [%]

SO BA others % ee

ꢀ

1

[b]

[d]

T [8C] 1/T [K

]

t

Conv.

R

ln [Ks/K ]

ꢀ

3

ꢂ10

[h] [%]

3

2

0

5

0

5

3.30

3.41

3.59

3.6

3.73

3.80

3.95

36

6.5

9.2

2.2 79.6 18.2

8.9 76.4 14.7

5

17

22

32

30

29

26

1.24

2.34

1.99

2.21

2.76

2.56

2.00

^o^f^H^O2 from 0.2 mL to 2 mL, two new additional peaks ap-

36 26.2

36 66.2

36 54.5

36 44.3

36 39.2

46

76

52

42

28

35

19

2

22.2 1.8

peared at retention times of 13.23 and 18.50 min correspond-

ing to epoxide and benzaldehyde (Figure 13). It was observed

ꢀ

32

38

51

16

20

11

ꢀ

10

20

[a] Reaction conditions: catalyst 10 mg, styrene 5 mmol, solvent 10 mL,

3

0% H 2 mL. [b] Determined by GC analysis; conversion (%)=[moles

2 2

O

of reactant converted]ꢂ100]/[moles of reactant in feed]. [c] Product se-

lectivity (%)=[moles of product formed]ꢂ100]/[moles of reactant con-

verted]. [d] Selectivity for the R product; determined by HPLC analysis on

a chiral stationary phase.

law. This law holds well only if the process continues to follow

the same reaction sequence. As the temperature increases, the

rate of the reaction will increase but the supply of molecules

to the surface will become slow, and accordingly the Arrhenius

plot will change slope. In the present case lowering tempera-

ture favors the reaction since adsorption and surface concen-

tration is a function of temperature, especially when the mole-

cule has to adsorb in a particular mode to give stereospecific-

ity. Further, when the temperature was lowered from 378C to

Figure 13. GC analysis showing the effect of peroxide concentration on the

oxidation of styrene catalyzed by complex 1.

0

3

8C, the partition coefficient (K) values increased from 3.13 to

.78. However, a further decrease in temperature decreases the

that as the concentration of H O was increased up to 2 mL,

2

the conversions of styrene and epoxidation selectivity also in-

creased. A further increase in H O did not lead to further im-

partition coefficient value. Thus the reaction was performed at

2

0

8C at pH 6.5–6.8.

In order to further substantiate this fact, the natural loga-

provements in catalytic conversion, rather the addition of

2.5 mL of H O resulted in some unwanted minor products.

2

rithm of the relative rate constant for the formation of (S)- and

R)-styrene oxide can be correlated with the ee values of the

reaction according to Equation (1), which can also be written

This indicates that excess oxidant does not have a beneficial

effect on oxidation of styrene. Thus, 2 mL of H O is considered

(

2

as the optimum amount of oxidant for the catalytic process.

Figure 14a is the cyclic voltammogram recorded for complex 1

with different concentrations of styrene. The concentration of

styrene has no any significant effect on the reduction potential

of complex 1. However, on addition of H O , the reduction

[35–37]

as the differential Eyring equation shown in Equation (2).

ꢀ

ꢁ

ꢀ

ꢁ

K_S

K_R

ð100 þ %eeÞ

ð100 ꢀ %eeÞ

ln

¼ ln

ð1Þ

2

III

II

peak corresponding to Mn !Mn vanishes and a new quasi-

reversible couple for Mn /Mn and a couple for Mn /Mn

III

IV

III

°

K_S

K_R

ꢀDDH

DDS

¼

ð2Þ

couple appear at 0.69 and 0.223 V, respectively (Figure 14b).

This indicates that the epoxidation of styrene proceeds

RT

R

IV

through the formation of Mn =O species as one of the inter-

[38]

Considering Equation (2), a linear plot would be expected if

mediates. Cyclic voltammetry studies also suggest that the

concentration of H O plays a dominant role in the catalytic ox-

the mechanism of the enantioselection remained constant

2

°

over a wide range of temperatures. In this case DDH and

idation of styrene and the epoxidation reaction proceeds

°

IV

DDS should remain unchanged. In contrast, a nonlinear curve

through the formation of Mn –oxo species.

will result if the enantiodetermining step changes at different

To investigate the effect of the catalyst amount, we per-

formed the catalytic oxidation of styrene using complex 1 as

the test catalyst and H O as the oxidant. Keeping the sub-

°

temperatures because of the variation in the DDH and DDS

values. In the present case the plot of ln(Ks/K ) vs 1/T shown in

R

2

Figure S3 is found to be nonlinear. Therefore it appears that

the enantiodetermining step is changing over the range of

temperatures examined. Thus, the observed reversal in abso-

lute chiral induction could arise from a temperature-dependent

interchange of the enantiodetermining mechanisms.

strate amount fixed (5 mmol) and oxidant amount at 2 mL, we

varied the catalyst loading from 0.5 to 25 mg. Increasing the

amount of catalyst from 0.5 to 15 mg leads to enhancement in

enantioselectivity from 23% to 46% ee. Styrene conversion

and epoxidation selectivity is found to be highest with 15 mg

ChemPlusChem 0000, 00, 0 – 0

9

ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

These are not the final page numbers! ÞÞ

Full Papers

Table 6. Oxidation of styrene using 15 mg loading of various catalysts

and H O (2 mL) as the oxidant in CH CN at 08C.

2 2 3

[

b]

Prod. sel. [%]

t [h] Conv. [%] SO BA Others % ee [R]

[

a]

[c]

Catalyst

H

2

O

2

26 20

36

24 30

–

5

–

76 24

–

Mn(OAc)

complex 1

complex 2

complex 1+H

complex 2+H

complex 1+H

2

.4H

+H

2

O

–

2

O

2

78 17

36

–

6

–

2

O

2

36 77

42 46

82 18

62 38

84 10

46

38

67

+pyridine 38 72

carboxaldehyde

complex 2+H

carboxaldehyde

2

O

2

+pyridine 40 63

61 28 11

44

complex 1/MCM-41+H

complex 2/MCM-41+H

2

O

2

27 68

34 57

88 12

67 33

–

71

52

2

[

a] Determined by GC analysis; conversion (%)=[moles of reactant con-

verted]ꢂ100]/[moles of reactant in feed]. [b] Product selectivity (%)=

moles of product formed]ꢂ100]/[moles of reactant converted]. [c] Selec-

[

tivity for the R product; determined by HPLC analysis on a chiral station-

ary phase.

better the environment of the Mn complexes in the solid. We

found that under these conditions the enantioselecitivity is

better than that obtained with the complexes alone and almost

comparable to that obtained with the heterogeneous system

(Table 6). In a control experiment with H O , MCM-41, pyridine

Figure 14. CV study to elucidate the effect of styrene on complex 1 and

d) the effect of peroxide.

2

carboxaldehyde modified MCM-41, Mn(OAc) , and H O , we ob-

2

served negligible catalytic activity without any selectivity

toward the epoxidation of styrene (Table 6). Hence we can con-

clude that the catalytic activity is entirely attributed to the

salen complex in both the homogeneous phase and also im-

mobilized on the pyridine carboxaldehyde modified MCM-41.

loading of the catalyst. However, no any enhancement in the

catalytic process nor in the enantioselectivity was observed on

increasing the catalyst loading to 25 mg. So we considered

15 mg as the effective amount of catalyst for the catalytic oxi-

dation of styrene.

Recyclability test

Comparison of catalytic activity of homogeneous and heter-

ogeneous catalyst

The homogeneous complexes could not be retrieved, whereas

we could successfully recover the MCM-41-supported chiral cat-

alyst. In recent years ionic liquids have been used recovering

After optimization of the solvent, temperature, and oxidant and

catalyst concentration, we the performed oxidation of styrene

using various catalysts under similar conditions All the chiral

catalysts give (R)-styrene oxide with some level of enantioselec-

tivity according to HPLC analysis (Table 6). Complex 1 and its

heterogeneous counterpart are found to be most effective for

catalytic epoxidation of styrene in the presence of hydrogen

peroxide. In comparison to complex 1, complex 2 and its heter-

ogeneous counterparts shows poor catalytic activity. The differ-

ences in the catalytic activity of the two complexes can be at-

tributed to structural changes cause by the nature of the

ligand. The conversion of styrene to styrene epoxide by com-

plex 1 and its heterogeneous counterpart is comparable. How-

ever, the reaction takes less time with immobilized catalysts

and the enantioselectivity is enhanced (71% ee vs. 46% ee). It

has been reported that the system composed of an aldehyde

and a transition metal can effect the aerobic epoxidation of al-

[41,42]

homogeneous catalysts.

We have also reported the suc-

II

cessful recovery of a chiral Cu Schiif base complex in an asym-

[16]

metric Henry reaction. So, in order to recover the chiral cata-

lyst we implemented ionic liquid 1-butyl-3-methylimidazolium

hexaflurophosphate [BMIM][PF ]. In the presence of the ionic

6

liquid we do not observe any change in catalytic conversion

and selectivity. But we could recover the catalyst and used it in

two consecutive cycles without any loss in catalytic activity.

After the third cycle sufficient metal leaching has probably oc-

curred; a UV/Vis study indicates that the metal complexes are

ruptured. The UV/Vis spectrum of the recovered catalyst does

not show any band characteristics of complex 1 and complex 2

and indicates that the structure has been destroyed (Figures S4

and S5). To study the recyclability of the immobilized catalyst

during epoxidation of styrene, we carried out a few epoxidation

reactions with complex 1/MCM-41 and complex 2/MCM-41.

After the first run, the catalyst was removed by centrifugation.

Fresh reactants were added to the supernatant. Analysis of the

[

39,40]

kenes.

So we performed the catalytic epoxidation by

adding pyridine carboxaldehyde as an additive to simulate

&

ChemPlusChem 0000, 00, 0 – 0

10

ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ÝÝ These are not the final page numbers!

Full Papers

reaction mixture by gas chromatography shows no further in-

crease in the conversion of styrene. UV/Vis analysis of the su-

pernatant also indicated an absence of of trace amounts of the

metal complex. After separation, the catalyst was washed thor-

oughly with dichloromethane (DCM), dried, and subjected to

another cycle with fresh reactants under similar epoxidation

conditions. Epoxide conversion, yield, and enantioselectivity

were found to be consistent. The above procedure was repeat-

ed for five cycles, and we did not observe any substantial loss

in the catalytic activity of the immobilized catalyst. This indi-

Conclusion

In conclusion we have successfully synthesized two chiral Mn

Schiff base complexes and they have been characterized by

various spectrochemical and physicochemical techniques. The

two complexes are immobilized on MCM-41 using linkers. The

homogeneous and heterogeneous catalysts are found to be ef-

fective catalysts for oxidation of styrene. The heterogeneous

catalysts show activity similar to that of the homogeneous cat-

alysts. Upon immobilization, the conversion and enantioselec-

tivity are enhanced. The use of an ionic liquid does not affect

the catalytic results but has provided a suitable means to recy-

cle the homogeneous catalyst up to three consecutive cycles.

Upon immobilization on MCM-41, the catalysts can be success-

fully recycled up to five times without loss of catalytic activity.

An EPR study indicates that the heterogeneous systems is de-

III

cates that the chiral Mn salen complex is strongly bonded to

the nitrogen atom of pyridine carboxaldehyde modified MCM-

41 through the axial coordination. However, after the fifth

cycle, that is, in sixth and seventh cycle, we observed a substan-

tial decrease in the catalytic activity.

To determine the reason for the loss in catalytic activity we

performed an EPR analysis. EPR spectra of the two Mn com-

plexes were recorded in the solid phase at 77 K after the sixth

2

+

activated due to loss of manganese as Mn

.

II

and seventh cycle (Figure 15a–d). They show the usual Mn

Experimental Methods

(

S=5/2) six-line hyperfine pattern assigned to the DM =0

I

spin-allowed transitions. Both chiral complexes exhibit an in-

tense signal at g=2.10 with the partially resolved hyperfine

splitting (10 mT or 10 G) from the manganese nucleus. The

presence of the six-line hyperfine signal at g=2.10 with A=

Preparation of Schiff base ligands

6

,6’’-(1,2-Diphenylethane-1,2-diyl)bis(azan-1-yl-1-ylidene)bis-

(

methan-1-yl-1-ylidene)bis(3-methoxyphenol) (ligand 1)

1

0 mT indicates that in both the complexes the metal ions

Ligand 1 was prepared by stirring a solution of (1R,2R)-(+)-1,2-

diphenylethylenediamine (424 mg, 2 mmol) and 2-hydroxy-4-

methoxybenzaldehyde (608 mg, 4 mmol) at room temperature

for 18 h under nitrogen atmosphere in ethanol, (Scheme 1).

After completion of the reaction, the reaction mixture was di-

luted with EtOAc and water. The organic layer was then sepa-

rated and washed sequentially

2

+

[43]

have leached out as Mn ions. It is interesting to see that

intensity of the EPR signal after the fifth run is less than after

sixth cycle. This further indicates that after sixth cycle amount

2

+

of Mn on the surface of MCM-41 increases due to metal

leaching.

with brine and dried over Na SO .

2

4

The solvent was removed in

vacuum and the resulting residue

was purified by chromatography.

TLC was conducted with 10%

EtOAc and hexane. Ligand 1 was

eluted with 6% EtOAc in hexane.

R_f=0.4.

1

H NMR (300 MHz, CDCl ): d=

3

1

3.83 (s, 1H), 8.17 (s, 1H), 7.19

m, 5H), 7.01 (d, J=9.06 Hz, 1H),

.45 (d, J=2.26 Hz, 1H), 6.35(dd,

J=2.26, 9.06 Hz, 1H), 4.65 (s,

(

6

1

3

1

H), 3.78 (s, 3H).

C NMR

d=165.06,

(

75 MHz, CDCl₃):

1

63.60, 163.28, 139.60, 132.81,

28.19, 127.72, 127.38, 112.36,

06.28, 100.96, 79.55, 55.15

2

,2’-(Cyclohexane-1,2-diyl)bis(a-

zan-1-yl-1-ylidene)bis(methan-1-

yl-1-ylidene)diphenol (ligand 2)

Ligand 2 was prepared analo-

Figure 15. EPR spectra of a) complex 1/MCM-41, b) complex 2/MCM-41 after the fifth cycle (c) and after the sixth

cycle (d), respectively.

gously using

a

solution of

ChemPlusChem 0000, 00, 0 – 0

Chem. Soc. 2009 , 131 , 8384–8385.

11

ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

These are not the final page numbers! ÞÞ

Full Papers

(

1R,2R)-(ꢀ)-1,2-diaminocyclohexane (456 mg, 8 mmol) and 2-

hydroxybenzaldehyde (1.952 g, 1.7 mL, 16 mmol) in ethanol

Scheme 1).

Acknowledgements

(

We thank Debabrata Singha (Debu), Research Scholar, IIT, Guwa-

hati, for providing ESR analysis and Prof. B. Viswanathan, Nation-

al Centre for Catalysis Research (NCCR), IIT Madra, for fruitful dis-

cussions.

1

H NMR (500 MHz, CDCl ): d=13.26 (s, 1H), 8.23 (s, 1H), 7.22 (t,

3

J=7.72, 6.62 Hz, 1H), 7.12 (d, J=7.72 Hz, 1H), 6.87 (d, J=

7

4

1

3

.72 Hz, 1H), 6.77 (t, J=7.72, 6.62 Hz, 1H), 3.29 (dd, J=9.93,

13

.41 Hz, 1H), 1.4–1.94(m, 4H). C NMR (75 MHz, CDCl ): d=

3

64.65, 160.92, 132.10, 131.42, 118.60, 118.53, 116.72, 72.57,

3.04, 24.13.

Keywords: ionic liquids · manganese · MCM-41 · styrene

oxidation · Schiff base complexes

[

1] H. Mihara, Y. Xu, N. E. Shepherd, S. Matsunaga, M. Shibasaki, J. Am.

III

Preparation of Mn Schiff base complexes (complex 1 and

complex 2)

[

2] T. Tanaka, Y. Yasuda, M. Hayashi, J. Org. Chem. 2006 , 71 , 7091–7093.

3] Y. Y. Kato, M. Furutachi, Z. Chen, H. Mitsunuma, S. Matsunaga, M. Shiba-

saki, J. Am. Chem. Soc. 2009 , 131 , 9168–9169.

To a well-stirred solution of 1 mmol of either ligand 1 or ligand

[

4] R. T. Ruck, E. N. Jacobsen, J. Am. Chem. Soc. 2002 , 124 , 2882–2883.

2

in 10 mL methanol, 1 mmol of Mn(OAc) ·4H O (0.245 g,

2 2

[5] P. S. Dixit, K. Srinivasan, Inorg. Chem. 1988 , 27 , 4507–4509.

1

mmol) was added. The homogeneous solution was stirred

[6] D. E. De Vos, M. Dams, B. F. Sels, P. A. Jacobs, Chem. Rev. 2002 , 102 ,

3615–3640.

for 3 h and kept for crystallization. The dark brown crystals so

obtained were washed with ethanol and diethyl ether and fi-

nally dried at room temperature (Scheme 1).

[

7] W. F. Beyer Jr., I. Fridovich, Manganese in Metabolism and Enzyme Func-

tion, Academic Press, New York, 1986, chap. 12, p. 193.

[

8] E. N. Jacobsen in Comprehensive Organometallic Chemistry II, Vol. 12

(

Eds.: E. W. Abel, F. G. A. Stone, E. Wilkinson), Pergamon, Oxford, 1995,

p. 1097.

Preparation of immobilized chiral metal complexes

[9] T. Katsuki, J. Mol. Catal. A: Chem. 1996 , 113 , 87–107.

10] D. D. Agarwal, R. P. Bhatnagar, R. R. Jain, S. Srivastava, J. Chem. Soc.

Perkin Trans. 2 1990 , 989–992.

[

MCM-41-supported chiral manganese complexes were synthe-

sized according to the stepwise process shown in Scheme 2. A

suspension of 2 g of MCM-41 and 0.3 g of 3-aminopropyltrime-

thoxysilane and in 20 mL of toluene was refluxed at 808C with

constant stirring for 24 h. The resulting solution was cooled to

room temperature and a white solid was collected after 1 h.

The powder was filtered and washed with diethyl ether. The

solid white product was finally dried under vacuum for 12 h.

To this amino-functionalized MCM-41, 2 mmol of 4-pyridine

carboxaldehyde was added and the reaction mixture stirred at

[11] M. D. Godbole, A. C. G. Hotze, R. Hage, A. M. Mills, H. H. Kooijman, A. L.

Spek, E. Bouwman, Inorg. Chem. 2005 , 44 , 9253–9266.

[

12] W. Adam, H. U. Humpf, K. J. Roschmann, C. R. S. Mçller, J. Org. Chem.

001 , 66 , 5796–5800.

13] J. G. de Vries, S. D. Jackson, Catal. Sci. Technol. 2012 , 2 , 2009–2009.

[

[14] P. W. N. M. van Leeuwen, Appl. Catal. A 2001 , 212 , 61–81.

15] C. Baleiz¼o, H. Garcia, Chem. Rev. 2006 , 106 , 3987–4043.

16] K. K. Bania, G. V. Karunakar, B. Sarma, R. C. Deka, ChemPlusChem 2014 ,

[

7

9 , 427–438.

[

17] C. Li, H. Zhang, D. Jiang, Q. Yang, Chem. Commun. 2007 , 547–558.

[18] K. K. Bania, D. Bharali, B. Viswanathan, R. C. Deka, Inorg. Chem. 2012 , 51 ,

657–1674.

758C for 16 h to obtain compound 2, pyridine carboxaldehyde

[

19] D. J. Cole-Hamilton, Science 2003 , 299 , 1702–1706.

modified MCM-41. In separate experiments solutions of com-

plex 1 and complex 2 (1 mmol) in ethanol (10 mL) were added

to pyridine carboxaldehyde modified MCM-41 (1 g), and the re-

sulting suspensions were refluxed for 48 h. The immobilized

catalysts 3 and 4 were filtered, washed thoroughly with anhy-

drous toluene, and extracted repeatedly with methanol and di-

chloromethane on a Soxhlet extractor until the extracted liquid

became colorless.

20] B. M. L. Dioos, P. A. Jacobs, Tetrahedron Lett. 2003 , 44 , 8815–8817.

21] P. Piaggio, P. McMorn, D. Murphy, D. Bethell, P. C. B. Page, F. E. Hancock,

C. Sly, O. J. Kerton, G. J. Hutchings, J. Chem. Soc. Perkin Trans. 2 2000 ,

2008–2015.

[22] D. J. Xuereb, R. Raja, Catal. Sci. Technol. 2011 , 1 , 517–534.

[23] J. M. Thomas, R. Raja, Acc. Chem. Res. 2008 , 41 , 708–720.

[24] M. E. Davis, Nature 2002 , 417 , 813–821.

[25] P. Piaggio, P. McMorn, C. Langham, D. Bethell, P. C. B. Page, F. E. Han-

cock, G. J. Hutching, New J. Chem. 1998, 1167–1169.

[

26] R. I. Kureshy, I. Ahmad, N. H. Khan, S. H. R. Abdi, K. Pathak, R. V. Jasra, J.

Catal. 2006 , 238 , 134.

27] D. Zhao, J. Zhao, S. Zhao, W. Wang, J. Inorg. Organomet. Polym. 2007 ,

The above-mentioned strategy to anchor the metal complexes

through apical co-ordination over MCM-41 was used taken be-

cause heterogenization can be also advantageous from the

point of view of catalyst stability, as immobilization frequently

improves catalyst stability compared to the homogeneous ana-

logues. This stabilization can be attributed to steric constraints

Am. Chem. Soc. 1995 , 117 , 4049–4056.

7 , 653–659.

28] R. I. Kureshy, I. Ahmad, N. H. Khan, S. H. R. Abdi, S. Singh, P. H. Pandia,

R. V. Jasra, J. Catal. 2005 , 235 , 28–34.

[29] J. A. Bonadies, M. L. Kirk, M. S. Lah, D. P. Kessisoglou, W. E. Hatfield, V. L.

Pecoraro, Inorg. Chem. 1989, 28, 2037–2044.

[

30] R. Schmidt, E. W. Hansen, M. Stoecker, D. Akporiaye, O. H. Ellestad, J.

31] J. A. Bonadies, M. J. Maroney, V. L. Pecoraro, Inorg. Chem. 1989 , 28 ,

[

15]

and site isolation that minimize complex degradation.

044–2051.

32] A. Domꢃnech, M. Alvaro, B. Ferrer, H. Garcꢄa, J. Phys. Chem. B 2003 , 107 ,

2781–12788.

General procedure for oxidation of styrene

[

33] J. Liu, F. Wang, S. Qi, Z. Gua, G. Wuc, New J. Chem. 2013 , 37 , 769–774.

34] A. N. Thadani, A. R. Stankovic, V. H. Rawal, Proc. Natl. Acad. Sci. USA

The enantioselective epoxidation of styrene was carried out

with synthesized homogeneous and heterogeneous catalysts

004 , 101 , 5846–5850.

(15 mg) using 5 mmol of the substrate in 10 mL of acetonitrile

[

35] J. Otera, K. Sakamoto, T. Tsukamoto, A. Orita, Tetrahedron Lett. 1998 , 39 ,

and 2 mL of 30% H O at 08C.

3201–3204.

2

&

ChemPlusChem 0000, 00, 0 – 0

Soc. 2000 , 122 , 9685–9691.

12

ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ÝÝ These are not the final page numbers!

Full Papers

[

36] R. Saito, S. Naruse, K. Takano, K. Fukuda, A. Katoh, Y. Inoue, Org. Lett.

006 , 8 , 2067–2070.

37] V. S. Chan, M. Chiu, R. G. Bergman, F. D. Toste, J. Am. Chem. Soc. 2009 ,

31 , 6021–6032.

38] W. Adam, C. M. Knoblauch, C. R. S. Mo1ler, M. Herderich, J. Am. Chem.

[42] a) P. Wasserscheid, W. Keim, Angew. Chem. Int. Ed. 2000 , 39 , 3772–3789;

Angew. Chem. 2000 , 112 , 3926–3945 ; b) J. Dupont, R. F. de Souza,

P. A. Z. Suarez, Chem. Rev. 2002 , 102 , 3667–3692.

[43] K. P. Bryliakov, D. E. Babushkin, E. P. Talsi, Mendeleev Commun. 1999 , 9 ,

29–31.

[39] T. Yamada, K. Imagawa, T. Mukaiyama, Chem. Lett. 1992 , 2109–2112.

[40] A. Dhakshinamoorthy, M. Alvaro, H. Garcia, ACS Catal. 2011 , 1 , 836–840.

[41] N. Audic, H. Clavier, M. Mauduit, J. C. Guillemin, J. Am. Chem. Soc. 2003 ,

Received: December 17, 2014

Revised: January 19, 2015

125 , 9248–9249.

Published online on && &&, 0000

ChemPlusChem 0000, 00, 0 – 0

13

ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

These are not the final page numbers! ÞÞ

FULL PAPERS

III

M. Mandal, V. Nagaraju, B. Sarma,

G. V. Karunakar, K. K. Bania*

Thanks for the support! Chiral Mn

Schiff base complexes were attached to

mesoporous silica MCM-41. Although

both the homogeneous and heteroge-

neous versions were effective catalysts

for the epoxidation of styrene, the sup-

ported catalysts retained their catalytic

activity for up to five cycles.

&& – &&

Enantioselective Epoxidation of

Styrene by Manganese Chiral Schiff

Base Complexes Immobilized on

MCM-41

&

ChemPlusChem 0000, 00, 0 – 0