Organosoluble zirconium phosphonate nanocomposites and their supported chiral ruthenium catalysts: The first example of homogenization of inorganic-supported catalyst in asymmetric hydrogenation

Article abstract of DOI:10.1039/c0dt00786b

In this article, we report the synthesis, structure, morphologies, and asymmetric catalytic properties of a series of novel organosoluble zirconium phosphonate nanocomposites and their supported chiral ruthenium catalysts, which have a good organosolubility (0.1-0.5 g mL^-1) in various solvents and mesoporous, filiform, and layered structures. Due to the organosoluble properties in various organic solvents, the first homogenization of zirconium phosphonate-supported catalyst was realized in the field of catalysis. In the asymmetric hydrogenation of substituted α-ketoesters, enantioselectivities (74.3-84.7% ee) and isolated yields (86.7-93.6%) were higher than the corresponding homogeneous Ru(p-cymene)(S-BINAP)Cl₂ due to the confinement effect caused by the remaining mesopores in the backbone of the zirconium phosphonate. After completing the reaction, the supported catalyst can be readily recovered in quantitative yield by adding cyclohexane and centrifugation, and reused for five consecutive runs without significant loss in catalytic activity.

Full text of DOI:10.1039/c0dt00786b

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Dalton
Transactions
Cite this: Dalton Trans., 2011, 40, 3325
www.rsc.org/dalton
PAPER
Organosoluble zirconium phosphonate nanocomposites and their supported
chiral ruthenium catalysts: The first example of homogenization of
inorganic-supported catalyst in asymmetric hydrogenation†
Taotao Chen, Xuebing Ma,* Xiaojia Wang, Qiang Wang, Jinqin Zhou and Qian Tang
Received 4th July 2010, Accepted 20th January 2011
DOI: 10.1039/c0dt00786b
In this article, we report the synthesis, structure, morphologies, and asymmetric catalytic properties of a
series of novel organosoluble zirconium phosphonate nanocomposites and their supported chiral
ruthenium catalysts, which have a good organosolubility (0.1–0.5 g mL-1) in various solvents and
mesoporous, filiform, and layered structures. Due to the organosoluble properties in various organic
solvents, the first homogenization of zirconium phosphonate-supported catalyst was realized in the field
of catalysis. In the asymmetric hydrogenation of substituted a-ketoesters, enantioselectivities
(74.3–84.7% ee) and isolated yields (86.7–93.6%) were higher than the corresponding homogeneous
Ru(p-cymene)(S-BINAP)Cl2 due to the confinement effect caused by the remaining mesopores in the
backbone of the zirconium phosphonate. After completing the reaction, the supported catalyst can be
readily recovered in quantitative yield by adding cyclohexane and centrifugation, and reused for five
consecutive runs without significant loss in catalytic activity.
counterparts.6 Based on the catalytic results, the reason that these
Introduction
two strategies, soluble polymer and dendrimer-supported catalysts
Homogeneousasymmetriccatalysiswasoneofthemostimportant
and inorganic porous solid-supported catalysts, were highlighted
developments in modern chemistry over the past several decades.
was the attractive homogeneous feature of soluble polymer and
However, the application of most homogeneous catalysts was
dendrimer-supported catalysts and the porous structure and
limited, partly due to problems of separation and recycling of
easy separation of inorganic porous solid-supported catalysts,
the catalysts. Several approaches have been used to heteroge-
respectively. The catalytic properties of heterogenized catalysts
nize homogeneous asymmetric catalysts, including attachment
sometimes improved compared with homogeneous ones, which
to porous inorganic-oxides, organic polymers, and dendrimers.1–6
was ascribed to the contribution of site-isolation and confinement
Two strategies include one using insoluble cross-linked polymers
effect owing to their ordered structures and uniform, suitable,
and tunable pore diameters.6,7–12 Therefore, from a practical point
of view, to seek the combination of the homogeneous manner
of soluble polymer-supported catalysts and the rigidly porous
feature of inorganic solid-supported catalysts with site-isolation
and confinement effect was interesting and important in the field
of catalysis.
Zirconium phosphonates [Zr(O3PR)2·nH2O, R = organic group]
are derived from the layered and porous structure of a-zirconium
phosphate [a-ZrP, Zr(HPO4)2·H2O], which is composed of the
uncharged lamellae.13 The appended organic moieties are arranged
in a dense, regularly ordered array that tailors the chemical
and steric environment between the inorganic layers. Due to
the versatile and tunable synthesis of appended phosphonates,
various zirconium phosphonates have been prepared to meet
different demands and are widely applied in probing DNA–protein
interactions,14–15 adsorption,16–17 membranes,18–20 catalysis,21–24 and
functional materials.25 However, zirconium phosphonates capable
of dissolving in polar or non-polar organic solvents are seldom
reported. We recently developed novel zirconium phosphonates
and inorganic porous materials as catalyst supports in the
heterogeneous phase and another using soluble polymers and
dendrimers to allow a catalytic reaction to be carried out under
homogeneous conditions. The key difference between these two
strategies is whether the catalytic process proceeds homogeneously.
Generally, most examples of supported catalysts tend to have
inferior catalytic properties to the corresponding homogeneous
College of Chemistry and Chemical Engineering Southwest University,
Chongqing, 400715, P. R. China. E-mail: zcj123@swu.edu.cn; Fax: +86-
23-68253237; Tel: +86-23-68252360
† Electronic supplementary information (ESI) available: 1H NMR, 13C
and 31P NMR of compounds 1a–e (Fig. S1–S5); the solid-state 13C and
31P NMR of compounds 2b (Fig. S6); the solution 13C and 31P NMR
of compounds 2b,c (Fig. S6); the thermogravimetric curves of the other
supports and their supported ruthenium catalysts (Fig. S9–S12); isotherms
and distributions of pore diameter of the other supports and their
supported ruthenium catalysts (Fig. S13–S17); particle analysis picture
of 2a (Fig. S18), the diameter of the particles (Table ST1) and the height
of the particles ranges (Table ST2); composition of as-prepared zirconium
phosphonate (Table ST3). See DOI: 10.1039/c0dt00786b
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 3325–3335 | 3325
©

View Article Online
Scheme 1 The synthetic route to organosoluble zirconium phosphonate-supported chiral ruthenium catalysts 3a–e.
with good organosolubility and layered, mesoporous structures
by covalently attaching aminoalcohol organic moieties,26 which
are considered suitable to meet the demand for homogeneous
inorganic support in the catalytic process. With this motivation,
we paved a way to prepare a series of organosoluble zirconium
phosphonate-supported ruthenium catalysts with different arm
lengths (n = 2–6) as the first homogeneous example of inorganic
solid-supported catalysts, from which a catalytic reaction can be
run under homogeneous (likely the best performing) conditions.
The catalyst itself can be easily precipitated by adding an
insoluble solvent and quantitatively recovered by solid–liquid
separation (Scheme 1). Such novel organosoluble, layered, and
porous functionalized zirconium phosphonate-supported chiral
ruthenium catalysts provided several key advantages.
Experimental
Chemicals were obtained from Aldrich or Alfa Aesar and used
as received unless otherwise stated. All solvents used were of
analytic grade. Br(CH2)nP(O)(OC2H5)2 (n = 2–6) were synthesized
according to ref. 27. TLC, where applicable, was performed
on pre-coated aluminium-backed plates and spots were made
visible by quenching ultraviolet (UV) fluorescence (l = 254 nm).
Fourier transform infrared spectra were recorded on a Perkin–
Elmer Model GX Spectrometer using a KBr pellet method with
polystyrene as a standard. Thermogravimetric analysis (TGA) was
performed on a SBTQ600 Thermal Analyzer (USA) with a heating
◦
rate of 20 C min-1 over a temperature range of 40–1200 ◦C
under flowing compressed N2 (100 mL min-1). H, 13C and 31P
1
NMR were performed on a Bruker AV-300 NMR instrument at
ambient temperature at 300, 75, and 121 MHz, respectively. Room
temperature 13C and 31P solid-state NMR spectra were recorded on
the same 300 MHz spectrometer equipped with a 7 mm variable-
angle sample spinning probe. All chemical shifts were reported
downfield in ppm relative to hydrogen, carbon, and phosphorus
resonance of TMS, chloroform-d1, and H3PO4 (85%), respectively.
The interlayer spacings were obtained on a DX-1000 automated
X-ray power diffractometer (XRD), using Cu Ka radiation and
internal silicon powder as a standard with all samples. The patterns
were measured between 2.00◦ and 20.00◦ (2q) with a step size of
1◦ min-1 and X-ray tube settings of 40 kV and 2.5 mA. C, H, and N
elemental analysis was obtained from an EATM 1112 automatic
elemental analyzer instrument (Thermo, USA). The morphologies
of as-synthesized samples were determined by a Hitachi model H-
800 transmission electron microscope (TEM) and a Multimode
SPM (Veeco Instruments Inc., USA) atomic force microscopy
(AFM). N2 adsorption–desorption analysis was carried out at
77 K on an Autosorb-1 apparatus (Quantachrome). The surface
areas of the zirconium phosphonates were determined by using the
Brunauer–Emmet–Teller (BET) equation, and the pore diameters
were estimated according to the BJH model. The ee values were
determined on an Agilent 6820 gas chromatograph (USA) using
First, unlike other insoluble inorganic supports, they are easily
soluble in various solvents such as THF, CH2Cl2, CHCl3, and
their alcohol-mixed solvents, although a lower organosolubility of
the supported ruthenium catalysts 3a–e than that of their parent
supports 2a–e was found. The organosolubility of the supported
ruthenium catalyst is beneficial for reactants to diffuse inside the
porous structure of the supporting matrix. Second, they possess
rigid porous, layered, andfiliformarchitecture, which are beneficial
in realizing the site-isolation and confinement effect in asymmetric
catalysis. They could retain their original rigidity of the porous
structure in the catalytic reaction process to protect the catalyt-
ically active sites from aggregation and degradation and permit
good access for reactants inside the matrix to the catalytically
active sites. Third, the as-synthesized zirconium phosphonate-
supported chiral ruthenium catalysts 3a–e are very insoluble
in solvents such as alcohol, cyclohexane, petroleum ether, and
hexane. Therefore, they could be easily separated from the reaction
mixture in quantitative yield via solvent precipitation by adding
insoluble solvents. Finally, enhancement of the enantioselectivity
and activity for the double C O bond, resulting from site-
isolation and a confinement effect owing to the immobilization
of the homogeneous ruthenium catalyst into the porous and
organosoluble solid support, was observed.
3326 | Dalton Trans., 2011, 40, 3325–3335
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
a Chiral Cyclodex-B capillary column (30 m ¥ 0.25 nm ¥ 0.15
128.6, 127.9, 127.3, 127.1, 126.7, 126.4 (Ph), 81.4 (C–OH), 74.3
(NCH), 61.7 (OCH2), 56.8 (NCH2), 39.4 (NCH3), 33.9 (PhCH2),
25.8 (PCH2, d, 1JC–P = 139.1 Hz), 20.5 (PCH2CH2), 16.8 (CH3).31P
NMR (CD3COCD3): d 32.8 (s).
◦
mm, Supelco), temperature program: 80 ◦C, 5 min; 5 C min-1,
◦
120 ◦C, 5 min; 5 C min-1, and 200 ◦C. The retention times of
(R) and (S)-ethyl 2-hydroxylpropionate were 7.10 and 7.32 min,
respectively.
◦
1d. 2.60 g, 63.1%, light yellow oily liquid. [a]D = +43.2 . 1H
20
NMR (CD3COCD3, TMS): d 7.89 (2 H, Ph, d, 3J = 7.4 Hz), 7.82
(2 H, Ph, d, J = 7.4 Hz), 7.56–7.33 (11 H, Ph, m), 4.25 (1 H,
NCH, d, J = 18.3 Hz), 4.20–4.18 (4 H, OCH2, m), 3.22 (1 H,
Synthesis of (S)-N-(x-alkyl diethyl phosphonate)-1,1,3-triphenyl-
N-methyl-2-aminopropanol (1a–e)
3
3
OH), 2.45–2.39 (2 H, NCH2, m), 2.36 (3 H, NCH3, s), 2.41–2.40
(2 H, PhCH2, d), 1.78–1.67 (2 H, NCH2CH2; 2 H, PCH2, m),
1.53–1.50 (2 H, PhCH2, m), 1.46–1.42 (6 H, CH3, m), 1.29–1.25 (2
H, CH2, m), 1.18–1.13 (2 H, CH2, m). 13C NMR (CD3COCD3): d
147.0, 146.5, 141.4, 129.2, 128.3, 128.1, 127.6, 127.4, 126.9, 126.3,
126.1, 125.7 (Ph), 80.6 (C–OH), 73.2 (NCH), 60.8 (OCH2), 56.5
(NCH2), 38.1 (NCH3), 33.0 (PhCH2), 28.0 (PCH2CH2, 2JC–P = 21.7
Hz), 27.4 (PCH2CH2CH2, JC–P = 16.9 Hz), 25.0 (PCH2, d, JC–P
= 139.3 Hz), 22.1 (PCH2CH2CH2CH2, 4JC–P = 4.8 Hz), 15.8 (CH3,
2JC-P = 5.1 Hz). 31P NMR (CD3COCD3): d 32.9–32.8 (m).
To a 30 mL acetonitrile solution of (S)-1,1,3-triphenyl-N-methyl-
2-aminopropanol (10.0 g, 31.6 mmol), potassium carbonate (4.4 g,
31.6 mmol) and catalytic amount of potassium iodide a 30 mL
acetonitrile solution of Br(CH2)nP(O)(OC2H5)2 (n = 2–6) (8.0
mmol) was added dropwise at four individual times (0 h, 10 mL;
24 h, 7 mL; 48 h, 7 mL; 60 h, 6 mL) and stirred at 55–60 ◦C
for 5–6 days with tracking by TLC. The reaction mixture was
evaporated under reduced pressure and 30 mL of deionized water,
and then extracted by CHCl3 (100 mL ¥ 3). The organic phase was
combined, washed with brine (30 mL ¥ 3), dried over anhydrous
Na2SO4, and concentrated under reduced pressure. The crude
product was purified by gradient silica gel column chromatography
using petroleum ether (60–90 ◦C)–ethyl acetate (v/v = 10 : 1) as
an eluent to remove the unreacted (S)-1,1,3-triphenyl-N-methy-
2-aminopropanol; then petroleum ether (60–90 ◦C)–ethyl acetate
(v/v = 5 : 1 and 3 : 1) and ethyl acetate as eluents to afford the oily
products 1a–e.
3
1
◦
20
1e. 2.50 g, 58.9%, light yellow oily liquid. [a]D = +55.8 . 1H
NMR (CD3COCD3, TMS): d 7.89 (2 H, Ph, d, 3J = 7.5 Hz), 7.83
(2 H, Ph, d, J = 7.5 Hz), 7.57–7.34 (11 H, Ph, m), 4.25 (1 H,
NCH, d, J = 3.4 Hz), 4.24–4.18 (4 H, OCH2, m), 3.24 (1 H,
3
3
OH), 2.46–2.39 (2 H, NCH2, m), 2.36 (3 H, NCH3, s), 1.86–1.52
(2 H, PCH2; 2 H, PhCH2; 2 H, N CH2CH2, m), 1.45 (6 H, CH3, t,
3JH–P = 1 Hz), 1.37–1.05 (4 H, CH2CH2, m), 1.18–0.95 (2 H, CH2,
m). 13C NMR (CD3COCD3, TMS): d 146.9, 146.5, 141.4, 129.1,
128.7, 128.3, 128.1, 127.6, 126.9, 126.3, 126.2, 125.7 (Ph), 80.5
(C–OH), 73.3 (NCH), 60.8 (OCH2), 56.6 (NCH2), 38.2 (NCH3),
1a. 2.08 g, 54.7%, light yellow oily liquid. [a]D = +30.0◦.
20
3
1H NMR (CD3COCD3, TMS): d 7.74 (2 H, Ph, d, J = 7.6 Hz),
3
7.68 (2 H, Ph, d, J = 7.6 Hz), 7.35–7.20 (9 H, Ph, m), 7.15–
1
33.0 (PhCH2), 30.1 (PCH2, d, JC–P = 16.3 Hz), 25.9, 25.0, 24.1,
7.10 (2 H, Ph, m), 4.35 (1 H, NCH, dd), 3.95–3.92 (2 H, NCH2,
22.2 (CH2CH2CH2CH2), 15.9 (CH3, d, 1JC–P = 5.5 Hz). 31P NMR
(CD3COCD3, 85% H3PO4): d 33.0–32.8 (m).
3
m), 3.44–3.26 (4 H, OCH2, m), 2.68–2.63 (2 H, PhCH2, d, J =
15.0 Hz), 2.24 (3 H, NCH3, s), 1.74–1.60 (2 H, PCH2, m), 1.21
3
3
(3 H, CH3, t, J = 6.9 Hz), 0.99 (CH3, 3 H, t, J = 6.9 Hz). 13C
NMR (CD3COCD3, TMS): d 149.2, 148.2, 142.6, 130.1, 128.9,
128.7, 128.4, 127.6, 126.9, 126.8, 126.7, 126.5 (Ph), 82.8 (C–OH),
72.5 (NCH), 61.4 (OCH2, d, 2JC–P = 27.8 Hz), 49.6 (NCH2), 41.3
Preparation of zirconium phosphonates (2a–e)
A mixture of phosphonates 1a–e (2.3 mmol), 40 mL of acetic acid
and 10 mL of hydrochloric acid (36%) was stirred at 80 ◦C for 24 h
then cooled to room temperature. Zirconium oxychloride (1.1 g,
3.4 mmol) was added dropwise to the reaction mixture in 10 mL of
deionized water and aged at room temperature for another 3 h. The
white solid was filtered, dispersed in 50 mL of water, and adjusted
to pH = 6–7 using sodium carbonate (0.1 mol L-1). The resulting
white mud cake was filtered and washed with deionized water,
until the chloride ion was not detected by ion chromatography,
then dried at 60 ◦C for 24 h under reduced pressure to afford the
zirconium phosphonates 2a–e in 50–60% yields.
1
(NCH3), 33.6 (PhCH2), 25.4 (PCH2, d, JC–P = 135.8 Hz), 16.7
(CH3, d, 3JC–P = 6.0 Hz). 31P NMR (CD3COCD3): d 33.6 (m).
◦
1b. 2.59 g, 66.3%, light yellow oily liquid. [a]D = +55.8 . 1H
20
NMR (CD3COCD3, TMS): d 7.92 (2 H, Ph, d, 3J = 7.7 Hz), 7.84
(2 H, Ph, d, 3J = 7.7 Hz), 7.51–7.37 (8 H, Ph, m), 7.33–7.28 (3 H,
Ph, m), 4.38 (1 H, NCH, dd), 4.15–4.01 (4 H, OCH2, m), 3.2 (1
H, OH), 2.93 (2 H, NCH2, m), 2.65–2.55 (2 H, PhCH2, m), 2.38
(3 H, NCH3, s), 1.59–1.57 (2 H, PCH2, m), 1.54–1.52 (2 H, CH2,
m), 1.48–1.31 (6 H, CH3, m). 13C NMR (CD3COCD3, TMS): d
147.7, 147.6, 141.9, 129.2, 128.0, 127.8, 127.3, 126.7, 126.0, 125.7,
125.6 (Ph), 81.9 (C–OH), 72.5 (NCH), 60.8 (OCH2), 53.9 (NCH2),
39.2 (NCH3), 32.3 (PhCH2), 20.8 (PCH2, d, 1JC–P = 138.8 Hz), 20.4
(CH2), 15.8 (CH3).31P NMR (CD3COCD3): d 34.2–34.0 (m).
Preparation of zirconium phosphonate-supported ruthenium chiral
catalysts (3a–e)
Ru(p-cymene)(S-BINAP)Cl2 (BINAP
=
2,2¢-bis(diphenyl-
1c. 2.61 g, 65.0%, light yellow oily liquid. [a]D = +155.8◦.
phosphino)-1,1¢-binaphthyl) was synthesized according to ref
28. A dried flask (10 mL) was charged with the zirconium
phosphonates 2a–e (34.9 mg) and Ru(p-cymene)(S-Binap)Cl2
(12.6 mg, 13.6 mmol), flushed three times with Ar2 atmosphere
and sealed. Then, 6 mL dichloromethane and 1 mL DMF were
added by syringe and stirred at 55–60 ◦C for 24 h with tracking by
TLC (chloroform–methane (v/v) = 20 : 1) until Ru(p-cymene)(S-
Binap)Cl2was not detected. The resulting reaction mixture was
evaporated under reduced pressure and 10 ml of methanol was
20
3
1H NMR (CD3COCD3, TMS): d 7.72 (2 H, Ph, d, J = 7.7 Hz),
3
7.66 (2 H, Ph, d, J = 7.7 Hz), 7.39–7.14 (11 H, Ph, m), 4.13 (1
H, NCH, d, 3J = 10.9 Hz), 4.03–3.96 (4 H, OCH2:, m), 3.04 (1 H,
OH), 2.28–2.26(2H, NCH2, m), 2.18 (3 H, NCH3, s), 1.56–1.45
3
(2H, PCH2, m), (2 H, PhCH2, d, J = 8.9 Hz), 1.56–1.51 (2 H,
PCH2, m), 1.47–1.45 (2 H, NCH2CH2, m), 1.28–1.24 (6 H, CH3;
2 H, CH2; 2 H, PhCH2, m;), 0.88–0.87 (2 H, CH2, m). 13C NMR
(CD3COCD3, TMS): d 147.8, 147.3, 142.3, 130.0, 129.7, 129.3,
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 3325–3335 | 3327
©

View Article Online
added, filtered, washed with methanol, and lyophilized at -50 ◦C
to give the brown powder solids 3a–e. About 3.0 wt% loading
amounts of ruthenium (II) in 3a–e were detected by ICP–AES
technique.
Table 1 Organosolubilities of zirconium phosphonates 2a–e and 3a–e in
various organic solvents (g mL-1)a
Entry CHCl3 CH2Cl2 THF Ethanol Hexane Mixed solventb
2a
3a
2b
3b
2c
3c
2d
3d
2e
3e
0.34
0.22
0.36
0.28
0.43
0.32
0.45
0.30
0.40
0.33
0.32
0.21
0.37
0.25
0.42
0.29
0.39
0.28
0.42
0.29
0.22
0.16
0.27
0.18
0.30
0.24
0.29
0.23
0.31
0.24
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.14
0.10
0.17
0.12
0.23
0.15
0.25
0.13
0.26
0.14
Asymmetric hydrogenation and recovery of catalyst
The supported ruthenium catalyst 3c (10.8 mg, 3.2 ¥ 10-3 mmol)
and PTSA (1.2 mg, 6.1 ¥ 10-3 mmol) was taken into a glovebox,
then 2 mL of CH2Cl2 and 13 mL of degassed MeOH was added
and stirred until the ruthenium catalyst 3c was absolutely dissolved
in the mixed solvents. Ethyl pyruvate (0.1484 g, 1.28 mmol) was
added to the resulting solution and transferred into a stainless-
steel autoclave. After purging four times with H2, the final H2
pressure was adjusted to 40 atm and stirred at 80 ◦C for 1 h. The
reaction mixture (about 15 mL) was concentrated under reduced
pressure to 2 mL and 8 mL of cyclohexane was added. Then
the yellow supported ruthenium catalyst 3c was quantitatively
recovered by centrifugation. The obtained centrifugate was further
purified by column chromatography to give the product, which was
directly analyzed by chiral GC (Chiral Cyclodex-B) to determine
the enantiomeric excess and conversion.
a To 1 mL of solvent was added 2a–e or 3a–e until no more dissolved, stirred
at 25 ◦C for 5 h, and filtered. The filtrate was evaporated and determined
by weight method. b CH3CH2OH–CH2Cl2 (v/v = 1/1)
chiral ruthenium catalysts 3a–e (0.16–0.33 g mL-1) were seen in
THF, CHCl3, and CH2Cl2 and immiscibility in ethanol, hexane,
petroleum ether, and cyclohexane. However, they could be soluble
in mixed ethanol–dichloromethane solvents, although ethanol was
a poor solvent. With the increase in the arm chain lengths (n)
from 2 to 6, the solubilities of 2a–e in various organic solvents
were strengthened. After immobilizing chiral Ru(p-cymene)(S-
BINAP)Cl2 into the backbone of the zirconium phosphonates
2a–e, the organosolubilities of the supported chiral ruthenium
catalysts 3a–e (0.16–0.33 g mL-1) decreased.
Results and discussion
1. Characterization of organosoluble zirconium phosphonate
nanocomposites and their supported ruthenium catalysts
Considering the different solubilities of the zirconium phospho-
nates 2a–e in various organic solvents, they were easily recovered
quantitatively by the “solubilization–precipitation” process. For
example, 0.21 g of zirconium phosphonate 2e was dissolved in
1 mL of a good solvent such as THF (v1), and a second poor
solvent such as petroleum ether (v2) was then added in a volume
ratio (v2 : v1) of 5 : 1. The zirconium phosphonate 2e (0.21 g)
precipitated out as a solid and was centrifugally separated from the
mixture. In the centrifugate, zirconium phosphonate 2e was not
determined by UV spectroscopy. For the precipitated zirconium
phosphonate 2e, the same chemical composition, d-spacing, and
surface morphology as its parent sample were observed, which
were ascertained by DTG, IR, PXRD, and TEM. However, the
surface area (21.1–24.6 m2 g-1), pore volume (0.13–0.15 cc g-1), and
average pore diameter (24.5–27.2 nm) of zirconium phosphonate
2e increased.
Zirconium phosphonates 2a–e (n = 2–6) with different x values
(0.44–1.50) were prepared by co-precipitation of zirconium (IV)
with the organophosphonate obtained from the hydrolysis of 1a–e
(Scheme 1). Due to the steric effects of bulky organic moieties, the
normal molar ratio of zirconium (IV) to organophosphonate (1 : 2),
whose theoretical formula was considered to be Zr(O3PR)2·nH2O
(R = organic group), cannot be obtained. The attached amount
of organic moieties (from x = 0.44 to 1.50) increased with the
increase in arm chain length (n) from 2 to 6, which could be
related to the flexibility and release of the interaction among the
organic moieties caused by the different arm chain lengths (n).
Due to the complexation of hydroxyl and amino groups, Ru(p-
cymene)(S-BINAP)Cl2 (1–5 wt%) could be immobilized onto the
backbone of the zirconium phosphonates 2a–e, which was verified
1
by 31P { H} NMR and determined by ICP.
1.1. Organosolubility. Generally, zirconium phosphonates
are seldom reported as being soluble in various organic solvents.29
Recently, a series of zirconium phosphonates organosoluble in
various solvents was reported for the first time, in which the
degree of solubility in organic solvents was significantly related
to the type and quantity of the appended functional groups such
as the hydroxyl, phenyl, and amino groups in the organic moieties
and the filiform morphology of the zirconium phosphonate.26 To
expand and support the interlayer d-spacings of the zirconium
planes to meet the demand for the good accessibility of the
catalytically active sites for reactants in the catalytic process,
zirconium phosphonates 2a–e [Zr(OH)4-2x(O3PR)x·nH2O, R =
organic group] with different organic arm chain lengths (n) were
prepared sequentially from n = 2 to 6.
1.2. Infrared spectroscopy and NMR. To gain an insight into
the covalent attachment of the functional organic moieties on the
frame of the zirconium phosphonates and immobilization of chiral
Ru(p-cymene) (S-BINAP)Cl2 on the supports, the infrared spectra
of 2a–e and 3a–e were recorded in the range of 4000–400 cm-1
(Fig. 1). The broad, strong and wide absorption band extending
from 3700 to 2500 cm-1 and centered at 3400 cm-1 was assigned to
the O–H stretching vibration, which was indicative of the presence
of hydrogen bonds between hydroxyl groups or adsorbed water
on the internal and external surfaces of 2a–e.30 The stretching
vibration of C–H in CH2 groups which emerged at 2940 cm-1 was
strengthened with the increase in the arm chain length (n) from 2 to
6. Themoderate absorption bands at 1640, 1478, and701 cm-1 were
attributed to the stretching vibration mode and flexural vibration
of the phenyl group, respectively. The strong and wide absorption
bands at about 1030 cm-1 were caused by Zr–O–P stretching
Table 1 showed that the excellent solubilities of zirconium
phosphonates 2a–e (0.22–0.45 g mL-1) and their supported
3328 | Dalton Trans., 2011, 40, 3325–3335
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
1
Fig. 1 IR and 31P{ H} NMR spectra of supports 2a–e and the supported ruthenium chiral catalyst 3c.
vibration.31 Zirconium phosphonate crystals of sufficient size for
structural determination by conventional X-ray diffraction (XRD)
techniques have yet to be grown. As such the solid-state 31 P and 13C
NMR are good tools aimed at resolving the structures of organic
moiety. With 2b as an example, the solid-state 31PNMR could be
easily obtained. However, the solid-state 13C NMR had very low
resolution power to allow the elucidation of the chemical structure
of 2b (in ESI†). Fortunately, the samples 2a–e were organosoluble
in various organic solvents such as CDCl3 and THF. As a result,
the first solution-phase 31P and 13C NMR spectra of zirconium
phosphonate were clearly observed, showing structural details of
the organic fragment. (shown in Fig. 1 and ESI†). From the data
of IR and NMR, it was concluded that the (S)-1,1,3-triphenyl-
N-methyl-2-aminopropanol organic moiety with different arm
lengths (n = 2–6), which acted as a functional complexing group,
was covalently immobilized on the backbone of the zirconium
phosphonate.
e were decomposed to the water-soluble mixture of complex
ZrF6 and organophosphonic acid. The content of zirconium
2-
(IV) and organophosphonic acid in the resulting solution could
be determined by quantitative colorimetric assay and solution 31
P
1
{ H} NMR, respectively. Therefore, based on the detected content
of zirconium (IV) ions and organophosphonic acid, the x values
in the empirical formulas of 2a–e [Zr(OH)4-2x (O3P R)x·nH2O, R
= organic group] were calculated to be 0.46, 0.91, 0.97, 1.34, and
1.48, respectively. The appended phosphonate (x) in the zirconium
phosphonate materials 2a–e can be usually evaluated by TGA,
although the incomplete combustion, which is suggested by the
typical pale gray color of the residue after TGA, leads to a
somewhat reduced weight loss and as a consequence a minor
under-prediction of the heavier phosphonate.25 From the DTG and
TG curves of the supports 2a–e (Fig. 2, just 2b as an example), it
can be deduced from the weight loss below 200 ◦C that 5.7, 3.5, 3.4,
2.5, and 1.3% of surface-bound or intercalated water was absorbed
in the internal pores, and the process was endothermic, verified by
the DSC curves. Accompanied by the endothermic peaks in the
DSC curves, the sharp and slow weight losses in the temperature
range of 200–700 ◦C and 700–1200 ◦C corresponded to the initial
and deep decomposition of the appended organic fragments,
respectively. The total weight losses of the appended organic
moieties in the supports 2a–e (n = 2–6) were 36.3, 56.2, 59.7, 67.4,
and 70.6% respectively. Based on the total weight losses of the
appended organic moieties in the temperature range 200–1200 ◦C,
the x values were calculated to be 0.44, 0.95, 1.00, 1.37, and 1.50,25
which were in the good agreement with the chemical analysis
results and those ascertained by elemental analysis. Furthermore,
from the total weight losses of the organic moieties of the supports
2a–e, the attachment of the organic moieties (x) increased with
organic arm length (n) from 2 to 6. It could be related to the
steric hindrance owing to the distance of the appended bulky b-
aminoalcohol group to the zirconium plane. When the organic
arm lengths (n = 2–6) become longer, they are more flexible and
it is easier to relieve the steric hindrance between neighboring b-
aminoalcohols, and then the large x value and even the formation
of Zr(O3PR)2 (R = bulky organic group) are more favored. After
After the immobilization of Ru(p-cymene)(S-BINAP)Cl2 (3.0
wt% Ru) on the zirconium phosphonates 2a–e, the characteristic
absorption of a phenyl group at 1640 cm-1 was significantly
strengthened, and the stretching frequency of the C–P bond at
1109 cm-1 re-emerged owing to the stretching vibration mode of
the C–P bond in the BINAP molecule. From the IR data of 2a–e, it
was demonstrated that Ru(p-cymene)(S-BINAP)Cl2 particles [1–5
wt% Ru (II)] could be successfully anchored into the porous frame
of zirconium phosphonates 2a–e, which was also readily verified
1
by solution 31P{ H}NMR spectroscopy and ICP determination.
1
In the solution 31P{ H} NMR spectra of sample 3c (see inset of
Fig. 1), the wide peaks in the range of d 5–25 ppm were caused by
the 31P absorption of the –PO3 group in zirconium phosphonate
2c, and the two doublets at d 25 and 42 ppm were attributed to
the supported Ru(p-cymene)(S-BINAP)Cl2.32
1.3. TGA. F- is a good ligand for transitional metals with a
very high oxidation state such as zirconium (IV). As such, upon
adding 2 mL of hydrofluoric acid (30%) and 2 mL of deionized
water to 2a–e (50 mg) and stirring for 6 h at room temperature
in a sealed plastic container, the zirconium phosphonates 2a–
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 3325–3335 | 3329
©

View Article Online
Fig. 2 The thermogravimetric curves of the support 2b and its supported ruthenium catalyst 3b.
immobilizing the chiral Ru(p-cymene)(S-BINAP)Cl2 (3.0 wt% Ru)
into the supports 2a–e, another 10.5 wt% organic weight loss
was observed, which corresponded to the composition of the
organic moieties in Ru(p-cymene)(S-BINAP)Cl2 (Fig. 2 just as an
example) and was near the theoretical amount of Ru(p-cymene)(S-
BINAP)Cl2.
to the farthest hydrogen atom in the organic functional group,
33
˚
and twice the hydrogen atomic radius (0.35 A). Comparing the
˚
experimental d-spacing values (17.1–20.2 A) with the calculated
˚
d-spacings, there were great deviations (above 1.0 A), which
illustrated that the organic chains were not perpendicular to the
◦
zirconium layer and were tilted by approximately 30 C.34 Upon
immobilizing of the Ru(p-cymene)(S-BINAP)Cl2 particles (3.0
wt% Ru), the d-spacings of the zirconium layers at the 0.4–2.0
A values unexpectedly decreased (Fig. 3).
1.4. Powder XRD. The layered structure and d-spacing of
the zirconium planes for crystalline and semi-crystalline zirconium
phosphonates can be determined from the 00n peaks in the powder
XRD pattern (via the Bragg equation, nl = 2d sinq). In Fig. 3, each
XRD pattern of the zirconium phosphonates at the small angle
range (q = 2–20◦) displayed a single and broad 001 peak (the lowest
angle diffraction peak) without the higher-order 00n peaks at the
larger angles. When the organic arm length (n) increased from 2 to
6, the d-spacings of the zirconium phosphonates in turn expanded
˚
1.5. Nitrogen adsorption–desorption. The BET surface areas,
average pore diameters and pore volumes of the zirconium
phosphonates and their supported ruthenium catalysts in solid
state are summarized in Table 2, and the representative nitrogen
adsorption–adsorption isotherms of 2d and 3d are shown in Fig. 4.
Like other zirconium phosphonates,35,36 these pore size distri-
butions (PSDs) of as-synthesized 2a–e suggested the existence
of large micropores of ca. 1.5 nm, together with a very non-
uniform mesopore, having several peaks at about 2–15 nm. Upon
immobilizing the chiral Ru(p-cymene) (S-BINAP)Cl2 particles
(3.0 wt% Ru), the nitrogen adsorption–desorption isotherms of
supported ruthenium catalysts 3a–e revealed relatively low surface
areas (SBET: 0.1–6.1 m2 g-1) and pore volumes (0.01–0.11 cc g-1)
˚
from 17.1 to 20.2 A. Assuming a perpendicular orientation and
bimolecular layers of the organic chains with an “end-to-end”
arrangement structure between neighboring zirconium planes, the
d-spacings of the zirconium phosphonates are predicted to be
˚
18.1, 24.5, 23.5, 30.7, and 34.5 A, by assumming the zirconium
˚
phosphate layer thickness (3.2 A) is twice the normal distance
(projection along the P–C bond axis) from the phosphorus atom
Table 2 Specific surface areas, average pore diameters, and pore volumes
of zirconium phosphonates 2a–e and their supported ruthenium catalysts
3a–ea
Surface area
(m2 g-1)b
Average pore diameter
(nm)c
Pore volume
(cc g-1)d
Entry
2a
3a
2b
3b
2c
3c
2d
3d
2e
3e
72.6
21.0
47.3
0.1
21.9
1.8
32.0
0.4
21.1
6.1
24.4
58.5
52.8
59.1
18.4
41.3
21.0
84.4
24.5
24.4
0.44
0.11
0.62
0.01
0.11
0.02
0.17
0.01
0.13
0.03
a The samples were degassed at 100 ◦C for 5 h. b Surface based on
multipoint BET method. c Pore size based on the desorption data using
BJH method. d Pore volume based on the desorption data of BJH method.
Fig. 3 The d-spacings of zirconium phosphonates 2a–e and supported
ruthenium catalyst 3b.
3330 | Dalton Trans., 2011, 40, 3325–3335
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Fig. 4 Isotherms and distributions of pore diameter of support 2d and supported ruthenium catalyst 3d.
compared to their parent supports 2a–e (SBET: 21–72.6 m2 g-1;
pore volumes: 0.11–0.62 cc g-1, respectively) in the solid state,
which demonstrated that the pores in 2a–e were occupied by
Ru(p-cymene)(S-BINAP)Cl2 particles. From Fig. 4 (support 2d
and catalyst 3d as a pair of comparative examples, others shown
in ESI†), micropores and some mesopores sharply disappeared
and only three typical mesopores at about 2–3, 4–6, and 10–
15 nm in 3a–e remained upon immobilizing the Ru(p-cymene)(S-
BINAP)Cl2 (3.0 wt% Ru) particles, which elucidated that chiral
Ru(p-cymene)(S-BINAP)Cl2 particles were successfully immo-
bilized into micropores and some mesopores in the zirconium
phosphonates 2a–e. Crucially the remaining mesopores in 3a–
e, at about 2–3, 4–6, and 10–15 nm, provided the diffusional
channels for substrates to smoothly access the catalytic sites of
the trapped Ru(p-cymene)(S-BINAP)Cl2 in the catalytic process.
However, data such as SBET and pore volume could not absolutely
represent the actual surface area and pore volume of the sample in
solution because the organosoluble zirconium phosphonates 2a–e
and ruthenium catalysts 3a–e in various organic solvents existed
in a fairly dispersed state owing to their organosolubilities.
ruthenium catalysts 3a–e (1.71–2.09 nm) determined by XRPD, it
can be deduced that each filiform chain consisted of the stacks of
4–5 zirconium layers. Filiform chains of ruthenium catalysts 3a–e
contained three typical mesopores at about 2–3, 4–6, and 10–
15 nm owing to the rigid backbone of the zirconium phosphonate
(Fig. 4). A similar structure could probably remain in the solution
state.
Due to the different organosolubility in various solvents, AFM
was used to illustrate the morphologies and particle sizes in
poor solvents such as alcohol and cyclohexane. The samples were
prepared as follows. A total of 1.0 mg of particles was dispersed
in 10 mL of ethanol, followed by ultrasonic stirring for 20 min. A
thin film of the suspension on mica was prepared by spin coating,
and the particles were left on the mica surface after evaporation
of the ethanol. The two and three-dimensional AFM images of
zirconium phosphonate 2a and supported ruthenium catalyst 3a
are shown in Fig. 6. Among the statistically analyzed 168 particles
of zirconium phosphonate 2a, the diameters ranged from 11.45–
242.39 nm, and the mean diameter was 87.82 nm with most particle
diameters in the range 50–110 nm. The height of the particles on
the mica surface ranged from 2.2–11.2 nm, and the mean height
was 6.7 nm. Upon immobilizing the Ru(p-cymene) (S-BINAP)Cl2
particles (3.0 wt% Ru), 16 particles of the supported ruthenium
catalyst 3a were statistically analyzed. Compared with its parent
support 2a, the larger diameters of the particles 3a in the 33.1–
384.1 nm range with a mean diameter 236.2 nm were obtained,
and the same phenomenon was observed in the height and mean
height of the particle ranging from 33.5–104.4 nm and 59.84 nm,
respectively. The TEM and AFM images confirmed that zirconium
phosphonate 2a and its supported ruthenium catalyst 3a existed
in different morphologies and particle sizes in good and poor
solvents owing to their different organosolubilities.
1.6. Surface morphology. Our aim in synthesizing this type of
organosoluble inorganic support was to realize the homogeneous
catalysis with supported chiral ruthenium catalysts 3a–e and their
recovery by solid–liquid separation. According to the relationship
between the catalytic activity and the physical property of
zirconium phosphonate, it was necessary that the morphologies
of ruthenium catalysts 3a–e be well clarified.26 Therefore, after
samples 2a–e and 3a–e were dissolved or dispersed in various
organic solvents such as cyclohexane, ethanol, CHCl3 and THF,
TEM and AFM were used to understand the surface morphologies
and particle sizes of 2a–e and 3a–e. After being well-dispersed in
THF for 3 h, sputtered over copper wire, and evaporated, the
TEM images were observed under an accelerating rate voltage of
200 keV. From Fig. 5, the spherical particles 2a–e, with uniform
a diameter of approximately 6–8 nm, were observed, and the
spherical particles closely connected each other to form the filiform
structure. Upon immobilizing the Ru(p-cymene)(S-BINAP)Cl2 (3
wt% Ru), the filiform morphologies of 3a–e were strengthened
to form filiform chains up to several hundred micrometers long
and 8–10 nm wide (Fig. 5). Based on the interlayer d-spacings of
2. Application in asymmetric hydrogenation of a-ketoesters
2.1. Optimized reaction conditions. To elucidate the applica-
tion of supported ruthenium catalysts 3a–e in homogeneous asym-
metric catalysis, the asymmetric hydrogenation of a-ketoesters
(R¢COCOR¢¢) was used as the model catalytic reaction to evaluate
the catalytic properties. Upon activation with p-toluene sulfonic
acid (0.5 mol%), ruthenium catalysts 3a–e (3.0 wt% Ru) were
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 3325–3335 | 3331
©

View Article Online
Fig. 5 TEM images of zirconium phosphonates 2b–e and their supported ruthenium catalysts 3b–e.
Table 3 The solvent effect on the catalytic performance in the homoge-
neous asymmetric hydrogenation of a-ketoestersa
In the following experimental, the supported chiral catalyst
3c was used as an example. Under reaction conditions such
as substrate/catalyst (based on Ru (II) mol%) = 130, 60 ◦C,
80 min, 40 atm H2, CH3OH–CH2Cl2 (v/v = 13/2) as the mixed
solvents and catalytic amount of PTSA, the loading amount
of Ru(p-cymene)(S-BINAP)Cl2 particles in the 1–8 wt% Ru (II)
range was screened. The best results with 36.7% conversion of
ethyl pyruvate, 29.7% yield and 58.8% ee of (S)-ethyl-2-hydroxyl-
propionate were obtained using ruthenium catalyst 3c at 3 wt%
loading amount of Ru(p-cymene)(S-BINAP)Cl2 particles. After
the reaction temperature and reaction time were optimized, it
afforded the best 54.8% conversion of ethyl pyruvate, 100%
Conv. Yield
(%)
Entry Solvent
R¢
R¢¢
(%)b % eec
1
2
3
4
5
6
7
8
CH2Cl2
CHCl3
THF
Me
Me
Me
Et
Et
Et
Et
Et
Et
Et
Et
Et
Et
Et
Me
i-Pr
10.5 9.9
58.2
56.0
—
5.3
—
4.4
—
MeOH–CH2Cl2 (v/v = 6.5) Me
EtOH–CH2Cl2 (v/v = 6.5) Me
2-PrOH–CH2Cl2 (v/v = 6.5) Me
EtOH–CH2Cl2 (v/v = 4.0) Me
EtOH–CH2Cl2 (v/v = 2.0) Me
EtOH–CH2Cl2 (v/v = 1.0) Me
83.8 10.2 68.1
81.6 81.6 71.6
74.5 74.5 71.9
85.0 60.5 73.4
97.2 87.4 73.5
9
100
92.6. 76.4
◦
10
11
12
13
14
15
16
EtOH–CHCl3 (v/v = 1.0)
EtOH–THF (v/v = 1.0)
Me
Me
78.3 78.1 69.8
56.8 45.4 68.7
selectivity, and 70.2% ee of ethyl-2-hydroxylpropionate at 80 C
for 1 h. The optimization studies of used amount of supported
catalyst 3c showed that it gave the best 85.5% conversion of ethyl
pyruvate, 100% selectivity, and 71.6% ee enantioselectivity of ethyl-
2-hydroxylpropionate at the ethyl pyruvate/catalyst 3c [based on
Ru (II) mol%] molar ratio (S/C) of 250 [0.4 mol% Ru (II)] (Fig. 7).
Under these optimized reaction conditions, the supported
ruthenium chiral catalysts 3a–e with different organic arm chain
lengths (n = 2–6) were examined for hydrogenation of ethyl
pyruvate (Fig. 8). Fig. 8 shows that the organic arm chain lengths
(n) played an important role in conversion and yield. When the arm
length (n) increased from 2 to 6, the conversion of ethyl pyruvate
increased from 19.3 to 99.4%, and the best 85.5% yield of ethyl-2-
hydroxylpropionate was obtained at n = 4. However, at n ≥ 3, the
arm lengths (n) of the appended organic moiety had no significant
influence on the enantioselectivities of ethyl-2-hydroxylpropionate
(70.2–74.2% ee), even at low yield and conversion (n = 2), which
was possibly related to the similar confinement effect owing to
EtOH–CH2Cl2 (v/v = 1.0) Me
EtOH–CH2Cl2 (v/v = 1.0) Me
EtOH–CH2Cl2 (v/v = 1.0) Me
100
100
t-Butyl 100
91.2 74.3
93.6 82.6
94.4 84.7
EtOH–CH2Cl2(v/v = 1.0)
EtOH–CH2Cl2 (v/v = 1.0) ClCH2 Et
BrCH2 Et
98.3 96.7 79.5
99.0 97.4 81.6
a Catalyst 3c, S/C = 250, 80 ◦C, 1 h, 40 atm. b isolated yield. c S-
configuration.
more active in homogeneous asymmetric hydrogenation of ethyl
pyruvate (92.6% yield and 76.4% ee, entry 9, Table 3) than their
parent Ru(p-cymene)(S-BINAP)Cl2 (58.1% yield and 69.0% ee).37
The effects of reaction temperature, reaction time, amount of
Ru(p-cymene)(S-BINAP)Cl2 used, organic arm length (n), and
reaction medium on the catalytic performance of the supported
homogeneous catalysts 3a–e in the asymmetric hydrogenation of
various a-ketoester derivatives were investigated.
3332 | Dalton Trans., 2011, 40, 3325–3335
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Fig. 6 The two-dimensional and three-dimensional AFM images of zirconium phosphonate 2a and its supported ruthenium catalyst 3a.
Fig. 8 The influence of the organic arm length (n) on the catalytic
performance.
Fig. 7 The influence of the amount of catalyst 3c used on the catalytic
performance in homogeneous asymmetric hydrogenation of ethyl pyruvate.
the similar remaining mesoporous structure determined by the
nitrogen adsorption–desorption isotherms.
obtained, in which the homogeneous catalysis was similarly carried
out as that in THF owing to organosolubility (0.15 g mL-1,
Table 1). Under the optimized reaction conditions (Table 3,
entry 9), a variety of substituted a-ketoesters were examined
for asymmetric hydrogenation with the supported ruthenium
catalyst 3c. All of the selected a-ketoesters (entries 12–16) were
hydrogenated to form chiral a-hydroxy esters with higher enantios-
electivities (74.3–84.7% ee) and isolated yields (86.7–93.6%) than
the homogeneous Ru(p-cymene)(S-BINAP)Cl2. The steric nature
of the substituted groups (R¢ or R¢¢) in the substrates had little
influence on the yields and enantioselectivities. The bulkier the
The effect of solvents on the catalytic property of ruthenium
catalyst 3c played an important role in the asymmetric hydrogena-
tion (Table 3).38,39 Low catalytic activities (below 10% conversion
and yield for 3a) in aprotic polar solvents such as CH2Cl2, CHCl3,
and THF were observed, in which ruthenium catalyst 3c can be
dissolved and homogeneous catalysis was realized. In entries 4–
11 in Table 3, the highest conversion of ethyl pyruvate (100%),
yield (92.6%) and enantioselectivity (76.4% ee) of (S)-ethyl-2-
hydroxylpropionate in the mixed ethanol–CH2Cl2 (v/v = 1) were
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 3325–3335 | 3333
©

View Article Online
R¢¢ groups (R¢¢ = Me, Et, i-Pr, and t-butyl) were, the higher the
enantioselectivities observed. This phenomenon could be related
to the confinement effect imposed by geometrical constraint
between the substrates and mesopores in 3c (Fig. 4), which
positively affected the yields and enantioselectivities (Table 3,
entries 13–16). To ensure that the reactivities and selectivities were
not produced by a mixture of zirconium phosphonate and Ru(p-
cymene)(S-BINAP)Cl2, Ru(p-cymene)(S-BINAP)Cl2 and zirco-
nium phosphonate 2c were added in situ in catalytic hydrogenation
under the same conditions. As expected, the inferior catalytic
property (62.4% yield and 70.4% ee) to that of ruthenium catalyst
3c was found in the asymmetric hydrogenation of ethyl pyruvate.
Furthermore, the recovered ruthenium catalyst significantly lost its
catalytic properties in the second run (14.2% yield and 65.4% ee),
indicating that homogeneous Ru(p-cymene)(S-BINAP)Cl2 was
not strongly bound to the support 2c.
2.2. The recovery and reuse of catalyst
Fig. 9 TEM image of recovered zirconium phosphonate 3c after fifth run.
Not only could the zirconium phosphonate-supported ruthenium
catalysts 3a–e be carried out in the homogeneous phase, they
could also be precipitated out and recovered quantitatively and
readily by adding poor solvents such as cyclohexane followed by
centrifugation. For example, in the catalytic hydrogenation of ethyl
pyruvate, the supported ruthenium catalyst 3c could be recovered
in quantitative yield and reused five times with conversions of
>99, 96, 92, 94, and 91% and enantioselectivities of 76.4, 76.0,
74.6, 71.3, and 72.2%, respectively for the five consecutive runs. To
elucidate the changes in the conversions and enantioselectivities,
the morphology and the leaching Ru (II) studies of the recovered
ruthenium catalysts 3c were investigated. After the ruthenium
catalyst 3c was precipitated out by adding a poor solvent such
as cyclohexane, it could be separated by centrifugation from the
reaction mixture. The resulting centrifugate was collected and the
leaching amount of Ru (II) in the centrifugate was monitored by
ICP–AES. No leaching of Ru (II) in the centrifugate was detected
by ICP–AES in the first run. However, there was a little leaching
of Ru (II) after the second run, and 98.4% of the loading Ru
(II) was still anchored in ruthenium catalyst 3c after the fifth
run, which indicated that the Ru (II) complex could be strongly
bound to the support. Unfortunately, the TEM image showed
that the filiform structure of zirconium phosphonate 3c after the
fifth run was tangled closer together and became wider (Fig. 9).
Therefore, it was concluded that the decrease in conversions and
enantioselectivities was mainly attributed to the change in the
morphology of the reused ruthenium catalysts 3c.
mixed ethanol solvents and the immiscibility in other solvents
such as petroleum ether and alcohol, the homogenization of an
inorganic supported ruthenium catalyst, which can be quantita-
tively recovered without significant loss of catalytic properties by
using a solid–liquid technique, was successfully realized for the first
time. In the asymmetric hydrogenation of substituted a-ketoesters,
higher enantioselectivities (74.3–84.7% ee) and isolated yields
(86.7–93.6%) than their parent homogeneous Ru(p-cymene)(S-
BINAP)Cl2 were observed, owing to the confinement effect of the
remaining mesopores in the zirconium phosphonate.
Acknowledgements
The work was supported by the National Science Foundation of
China (grants. 21071116) and Chongqing Scientific Foundation,
P.R. China (CSTC, 2007BB4368).
Notes and references
1 E. D. Jesus and J. C. Flores, Ind. Eng. Chem. Res., 2008, 47, 7968–7981.
2 A. F. Trindade, P. M. P. Gois and C. A. M. Afonso, Chem. Rev., 2009,
109, 418–514.
3 D. E. Bergbreiter, J. Tian and C. Hongfa, Chem. Rev., 2009, 109, 530–
582.
4 J. Lu and P. H. Toy, Chem. Rev., 2009, 109, 815–838.
5 M. Heitbaum, F. Glorius and I. Escher, Angew. Chem., Int. Ed., 2006,
45, 4732–4762.
6 C. E. Song, Annu. Rep. Prog. Chem., Sect. C, 2005, 101, 143–173.
7 D. J. Mihalcik and W. Lin, Angew. Chem., Int. Ed., 2008, 47, 6229–6232.
8 P. Yu, J. He and C. Guo, Chem. Commun., 2008, 2355–2357.
9 H. Zhang, Y. Wangb, L. Zhang, G. Gerritsen, H. C. L. Abbenhuis,
R. A. V. Santen and C. Li, J. Catal., 2008, 256, 226–236.
10 P. Yu, J. He, L. Yang, M. Pu and X. Guo, J. Catal., 2008, 260, 81–85.
11 M. Holbach and M. Weck, J. Org. Chem., 2006, 71, 1825–1836.
12 C. E. Song, D. H. Kim and D. S. Choi, Eur. J. Inorg. Chem., 2006, (15),
2927–2935.
Conclusion
Generally, zirconium phosphonates are not soluble in organic
solvents, which limits their applications in many fields. In this
paper, a series of organosoluble zirconium phosphonate nanocom-
posites and their supported ruthenium catalysts were readily
prepared by the reaction of various alkylphosphonates containing
a suitable quantity of amino and hydroxyl complex groups with
ZrOCl2 in acetic acid and hydrochloric acid medium. This novel
zirconium phosphonate nanocomposite possessed a layered and
porous structure and good organosolubility in organic solvents.
Based on the organosolubilities in THF, CH2Cl2, CHCl3, and
13 A. Clearfield and Z. Wang, J. Chem. Soc., Dalton Trans., 2002, 2937–
2947.
14 J. Monot, M. Petit, Sarah M. Lane, I. Guisle, J. Le´ger, C. Tellier, D. R.
Talham and B. Bujoli, J. Am. Chem. Soc., 2008, 130, 6243–6251.
15 B. Bujoli, S. M. Lane, G. Nonglaton, M. Pipelier, J. Le´ger, D. R. Talham
and C. Tellier, Chem.–Eur. J., 2005, 11, 1980–1988.
16 X. Shi, J. Liu, C. Li and Q. Yang, Inorg. Chem., 2007, 46, 7944–7952.
17 E. Brunet, M. J. d. l. Mata, O. Juanes and J. C. Rodr´ıguez-Ubis, Chem.
Mater., 2004, 16, 1517–1522.
3334 | Dalton Trans., 2011, 40, 3325–3335
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
18 Y. Li, T. Kunitake, Y. Aoki and E. Muto, Adv. Mater., 2008, 20, 2398–
2404.
19 R. O. Marcon and S. Brochsztain, Langmuir, 2007, 23, 11972–11976.
20 M. A. Petruska, B. C. Watson, M. W. Meisel and D. R. Talham, Chem.
Mater., 2002, 14, 2011–2019.
29 D. Chakraborty, V. Chandrasekhar, M. Bhattacharjee, R. Kra1tzner,
H. W. Roesky, M. Noltemeyer and H. G. Schmidt, Inorg. Chem., 2000,
39, 23–26.
30 F. Odobel, B. Bujoli and D. Massiot, Chem. Mater., 2001, 13, 163–173.
31 L. A. Vermeulen1and and S. J. Burgmeyer, J. Solid State Chem., 1999,
147, 520–526.
32 W. Baratta, G. Chelucci, S. Magnolia, K. Siega and P. Rigo, Chem.–
Eur. J., 2009, 15, 726–732.
33 J. C. Amicangelo and W. R. Leenstra, Inorg. Chem., 2005, 44, 2067.
34 W. R. Leenstra and J. C. Amicangelo, Inorg. Chem., 1998, 37, 5317–
5323.
21 A. Hu, H. L. Ngo and W. Lin, J. Am. Chem. Soc., 2003, 125, 11490–
11491.
22 X. Ma, Y. Wang, W. Wang and J. Cao, Catal. Commun., 2010, 11,
401–407.
23 X. Wu, X. Ma, Y. Ji, Q. Wang, X. Jia and X. Fu, J. Mol. Catal. A:
Chem., 2007, 265, 316–322.
24 B. Gong, X. Fu, J. Chen, Y. Li, X. Zou, X. Tu, P. Ding and L. Ma,
J. Catal., 2009, 262, 9–17.
35 K. Sarkar, T. Yokoi, T. Tatsumi and A. Bhaumik, Microporous
Mesoporous Mater., 2008, 110, 405–412.
25 S. L. Burkett, N. Ko, N. D. Stern, J. A. Caissie and D. Sengupta, Chem.
Mater., 2006, 18, 5137–5143.
26 X. Ma, J. Liu, L. Xiao, R. Chen, J. Zhou and X. Fu, J. Mater. Chem.,
2009, 19, 1098–1104.
36 J. Dong, L. Liu, J. Li, Y. Li, C. Baerlocher and L. B. McCusker,
Microporous Mesoporous Mater., 2007, 104, 185–191.
37 A. Wolfson, I. F. J. Vankelecom, S. Geresh and P. A. Jacobs, J. Mol.
Catal. A: Chem., 2004, 217, 21–26.
27 Z. Xu, X. Ma, Y. Ma, Q. Wang and J. Zhou, Catal. Commun., 2009,
10, 1261–1266.
28 K. Mashima, K. Kusano, T. Ohta, R. Noyori and H. Takaya, J. Chem.
Soc., Chem. Commun., 1989, 1208–1210.
38 T. Ohkuma, N. Utsumi, K. Tsutsumi, K. Murata, C. Sandoval and R.
Noyori, J. Am. Chem. Soc., 2006, 128, 8724.
39 X. Sun, W. Li, L. Zhou and X. Zhang, Chem.–Eur. J., 2009, 15, 7302–
7305.
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 3325–3335 | 3335
©