Selective synthesis of primary methoxypropanol using clay supported tris(2,4-pentanedionato)zirconium(IV)

Article abstract of DOI:10.1016/j.molcata.2004.06.020

2-Methoxy-1-propanol has been selectively synthesized using clay supported tris(2,4-pentanedionato)zirconium(IV), prepared by facile ion-exchange process. The tris(2,4-pentanedionato)zirconium(IV) cation was synthesized by refluxing zirconium chloride with acetylacetone in chloroform. The heterogenized complex showed an increase in conversion rate and selectivity compared to the homogeneous one. Selectivity of 58% towards primary methoxypropanol was achieved under mild reaction condition. The clays selected as supporting materials were synthetic taeniolite and natural bentonite. It was observed that natural bentonite showed similar activity to synthetic taeniolite as a catalyst support.

Full text of DOI:10.1016/j.molcata.2004.06.020

Journal of Molecular Catalysis A: Chemical 221 (2004) 141–144
Selective synthesis of primary methoxypropanol using clay supported
tris(2,4-pentanedionato)zirconium(IV)
Yessi Permana, Shogo Shimazu∗, Nobuyuki Ichikuni, Takayoshi Uematsu
Department of Applied Chemistry and Biotechnology, Chiba University, 1-33 Yayoi, Inage, Chiba 263-8522, Japan
Received 5 February 2004; received in revised form 24 June 2004; accepted 28 June 2004
Available online 31 July 2004
Abstract
2-Methoxy-1-propanol has been selectively synthesized using clay supported tris(2,4-pentanedionato)zirconium(IV), prepared by facile ion-
exchange process. The tris(2,4-pentanedionato)zirconium(IV) cation was synthesized by refluxing zirconium chloride with acetylacetone in
chloroform. The heterogenized complex showed an increase in conversion rate and selectivity compared to the homogeneous one. Selectivity
of 58% towards primary methoxypropanol was achieved under mild reaction condition. The clays selected as supporting materials were
synthetic taeniolite and natural bentonite. It was observed that natural bentonite showed similar activity to synthetic taeniolite as a catalyst
support.
© 2004 Elsevier B.V. All rights reserved.
Keywords: Primary methoxypropanol; Tris(2,4-pentanedionato)zirconium(IV); Taeniolite; Bentonite; Clay supported catalysts.
1. Introduction
A previous work was able to achieve 51.1% of 2MP
using Co2(CO)8 catalyst [7]. Here, selectivity of 58%
towards 2MP was achieved under milder condition by
immobilized zirconium complex. The catalyst precursor
used in this work was tris(2,4-pentanedionato)zirconium(IV)
([Zr(acac)3]+, denoted as Zr-acac3) cation. The selected
clays were lithium taeniolite (Li[(Mg2Li)(Si4O10)F2])
and natural bentonite taken from East Java, Indone-
sia (K0.022Ca0.180Na0.008[(Si3.626Al0.374O10)(Al1.589Fe0.296
Ti0.023Mg0.284)(OH)2]), which is close to beidellite struc-
ture [8]. The synthesized zirconium complex was then im-
mobilized into taeniolite (C.E.C. = 268.2 meq./100 g) and
natural bentonite (C.E.C. = 82.0 meq./100 g), denoted as Zr-
acac3/TN and Zr-acac3/BNT, respectively by simple ion-
exchange method.
Acid catalysis has been of great interest in organic
synthesis. One of the challenges is designing acid catalysts
to synthesize primary methoxypropanol, which is used as a
versatile synthetic intermediate. Clay has drawn much atten-
tion as catalyst support due to its layered structure [1–6]. This
workdealswiththepillarizationofsyntheticandnaturalclays
with zirconium metal complex, forming organic/inorganic
hybrid acid catalyst for synthesizing 2-methoxy-1-propanol
(2MP) from propylene oxide (PO) and methanol (Scheme 1).
Theoretically, the use of any acid catalyst is supposed to result
in primary methoxypropanol. But in practice, the use of con-
ventional acid catalysts (e.g. TiCl4, FeCl3 and ZrCl4) always
results in higher selectivity towards 1-methoxy-2-propanol
(1MP)insteadof2MP. Theproblemofseparationalsomakes
those catalysts less favorable in industry.
∗
Corresponding author. Tel.: +81 43 290 3379; fax: +81 43 290 3401.
E-mail address: shimazu@faculty.chiba-u.jp (S. Shimazu).
Scheme 1.
1381-1169/$ – see front matter © 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2004.06.020

142
Y. Permana et al. / Journal of Molecular Catalysis A: Chemical 221 (2004) 141–144
Table 1
2. Experimental
Synthesis of primary methoxypropanol from propylene oxide and methanola
2.1. Synthesis of tris(2,4-pentanedionato)zirconium(IV)
chloride
Entry Catalyst Host
Conversion
(%PO)
Yield
(%2MP)
Selectivity
(%2MP)
b
1
2
3
4
5
6
7
8
9
TiCl4
–
–
–
–
–
–
17
64
21
23
50
3
30
5
5
8
11
57
58
31
5
45
50
32
26
39
34
58
58
42
20
38
A solution of acetylacetone (3.0 ml, 29.1 mmol) in chlo-
roform (30 ml) was added dropwise under inert atmosphere
into a solution of zirconium tetrachloride (2 g, 8.58 mmol) in
chloroform (30 ml) at 65 ◦C. The reaction mixture was re-
fluxed for 15 h at 65 ◦C. After the reflux, the solution was
immediately cooled down at 0 ◦C to let a crystallization pro-
cess occur and the chloroform was removed under vacuum.
Faint yellow powder of the product was recrystalized to re-
lease impurities and dried in vacuo overnight before analysis.
Analysis was conducted by UV–vis, FT-IR, 1H NMR and CH
Analysis.
FeCl3
NiCl2
CoCl2
ZrCl4
Zr-acac3
Zr-acac3 TN
41
>99
>99
74
24
29
Zr-acac3 BNT
Zr-acac3 Zeolitec
10
11
–
–
TN
BNT
11
a
Reaction condition: catalyst 1 mol%; propylene oxide 3.55 mmol;
methanol 0.247 mol; 60 ◦C; 1 atm; 24 h.
b
Room temperature; catalyst 10 mol%.
Natural zeolite containing 18% clinoptillolite.
[Zr(acac)3]Cl: yield 20%; UV λmax (MeOH), 308.0 (ε
8869) nm; δH (400 MHz, DMSO), 2.004 (s, 6H); 5.872
c
(s, 1H); FT-IR νmax, νC H (methyl), 2995–2922 (w); νC O
,
dition to 19% compared to the homogeneous state or 7%
compared to the previous work, of which the reaction was
run under the following conditions: 160 ◦C; 1 h; 40 kg/cm2
CO and 80 kg/cm2 H2 using Co2(CO)8 catalyst [7]. Actu-
ally, unless the heterogeneous and homogeneous catalysts
reached the same conversion rate, we could not directly com-
pare the selectivity between them. To ascertain the fact, the
time course of catalytic reactions which were run for 90 min
using Zr-acac3/TN and Zr-acac3/BNT was observed. The
length of time was sufficient to observe the change of selec-
tivity towards conversion since the catalytic reaction using
Zr-acac3/TN fully completed in 90 min (Fig. 1).
1572 (s); ␯
(enol form), 1533 (s) cm−1; Anal. Calc. for
C C
[Zr(acac)3]Cl: C, 42.49; H, 4.99. Found: C, 41.51; H, 5.13.
2.2. Catalyst preparation
Clays (lithium taeniolite from Topy Industry Co., Ltd.
and natural bentonite) of 0.1 g respectively were swollen
by deionized water (20 ml). The synthesized complex,
[Zr(acac)3]Cl (50% of the C.E.C.), was dissolved in deion-
ized water (20 ml) and added into the swollen clay followed
by stirring for 48 h at ambient atmosphere. The catalysts were
characterized using XRD, FT-R and UV–vis.
As we can observe from Fig. 1, both Zr-acac3/TN and
Zr-acac3/BNT reached selectivity of around 60% towards
2MP when their conversions were about 40% although the
selectivity of the homogeneous catalyst, Zr-acac3, was only
34% at the same conversion (Table 1, entry 6). Therefore we
2.3. Synthesis of 2-methoxy-1-propanol
Propylene oxide (0.25 ml, 3.55 mmol) was added into
methanol in excess (10 ml, 0.274 mol) containing 1 mol%
of the active catalyst. The catalyst was dried under vacuum
overnight prior to the catalytic experiment. The reaction mix-
ture was refluxed at 60 ◦C. Analysis was conducted by GC
and GC-MS.
3. Results and discussion
As shown in Table 1 (entries 1–5), conventional acid cat-
alysts (i.e. TiCl4, FeCl3, NiCl2, CoCl2 and ZrCl4) demon-
strated relatively low catalytic activity for the reaction of
propylene oxide with methanol. The low catalytic activity
was also observed in the case of the homogeneous zirconium
complex (Table 1, entry 6). Interestingly, the immobilization
of the zirconium complex into layered structures, such as
taeniolite and bentonite, was able to increase the selectivity
of catalysts as well as the conversion rate of the reaction as
they were employed in the catalytic reactions run for 24 h
(Table 1, entries 7 and 8). The selectivity increased to 58% in
the reaction using zirconium complex either supported into
taeniolite or bentonite, raising up under milder reaction con-
Fig. 1. Time course of the reactions. The selectivity of the reaction using
either Zr-acac3/TN or Zr-acac3/BNT was approximately 60% at which the
conversion rate was about 40%.

Y. Permana et al. / Journal of Molecular Catalysis A: Chemical 221 (2004) 141–144
143
conclude that the selectivity of the heterogenized complex
into clay was higher than that of the homogeneous one. The
enhancement in conversion rate can be explained from a des-
olvation phenomenon during an ion-exchange reaction of an
intercalated metal, most likely due to the ionic interaction be-
tween the metal and the negative charge of the layers. Such a
phenomenon was previously reported by Suzuki et al. [9,10]
in their work on ion-exchange properties on layered mate-
rials, such as hectorite, taeniolite and ␥-titanium phosphate.
Here, the desolvation property explained that during the cat-
alytic reaction, the number of methanol inside the interstitial
layer, solvating the zirconium complex, was lower than that
in homogeneous state due to weakened solvation ability of
methanol. However, it augmented the contact of the catalyst
precursor with the substrate, enhancing the conversion rate
of the reaction.
The use of natural zeolite as a catalyst support has also
been observed to evaluate the performance of selected clays
towards other supporting material (Table 1, entry 9). It was
found that in this work layered structures such as taeniolite
and bentonite were superior to the selected zeolite.
Fig. 2. Powder XRD patterns of zirconium complex intercalated taeniolite
and bentonite. (A) taeniolite (TN), (B) Zr-acac3/TN, (C) bentonite (BNT),
(D) Zr-acac3/BNT. (A) and (B) were measured at 20 mA, 25 kV; while (C)
and (D) at 20 mA, 50 kV.
To assure the better performance of the synthesized solid
acid catalyst compared to others, an experiment using well-
knownproticacidcatalyst, p-toluenesulfonicacidwascarried
out. The catalyst was able to achieve >99% conversion after
24 h reaction, showing equal activity to the synthesized solid
zirconium acid catalyst. However, the selectivity was still
lower than that of the clay supported zirconium complex. It
only achieved 55% of selectivity towards 2MP, and also left
a separation problem.
Fig. 3. Proposed arrangement of Zr-acac3 in the interlayer of taeniolite.
The clearance space of interstitial layer expanded to
0.57 nm in taeniolite and 1.0 nm in bentonite (Table 2), indi-
cated by the shift of (0 0 1) reflection to lower angle (Fig. 2),
due to the catalyst immobilization. In the case of taeniolite
pillarization, it was moderately close to the molecular size
of Zr-acac3 with the assumption that the complex was ar-
ranged as the structure model in Fig. 3. Such a limited in-
terlayer space can enforce the interaction between the active
site (zirconium atom) and substrate (PO), resulting in an in-
crease in the product selectivity. The catalysts swelled dur-
ing the catalytic reaction. The swelling expansion depended
on the solvent. Experiments of swelling the lithium taenio-
lite with various solvents showed that expansion was pro-
portional to the dielectric constant of the solvent (Table 3).
Lithium taeniolite was used instead of taeniolite supported
zirconium complex due to their expansion proportionality
and the ease of observing the change of expansion accurately
[4].
Among the solvents used here, methanol resulted in the
highest expansion. The expansion of the interlayer clay was
important since the molecular length of zirconium complex
was longer than the interlayer size of the clay in the dry state
(molecular length of Zr-acac3 is 0.98 nm). Several angstrom
units of expansion could increase mobility of the zirconium
complex slightly, which was important for the complex to get
the optimal arrangement in the interlayer. This would then
enable the active site to be easily reached by the substrate.
The formation of primary methoxypropanol occurred by
an SN1 mechanism, in which the catalyst first activated the
substrate through PO cleavage, forming a secondary car-
Table 2
Table 3
Loading amounts and XRD parametersa of clay supported Zr-acac3
XRD parametersa of lithium taeniolite swollen with various solvents
Sample
L.A.b (meq./100 g)
d001 (nm)
C.S.c (nm)
Solvent
εb
d001 (nm)
C.S. (nm)
Expansion (nm)
TN
Zr-acac3/TN
BNT
–
112
–
1.21
1.53
1.56
1.96
0.25
0.57
0.60
1.00
None
–
1.21
1.37
1.33
1.22
1.21
0.25
0.41
0.37
0.26
0.25
–
Methanol
Ethanol
Toluene
n-Hexane
a
33.0
25.3
2.4
0.16
0.12
0.01
–
Zr-acac3/BNT
78
1.9
a
Cu-K␣.
b
Loading amount.
Clearance space = d001 − 0.96 nm (thickness of layer).
Cu-K␣.
c
b
Dielectric constant of solvents [11].

144
Y. Permana et al. / Journal of Molecular Catalysis A: Chemical 221 (2004) 141–144
structures, such as taeniolite and bentonite. The synthesized
catalysts were suitable for selective synthesis of primary
methoxypropanol, which was difficult to be synthesized by
conventional acid catalysts. The immobilization of the com-
plex into the clays has increased catalytic performance as
well as the selectivity. The catalyst showed better perfor-
mance even under mild reaction condition compared to the
previous work. It was also observed that natural bentonite
possessed similar activity to taeniolite as a catalyst support
when they both were employed in the catalytic reaction run
for 24 h, indicating the potential use of natural bentonite from
Indonesia as a supporting material.
Acknowledgment
Scheme 2. Proposed catalytic cycle.
Table 4
This work has been supported by Grant-in-Aid for the
Scientific Research from the Ministry of Education, Culture,
Sports, Science and Technology, Japan (c) (No. 15550137).
We are indebted to Dr. Yateman Arryanto (Chemistry Depart-
ment, Gadjah Mada University, Indonesia) for permission of
using the natural bentonite and zeolite samples as well as the
basic data concerning the materials.
Solvent effect examined on homogeneous and heterogeneous zirconium acid
catalyzed reactiona
Entry
Catalyst
Solvent
Conversion
(%PO)
Selectivity
(%2MP)
1
2
3
4
5
6
7
8
9
a
Zr-acac3
Methanol
Ethanol
Toluene
n-Hexane
Methanol
Ethanol
Toluene
n-Hexane
Methanol
11
49
0
0
0
56
0
38
57
60
Trace
Trace
0
>99
2
Zr-acac3/TN
References
4
Trace
55
[1] E.M.F. Torres, E. Sham, P. Grange, Catal. Today 15 (1992) 515–526.
[2] S. Shimazu, W. Teramoto, T. Iba, M. Miura, T. Uematsu, Catal.
Today 6 (1989) 141–146.
[3] S. Shimazu, N. Baba, N. Ichikuni, T. Uematsu, J. Mol. Catal. A
182/183 (2002) 343–350.
Zr-acac3/BNT
Reaction condition: 60 ◦C; 1 atm; 90 min; catalyst 1 mol%; propylene
oxide 3.55 mmol; methanol 6.17 mmol (ignored if used as solvent); solvent
10 ml.
[4] S. Shimazu, K. Ro, T. Sento, N. Ichikuni, T. Uematsu, J. Mol. Catal.
A 107 (1996) 297–303.
bocation (Scheme 2). The other way of PO cleavage led
to a primary carbocation, which was less stable than the
secondary one. The selection of proper solvent was impor-
tant since protic and polar solvent would lead to the sol-
vation of methoxy nucleophile. The use of less polar sol-
vent also demonstrated low performance since the clay was
hardly swelled (Table 4), resulting in difficult re-arrangement
of the catalyst inside the layer to be easily reached by the
substrate.
[5] T. Sento, S. Shimazu, N. Ichikuni, T. Uematsu, J. Mol. Catal. A 137
(1999) 263–267.
[6] K. Ebitani, T. Kawabata, K. Nagashima, T. Mizugaki, K. Kaneda,
Green Chem. 2 (2000) 157–160.
[7] Mitsubishi Gas Chemical Co., Inc., Jpn. Kokai Tokkyo Koho, JP
58134044, 1983.
[8] For further reading on clay, see: A.C.D. Newman, Chemistry of Clay
and Clay Minerals, Harlow, Longman, 1987, pp. 81–83 (taeniolite);
T. Sudo, S. Shimoda, Clays and Clay Minerals of Japan, Elsevier,
1978, pp. 227–229 (bentonite); D.M. Moore, R.C. Reynolds Jr., X-
Ray Diffraction and the Identification and Analysis of Clay Minerals,
second ed., Oxford, 1997, pp. 138–197 (individual clay minerals).
[9] N. Suzuki, D. Yamamoto, N. Anaguchi, H. Tsuchiya, K. Aoki, Y.
Kanzaki, Bull. Chem. Soc. Jpn. 73 (2000) 2599–2603.
[10] N. Suzuki, M. Itoh, M. Igarashi, Y. Kanzaki, J. Ion Exchange 14
(2003) 169–172.
4. Conclusion
New organic/inorganic hybrid acid catalysts have been
prepared by immobilization of tris(2,4-pentanedionato) zir-
conium(IV) cation as the catalyst precursor into layered
[11] D.R. Lide (Ed.), CRC Handbook of Chemistry and Physics, 82nd
ed., CRC, 2001, pp. 6–153, 154, 162, 163.