Cyclization of monoethanolamine to aziridine over Cs2O-P 2O5/SiO2

Article abstract of DOI:10.1007/s11164-012-0499-8

Several commercially available supports were examined for cyclization of monoethanolamine to aziridine, and SiO₂ was found to yield the best results. The obtained results indicated that selectivity of aziridine was mainly influenced by support. The catalysts were characterized by NH₃-TPD and XRD. It was found that SiO₂ with lower acidity could inhibit the intermolecular condensations, and thus favored the formation of aziridine. The Cs₄P₂O₇ phase was confirmed as the active site in the supported cesium phosphate catalyst. The reaction parameters were also optimized and a yield of 52% aziridine was obtained over 200 h. Thus, a continuous process for the cyclization of monoethanolamine to aziridine has been established. Springer Science+Business Media B.V. 2012.

Full text of DOI:10.1007/s11164-012-0499-8

Res Chem Intermed (2012) 38:1743–1750
DOI 10.1007/s11164-012-0499-8
Cyclization of monoethanolamine to aziridine
over Cs O–P O /SiO
2
2
5
2
Xiangjin Kong
•
•
Guangyuan Wang
•
•
Lei Li
•
Meng Sun
Xiaobao Du
Ligong Chen
Received: 9 January 2012 / Accepted: 20 January 2012 / Published online: 2 February 2012
Ó Springer Science+Business Media B.V. 2012
Abstract Several commercially available supports were examined for cyclization
of monoethanolamine to aziridine, and SiO was found to yield the best results. The
2
obtained results indicated that selectivity of aziridine was mainly influenced by
support. The catalysts were characterized by NH -TPD and XRD. It was found that
3
SiO with lower acidity could inhibit the intermolecular condensations, and thus
2
favored the formation of aziridine. The Cs P O phase was confirmed as the active
4
2 7
site in the supported cesium phosphate catalyst. The reaction parameters were also
optimized and a yield of 52% aziridine was obtained over 200 h. Thus, a continuous
process for the cyclization of monoethanolamine to aziridine has been established.
Keywords Cyclization Á Monoethanolamine Á Aziridine Á
Cs O–P O /SiO catalyst
2
2
5
2
Introduction
Aziridine is an important intermediate for the manufacture of fine chemicals and
pharmaceuticals [1–3]. Nowadays, most aziridine is still manufactured by the
Wenker reaction from monoethanolamine (MEA) [4], in which large amounts of
sulfuric acid and alkaline substances have to be employed leading to a serious
impact on the environment.
X. Kong
School of Pharmaceutical Science and Technology, Tianjin University,
Tianjin 300072, People’s Republic of China
G. Wang Á L. Li Á M. Sun Á X. Du Á L. Chen (&)
School of Chemical Engineering and Technology, Tianjin University,
Tianjin 300072, People’s Republic of China
e-mail: lgchen@tju.edu.cn
123

1744
X. Kong et al.
Recently, vapor phase cyclization of MEA to aziridine has been extensively
studied. WO - and Nb O -based catalysts have been developed for the reaction [5–
3
2 5
7
]; however, these catalysts did not exhibit high catalytic performance either on the
selectivity or lifetime. Some reports have proved that phosphates, especially cesium
phosphate, show a better catalytic performance than the above-mentioned catalyst
[
8–10]. Even so, the yield of aziridine over the cesium phosphate catalyst was not
perfect. Generally, a catalyst support or binder is always used in an industrial
process. It has been shown in the literature that these supports or binders could have
a key impact on the overall acidic properties of the catalyst and lead to different
catalytic performances [11]. So far, the influences of the support on the catalysts for
cyclization of MEA to aziridine have seldom been researched.
Thus, in this work, several commercially available supports were examined for
cyclization of MEA to aziridine, and the most satisfactory catalytic performance
was displayed by SiO . The catalysts were studied by NH -TPD and XRD, and the
2
3
process parameters were optimized. Thus, a continuous process for the cyclization
of MEA to aziridine has been established.
Experimental
Catalyst preparation
SiO pellets with a diameter of about 3 mm were purchased from Hailang silica-gel
2
drier factory, Qingdao, China. Shaped H-ZSM-5 support with a molar Si/Al ratio of
1
00 was obtained from Nankai University Catalyst, Tianjin, China. Al O pellets
2 3
with a diameter of about 3 mm were provided by the Aluminum Corporation of
China, Zibo, China. The other commercially available reagents were used without
further purification.
The catalysts used in this study were prepared by the impregnation method. For
example, (Cs O) –(P O ) /SiO2 (the suffix = the weight percentage of this
component in the catalyst) was prepared as follows. SiO pellets (100 g) were
2
5.5
2 5 1.8
2
impregnated with 180.0 mL aqueous solution of 5.4 g (NH ) PO Á3H O and 8.2 g
4
3
4
2
CsNO . After being impregnated at ambient temperature for 4 h, the catalyst was
3
dried at 120 °C for 8 h and calcined at 800 °C for 4 h. These catalysts were
activated at 400 °C in a stream of nitrogen (0.8 L/min) for 2 h before use.
Catalytic experiments
The reaction was performed in a fixed-bed reactor at atmospheric pressure, which
was loaded with 40.0 mL catalyst. MEA was dosed into the reactor with a syringe
pump. Nitrogen was employed as the dilution gas. The reaction mixture was
condensed in a cold trap (-2 °C) and analyzed without preliminary separation by
offline gas chromatograph (SE-30 capillary column: 60 m 9 0.32 mm, 1.0 lm film
thickness). Moreover, the composition of samples were identified by GC/MS (HP-1
capillary column: 30 m 9 0.25 mm, 0.2 lm film thickness) equipped with an ion
trap MS detector.
1
23

Cyclization of monoethanolamine to aziridine
1745
Catalyst characterization
Ammonia temperature programmed desorption (NH -TPD) was performed on a TP-
3
5
000 instrument with a thermal conductivity detector (TCD). X-ray diffraction
XRD) was recorded on a Rigaka D/max 2500 X-ray diffractometer with Cu-Ka
(
radiation (40 kV, 100 mA) in the range of 5–60°.
Results and discussion
Catalyst activity
As mentioned above, cesium phosphate has been reported to be effective for
cyclization of MEA to aziridine, and c-Al O was widely used as support in our
2
3
previous works [12, 13]. This encouraged us to carry out this reaction over (Cs O)–
2
(
(
P O )/c-Al O . The reaction mixture was identified by GC–MS, while aziridine
AZD), piperazine (PIP), pyrazine(PY), triethylenediamine (TEDA), 1-aziridineeth-
2
5
2 3
anol (AE) and amounts of C chemicals (acetaldehyde, ethylamine and acetonitrile)
2
were also detected. Based on the obtained results and the reported previous work [8, 9],
the reaction pathway was deduced and is shown in Scheme 1. It is obvious that the
selectivity of aziridine is quite poor due to the complicated reactions.
-2 H2
NH2
N
EA
ACN
NH
O
N
OH
AZD
AcH
AE
-
NH3
-
H O
2
R
N
H
N
H
N
OH
-
H O
-H O
2
2
H N
2
N
H
N
H
H N
OH
OH
2
MEA
N
PIP
-
3 H2
N
N
N
N
R
-
H O
2
N
H
N
TEDA
N
PY
R= methyl or ethyl
Scheme 1 The cyclization of monoethanolamine and side reactions
123

1746
X. Kong et al.
In order to improve the selectivity of aziridine, some commercially available
supports, such as HZSM-5 and SiO , were examined, and the results are presented
2
in Table 1. The experimental results showed that SiO displayed much better
2
selectivity of aziridine than the other supports. Therefore, SiO was chosen as the
2
support for the following study. To further understand the reasons for these results,
the catalysts were studied by NH -TPD and XRD.
3
Catalysts characterization
NH -TPD
3
In this study, extensive TPD of ammonia studies were performed to compare the
acidity of different supports and supported catalysts. As shown in Fig. 1, HZSM-5
and c-Al O exhibit both the weak (120–300 °C) and strong acid sites peak (above
2
3
4
50 °C), but for SiO , only the weak acid sites peak is observed. These indicated
2
that SiO possess the least acidity in the three kinds of support.
The accurate calculation of total acidity of these catalysts was difficult, since not
all TPD profiles displayed distinct NH desorption peaks. However, according to the
2
3
acidity of the supports, it is clear that the total acidic of the supported catalysts can be
deduced as (Cs O) –(P O ) /SiO \ (Cs O) –(P O ) /c-Al O \ (Cs O)5.5–
2
5.5
2
5 1.8
2
2
5.5
2
5 1.8
2
3
2
(
P2O5)1.8/HZSM-5. This result implied that. with the decrease of acidity in the
catalyst, selectivity of aziridine drastically increased from 0.5 to 51.0%; simulta-
neously, selectivity of PIP, TEDA, and PP declined sharply (see Table 1).
It has been reported that the desorption of amino compounds from the surface of
catalyst could be restrained by the strong acidity sites [14], and this may be the
reason why the multi-molecular condensation products such as PIP and TEDA were
readily formed over the strong acidity HZSM-5 and c-Al O -supported catalyst, and
2
3
aziridine was smoothly obtained over the weak acidity SiO -supported catalyst. In
2
Table 1 Effect of different supports on the catalytic performance
Catalyst
Conversion/ Selectivity/%
%
a
2
b
c
Aziridine 1-Aziridine-
ethanol
C
PIP
PY
Triethy-
lenediamine
Unknown
(
(
(
Cs
HZSM-5
Cs O)5.5–(P
c-Al
Cs O)5.5–(P
SiO
2
O)5.5–(P
2
O
2
O
2
O
5
5
5
) /
1.8
) /
1.8
) /
1.8
97.6
98.7
84.0
0.5
2.6
2.7
2.5 28.0
6.0 22.3 13.0 10.1
6.8 2.2 6.1 0.7
7.5 18.6
40.2
42.3
1.6
2
3.7
2 3
O
2
51.0
31.6
2
Reaction conditions: 420 °C; feeding rate: 0.1 mL/min; N
2
flow rate: 0.8 mL/min
a
The sum of acetaldehyde, ethylamine and acetonitrile
b
The sum of piperazine and N-alkylpiperazine
c
The sum of pyrazine and N-alkylpyrazine
1
23

Cyclization of monoethanolamine to aziridine
1747
Fig. 1 NH
a HZSM-5, b c-Al
d (Cs O)5.5–(P
e (Cs O)5.5–(P
f (Cs O)5.5–(P
3
-TPD curves for
, c SiO
1.8/HZSM-5,
1.8/c-Al
1.8/SiO
2
O
3
2
,
2
2 5
O )
2
2
O
5
)
2
O
3
,
(a)
2
2
O
5
)
2
(b)
(c)
(d)
(e)
(f)
200
300
400
500
600
700
800
Temperature(°C)
Fig. 2 X-ray diffraction
patterns of a (Cs
1.8/SiO , b SiO
2
O)5.5–
(P
2
O
5
)
2
2
*
o
Cristobalite low SiO
Cs P O
2
4
2 7
o
o
(a)
*
*
*
*
(b)
20
40
60
Two-Theta (deg)
other words, the SiO support with lower acidity favored the formation of aziridine
2
more than other supports.
XRD
The XRD patterns for SiO support and (Cs O) –(P O ) /SiO are shown in
5 1.8
2
2
5.5
2
2
Fig. 2. The typical diffraction lines at about 22.0°, 28.5°, 31.3°, and 36.1° in both of
the XRD curves can be assigned to cristobalite low SiO phase. Only diffraction
peaks of pyrophosphate (Cs P O ) crystals were detected for (Cs O) –(P O ) /
2 5 1.8
2
4
2
7
2
5.5
SiO . This demonstrated that most of the initial impregnated orthophosphate
2
(
calcination treatment, and thus pyrophosphate (Cs P O ) crystals is believed to be
Cs PO ) was transformed to the more stable pyrophosphate phase during the
3 4
4
2 7
the main active species in (Cs O) –(P O ) /SiO . This result was in accordance
2
5.5
2
5 1.8
2
with a previous report [15].
123

1748
X. Kong et al.
Optimization of the process conditions
Reaction temperature
The cyclization reaction was examined at a temperature range from 400 to 460 °C
and the results are listed in Table 2. As the temperature increased, selectivity of
aziridine slightly decreased from 52.2 to 46.4%, but the conversion of MEA
increased steadily from 74.4 to 98.0%. It was unexpectedly found that selectivity of
1
-aziridineethanol sharply decreased at the same time. This may be because the
higher temperatures favor desorption of aziridine from the weak acid sites of
catalyst so preventing the unexpected nucleophilic addition of aziridine to
acetaldehyde as mentioned in Scheme 1. However, higher temperatures may result
in other side reactions, thus 440 °C was chosen to be the reaction temperature.
Effect of N flow rate
2
Nitrogen was applied as the dilution gas in this study. The influences of dilution gas
on the catalytic performance are listed in Table 3. With an increase of N flow rate
2
from 0.8 to 1.6 L/min, selectivity of aziridine jumped from 50.7 to 59.0%, while
conversion of MEA reduced to 78.2 from 90.5%, making it obvious that 1.2 L/min
could be selected as the optimum N flow rate.
2
Lifetime of (Cs O) –(P O ) /SiO2
2
5.5
2 5 1.8
The lifetime test was carried out under the optimum conditions and the results are
shown in Fig. 3. During this process, conversion of MEA and selectivity of aziridine
remained at about 84.0 and 62.0%, respectively. These results demonstrated that
(Cs O) –(P O ) /SiO showed good stability for the cyclization of MEA to
2 5.5 2 5 1.8 2
aziridine.
Table 2 Effect of reaction temperature on catalytic performance
Temperature/°C Conversion/% Selectivity/%
a
2
b
c
Aziridine 1-Aziridine-
ethanol
C
PIP
PY
Triethy-
lenediamine
Unknown
4
4
4
4
00
20
40
60
74.4
84.0
90.5
98.0
52.2
51.0
50.7
46.4
33.4
31.6
24.6
16.7
7.2 1.4
6.8 2.2
14.1 2.2
24.7 1.1
4.3
6.1
5.5
6.5
0.3
0.7
0.2
0.9
1.2
1.6
2.7
3.7
Reaction conditions: feeding rate: 0.1 mL/min; N
2
flow rate: 0.8 mL/min
a
The sum of acetaldehyde, ethylamine and acetonitrile
The sum of piperazine and N-alkylpiperazine
The sum of pyrazine and N-alkylpyrazine
b
c
1
23

Cyclization of monoethanolamine to aziridine
1749
Table 3 Effect of N
2
flow rate on catalytic performance
Conversion/% Selectivity/%
Aziridine 1-Aziridine-
N
rate/L/min
2
flow
a
2
b
PIP
c
PY
C
Triethy-
lenediamine
Unknown
ethanol
0
1
1
.8
.2
.6
90.5
83.7
78.2
50.7
61.8
59.0
24.6
18.6
25.6
14.1 2.2
9.6 3.1
9.2 0.0
5.5
6.8
6.2
0.2
0.0
0.0
2.7
0.1
0.0
Reaction conditions: 440 °C; feeding rate: 0.1 mL/min
a
The sum of acetaldehyde, ethylamine and acetonitrile
b
The sum of piperazine and N-alkylpiperazine
c
The sum of pyrazine and N-alkylpyrazine
100
80
60
40
20
0
Conversion of MEA
Selectivity of aziridine
0
50
100
150
200
Time on stream (h)
Fig. 3 Service life of (Cs
2
2 5 2
O)5.5–(P O )1.8/SiO
Conclusion
(
aziridine. It was found that SiO with lower acidity favors the formation of aziridine
Cs O) –(P O ) /SiO2 was certified as suitable for cyclization of MEA to
2 5.5 2 5 1.8
2
more than other supports, and that the Cs P O crystal is believed to be the main
4 2 7
active species in (Cs O) –(P O ) /SiO . Conversion of MEA and selectivity of
2
2
5.5
5 1.8
2
the aziridine are deeply affected by reaction temperature and N flow rate. The
2
reaction performed well for 200 h. Therefore, this process may offer interesting
prospects for large-scale industrial production of aziridine from MEA.
Acknowledgment Financial support was provided by the National Natural Science Foundation of
China (Grant No. 20976123).
123

1750
X. Kong et al.
References
1. K. Park, S.W. Hong, W.H. Hur, M.Y. Lee, J.A. Yang, S.W. Kim, S.K. Yoon, S.K. Hahn, Bioma-
terials 32, 4951 (2011)
2
3
4
5
6
7
8
9
. Q.D. Hu, H. Fan, Y. Ping, W.Q. Liang, G.P. Tang, J. Li, Chem. Commun. 47, 5572 (2011)
. S. Ballereau, N. Andrieu-Abadie, N. Saffon, Y. Genisson, Tetrahedron 67, 2570 (2011)
. H. Wenker, J. Am. Chem. Soc. 57, 2328 (1935)
. D.L. Childress, W.V. Hayes. US 4301036 (1981)
. W.V. Hayesand, D.L. Childress. US 4289656 (1981)
. E.G. Ramirez. US 4337175 (1982)
. H. Tsuneki, Appl. Catal. A. 221, 209 (2001)
. Y.H. Sun, H.S. Yan, D.X. Liu, D.F. Zhao, Catal. Commun. 9, 924 (2008)
1
1
1
1
0. T. Oku, Y. Arita, H. Tsuneki, T. Ikariya, J. Am. Chem. Soc. 126, 7368 (2004)
1. H. L, Y.M. Zhou, Y.W. Zhang, L.Y. Bai, M.H. Tang, Ind. Eng. Chem. Res. 47, 8142 (2008)
2. M. Sun, X.B. Du, H.B. Wang, Z.W. Wu, Y. Li, L.G. Chen, Catal. Lett. 141, 1703 (2011)
3. J.C. Ma, S. Liu, X.J. Kong, X.P. Fan, X.L. Yan, L.G. Chen, Res. Chem. Intermed. (2011). doi:
10.1007/s11164-011-0454-0
1
1
4. C.H. Bartholomew, Appl. Catal. A 212, 17 (2001)
5. Y.H. Sun, H.S. Yan, D.X. Liu, D.F. Zhao, J. Mol. Catal. (China) 22, 142 (2008)
1
23