Direct Synthesis of Diallyl Carbonate Via Urea Transesterification with Allyl Alcohol Over Metal Chlorides

Article abstract of DOI:10.1007/s10562-019-02693-y

A promising technique was studied towards diallyl carbonate (DAC) manufacture via urea transesterification with allyl alcohol over several metallic chlorides. It was revealed that the intermediate allyl carbamate (AC) was first generated via urea mono-alcoholysis with high yield, and subsequently AC further reacted with another allyl alcohol molecule to produce DAC. All the metal halides were evaluated for the overall reaction and the reaction of AC and allyl alcohol, respectively. This work found that the catalytic ability of catalysts originated from metal ions. Importantly, the highest yield of DAC could be achieved using LaCl₃ as catalyst. In addition, the activation patterns of reactants were investigated using in situ FT-IR measurement. Based on the characterization results of X-ray power diffraction and elemental analysis of the lanthanum species, separated from the reaction conduced over LaCl₃, a possible reaction mechanism was speculated. Graphical Abstract: [Figure not available: see fulltext.].

Full text of DOI:10.1007/s10562-019-02693-y

Catalysis Letters (2019) 149:1067–1074
https://doi.org/10.1007/s10562-019-02693-y
Direct Synthesis of Diallyl Carbonate Via Urea Transesterification
with Allyl Alcohol Over Metal Chlorides
Dengfeng Wang1 · Xuelan Zhang1 · Hainan Luo1 · Shuwei Wei1 · Xueying Zhao1
Received: 12 October 2018 / Accepted: 26 January 2019 / Published online: 11 February 2019
© Springer Science+Business Media, LLC, part of Springer Nature 2019
Abstract
A promising technique was studied towards diallyl carbonate (DAC) manufacture via urea transesterification with allyl alcohol
over several metallic chlorides. It was revealed that the intermediate allyl carbamate (AC) was first generated via urea mono-
alcoholysis with high yield, and subsequently AC further reacted with another allyl alcohol molecule to produce DAC. All the
metal halides were evaluated for the overall reaction and the reaction of AC and allyl alcohol, respectively. This work found
that the catalytic ability of catalysts originated from metal ions. Importantly, the highest yield of DAC could be achieved
using LaCl3 as catalyst. In addition, the activation patterns of reactants were investigated using in situ FT-IR measurement.
Based on the characterization results of X-ray power diffraction and elemental analysis of the lanthanum species, separated
from the reaction conduced over LaCl3, a possible reaction mechanism was speculated.
Graphical Abstract
Without
catalyst
57.9% yield
O
O
+
OH
NH3
+
H2N
NH2
O
O
: LaCl3
+
CO2
Keywords Dially carbonate · Urea · Allyl alcohol · LaCl3
1 Introduction
CO2 as starting feedstock, have attracted increasing attention
from researchers [4, 5]. To date, several methodologies for
the synthesis of organic carbonates have been proposed. As
the oldest method, the process based on the use of phosgene
was phased out owing to high toxicity of phosgene [6]. Oxi-
dative carbonylation of alcohol [7] and transesterification
alcohol with alkyl carbonate [8–10] were well known routes
with high selectivity to the target product. However, they
were limited severely by corresponding disadvantages such
as hash reaction conditions or high investment cost which
inhibit their expansion of production to desired scale. Cur-
rently, an attractive method that synthesis of organic car-
bonates using CO2 as carbonylation agent is investigated
widely because that CO2 is a readily available and environ-
mentally acceptable material with low-price [11]. Neverthe-
less, this process was far from industrial application owing
Recently, owing to the continuous depletion of fossil based
resources and increasing CO2 concentration in the atmos-
phere, it is highly desirable to explore a novel strategy to
produce more renewable fuels and chemicals [1–3]. In this
aspect, organic carbonates that could be synthesized using
* Dengfeng Wang
bitwdf@163.com
* Xuelan Zhang
mhszxl@sxicc.ac.cn
1
College of Chemistry, Chemical Engineering and Materials
Science, Zaozhuang University, Zaozhuang 277160,
People’s Republic of China
Vol.:(0123456789)
1 3

1068
D. Wang et al.
to its poor thermodynamics because of highly stable struc-
ture of CO2 [2]. In order to avoid such negative effect, an
alternative route is the transesterification urea with alcohol.
Urea was mainly produced by the reaction of CO2 and NH3,
namely Bosch–Meiser process [12]. Therefore, it is gener-
ally reported that the production of organic carbonate by
urea alcoholysis is a green process of indirect utilization
CO2. Furthermore, the by-product NH3 can be recycled for
the production of urea via reaction with CO2, which ren-
ders urea alcoholysis an atom-efficient technique [5, 13].
So far, the manufacture of various kinds of alkyl and cyclic
carbonates like dimethyl carbonate (DMC) [14], diethyl
carbonate (DEC) [15], ethylene carbonate [16], propylene
carbonate [17] and glycerol carbonate [18] has been suc-
cessfully achieved via this method. Unfortunately, there is
very few document reported on the green synthesis of dially
carbonate (DAC), including several active basic groups such
as CH2=CH–CH2–, CH2=CH–CH2–O–, and –CO–. Thus,
DAC could be acted as green agent for manufacture of poly-
carbonates and nontoxic intermediates for organic synthesis.
Moreover, its potential application in solvent, agrochemical
and pharmaceutical industry has been extensively demon-
strated [19]. In addition, thanks to its higher flash point and
larger molecular weight than that of DMC and DEC, DAC
can be applied as an ideal electrolyte for lithium ion bat-
teries at relative high temperature [20]. Considering these
aspects and the advantage of urea alcoholysis, it is of great
importance to explore catalyst to promote the DAC produc-
tion from urea and allyl alcohol effectively.
DAC with low-price, but also make the catalytic mechanism
of urea alcoholysis over LaCl3 clear.
2 Experimental
2.1 Catalyst Preparation
MgCl2, CuCl2, Cu(NO3)2, LaCl3, CeCl3 and La(NO3)3
catalysts were acquired by evacuating MgCl2·6H2O,
CuCl2·2H2O, Cu(NO3)2·6H2O, LaCl3·7H2O, CeCl3·7H2O
and La(NO3)3·6H2O at 80 °C in vacuum oven under 5 kPa
overnight, respectively. Other catalytic samples were bought
commercially without pre-treatment.
2.2 Catalytic Activity Test
The reaction of allyl alcohol and urea were evaluated in a
50 mL autoclave with a reflux column. Typically, after 1.0 mol
allyl alcohol, 0.1 mol urea and 0.2 g catalyst were charged into
the reactor, the system was heated to 180 °C in a short time and
the reaction was conducted for 6 h under magnetically stirring.
When the reaction was finished, the product was taken out
from the autoclave. Finally, after the catalyst was separated by
centrifugation, the liquid product was analyzed on a GC-950
gas chromatograph with an FID detector and a capillary col-
umn. Moreover, the chemical structures of the liquid product
were measured using a GC–MS (Agilent 7890-700B) with
HP-5MS capillary column (30 m). Here, the yields of DAC (or
AC) were calculated by the following formulas, respectively:
In recent years, several heterogonous catalytic proce-
dures have been employed to realize the effective synthesis
of organic carbonates with pure metal oxides [21], mixed
oxides [22] and supported zinc catalysts [23]. On the other
hand, it was also reported that the urea alcoholysis towards
organic carbonates synthesis took place in two steps. In the
first step, the carbamate was generated fast; nevertheless,
the further reaction to the carbonate was very difficult [21,
24]. To the best of our knowledge, the process for production
of DAC via trasesterification between urea and allyl alco-
hol was not clarified yet. Furthermore, the reaction mecha-
nism and characterization of the active species, especially
in homogeneous catalytic systems for the DAC production
was still unknown.
DAC (mol)
DAC yield =
× 100%
starting urea (mol)
AC (mol)
AC yield =
× 100%
starting urea (mol)
In case of the reaction between allyl carbamate (AC) and
allyl alcohol, 0.2 g catalyst, 1.0 mol allyl alcohol and 0.1 mol
AC were added into the reactor. The reaction procedure was
similar to that of reaction between urea and allyl alcohol. After
the desired time of reaction, the liquid products were deter-
mined on GC. Accordingly, the AC conversion and DAC yield
were determined by these two equations, separately:
(starting AC − AC) (mol)
Herein, we experimentally studied the homogeneous syn-
thesis of DAC from urea and allyl alcohol using various metal
chlorides as catalysts. The reaction process was confirmed over
LaCl3 owing to its high catalytic ability. Then, the model cata-
lyst before and after the reactions was characterized by X-ray
power diffraction (XRD) and elemental analysis techniques.
To gain precise insights into the role of LaCl3 in this reaction,
the in situ FT-IR experiments were conducted. Finally, a pos-
sible reaction mechanism was disclosed. Therefore, this work
not only open a new way for green and effective synthesis of
AC convertion =
× 100%
starting AC (mol)
DAC (mol)
DAC yield =
starting AC (mol)
2.3 Characterization
× 100%
XRD patterns were collected using a Rigaku Miniflex dif-
fractiometer at 40 kV and 30 mA with Cu Kα radiation.
1 3

Direct Synthesis of Diallyl Carbonate Via Urea Transesterification with Allyl Alcohol Over…
1069
The quantitative analysis of lanthanum was performed on
an inductively coupled plasma (ICP) emission spectroscopy
(Perkin-Elmer ICP OPTIMA-3000). With the respect to
the nitrogen and hydrogen content in catalyst, the data was
acquired on an Elemental Analyzer (Vario EL III).
90
80
70
60
50
a
AC
DAC
The in situ FT-IR measurements were conducted using a
closed stainless steel FT-IR cell. The spectra were obtained
using a Nicolet Magna 550 Fourier-transform infrared spec-
trometer. The scanning region was 4000–400 cm−1 with a
resolution of 4 cm−1. Here, four kinds of experiments were
conducted, described as follows, respectively: a), allyl alco-
hol adsorption on LaCl3: 5 mg LaCl3 and 100 mg KBr were
ground finely, and then 4 mg of mixing sample was pressed
into 13 mm discs and placed into the cell. Subsequently,
20 µL allyl alcohol was added into the cell. In advance, the
sample should be pre-treated in vacuum for 2 h to elimi-
nate the adsorbed gaseous impurities. The cell was heated
to the desired temperature for FT-IR characterization. b),
urea decomposition on LaCl3: first, 13 mg LaCl3 and 2 mg
urea were mixed together. Then, 5 mg of mixture was made
into a ϕ 13 mm pellet with 100 mg KBr, and finally put
into the apparatus. The FT-IR spectra were obtained as the
cell was heated from room temperature to 180 °C. c), AC
adsorption on LaCl3: the analysis method and procedure of
AC adsorption was similar to that of urea, except that AC
was used instead of urea. d), co-adsorption of AC and allyl
alcohol on LaCl3: 13 g catalyst and 2 mg AC were mixed
together. Afterwards, a disc with 13 mm was obtained by
grounding 5 mg such mixture and 100 mg fresh dried KBr.
Subsequently, the sample was placed into the FT-IR cell and
pretreated under vacuum for 3 h. After various volumes of
allyl alcohol were introduced into the apparatus, the FT-IR
spectra were recorded at 180 °C.
4
2
0
1
2
3
4
5
6
7
Reaction time (h)
b
AC
DAC
70
60
50
40
30
20
10
1
2
3
4
5
6
7
Reaction time (h)
Fig. 1 Influence of reaction time without (a), and with LaCl3 catalyst
(b). Reaction conditions: 0.1 mol urea, 1 mol allyl alcohol; catalyst
amount 0.2 g; 180 °C; 6 h. Product selectivies were based on urea
production of DAC between urea and allyl alcohol could be
divided into two steps (Scheme 1). Initially, a large amount
of the intermediate AC was formed, and this step was fast.
Nevertheless, the conversion of AC to DAC was difficult,
and it only occurred efficiently over proper catalyst. Herein,
the second step was the rate-control step for the DAC syn-
thesis. The remarkable catalytic performance of LaCl3 could
be ascribed to its high ability towards conversion of AC to
DAC.
3 Results and Disscusion
3.1 Catalytic Test
Here, the products in liquid phase were measured using
GC–MS and quantified by means of GC. The experimental
results displayed that the major components were allyl alco-
hol, AC and DAC. In order to identify the reaction pathway
between urea and allyl alcohol, the effect of reaction time
was evaluated without and with LaCl3, respectively, and the
comparative results are shown in Fig. 1. Evidently, AC could
be synthesized fast from urea and allyl alcohol even without
catalyst. By contrast, only trace amount of desired prod-
uct of DAC was generated although the reaction proceeded
more than 6 h. When the reaction was catalyzed by LaCl3,
the selectivity of AC decreased rapidly. During this proce-
dure, the selectivity of DAC rose sharply at the cost of AC
consumption. Therefore, it was reasonable to state that the
Table 1 summarizes the reactivity comparison of vari-
ous samples for the DAC manufacture from urea or from
AC. Only trace amount of DAC was formed in absence of
catalyst no matter the staring material was urea or AC. For
all the catalysts, the deviation of DAC yield from transes-
terification of urea and that from AC alcoholysis could be
neglected. As a result, the following discussion regarding
these catalysts was based on their catalytic performance
towards DAC synthesis from AC and allyl alcohol. It was
important to note that the DAC yield differed significantly
among these metal chlorides after 6 h reaction. NaCl, KCl,
MgCl2, CaCl2 and CuCl2 showed rather low activity. At the
same reaction conditions, the DAC yield increased sharply
1 3

1070
D. Wang et al.
O
H2
C
O
C
+
OH
-NH3
CH2
(1)
HN
C
O
H2N
O
C
H
H2N
H2N
NH2
AC
Urea
O
O
H2
C
H2
Catalyst
H2
C
H2
C
+
CH2
(2)
CH2
C
H
OH
C
H2C
CH2
O
C
H
C
H
O
O
C
H
DAC
Scheme 1 The reaction procedure for the synthesis of DAC from thansesterification between urea and allyl alcohol
Table 1 Yield comparison of DAC synthesized from urea and that
that the reactivity of these homogeneous catalysts closely
depended on their metal cations. To verify this hypoth-
esis, the catalytic abilities of La(NO3)3 were further inves-
tigated. As expected, this compound also gave a rather
high yield of DAC (54.8%). Meanwhile, the possibility of
NO3− devotion to the reactivity was also ruled out because
that both NaNO3 and Cu(NO3)2 were inactive for the DAC
synthesis (Entries 21–24). Therefore, it was ascertained
that the reactivity of these La-containing samples only
stemmed from trivalent lanthanum ions, and the reaction
hardly took place without La3+ species.
from AC
Entry
Catalyst
AC conversion/
yield (%)
DAC yield (%)
1
2
3
4
5
6
7
8
Nonea
Noneb
NaCla
NaClb
KCla
81.9
6.2
80.2
8.4
83.1
6.8
4.8
5.3
6.2
6.9
5.2
KClb
5.7
a
MgCl2b
69.2
19.3
66.3
14.5
77.6
9.4
53.7
48.4
43.2
64.6
32.8
74.1
33.7
70.5
77.4
4.3
12.4
15.5
10.4
11.8
6.8
MgCl2
a
9
CaCl2b
3.2 FT‑IR Spectra of Product of Interaction Between
Reactants and LaCl3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
CaCl2a
CuCl2b
CuCl2a
ZnCl2b
ZnCl2a
CeCl3b
CeCl3a
LaCl3b
LaCl3
7.2
To explore the possible reaction mechanism of the trans-
esterification between urea and allyl alcohol over LaCl3
catalyst, in situ FT-IR experiments were performed under
different conditions (see Fig. 2).
28.5
34.9
45.8
49.0
57.9
60.7
52.6
54.8
3.6
3.7
5.3
6.0
59.1
61.6
3.2.1 Allyl Alcohol Absorption Over LaCl3
a
La(NO3)3b
La(NO3)3
At first, the adsorption of allyl alcohol over LaCl3 was
investigated and the FT-IR spectra are given in Fig. 2a.
It was showed that there was no obvious change for allyl
alcohol at various temperatures, implying that allyl alcohol
did not seem to interact with LaCl3. Allyl alcohol gave
its characteristic bands at 3310 cm−1, which was associ-
ated with O–H stretching vibration. The bands at 2810 and
1646 cm−1 were ascribed to the C–H stretching vibrations.
Besides, the deformation vibrations attributed to –CH3
could be seen at 1420 and 1115 cm−1. The C–O starching
vibration at 1025 cm−1, as well as the bands attributed to
the vinyl groups (3086, 992 and 920 cm−1) were all in line
well with the standard infrared spectrum of allyl alcohol
[25]. Therefore, it was likely that allyl alcohol could not
be activated over LaCl3 in the present case.
a
NaNO3b
NaNO3
a
Cu(NO3)2b
75.7
7.4
26.5
76.2
Cu(NO3)2
Precipitatea
Precipitateb
aReaction conditions: 0.1 mol urea, 1 mol allyl alcohol; catalyst
amount 0.2 g; 180 °C; 6 h. DAC yields were based on urea
bReaction conditions: 0.1 mol AC, 1 mol allyl alcohol; catalyst
amount 0.2 g; 180 °C; 6 h. DAC yields were based on AC
to 34.9% and 49.0%, when ZnCl2 and CeCl3 were sepa-
rately applied as catalyst (Entries 14, 16). Notably, LaCl3
gave the best catalytic performance with 60.7% of DAC
yield (Entry 18). So, these catalytic test results indicated
1 3

Direct Synthesis of Diallyl Carbonate Via Urea Transesterification with Allyl Alcohol Over…
1071
920
1630
1670
a
992
1025
1115
b
2210
180oC
150oC
3310
3086
1420
130oC
100oC
80oC
1646
2810
180oC
150oC
120oC
80oC
30oC
30oC
3357
3458
3210
3600 3200 2800
1600
1200
800 3600 3450 3300 3150 2400 2100 1800 1500
-1
-1
Wavenumbers (cm
)
Wavenumbers (cm )
d
1760
c
1592
1737
1620
1710
2210
2220
180oC
150oC
100µL
80µL
60
µ
L
80oC
L
µ
40
3350
3455
20µL
1710
30oC
0 µL
1620
3600
3400
3200
2100 1750 1400
-1
2400
2200
1800
1600
-1
Wavenumbers (cm )
Wavenumbers (cm )
Fig. 2 FT-IR spectra evolution of allyl alcohol absorbed on LaCl3 (a), urea absorbed on LaCl3 (b), AC absorbed on LaCl3 (c) at different tem-
peratures; and AC with different amounts of allyl alcohol co-adsorbed on LaCl3 at 180 °C (d)
3.2.2 Urea Decomposition
that of urea almost disappeared. Considering the structure of
urea, it was reasonable to state that urea could be activated
and decomposition into isocyanic acid, releasing a NH3 mol-
ecule over LaCl3 (See Scheme 1). The similar discovery and
conclusion were also proposed by other researchers [28]. In
addition, it should be mentioned that with further increase
temperature, the intensity of HN=C=O decreased, imply-
ing that it was decomposed at a relative higher temperature.
To figure out how the reactant of urea was activated over
LaCl3, the FT-IR spectra of urea alternation over LaCl3 were
collected (see Fig. 2b). At the room temperature, the feature
bands of urea could be seen at 3458, 3357 and 3210 cm−1,
which were ascribed to N–H symmetry and anti-symme-
try stretching vibration, respectively. Moreover, the bands
assigned to C=O stretching and N–H bending vibration were
separately observed at 1670 and 1630 cm−1 [26, 27]. With
the rise of temperature, the intensities of aforementioned
bands tended to be weaker and weaker. Interestingly, a new
band related to N=C=O asymmetric stretching vibration
of isocyanate species, appeared at around 2210 cm−1 [28].
Moreover, its intensity increased gradually with the increas-
ing temperature. When the temperature reached 150 °C, the
band of isocyanate species reached its peak value, while
3.2.3 Interaction of AC and LaCl3
Figure 2c illustrates the FT-IR spectra of AC adsorption over
LaCl3 with increasing temperature. For pure AC, it exhibited
the absorptions at 3455 and 3350 cm−1, which were ascribed
to N–H antisymmetry and symmetry stretching vibrations,
respectively [29]. Because that some AC dimer could be
formed through connection of hydrogen bonds of monomer,
1 3

1072
D. Wang et al.
they showed red shifts to 3285 and 3220 cm−1, respectively
[30]. Besides, the C=O stretching vibration, denoted as
amide I band, was found at 1710 cm−1. Amide II band,
caused by NH2 bending vibration, appeared at 1620 cm−1.
As the temperature increased, blue shifts towards 3490 and
3370 cm−1 occurred for N–H stretching vibrations antisym-
metry and symmetry stretching vibrations, respectively. At
the same time, the bands attributed to AC dimer decreased,
which was an indicator of break of intermolecular hydro-
gen bonds. More importantly, the amide I and II bands had
a clearly blue shift towards 1737 cm−1 and a red shift to
1592 cm−1, respectively. These changes suggested that the
La3+ had coordinated with the nitrogen atom of AC [14, 31,
32]. In addition, it was interesting to see that a new absorp-
tion corresponding to isocyanate species was detected at
about 2220 cm−1 [28], accompanying a gradual increase
in intensity with increasing temperature. Meanwhile, the
feature bands above-mentioned attributed to pure AC all
showed a gradual decrease in intensity. All these phenomena
provided evidences that AC could be activated and decom-
posed into isocyanic acid over LaCl3 (see Scheme 2).
co-adsorption of them over LaCl3 was also studied to
elucidate the reaction procedure more clearly and reason-
ably. Here, we mainly focused on the evolution of the band
belonged to N=C=O coordinated to LaCl3, the absorption
at around 1760 cm−1 corresponding to the C=O stretching
vibration of carbonate [33], as well as the amide I and II
bands of AC. Absence of allyl alcohol, AC was decom-
posed into isocyanic acid species at 180 °C to some extent.
Upon the addition of allyl alcohol, some obvious alterna-
tions could be observed. The absorption of La–N=C=O
decreased rapidly with the increasing amount of allyl
alcohol, and vanished as it reached 60 µL. Interestingly,
the bands attributed to AC at 1710 and 1620 cm−1 also
appeared. Thus, it was likely that the reaction for AC con-
version to isocyanic acid was reversible [14]. Therefore,
allyl alcohol would easily react with isocyanic acid to re-
generate AC because that there was a large amount of allyl
alcohol available in the system. Besides, amide I and of
amide II showed a red shift and a blue shift, respectively,
unambiguously demonstrating that the AC amino group
had be coordinated with La3+ via nitrogen atom [14, 31,
32]. Importantly, it was noteworthy that the characteristic
band of DAC appeared at around 1760 cm−1 when 60 µL
allyl alcohol was introduced into system. As the amount
of allyl alcohol further increased to 100 µL, the intensity
of DAC always monotonically rose, whereas that of AC
3.2.4 Co‑Adsorption of AC and Allyl Alcohol on LaCl3
Catalyst
The above results confirmed that the reaction of AC
and allyl alcohol was the rate-control step. Thus, the
O
C
NH
LaCl3
OH
- 3
HN
C
O
O
C
NH
3
OH
+ 3
O
O
OH
O
3NH3
O
N
N
O
O
N
LaCl3
NH3
O
H3N
3NH2COOCH2CH=CH2
LaCl3
NH3
3
O
O
AC
DAC
Scheme 2 The possible reaction mechanism for the DAC synthesis from AC and allyl alcohol over LaCl3
1 3

Direct Synthesis of Diallyl Carbonate Via Urea Transesterification with Allyl Alcohol Over…
1073
Table 2 Element analysis of the precipitate from the transesterifica-
consistently decreased, revealing the manufacture of DAC
from AC and allyl alcohol over the LaCl3 catalytic sample.
tion of urea and allyl alcohol over LaCl3
La
N
H
Determined (%)
47.2
46.8
14.1
14.2
2.9
3.0
3.3 Structure Change of LaCl3 Before and After
Reaction
Formula of La(NH3)3Cl3
(%)
It should be mentioned that LaCl3 could not dissolved
into alcohol solution completely at room temperate. Sur-
prisingly, it could dissolve into reaction system and the
mixture solution became transparent when reaction tem-
perature was over 100 °C. However, some yellow crystals
were generated when the reactor was put in the ice bath.
To gain the clue on the nature of active species, the yellow
crystal was recovered by centrifugation, and then washed
with ethanol several times to eliminate the organic spe-
cies. Figure 3 demonstrates the XRD patterns of the fresh
and recovered LaCl3. To our surprise, after one reaction
run, LaCl3 had converted to another species, which was
proved to be La(NH3)Cl3, according to the standard XRD
data base (PDF# 32-0429). Additionally, it was found that
this precipitant possessed greater quality than that of fresh
LaCl3. Furthermore, the experiment data of element analy-
sis of this sample was highly in agreement with the stand-
ard result based on the chemical formula (see Table 2).
The catalytic test showed that this yellow sample was also
active for the DAC synthesis (see Table 1, Entries 25, 26).
Thus, the above results was strong evidenced that LaCl3
had changed to La(NH3)Cl3 completely, and it might be
the active intermediate during the catalysis.
3.4 Possible Reaction Mechanism
for Transesterification of Urea and Allyl Alcohol
According to the experimental results and forgoing discus-
sion, the production of DAC from urea and allyl alcohol
over LaCl3 could be divided into two steps. It was obviously
showed that the conversion of urea to AC was easier than the
second step. The in situ FT-IR characterization meant that
isocyanic acid was generated from urea decomposition. It
was generally considered that this compound was an active
intermediate species and it could not exist in nature sepa-
rately [21]. When there was abundant alcohol present in sys-
tem, it would react with allyl alcohol to form AC very easily
[14]. Herein, AC could be obtained with high yield without
catalyst. However, although AC could be converted into
isocyanic acid over LaCl3, the generation of DAC via the
isocyanic acid from AC was proved to be hardly achieved.
The reason might be as follows: as the discussion above-
mentioned, the decomposition of AC into allyl alcohol and
isocyanic acid could be reversible. Considering that the allyl
alcohol/AC molar ratio was 10 in the reaction system, the
excessive allyl alcohol would drive the reaction equilibrium
−
towards the AC production. Simultaneously, NH2 was a
stronger nucleophile compared to CH2=CH–CH2–O−, there-
fore, it was required high energy for replacement of amino
group by allyl group. This made the reaction of AC and allyl
alcohol thermodynamically unfavorable [21]. Furthermore,
the NH3 accumulated in the reactor due to the conversion
of urea to AC unambiguously prevented the generation of
DAC, for this side-product also produced in the reaction of
AC and allyl alcohol. Thus, the conversion of AC to DAC
was very difficult and hardly occurred without catalyst.
In the other hand, it was well known that LaCl3 was a
hard Lewis acid, and it could be easily coordinated to nitro-
gen or oxygen atom with lone pair electrons. In the present
case, AC molecule had one nitrogen atom in amino group,
two oxygen atoms in allyl and carbonyl group. All of these
hetero atoms were likely to be coordinated with La3+ [34].
Zhao et al. proposed that the electron densities of nitrogen
were higher than that of oxygen atom in the carbamate mol-
ecule [30], implying that the former more likely bound to
LaCl3. Indeed, this conclusion was strongly proved by the
measurement of FT-IR, which showed that AC bounded
La3+ via the nitrogen atom of amino group. So, a plausible
1 Pattern of LaCl before used
2 Pattern of yellow3precipitate
) Cl
3 Standard pattern of La(NH3 3
3
1
2
3
10
20
30
40
50
60
70
2 Theta (degree)
Fig. 3 XRD patterns of the fresh LaCl3 catalyst and the precipitant
obtained after one reaction
1 3

1074
D. Wang et al.
catalytic mechanism is depicted in Scheme 2 for the synthe-
sis of DAC from AC and allyl alcohol over LaCl3. First, a
coordination complex was generated as the result of three
equiv AC bound to one equiv LaCl3 via nitrogen atom in
AC. Consequently, the La3+ ion reached a 6-coordinated
stable state (the three chloride atom would give one pair of
electrons, respectively). Meanwhile, as the nitrogen atom
bounded with La3+, its hybridized orbital might altered from
sp2 to sp3. As the result, the π-delocalized bond of amide
would disappear completely. This undoubtedly led to the
weakness of amino bond of AC, which was in turn benefit
for the attack of allyl alcohol. Subsequently, the lone pair of
oxygen of allyl alcohol could attack the carbonyl carbon of
the activated AC. During this process, the proton of O–H
in allyl alcohol would bind to nitrogen atom, giving rise to
the break of C–N bond. At the same time, the target product
DAC and the active intermediate species La(NH)3Cl3 was
produced. Finally, due to the liberation of NH3, AC, which
possessed high nucleophilic ability, would replace the NH3
in this complex to start next catalytic circle.
References
1. Liu Q, Wu L, Jackstell R, Beller M (2015) Nat Commun 6:5933
2. Sakakura T, Choi J-C, Yasuda H (2007) Chem Rev 107:2365–2387
3. Mikkelsen M, Jørgensen M, Krebs FC (2010) Energy Environ Sci
3:43–81
4. Martín C, Fiorani G, Kleij AW (2015) ACS Catal 5:1353–1370
5. Shukla K, Srivastava VC (2017) Catal Rev 59:1–43
6. Shaikh AG, Sivaram S (1996) Chem Rev 96:951–976
7. Sun W, Shi R, Wang X, Liu S, Han X, Zhao C, Li Z, Ren J (2017)
Appl Surf Sci 425:291–300
8. Zhang X, Wang D, Wu G, Wang X, Jiang X, Liu S, Zhou D, Xu
D, Gao J (2018) Appl Catal A 555:130–137
9. Wu Y, Song X, Cai F, Xiao G (2017) J Alloys Compd
720:360–368
10. Wu Y, Song X, Zhang J, Li Sh, Yang X, Wang H, Wei R, Guo L,
Zhang J, Xiao G (2018) J Taiwan Inst Chem E 87:131–139
11. Stoian D, Medina F, Urakawa A (2018) ACS Catal 8:3181–3193
12. Petersen JH (2000) Urea. Wiley, New Jersey
13. Fujita S, Yamanishi Y, Arai M (2013) J Catal 297:137–141
14. Wang D, Zhang X, Ma J, Yu H, Shen J, Wei W (2016) Catal Sci
Technol 6:1530–1545
15. Shukla K, Srivastava VC (2017) Fuel Process Technol
161:116–124
16. Fakhrnasova D, Chimentão RJ, Medina F, Urakawa A (2015) ACS
Catal 5:6284–6295
17. An H, Ma Y, Zhao X, Wang Y (2016) Catal Today 264:136–143
18. Nguyen-Phu H, Park C, Shin EW (2018) Appl Catal A 552:1–10
19. Over LC, Meier MAR (2016) Green Chem 18:197–207
20. Ma Q, Zhao T, Wang D, Niu W, Lv P, Tsubaki N (2013) Appl
Catal A 464–465:142–148
4 Conclusion
In summary, we have investigated the catalytic ability of
various metal chlorides for the green production of DAC
via transesterification of urea and allyl alcohol. A highest
DAC yield of 57.9% was obtained applying LaCl3 as cata-
lyst. Importantly, it was demonstrated that this process was
a two-step reaction, and the second step difficulty occurred
in the absence of catalysts. The excellent catalytic ability of
LaCl3 could be ascribed the fact that AC could be activated
by coordination of La3+ via its nitrogen atom. In addition,
we discovered that LaCl3 had been changed to La(NH3)3Cl3,
which was might the active intermediate for the efficient
conversion of AC to DAC. Considering the variety applica-
tion of organic carbonates, this new findings may provide a
new way to achieve green synthesis of DAC with high yield
and low-price.
21. Wang M, Wang H, Zhao N, Wei W, Sun Y (2006) Catal Commun
7:6–10
22. Manjunatha n P, Ravishankar R, Shanbhag GV (2016) Chem-
CatChem 8:631–639
23. Jagadeeswaraiah K, Kumar CR, Prasad PSS, Lingaiah N (2014)
Catal Sci Technol 4:2969–2977
24. Wang D, Zhang X, Liu C, Cheng T (2015) React Kinet Mech
Catal 115:597–609
25. Iio K, Kobayashi K, Matsunaga M (2007) Polym Adv Technol
18:953–958
26. Keuleers R, Desseyn HO, Rousseau B, Van Alsenoy C (1999) J
Phys Chem A 103:4621–4630
27. Li Q, Zhao N, Wei W, Sun Y (2007) J Mol Catal A 270:44–49
28. Li F, Li H, Wang L, He P, Cao Y (2015) Catal Sci Technol
5:1021–1034
29. Carter JC, Devia JE (1973) Spectrochim Acta A 29:623–632
30. Furer VL (1998) J Mol Struct 449:53–59
31. Sahni SK (1979) Transit Metal Chem 4:73–76
32. Zhao W, Wang F, Peng W, Zhao N, Li J, Xiao F, Wei W, Sun Y
(2008) Ind Eng Chem Res 47:5913–5917
Acknowledgements This work is subsidized by Key Research and
Developement Program of Shandong Province (2018GGX107010,
2017GGX70103), National Natural Science Foundation of China
(51674223), Natural Science Foundation of Shandong Province
(ZR2016BL21), Science and Technology Research Program for Col-
leges and Universities in Shandong Province (J18KA107), and National
Training Programs of Innovation and Entrepreneurship for Undergradu-
ates (201710904063 and 201710904064).
33. Beutel T (1998) J Chem Soc Faraday Trans 94:985–993
34. Wang D, Zhang X, Gao Y, Xiao F, Wei W, Sun Y (2010) Fuel
Process Technol 91:1081–1086
Publisher’s Note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.
Compliance with Ethical Standards
Conflict of interest The authors declare that they have no conflict of
interest.
1 3