Amine Functionalized Graphene Oxide Stabilized Pickering Emulsion for Highly Efficient Knoevenagel Condensation in Aqueous Medium

Article abstract of DOI:10.1007/s10562-020-03103-4

Abstract: A Pickering emulsion catalytic system was developed using amphiphilic amine-functionalized graphene oxide (GO-NH₂) for a Knoevenagel condensation in aqueous medium instead of organic solvent. The properties of GO-NH₂ and th

Full text of DOI:10.1007/s10562-020-03103-4

Catalysis Letters
https://doi.org/10.1007/s10562-020-03103-4
Amine Functionalized Graphene Oxide Stabilized Pickering Emulsion
for Highly Eꢀcient Knoevenagel Condensation in Aqueous Medium
1
1
1
1
1
1
1
Jie Zhu · Fei Wang · Dongsheng Li · Jie Zhai · Ping Liu · Weihong Zhang · Yongxin Li
Received: 23 August 2019 / Accepted: 9 January 2020
©
Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract
A Pickering emulsion catalytic system was developed using amphiphilic amine-functionalized graphene oxide (GO-NH )
2
for a Knoevenagel condensation in aqueous medium instead of organic solvent. The properties of GO-NH and the Pickering
2
emulsion stabilized by GO-NH were characterized by Fourier-transform infrared (FT-IR) spectroscopy, thermogravimetric
2
analysis (TGA), X-ray diꢀraction (XRD) analysis, X-ray photoelectron spectroscopy (XPS), organic elemental analysis
(
OEA), contact angle (CA), Brunauer–Emmett–Teller surface area, and optical microscopy. The results indicated that
the properties of the as-prepared Pickering emulsion system, such as droplet density, size, and distribution, were greatly
aꢀected by the amount of GO-NH and the oil–water ratio, which were closely related to the catalytic performance for
2
Knoevenagel condensation reactions. Furthermore, this Pickering emulsion system for the Knoevenagel condensation in an
aqueous medium exhibited a much higher catalytic activity than those of GO-NH in organic solvents under the same reac-
2
−
1
tion conditions. The reaction rate constant (k) in the Pickering emulsion system was calculated to be 0.043 min at 50 °C.
The enhanced activity of the Pickering emulsion was mainly attributed to large interfaces between GO-NH and reactants
2
in the emulsion, and the clear conꢁnement eꢀect. Furthermore, GO-NH remained stable for preparing Pickering emulsion
2
systems after recycling in six runs, exhibiting its good reusability.
Graphic Abstract
A novel Pickering emulsion catalytic system was developed using amphiphilic amine functionalized graphene oxide (GO-
NH2) for Knoevenagel condensation in aqueous medium instead of organic solvents. It exhibited excellent catalytic perfor-
mance in the reaction.
Keywords Pickering emulsion · Graphene oxide · Amine functionalization · Knoevenagel condensation
Electronic supplementary material The online version of this
article (https://doi.org/10.1007/s10562-020-03103-4) contains
supplementary material, which is available to authorized users.
Extended author information available on the last page of the article
Vol.:(0
1
123456789)
3

J. Zhu et al.
1
Introduction
This favors the formation of a Pickering emulsion in two-
liquid-phase systems, resulting in high catalytic activities
[23, 24]. For example, He et al. [25] reported the prepara-
tion of a stable Pickering emulsion, employing GO and
GO-supported Ag (Ag/GO) as emulsiꢁers. The Ag/GO-
stabilized emulsion exhibited good catalytic performance
for the reduction of 4-nitrophenol.
The Knoevenagel condensation is an important reaction
for forming carbon–carbon double bonds (C=C). The
resulting α,β-unsaturated carbonyl compounds are valu-
able intermediates in the pharmaceutical, cosmetic, and
pesticide industries [1–3]. Generally, the Knoevenagel
condensation occurs in the presence of base catalysts, such
as organic amines and piperidine [4, 5]. Despite the high
catalytic activity of these catalysts, the required separation
of products and catalysts, and catalyst recovery, owing to
their inherently homogeneous nature, limits their applica-
tions in industrial production [6, 7]. Therefore, organic
amines can be grafted on solid particles (such as polymers,
zeolites, and silica) to serve as heterogeneous catalysts for
this reaction [8–11].
Herein, we have successfully developed a Pickering emul-
sion system prepared from amine-functionalized graphene
oxide (GO-NH2) for catalysis of the Knoevenagel conden-
sation for the ꢁrst time. This Pickering emulsion greatly
enlarged the interfaces between catalyst GO-NH2 and the
reactants, resulting in a small mass transfer limitation in the
reaction and, therefore, high catalytic activity. The Pickering
emulsion system stabilized by GO-NH2 provided a highly
eꢂcient and environmentally friendly catalytic system for
aqueous Knoevenagel condensation.
The Knoevenagel condensation traditionally proceeds
in organic solvents [12–14]. Therefore, some disadvan-
tages of organic solvents, such as high cost and environ-
ment pollution, restrict their further practical applications
2 Experimental
[
15]. In comparison, water, as a green solvent, has received
increasing interest recently in catalysis and organic synthe-
sis. Unfortunately, most weakly polar reactants are poorly
soluble in water, which can produce serious mass transfer
limitations in aqueous–organic biphasic systems. Perform-
ing the Knoevenagel condensation in aqueous medium is
clearly unfavorable owing to the weakly polar nature of the
reactants, namely, benzaldehyde (BAL) and its derivatives.
Recently, Pickering emulsions stabilized by amphiphilic
solid particles have attracted much attention owing to their
innovative applications in designing eꢂcient two-liquid-
phase catalytic systems [16–19]. Pickering emulsions can
greatly decrease mass transfer limitations in a reaction
by enlarging the interfaces between catalysts and reac-
tants, obtaining high catalytic activities. To achieve this
objective, some solid particles (such as silica and zeolites)
are often functionalized by grafting with some moieties
to construct amphiphilic surfaces. For example, Zapata
et al. [20] synthesized interface-active zeolites by grafting
organosilanes on the surface of HY zeolites. The resulting
emulsions stabilized by interface-active zeolites can sig-
niꢁcantly improve catalytic eꢂciency in biofuel upgrading
reactions in water/oil biphasic systems.
2.1 Preparation of GO‑NH2 Catalyst
Graphene oxide was ꢁrst prepared by the modiꢁed Hum-
mers method [26]. GO (0.2 g) was then dispersed in DMSO
(100 mL) and 3-aminopropyltrimethoxysilane (APTMS;
2.0 g) was added dropwise, followed by ultrasonication for
30 min three times. The mixture was then reꢃuxed at 100 °C
for 24 h. At the end of the grafting reaction, as-synthesized
GO-NH2 was collected by ꢁltration and rinsed with absolute
ethanol three times at room temperature to remove unre-
acted impurities. Finally, GO-NH2 was dried under vacuum
at 80 °C overnight.
The preparation process of GO-NH2 was shown in
Scheme 1. In addition, APTMS was grafted on SBA-15,
OH OH
OH
COOH
OH
OH
O
,
HO
modified Hummer s method
COOH
HOOC
O
Graphene Oxide
Graphite
24h
NH2
DMSO,
MeO Si
MeO
Presently, graphene oxide (GO) is often employed as
a promising catalyst support. Compared with traditional
supports (such as silica and zeolites), GO has several sig-
niꢁcant advantages, including high hydrothermal stability,
strong resistance to acid and base, good electronic transfer
properties, and minimal mass transfer limitations [21, 22].
More importantly, GO and its derivatives (such as amine-
functionalized GO in this study) have been shown in previ-
ous studies to possess predominant amphiphilic properties.
OMe
H2N
H2N
H2N
Si
MeO
Si
Si
MeO
O
O
O O
OMe
COOH
O
O
O
COOH
HOOC
O
GO-NH2
Scheme 1 Preparation process of GO-NH2
1
3

Amine Functionalized Graphene Oxide Stabilized Pickering Emulsion for Highly Eꢀcient…
using the same process as GO-NH . The resulting com-
2.2.3 Recovery of GO‑NH2
2
posite was named SBA-15-NH .
2
The recycled GO-NH was ꢁltered and rinsed with pure etha-
2
nol for three times at the room temperature, followed by drying
2
.2 Catalytic Performance of Pickering Emulsion
under vacuum at 80 °C overnight.
System on Knoevenagel Condensation
2
.3 Characterization of GO‑NH and Pickering
2
2.2.1 Preparation of Pickering Emulsion
Emulsion
Catalyst GO-NH (10–25 mg), water (3–6 mL), and benzal-
Fourier-transform infrared (FT-IR) spectra of GO-NH were
2
2
dehyde (approximately 1.5 mL, 15 mmol) were successively
added to a 50-mL round-bottom ꢃask, followed by ultrasoni-
cation for 5 min. The mixture was then vigorously stirred at
obtained using a Bruker TENSOR 27 spectrometer (Ger-
many). Thermogravimetric analysis (TGA) was performed at
−
1
room temperature to 650 °C with a heating rate of 20 °C·min
8
00 rpm for 5 min to obtain the Pickering emulsion.
under nitrogen using a Perkin-Elmer TGA7 analyzer. The crys-
tal composition of the sample was analyzed using a D/Max
2
500PC X-ray diꢀractometer. (OEA) was performed using
2.2.2 Knoevenagel Condensation
Thermo Scientiꢁc™ FLASH 2000 CHNS/O. The content
of C, H and N elements in the sample could be accurately
determined by the method of high temperature combustion.
Contact angle tests were conducted using a DSA25 optical
contact angle measuring device. Nitrogen adsorption–desorp-
tion isotherms were measured at −196 °C using a TriStar3020
sorptometer (Micromeritics). The speciꢁc surface areas were
calculated according to the Brunauer-Emmet-Teller (BET)
Knoevenagel condensation in this study (Scheme 2) was
carried out in a 50 mL round-bottom flask.
Malononitrile (approximately 0.99 g, 15 mmol) and
n-dodecane (internal standard substance, 0.3 mL) were
successively added dropwise to the prepared Pickering
emulsion. The reaction proceeded at 50 °C for 1 h. After
the reaction was complete, the products were extracted
with chloroform, followed by centrifugation to remove
the catalyst and water. The reaction sample was analyzed
by gas chromatography (SP-6850) equipped with a flame-
ionization detector (FID) and a capillary column (SE-54)
using nitrogen as the carrier gas. BAL conversion (X) and
selectivity to benzalmalononitrile (BMA) were calculated
through Eqs. (1), (2).
o
method using adsorption data at relative pressure (p/p ) in the
range of 0.05–0.30. Morphologies of emulsions were char-
acterized using an XSP-8CAE optical microscope (Shanghai
Bim Co., Ltd).
2.4 Methodology of Processing XPS Data
X-ray photoelectron spectroscopy (XPS) measurements were
performed using a Mulꢁlab 2000 spectrometer working in the
constant analyzer energy mode with Al Kα radiation as the
excitation source. The carbonaceous C 1 s line (284.6 eV) was
used as the reference to calibrate the binding energies (B.E.).
0
BAL
C
− CBAL
X =
× 100%
(1)
(2)
0
BAL
C
CBMA
Si =
× 100%
2.4.1 Quantitative XPS Peak Analysis
0
BAL
C
− CBAL
Background removal methods have been proposed and used
in the literature [27]. The Shirley background subtraction is
one of the most widely used and was adopted in this study. To
measure the area of spectral peaks correctly, the energy range
of the spectrum (Emax and Emin) needs to be determined and
where C
and C
are the instant molar concentration
BAL
BMA
of the reactant BAL and the product BMA in the reaction;
0
C
is the initial molar concentration of BAL. The mass
BAL
balance was close to or even better than 97%.
In addition, SBA-15-NH was employed in Knoev-
2
ꢁ
xed in a manner consistent for comparing diꢀerent spectra of
enagel condensation for comparison.
the same element. The tools of this study, including a graphical
Scheme 2 Knoevenagel con-
densation between benzalde-
hyde and malononitrile
O
CN
CN
Cat.
H O, 1h
H
+
+
H O
2
2
CN
CN
1
3

J. Zhu et al.
interface facilitating preview, were built with the MATLAB
software from Math Works. The analysis consisted of a manual
or automatic binding energy scale correction to remove the
possible charging eꢀect, a normalization to take diꢀerences
in spectrometer response with time into account, a splitting
of peaks and background if necessary and the multivariate
procedure in itself.
corresponding to the stretching and wagging vibrations
of N–H bonds, respectively [29, 30] (Fig. 1b), which
clearly originated from APTMS. Notably, the appearance
−
1
of a C–O–Si peak (1120 cm ) and a weaker C–OH peak
−
1
(1068 cm ) hinted that APTMS was grafted to GO through
covalent linkage via C–O–Si bonds [20]. In addition, the
peak at 1650 cm-1 corresponding to the stretching vibra-
tion of amide group [29] was found in FT-IR spectrum of
GO-NH2. It was attributed to the possible covalent combi-
nation of amino groups (–NH2) on APTMS with carboxyl
groups (COOH) on GO in the process of grafting. Therefore,
FT-IR results provided the direct evidence for the successful
immobilization of APTMS on GO.
2.4.2 Binding Energy Scale Correction
Insulating samples such as organic surfaces undergo positive
charging when exposed to X-ray radiation, and as a result,
the energy scale is shifted. Peak ꢁtting of XPS spectra was
performed using commercially available software, ESCA
tools developed by Surface/Interface Inc. Binding energy
scales are usually corrected using a reference peak of known
binding energy (e.g., C–OH component of C 1s peak of con-
tamination carbon) [27].
TGA of GO and GO-NH was conducted to determine
2
their thermal stability. The TG curve of GO (Fig. 2a) showed
two signiꢁcant weight losses before 250 °C. The ꢁrst, below
125 °C, was derived from the evaporation of adsorbed water
on GO. The second was found in the temperature range of
1
90–250 °C, accounting for approximately 25% of the total
weight. This was reasonably attributed to the thermolysis
of labile oxygen-containing functional groups (such as
OH, C=O, and C–O–C) on GO. Continued weight loss was
observed above 250 °C owing to thermolysis of the car-
bon skeleton [31]. The ꢁnal weight loss was approximately
3
Results and Discussion
3.1 Characterizations of Synthesized GO‑NH2
The FT-IR spectra in Fig. 1 identified some functional
64% at 650 °C. In comparison, GO-NH (Fig. 2b) showed
2
groups on GO and GO-NH . Speciꢁcally, the FT-IR spec-
gradual weight loss throughout the temperature range. The
overall weight loss at 650 °C was approximately 41%, which
was less than that of GO. Although APTMS was thermally
unstable, the grafted APTMS on GO exhibits the higher
thermal stability than GO in some temperature range due to
the formation of stable C–O–Si bonds. Hence, the mass loss
of GO-NH2 below 650 oC is less than that of GO, in accord-
ance with some reported references [29, 32, 33]. Notably,
2
trum of GO (Fig. 1a) exhibited typical groups includ-
−
1
2
ing carbonyl groups (C=O, 1724 cm ), sp carbon rings
−1
−1
(
C=C, 1616 cm ), epoxy groups (C–O–C, 1259 cm ), and
–
1
hydroxyl groups (C–OH, 1068 cm ), in accordance with
previous reports [20, 21, 28]. After grafting APTMS onto
the GO surface, the spectrum of the sample showed distinct
−
1
peaks for amino groups (C–NH ) at 1569 and 773 cm ,
2
the insigniꢁcant weight loss of GO-NH in the temperature
2
Fig. 1 FT-IR spectra of GO (a) and GO-NH (b)
Fig. 2 TG curve of GO (a) and GO-NH (b)
2
2
1
3

Amine Functionalized Graphene Oxide Stabilized Pickering Emulsion for Highly Eꢀcient…
range of 190–250 °C indicated that a large amount of liable
oxygen-containing functional groups, such as C–OH, had
been employed to graft APTMS, resulting in the formation
of comparatively stable C–O–Si groups. These results indi-
cated the improved thermal stability of GO after grafting
with APTMS.
carboxyl groups (COOH, 288.6 eV) [37] were observed
in the C1s spectrum of GO-NH (Fig. 4d) compared with
2
those of GO (Fig. 4c). The decrease in epoxy groups on
GO-NH might be due to amino groups (–NH ) on APTMS
2
2
attacking epoxy groups on GO [13], resulting in the ring
opening of epoxy groups for APTMS grafting. Meanwhile,
Figure 3 shows the XRD spectra of graphene, GO, and
the decreased content of carboxyl groups on GO-NH was
2
GO-NH . The XRD spectrum of graphene exhibited a nar-
attributed to the possible covalent combination of amino
groups on APTMS with carboxyl groups on GO during the
grafting process, in accordance with FT-IR analysis.
In addition, the eꢀects of additive amount of APTMS
on amino groups grafted on GO were studied here. OEA
2
row and strong diꢀraction peak at about 26.5° attributed
to graphene (002). The crystalline interplanar spacing was
calculated to be 0.34 nm. However, a new strong character-
istic peak at about 12° instead of 26.5° appeared in the XRD
spectra of both GO and GO-NH , corresponding to graphene
analysis of GO-NH samples were summarized in Table S1.
2
2
oxide (002) [33–35]. A large number of oxygen-containing
functional groups produced in the preparation of graphene
oxide led to expansion of the graphene layer spacing. Fur-
thermore, a comparatively small 2θ value (11.7°) in the
Clearly, the more the additive amount of APTMS for graft-
ing, the higher the amino content on GO materials. When
additive amount of APTMS reached 2.0 g, the content of
amino groups grafted on GO was calculated to be the highest
(5.98 mmol/g GO) among 4 samples. Therefore, the sample
XRD spectrum of GO-NH compared with that of graphene
2
oxide (12.4°) hinted the successful grafting of APTMS on
the GO sheets, leading to further expansion of the GO layers.
Meanwhile, XRD of the GO modiꢁed with diꢀerent amounts
of APTMS are listed in supporting information (Figure S3).
Figure 4a shows only two characteristic peaks for C1s and
O1s in the XPS spectrum of GO. However, after grafting
APTMS onto GO, two new elements (nitrogen and silicon)
derived from APTMS were found in the spectrum (Fig. 4b),
indicating the successful synthesis of amine-functionalized
GO, in accordance with the FT-IR results. Furthermore,
compared with the C1s spectrum of GO (Fig. 4c), that
GO-NH 2.0 was further employed for the following prepara-
2
tion of Pickering emulsion.
3.2 Characterizations of Pickering Emulsion System
Prepared by GO‑NH2
The Pickering emulsion system stabilized by GO-NH was
2
observed by optical microscopy. For example, the Pickering
emulsion prepared from GO-NH (20 mg), BAL (1.5 mL),
2
and H O (4 mL) had a high density of small droplets, as
2
shown in Fig. 5. Droplets in this emulsion had a size distri-
bution of 5–75 μm. According to statistical analysis, droplets
with the sizes of 15–20 μm accounted for about 20% of all
counted droplets, which was the highest relative frequency
in the size distribution. Small droplets with sizes less than
40 μm accounted for about 84% of the total. The mean size
of droplets in this Pickering emulsion was calculated to be
about 28 μm.
of GO-NH (Fig. 4d) exhibited a new peak for C–O–Si
2
groups (285.6 eV) and a weakened peak for C–OH groups
(
286.5 eV) [30, 36]. This clearly indicated the covalent com-
bination of APTMS with hydroxyl groups on GO through
C–O–Si bonds produced by dealcoholization. Furthermore,
weakened peaks for epoxy groups (C–O–C, 287.4 eV) and
The as-prepared Pickering emulsion was further investi-
gated by comparing its dispersion in water and BAL, respec-
tively. As shown in Fig. 6, the Pickering emulsion stabilized
by GO-NH dispersed well in BAL (Fig. 6b), but showed
2
an obvious interface between water and BAL (Fig. 6a).
This indicated that the water-in-BAL (W/O) type of Picker-
ing emulsion was stabilized by GO-NH , as proven by the
2
smaller contact angle of BAL (about 22°) compared with
that of water (about 33°) on the GO-NH surface (Fig. 7)
2
[
23, 38].
The surface free energies of GO and GO-NH were
2
further calculated to investigate their wettability. Gener-
ally, the higher the surface free energy of a material, the
stronger its hydrophilicity. Based on the mathematical
model proposed by Owens [39], the surface free energy
d
p
of a material is composed of dispersion (γ ) and polar (γ )
Fig. 3 XRD curve of graphite (a), graphite oxide (b) and GO-NH (c)
components. The surface free energy (γ ) of the solid at
2
s
1
3

J. Zhu et al.
Fig. 4 Survey XPS spectra and C 1s ones of GO (a, c) and GO-NH (b, d)
2
Fig. 5 Optical micrograph of
GO-NH stabilized Pickering
2
emulsions
d
p
ꢀ
ꢀ
dꢁ
s
ꢀ p
p
p
pꢁꢁ
a solid–liquid interface can be calculated from γ + γ ,
d
l
d
s
d
l
s
s
(1 + cosꢀ) ꢁ = 4 ꢁ ꢁ ∕ ꢁ + ꢁ + ꢁ ꢁ ∕ ꢁ + ꢁ
s
(3)
l
l
s
l
while that of the liquid (γ ) has a similar expression of
l
d
p
γ + γ . In combination with Young’s equation, the rela-
In this study, water and diiodomethane were employed
as test liquids to calculate the surface energy of GO and
l
l
tionship between surface free energy (γ) and contact angle
(
θ) can be expressed as shown in Eq. (3).
GO-NH . Based on the reference [40], dispersion/polar
2
1
3

Amine Functionalized Graphene Oxide Stabilized Pickering Emulsion for Highly Eꢀcient…
Table 2 Catalytic performance of Pickering emulsion for Knoev-
enagel condensation
Entry
Catalysts
Conv. (%)
Sel. (%)
1
2
3
4
Blank
GO
SBA-15-NH2
25.2
32.6
73.7
>99.5
>99.5
93.2
GO-NH stabilized Pick- 96.1
>99.5
2
ering Emulsion
Reaction conditions: benzaldehyde (15 mmol), malononitrile
(
15 mmol), H O (4 mL), catalyst (20 mg), n-dodecane (0.3 mL); reac-
2
tion temperature 50 °C, reaction time 1 h
Fig. 6 GO-NH stabilized Pickering emulsion dispersion in water (a)
2
and benzaldehyde (b)
3.3 Performance of Pickering Emulsion System
for Knoevenagel Condensation
d
p
components (γ /γ ) of water and diiodomethane are
The as-prepared Pickering emulsion stabilized by GO-NH
2
l
l
2
2
2.1/50.7 and 44.1/6.7 mJ/m , respectively.
was further employed in the Knoevenagel condensation,
with the results compared with those of GO and APTMS-
Contact angles of water and diiodomethane on both
GO and GO-NH surfaces were measured, as shown in
grafted SBA-15 (SBA-15-NH ), as shown in Table 2. The
2
2
Figure S1. Thus, surface free energy of GO-NH was cal-
GO-NH -stabilized Pickering emulsion exhibited excellent
2
2
2
culated in Table 1 to be 62.24 mJ/m , higher than that of
catalytic performance. Using this system, the BAL conver-
sion reached about 96.1% (Table 2, entry 4) at 50 °C in 1 h,
which was much higher than those over GO (about 32.6%,
2
GO (59.40 mJ/m ). Clearly, the increased surface free
energy of GO-NH hinted its enhanced hydrophilicity,
2
attributing to the grafting of APTMS on GO. In addi-
entry 2) and SBA-15-NH (about 73.7%, entry 3) under the
2
tion, GO-NH surface exhibited small contact angles of
same reaction conditions. Notably, GO itself possesses a few
weak basic sites (such as OH and C–O–C groups) on its
sheets [41] that result in quite high catalytic activity in the
Knoevenagel condensation. However, after grafting APTMS
onto GO, the enhancement of strong basic sites (amino
2
water and benzaldehyde (Fig. 7), proving its amphiphilic
property.
groups) on GO-NH favored improved catalytic activity
2
Fig. 7 Contact angles of water
(
a) and BAL (b) on GO-NH
2
surface
Table 1 The surface wettability
Surface free energya,b (mJ/m2)
Catalysts
Contact angle (°)
of GO and GO-NH2
θ H O
θ CH2I2
γ
γd
γp
Xp
2
GO
GO-NH2
37
33
29
21
59.40
62.24
23.32
25.25
36.08
36.99
0.61
0.59
a
d
p
d
p
γ: total surface free energy; γ=γ +γ ; γ : dispersion component of the surface free energy; γ : polar com-
p
p
ponent of the surface free energy; X : percentage of the polar component in total surface free energy, γ /γ
b
Surface free energy calculated with the contact angles of the catalysts
1
3

J. Zhu et al.
in the reaction. In contrast, the greatly decreased OH and
reaction. Therefore, SBA-15-NH exhibited a lower activity
2
C–O–C groups on GO-NH (Fig. 4d) compared with those
than the GO-NH -stabilized Pickering emulsion under the
2
2
on GO (Fig. 4c) indicated a minor eꢀect of those groups on
same reaction conditions.
the catalytic activity. Therefore, grafted amino groups on
GO-NH were an important factor in enhancing catalytic
3.4 Comparison of Pickering Emulsion and Other
Solvents
2
activity toward the Knoevenagel condensation. In contrast,
the amphiphilic nature of GO-NH helped with formation
2
of the Pickering emulsion, which showed a high density of
small droplets (Fig. 5). As described in Sect. 3.2, the small
droplets with sizes of less than 40 μm accounted for about
Table 3 compares the catalytic performance in the Knoev-
enagel condensation of the Pickering emulsion system and
GO-NH in several organic solvents. In nonemulsion cata-
2
8
4% of the total, and the mean droplet size was calculated to
lytic systems, GO-NH achieved the highest catalytic activ-
2
be about 28 μm. Therefore, the GO-NH -stabilized Picker-
ity in THF among all organic solvents. The BAL conversion
into THF was about 78.5% (Table 3, entry 2) after 1 h at
50 °C, which was higher than that in other organic solvents,
such as toluene (about 45.8%, entry 3) and ethyl acetate
(about 23.8%, entry 7) under the same reaction conditions.
2
ing emulsion signiꢁcantly enlarged the interfaces between
GO-NH and the reactants, eꢀectively decreasing the mass
2
transfer limitation in the aqueous–organic biphasic system.
In contrast, forming a Pickering emulsion with nonamphi-
philic SBA-15-NH in the aqueous–organic biphasic system
This was attributed to the excellent dispersion of GO-NH
2
2
was hard, leading to serious mass transfer limitations in the
in THF [42]. The contact angle of THF on GO-NH (Fig. 8f)
2
was measured to be 12°, which was far smaller than those of
other solvents, such as ethyl acetate (25°, Fig. 8a) and tolu-
ene (17°, Fig. 8e), conꢁrming this opinion. Therefore, the
Table 3 Catalytic performance in Knoevenagel condensation over
Pickering emulsion and GO-NH in organic solvents
2
excellent dispersion of GO-NH in THF favored the decline
2
Entry
1
Medium
Conv. (%)b
Sel. (%)
>99.5
of mass transfer limitations in the reaction, resulting in a
high activity.
GO-NH stabilized Pick- 96.1
2
ering emulsion
More importantly, compared with nonemulsion catalytic
systems, the Pickering emulsion system exhibited a higher
activity in the reaction. BAL conversion in the emulsion was
96.1% after 1 h at 50 °C (Table 3, entry 1), while that over
a
2
3
4
5
THF
Toluene
78.5
45.8
33.9
31.5
26.6
23.8
>99.5
>99.5
>99.5
>99.5
>99.5
>99.5
a
a
CH CN
3
a
CHCl3
GO-NH in THF was only 78.5% under the same reaction
2
a
6
CH Cl2
conditions. The rate constant, k, in the emulsion was further
2
a
−1
7
Ethyl acetate
calculated to be 0.043 min (Figure 9b), the highest found
in this work. This high activity originated from the large
Reaction conditions: benzaldehyde (15 mmol), malononitrile
15 mmol), solvent (4 mL), catalyst (20 mg), n-dodecane (0.3 mL);
interfaces between GO-NH and reactants in the emulsion,
2
(
as well as the clear conꢁnement eꢀect. Furthermore, as a
green solvent, employing water to prepare the Pickering
emulsion greatly reduced environmental pollution. There-
reaction temperature 50 °C, reaction time 1 h;
Non-emulsion system
Detailed kinetics data seen in Fig. 9 and Figure S2 (Supporting
a
b
Information)
fore, the Pickering emulsion stabilized by GO-NH showed
2
Fig. 8 Contact angles of diꢀer-
ent organic solvents on GO-NH
surface: a Ethyl acetate, b
2
CH Cl , c CHCl , d CH CN, e
2
2
3
3
Toluene and f THF
1
3

Amine Functionalized Graphene Oxide Stabilized Pickering Emulsion for Highly Eꢀcient…
(
a)
1
(b)
1.4
00
Equation
Solvent
y = a + b*x
Water
THF
Toluene
Ethyl acetate
0.00448
1
.2
.0
.8
.6
.4
k value (min-1)
0.04332
0.02473
0.0012
0.00986
80
60
40
20
0
Standard Error 8.39298E-5
3.16261E-4 6.19662E-5
1
0
0
0
Emulsion
THF
Toluene
Ethyl acetate
Emulsion
THF
Toluene
Ethyl acetate
0.2
0.0
0
20
40
60
80 100 120
0
20
40
60
80 100 120
Reaction Time (min)
Reaction Time (min)
Fig. 9 Catalytic activity over GO-NH in several solvents: a Benzaldehyde conversion versus reaction time; b Reaction rate constants for the
2
reaction
Fig. 10 Optical micrograph of Pickering emulsions prepared using diꢀerent mass of GO-NH (a: 10 mg, b: 15 mg, c: 20 mg, d: 25 mg)
2
many advantages over the traditional GO-NH catalyst in
2
Table 4 Catalytic performance in Knoevenagel condensation over
organic solvents for the Knoevenagel condensation.
Pickering emulsions prepared using diꢀerent amount of GO-NH2
Entry
Catalyst weight
mg)
Conv. (%)
Sel. (%)
3
.5 Eꢀect of Emulsion Preparation‑Conditions
(
on Its Catalytic Performance
1
2
3
4
5
Blank
10
15
20
25
25.2
76.8
82.9
96.1
99.0
>99.5
>99.5
>99.5
>99.5
>99.5
The mass of GO-NH and oil–water ratio greatly aꢀected the
2
properties of the as-prepared Pickering emulsion, including
its catalytic performance.
Initially, as shown in Fig. 10, as the mass of GO-NH was
2
increased, the as-prepared emulsion exhibited continuously
Pickering emulsion system was composed of benzaldehyde
enhanced catalytic activity. BAL conversion was only 76.8%
(
15 mmol), malononitrile (15 mmol), H O (4 mL), n-dodecane
2
after 1 h over the sample with 10 mg of GO-NH (Table 4,
2
(0.3 mL) and GO-NH (10–25 mg); Reaction conditions: 50 °C for
2
entry 2), but climbed to 96% when the mass of GO-NH
1 h
2
1
3

J. Zhu et al.
was increased to 20 mg (entry 4). The high activity of this
Pickering emulsion system was strongly correlated with its
excellent properties. For instance, the sample with 10 mg of
sample with 10 mg of GO-NH . Meanwhile, no droplets
2
with large sizes above 80 μm were found in this emulsion.
The mean droplet size was further calculated to be about
28 μm. This emulsion had comparatively large interfaces
GO-NH (Fig. 10a) had comparatively large droplets with a
2
low density. Droplets in this emulsion had a size distribution
of 10–120 μm in width. Based on statistical analysis, small
droplets with sizes of less than 40 μm accounted for about
between GO-NH and the reactants, which helped enhance
2
the catalytic activity. When the mass of the catalyst was
beyond 20 mg, the unchanged emulsion properties (Fig. 10c,
d) led to a slight increase in conversion from about 96% to
99% (Table 4, entry 5). Therefore, the Pickering emulsion
7
1% of all droplets counted. Meanwhile, a certain number
of large droplets with sizes above 100 μm were also discov-
ered in this emulsion, accounting for about 8% of the total
droplets. The mean size was calculated to be about 37 μm.
prepared from 20 mg of GO-NH possessed high catalytic
2
eꢂciency in the reaction. Furthermore, all samples pos-
sessed high selectivity toward benzylidenemalononitrile
(>99.5%), which was beneꢁcial to product puriꢁcation after
the reaction.
Therefore, comparatively small interfaces between GO-NH
2
and reactants in this emulsion led to low catalytic activity in
the reaction. In contrast, when the mass of GO-NH reached
2
2
0 mg, the prepared Pickering emulsion exhibited enhanced
Furthermore, the eꢀect of the oil–water ratio on emulsion
formation and catalytic activity was investigated. When the
amount of water added reached 4 mL, the prepared emulsion
exhibited the highest BAL conversion among all samples
(about 96%; Table 5, entry 2) owing to its excellent proper-
ties (Fig. 11b). In contrast, three other samples exhibited
comparatively low activities in the reaction owing to their
low densities of large droplets. For example, the sample with
properties. A high density of small droplets was observed
in this emulsion through optical microscopy (Fig. 10c).
Droplets in this emulsion exhibited a size distribution of
5
–75 μm. Those with small sizes less than 40 μm accounted
for about 84% of the total, which was much higher than the
Table 5 Catalytic performance in Knoevenagel condensation over
6
mL of water (entry 4) had the lowest BAL conversion
of about 83%, which was about 13% lower than that with
mL of water. Droplets in this emulsion exhibited a wide-
Pickering emulsions prepared using diꢀerent amount of water
Entry
V (mL)
Conv. (%)
Sel. (%)
4
1
2
3
4
3
4
5
6
87.0
96.1
86.8
83.2
>99.5
>99.5
>99.5
>99.5
ranging size distribution of 20–200 μm. The small droplets
with sizes less than 40 μm only accounted for about 16% of
the total, which was far lower than that of the sample with
4 mL of water (about 84% of the total, Fig. 11b). Meanwhile,
a number of large droplets with sizes above 100 μm were
discovered in this emulsion, accounting for about 15% of
the total. The mean size was calculated to be about 67 μm.
Pickering emulsion system was composed of benzaldehyde
(
(
15 mmol), malononitrile (15 mmol), H O (3–6 mL), n-dodecane
2
0.3 mL) and GO-NH (20 mg); Reaction conditions: 50 °C for 1 h
2
Fig. 11 Optical micrograph of Pickering emulsions prepared using 3 mL (a), 4 mL (b), 5 mL (c) and 6 mL (d) of water and size distributions of
droplets in them
1
3

Amine Functionalized Graphene Oxide Stabilized Pickering Emulsion for Highly Eꢀcient…
Logically, the large sizes of droplets in the emulsion shrank
emulsion produced a high density of small droplets, which
the interfaces between GO-NH and reactants, leading to the
greatly enlarged interfaces between GO-NH and the reac-
2
2
comparatively low catalytic activity.
tants. Therefore, the Pickering emulsion overcame mass
transfer limitations between oil–water interfaces, eꢂ-
ciently enhancing the catalytic activity of active sites on
In summary, the Pickering emulsion prepared from
GO-NH (20 mg) and H O (4 mL) possessed the smallest
2
2
droplets with the highest density. This emulsion exhibited
GO-NH , as proven by the high conversions of aldehydes
2
comparatively large interfaces between GO-NH and BAL
in the Pickering emulsion catalytic system (Table 6).
Furthermore, compared with substrates bearing electron-
donating groups, such as p-tolualdehyde (entry 3) and
p-methoxybenzaldehyde (entry 4), those with electron-
withdrawing groups, such as 2-ꢃuorobenzaldehyde (entry
2
for the reaction, resulting in a high catalytic activity and was,
therefore, used for subsequent reactions.
3.6 Exploration of Substrates Scope
1
) and p-ꢃuorobenzaldehyde (entry 2), achieved a slightly
To explore the substrate scope of the GO-NH -based Pick-
higher conversion. This was attributed to strong electron-
positive carbonyl carbon in the compound with electron-
withdrawing groups facilitating the attack of carbanions
of active methylene compounds, such as malononitrile.
Furthermore, the Pickering emulsion had a higher cata-
lytic activity toward aromatic aldehydes as substrates than
aliphatic aldehydes. For example, the p-tolualdehyde con-
version over the Pickering emulsion reached 98.2% in 1 h
2
ering emulsion, several aldehydes, including aromatic
aldehydes with electron-withdrawing groups (Table 6,
entries 1 and 2) and electron-donating groups (entries 3
and 4), a heterocyclic aldehyde (entry 5), and a long-chain
aliphatic aldehyde (entry 6) were employed for the Kno-
evenagel condensation with malononitrile in this catalytic
system. The results showed that the GO-NH -stabilized
2
Table 6 Catalytic performance
Entry
1
Aldehyde
Nitrile
Conv. (%)
99.1
Sel. (%)
>99.5
in Knoevenagel condensation
over Pickering emulsions
prepared using several
aldehydes
2
3
4
5
6
98.6
98.2
95.8
89.1
85.2
>99.5
>99.5
>99.5
>99.5
98.6
Pickering emulsion system was composed of aldehyde (15 mmol), malononitrile (15 mmol), H O (4 mL),
2
n-dodecane (0.3 mL) and GO-NH (20 mg); Reaction conditions: 50 °C for 1 h
2
1
3

J. Zhu et al.
(
(
entry 3), which was 13% higher than that of valeraldehyde
entry 6). The high conversions of aromatic aldehydes over
3.7 Catalyst Recycling
the catalytic system were possibly due to the strong inter-
action between aromatic rings and GO. This aided forma-
tion of the Pickering emulsion possessing a high density
of droplets in the aqueous phase.
Finally, the reusability of catalyst GO-NH was examined
2
owing to its heterogeneous nature. The results showed
that GO-NH maintained high activity across six runs.
2
The BAL conversion slightly declined from about 96%
to 90% (Fig. 12), which was mainly ascribed to some
mass loss of the catalyst during recycling. The proper-
ties of recycled GO-NH were further studied by optical
2
microscope, XRD, FT-IR and XPS. XRD spectrum of
recycled GO-NH (Fig. 13a) exhibited the characteris-
2
tic peak at about 12° on behalf of GO (002), similar to
that of fresh one. Its FT-IR spectrum (Fig. 13b) and XPS
ones (Fig. 13c, d) hinted the presence of C-O-Si groups,
proving the stability of GO-NH . The Pickering emulsion
2
after 6 runs exhibited good properties as well (Fig. 14).
Droplets in this emulsion exhibited the size distribution in
the range of 10–90 μm (Fig. 14). Those with small sizes
less than 40 μm accounted for about 80% of the total. The
mean size of droplets in this Pickering emulsion was about
3
2 μm. In summary, the synthesized GO-NH in this work
2
had good recyclability and reusability.
Fig. 12 Recycling of GO-NH across 6 running cycles
2
Fig. 13 XRD (a), FT-IR (b), XPS survey (c) and C 1s (d) spectra of GO-NH after 6 runs
2
1
3

Amine Functionalized Graphene Oxide Stabilized Pickering Emulsion for Highly Eꢀcient…
Fig. 14 Optical micrograph of
Pickering emulsion after 6 runs
1
1
1
3. Xue B, Zhu J, Liu N, Li Y (2015) Catal Commun 64:105–109
4
Conclusions
4. Li G, Xiao J, Zhang W (2011) Green Chem 13:1828–1836
5. Tobiszewsk M, Tsakovsk S, Simeonov V, Namieśnik J, Pena-
Pereira F (2015) Green Chem 17:4773–4785
An amphiphilic amine-functionalized graphene oxide (GO-
NH ) was successfully synthesized for the preparation of
16. Wei L, Zhang M, Zhang X, Xin H, Yang H (2016) ACS Sustain
Chem Eng 4:6838–6843
2
a W/O-type Pickering emulsion. This Pickering emulsion
achieved high catalytic activity in the aqueous Knoevenagel
condensation under a mild reaction temperature for a short
time owing to its large interfaces between catalysts and reac-
tants, and the conꢁnement eꢀect in the Pickering emulsion.
1
7. Xue F, Zhang Y, Zhang F, Ren X, Yang H (2017) ACS Appl Mater
Interfaces 9:8403–8412
1
8. Huang J, Yang H (2015) Chem Commun 51:7333–7336
19. Yang H, Fu L, Wei L, Liang J, Binks BP (2015) J Am Chem Soc
1
37:1362–1371
2
0. Zapata PA, Faria J, Ruiz MP, Jentoft RE, Resasco DE (2012) J
Am Chem Soc 134:8570–8578
GO-NH showed excellent recyclability and reusability.
2
Pickering emulsions prepared using amphiphilic GO-based
materials have broad application prospects in heterogeneous
interfacial catalysis.
2
1. Wang Y, Rong Z, Wang Y, Qu J (2016) J Catal 333:8–16
22. Sun Z, Rong Z, Wang Y, Xia Y, Du W, Wang Y (2014) RSC Adv
4
:1874–1878
2
2
3. Zhang F, Jiang H, Li X, Wu X, Li H (2014) ACS Catal 4:394–401
4. Luo Q, Wang Y, Yoo E, Wei P, Pentzer E (2018) Langmuir
Acknowledgements This work was supported by the National Natural
3
4:10114–10122
Science Foundation of China (No. 21673024 and No. 21676029). The
2
2
5. He Y, Wu F, Sun X, Li R, Guo Y, Li C, Zhang L, Xing F, Wang
W, Gao J (2013) ACS Appl Mater Interfaces 5:4843–4855
ꢁ
nancial support from the Priority Academic Program Development of
Jiangsu Higher Education Institutions (PAPD) is also acknowledged.
6. Marcano DC, Kosynkin DV, Berlin JM, Sinitskii A, Sun ZZ,
Slesarev A, Alemany LB, Lu W, Tour JM (2010) ACS Nano
4
:4806–4814
2
2
7. Burrell MC, Armstrong NR (1983) Appl Surf Sci 17:53–69
8. Liu F, Sun J, Zhu L, Meng X, Qi C, Xiao F-S (2012) J Mater
Chem 22:5495–5502
References
2
9. Valentini L, Bon SB, Monticell O, Kenny JM (2012) J Mater
Chem 22:6213–6217
1
2
.
.
Saravanamurugan S, Palanichamy M, Hartmann M, Murugesan
V (2006) Appl Catal A 298:8–15
3
3
0. Chong ASM, Zhao XS (2003) J Phys Chem B 107:12650–12657
1. Yang H, Li F, Shan C, Han D, Zhang Q, Niu L, Ivaska A (2009)
J Mater Chem 19:4632–4638
Zhang X, Lai ESM, Martin-Aranda R, Yeung KL (2004) Appl
Catal A 261:109–118
3
4
.
.
Chen X, Arruebo M, Yeung KL (2013) Catal Today 204:140–147
Ansari MB, Jin H, Parvin MN, Park S-E (2012) Catal Today
3
3
3
3
3
2. Matsuo Y, Nishino Y, Fukutsuka T, Sugie Y (2007) Carbon
4
5:1384–1390
1
85:211–216
3. Huang J, Ding S, Xiao W, Peng Y, Deng S, Zhang N (2015) Catal
Lett 145:1000–1007
5
6
7
.
.
.
Dalessandro EV, Collin HP, Guimarães LGL, Valle MS, Pliego
JR (2017) J Phys Chem B 212:5300–5307
4. Wu T, Wang X, Qiu H, Gao J, Wang W, Liu Y (2012) J Mater
Chem 22:4772–4779
Li X, Lin B, Li H, Yu Q, Ge Y, Jin X, Liu X, Zhou Y, Xiao J
(
2018) Appl Catal B 239:254–259
5. Li Y, Gao W, Ci L, Wang C, Ajayan PM (2010) Carbon
Kan-nari N, Okamura S, Fujita S, Ozaki J, Arai M (2010) Adv
Synth Catal 352:1476–1484
4
8:1124–1130
6. Pokhrela J, Bhoriaa N, Anastasioua S, Tsouꢁs T, Gournisc D,
Romanos G, Karanikolos GN (2018) Microporous Mesoporous
Mater 267:53–67
8
9
.
.
Zheng L, Li P, Tao M, Zhang W (2019) Catal Commun 118:19–24
Zhu F, Sun X, Lou F, An L, Zhao P (2015) Catal Lett
1
45:1072–1079
3
3
3
7. Ramanathan T, Fisher FT, Ruoꢀ RS, Brinson LC (2005) Chem
Mater 17:1290–1295
1
1
0. Wang S-G (2003) Catal Commun 4:469–470
1. Angeletti E, Canepa C, Martinetti G, Venturello P (1988) Tetra-
hedron Lett 29:2261–2264
8. Wan Y-J, Tang L-C, Gong L-X, Yan D, Li Y-B, Wu L-B, Jiang
J-X, Lai G-Q (2014) Carbon 69:467–480
1
2. Parida KM, Mallick S, Sahoo PC, Rana SK (2010) Appl Catal A
9. Owens DK, Wendt RC (1969) J Appl Polym Sci 13:1741–1747
3
81:226–232
1
3

J. Zhu et al.
4
4
0. Su ZF, Zuo BQ (2011) Silk 48:13–15
Publisher’s Note Springer Nature remains neutral with regard to
1. Islam SM, Roy AS, Dey RC, Paul S (2014) J Mol Catal A
jurisdictional claims in published maps and institutional aꢂliations.
3
94:66–73
4
2. Paredes JI, Villar-Rodil S, Martínez-Alonso A, Tascón JMD
2008) Langmuir 24:10560–10564
(
Aꢀliations
1
1
1
1
1
1
1
Jie Zhu · Fei Wang · Dongsheng Li · Jie Zhai · Ping Liu · Weihong Zhang · Yongxin Li
1
*
Yongxin Li
Jiangsu Key Laboratory of Advanced Catalytic Materials
and Technology, School of Petrochemical Engineering,
Changzhou University, Changzhou 213164, China
liyxluck@163.com
1
3