Journal of Molecular Liquids 261 (2018) 540–549

Contents lists available at ScienceDirect

Journal of Molecular Liquids

journal homepage: www.elsevier.com/locate/molliq

MCM-48 encapsulated with reduced graphene oxide/graphene oxide and

as-synthesised MCM-48 application in remediation of pharmaceuticals
from aqueous system
Samson O. Akpotu, Brenda Moodley ⁎
School of Chemistry and Physics, University of KwaZulu-Natal, Westville Campus, Durban 4000, South Africa

a r t i c l e i n f o a b s t r a c t

Article history: As-synthesised MCM-48 (ASM48), graphene oxide encapsulated MCM-48 (M48GO) and reduced graphene
Received 26 November 2017 oxide encapsulated MCM-48 (M48G) was synthesised and applied in remediation of caffeine (CAF) and phenac-
Received in revised form 20 March 2018 etin (PHE) from aqueous solution. Effects of surfactant template (cetyltrimethylammonium bromide) on MCM-
Accepted 9 April 2018
48 on the removal of CAF and PHE from wastewater samples was also investigated. Adsorbents were character-
Available online 13 April 2018
ized with the following techniques; FTIR, TGA, XRD, SEM, TEM, BET, EA. Elemental analysis revealed ASM48 had
Keywords:
improved hydrophobic properties and the successful synthesis of M48GO and M48G. All adsorbents were applied
As-synthesised MCM-48 in batch sorption studies of CAF and PHE to determine the effects of adsorbent dose, pH and initial concentration.
Graphene oxide Adsorption mechanism was through π-π interaction, EDA interaction and hydrophobic effects. Sorption kinetics
Reduced-graphene oxide showed that the data fitted pseudo-second order and intraparticle diffusion models. Equilibrium data obtained
Pharmaceuticals fitted the Freundlich and Langmuir adsorption isotherm. M48G had the best adsorption capacity of 153.8 and
Adsorption 212.7 mg/g for CAF and PHE, respectively and ASM48 showed significant adsorption which was attributed to in-
creased hydrophobicity created by the surfactant template in ASM48. Thermodynamic studies revealed adsorp-
tion process as spontaneous and exothermic. Desorption studies revealed that adsorbents could be regenerated
and reused for adsorption.
© 2018 Elsevier B.V. All rights reserved.

1. Introduction synthesised in 1992 [12], has enormous potential in remediating phar-

maceuticals from wastewater. MCM-48, a class of OMS with properties
Over the last decade pharmaceuticals have gained much attention as such as large surface area (N1000 m2/g), homogenous pores, narrow
emerging contaminants (EC) [1,2]. Pharmaceuticals are found in trace pore size distribution, cubic Ia3d symmetry and thermal stability [13].
concentrations in environmental media [3]. They have been detected However, due to its strong hydrophilic properties, it has low adsorptive
in aqueous systems and in soil at concentration levels from a few capacity for pharmaceuticals. Hence, to exploit its true adsorptive capa-
μg/L-100 μg/L, particularly in pharmaceutical industry effluents [4]. bilities, MCM-48 is made hydrophobic by coating with a hydrophobic
Pharmaceuticals are suspected to have chronic toxic effects in the envi- material such as graphene oxide (GO)/reduced graphene oxide (RGO)
ronment and may cause resistance of microorganisms against antibiotic or used as-synthesised (ASM48).
drug usage [5]. Hence, there is a need to remove EC from environmental ASM48 is made by retaining the surfactant template
media. However, conventional water treatment systems do not effec- (cetyltrimethylammonium bromide) in MCM-48. The surfactant is
tively remove pharmaceuticals from water systems [3,6]. It is important often times a long chain carbon and confers hydrophobicity on an oth-
to develop efficient water treatment systems for pharmaceuticals erwise strongly hydrophilic material. Mangrulkar, et al. [14] applied
removal. as-synthesised MCM-41, which showed increased adsorptive ability
Amongst the technologies available for pharmaceutical removal, ad- for phenols. GO/RGO are hydrophobic materials and are good adsor-
sorption is preferred because of ease of use, uncomplicated technology, bents for organic pollutants but do not achieve their true adsorption po-
and no toxic by-products. Materials such as activated carbon [7], zeolites tential. This is because G forms aggregated nanosheets [15] and GO is
[8], graphene oxide [9], graphene [10] and ordered mesoporous silica extremely difficult to separate from the adsorbate [16]. To overcome
(OMS) [11] have been used to adsorb pharmaceuticals. OMS these challenges, GO/RGO are anchored on OMS to produce hybrid ma-
terials. This material combines the uniqueness of the individual mate-
rials such as large surface area of the OMS and high hydrophobicity of
⁎ Corresponding author. the GO/RGO. GO/G has been used previously in adsorption of pharma-
E-mail address: [email protected]. (B. Moodley). ceuticals [10].

https://doi.org/10.1016/j.molliq.2018.04.046
0167-7322/© 2018 Elsevier B.V. All rights reserved.
 S.O. Akpotu, B. Moodley / Journal of Molecular Liquids 261 (2018) 540–549 541

Caffeine (CAF) is a basic pharmaceutical with a pKa value N14, pres- and sonicated in water. GO (10 mg/L), previously sonicated for 2 h
ent in drugs and beverages and affect central nervous system. Phenace- was added to the suspension and the pH adjusted to 7 by adding
tin (PHE) is neutral with pKa value N14, a metabolite of paracetamol NH4OH whilst stirring. It was stirred continuously for 8 h and left to set-
banned in USA and India because of long term kidney and liver damage tle for an hour. The product was filtered, washed with water and dried
associated with ingestion [17]. Their structures are shown in Fig. SI 1. in vacuum for 6 h at 50 °C and labelled as M48GO. In the synthesis of
In this study, novel ASM48, GO and RGO encapsulated MCM-48 were M48G, a similar procedure was adopted to M48GO but the temperature
fabricated and their removal efficiency for CAF and PHE from wastewa- of synthesis was 80 °C and hydrazine monohydrate was added to the
ter was investigated. Also, the adsorption capacities, adsorption kinetics mixture to reduce GO to RGO. The mixture was stirred for 8 h, left to set-
and the reusability of all adsorbents were further studied. This research tle for an hour, filtered, washed with water and dried under vacuum.
will be helpful in determining effective conditions, optimising adsorp- The solid black product was oven dried in a vacuum at 80 °C for 6 h.
tion conditions and provide an in-depth understanding of adsorption This was labelled M48G.
mechanism.
2.6. Characterization of samples
2. Experimental
FTIR spectra of the adsorbents were recorded on a Perkin Elmer se-
2.1. Materials and chemicals ries 100 spectrometer. Thermogravimetric (TGA) profiles were ob-
tained using 5 mg samples in a platinum pan on an SDT Q 600 V 209
Natural graphite powder, methanol (HPLC grade), hydrazine Build 20 between 25 and 900 °C at 10 °C/min in a N2 environment. Sur-
monohydrate (80%), 3-aminopropyl triethoxysilane (APTES, 99%), HCl face morphology and microstructure of adsorbents were obtained by
(37%), KMnO4, tetraethylorthosilicate (TEOS, 98%) and the pharmaceu- analysis on a field emission scanning electron microscope (FESEM,
ticals, caffeine (CAF) and phenacetin (PHE) were obtained from Sigma- Zeiss instrument, 10 kV operating voltage), and high-resolution trans-
Aldrich. Absolute ethanol (Merck), cetyltrimethylammonium bromide mission electron microscope (HRTEM, JEOL). X-ray diffraction (XRD)
(CTAB 99%+) (Calbiochem), H2SO4 (98%), H2O2 (35%) and H3PO4 experiments were carried out on an x-ray diffractometer (Bruker D8
(80%) were obtained from Promark. All the reagents were analytically Advance, Bruker, AXS, Germany) operating at 45 kV and 40 mA with
pure and were used without further purification. MilliQ water was Cu. The carbon content of the samples was determined using an ele-
used in the preparation of the standard solutions and the mobile mental analyser (Thermo Scientific CHNS/O analyser). N2 adsorption–
phase used in chromatography analysis. desorption isotherms were obtained using a nitrogen adsorption
analyser (Micrometrics Tristar II 3020). Samples were degassed at 90
2.2. Synthesis of MCM-48 and as-synthesised MCM-48 °C with a Micromeritics VacPrep 061 for an hour and then ramped up
to 200 °C overnight. Surface area and pore size measurements of sam-
MCM-48 was synthesised using a modified method from [18]. De- ples were obtained using BET method. Pore volume was calculated
tails in SI. from adsorption branch based on volume of nitrogen adsorbed at a rel-
ative pressure (P/Po) of approximately 0.99. Point of zero charge
2.3. Activation of MCM-48 and synthesis of NH2-MCM-48 (MCM-48 sur- (pHpzc) of adsorbents were determined by an acid-base titration
face functionalised with APTES) method by [23].

MCM-48 surface modification was achieved by using a method from 2.7. Analytical method
our previous publications [19,20]. A-MCM-48 is MCM-48 after acid
treatment and thermal treatment. Activation of MCM-48 was carried Concentrations of caffeine and phenacetin were determined by
out to serve the following purposes; exposing the oxygen on the silica HPLC measurements at a wavelength of 210 nm and 254 nm, respec-
surface making it very reactive with graphene oxide and reduced tively. Adsorption experiments were carried out using PTFE screw cap
graphene oxide; bringing the surface groups of the MCM-48 to their vials covered with aluminium foil at room temperature to determine
polar state and significant reduction of the water content of MCM-48. the effects of adsorbent dose, adsorption time, pH and temperature. A
About 1 g of MCM-48 was added to approximately 40 mL of 6 M HCl 1000 mg/L stock solution was prepared, from which several 15.0 mL
under stirring at a temperature of 150 °C and the mixture was refluxed working solutions of 12.5 mg/L (CAF and PHE) were prepared. Caffeine
for 4 h and cooled to room temperature. The product was filtered and and phenacetin concentrations were quantified using a HPLC system
washed with double distilled deionised water until pH 7. It was dried consisting of an Agilent 1200 SL (Agilent Technology, USA) with a pres-
in an oven at 150 °C for 5 h and was labelled as A-MCM-48. sure grade pump and a UV–Vis detector at a wavelength of 210 nm for
The surface of the A-MCM-48 obtained was modified by an aqueous CAF and 254 nm for PHE. Separation was carried out at room tempera-
easy one-step synthesis. Approximately 1 g of A-MCM-48 was dispersed ture with an Agilent Eclispe XDB-C-18 (4.6 × 150 mm, 5 μm particle
in 30 mL of ethanol, which was sonicated for an hour and stirred for an- size) column at a mobile phase flow rate of 0.6 mL/min under isocratic
other hour. Approximately 1000 μL of APTES was added to bind –NH2 to conditions with a mixture of methanol/water (60:40; v/v). The pH of
MCM-48. The reaction was stirred at 50 °C for 8 h to obtain a monolayer the water was adjusted with H3PO4 to 2.3. The sample volume injected
and a positively charged MCM-48. The product was collected by filtra- was 5 μL.
tion and vacuum dried at 60 °C and labelled as NH2-MCM-48.
2.8. Adsorption experiments
2.4. Synthesis of GO
Batch adsorption experiments were determined by adding
GO was synthesised using a modified Tour method [21,22]. Details in about 40 mg of the adsorbent to 15 mL of 12.5 mg/L (CAF and
SI. PHE solution) with a pH of 2–10 adjusted using 0.01 M NaOH/HCl
in a series of beakers covered with aluminium foil. The mixture
2.5. Synthesis of MCM-48-graphene oxide (M48GO) and MCM-48-reduced was agitated on a shaker at 25 °C for 180 mins. At set times, the so-
graphene oxide (M48G) lutions were centrifuged and filtered through a 0.45 μm filter and
the residual concentration of the filtrate determined by HPLC-UV
The procedure of Akpotu and Moodley was used for the encapsula- analysis. The concentration of CAF and PHE in the supernatant
tion [22]. In this encapsulation process, NH2-MCM-48 was dispersed was calculated from their calibration curves. The amount of
542 S.O. Akpotu, B. Moodley / Journal of Molecular Liquids 261 (2018) 540–549

adsorbate adsorbed on to the adsorbents at time t, and qt (mg/g) for 180 min and dried in a vacuum oven at 60 °C. The solutions were fil-
were calculated using Eq. 1: tered off after agitation and final concentration of pharmaceuticals were
determined with HPLC. Washing and recycling were carried out 4 times.
ðC o −C e ÞV
qt ¼ ð1Þ
m 3. Results and discussion

The removal efficiency of the pharmaceuticals in solution, and the 3.1. Characterization of adsorbents
percentage adsorption were calculated as:
FTIR spectroscopy was used to determine the functional groups
ðC o −C e Þ present on the surface of the adsorbents (Fig. 1). On all adsorbents, a
%adsorption ¼  100 ð2Þ
Co broad peak around 3230 cm−1 was prominent and was due to –OH vi-
brations. In GO, the peaks observed at 1729 cm−1 and 1634 cm−1 were
where C0 and Ce are the initial and equilibrium concentrations of the due to C_O stretching and the peak at 1364 cm−1 was due to the car-
pharmaceutical solutions, respectively, V is the solution volume bonyl group, respectively. A similar observation was recorded in studies
(L) and m is the mass of the adsorbent (g). by [21]. ASM48, MCM-48, M48GO and M48G had peaks that typically
Kinetic studies were carried at 25 °C, using 150 mg of adsorbents, define siliceous material. This implied that the synthesised MCM-48
with 200 mL of 25, 50, 75 and 100 mg/L of CAF and PHE solution. Ali- was successfully encapsulated with GO and RGO. In M48GO, the peaks
quots of the solution were withdrawn periodically (2–1440 min), fil- at 1622 cm−1 and 1415 cm−1 are due to \\C_O stretching vibration.
tered off, and the concentration of (CAF and PHE) adsorbed was M48GO had a profile like GO and this shows that the encapsulation of
determined. GO on MCM-48 was successful. M48G had no carbonyl, C\\O\\C and
epoxy group peaks. The band around 1048 cm−1 for M48G was due to
2.8.1. Analysis of real water samples the Si\\O\\Si/Si\\O\\C asymmetric vibration and Si\\O\\C was linked
Water samples (from wastewater treatment plants and surface by covalent bonds. A similar observation was also recorded by Liu, et al.
(river) water without laboratory simulation or pollutant spiking) were [16]. This implied that the –COOH group in the GO was converted to a
collected from Blue Lagoon (BLG) (mouth of the Umgeni River into Si\\O\\C bond as was also observed by [30]. The peaks around 780
the Indian Ocean) with co-ordinates of 29° 48′ 41″ S and 31° 02′ 12″ and 430 cm−1 in the M48G spectrum were assigned as Si\\O\\Si sym-
E, and the Northern Wastewater Treatment plant (WWTP) inlet point metric and bending vibrations, respectively. The peak around
with co-ordinates of 29° 48′ 27″ S and 30° 59′ 51″ E, in Durban, South 1555 cm−1 seen in M48G is a \\C_C\\ which is assigned to the
Africa. The samples collected were filtered with a 0.45 μm filter disc amide group stretching vibration which means the vinyl group was suc-
and were kept in amber bottles, wrapped in aluminium foil and stored cessfully introduced into M48G [30]. The M48G spectrum revealed that
at 4 °C. The sample properties are shown in Table 1. The concentration MCM-48 was successfully encapsulated with graphene. In ASM48, there
of CAF and PHE in the real samples was determined initially and was were peaks present which were typical of ammonia at 1415–1533 cm−1
found to be below the detection limit of the HPLC. A 0.125 mL also observed [14]. Peaks at 2925 and 2850 cm−1 were due to the bend-
(1000 mg/L) CAF and PHE was added to 10 mL aliquot of real water ing of \\CH2 and\\CH3 of the CTAB, respectively which impacted the
samples to obtain a concentration of 12.5 mg/L. The samples were equil- hydrophobicity of ASM48. Similar vibrations were observed in the syn-
ibrated overnight, and 0.04 g of the adsorbents were added to the sam- thesis of as-synthesised MCM-41 [14].
ples and shaken for 60 min. The suspension was centrifuged and filtered Thermogravimetric analysis was carried out on all the adsorbents in
and the removal efficiency and adsorption capacity for CAF and PHE was the temperature range between 25 °C and 900 °C (Fig. 2). All the adsor-
evaluated with Eqs. (1) and (2), respectively. bents exhibited weight loss below 120 °C and this was due to the expul-
sion of physisorbed water. MCM-48 and ASM48 had 3 stages of weight
2.8.2. Adsorption isotherms loss. At stage one below 120 °C, approximately 10% weight loss
Langmuir [24], Freundlich [25] and Temkin [26] models were used to occurred. Stage had weight losses of 41.5% and 2% between 150 and
explain isothermal relationship between adsorbents and adsorbates
(details in SI).

2.8.3. Adsorption kinetics ASM48

Adsorption kinetics were carried out to determine the rate deter-
mining step and adsorption mechanism. Data obtained from the ad-
sorption studies were fitted and analysed using pseudo-first order M48G
(PFO) [27], pseudo-second order (PSO) [28], and intraparticle diffusion
Transmittance/a.u

(IPD) [29], models (equations in SI). Thermodynamics studies were M48GO

performed.

2.8.4. Adsorbent regeneration GO

To regenerate the adsorbents used, preloaded adsorbents were
washed with acidic ethanol at a pH of 2 and shaken in a water bath MCM-48
-1
Table 1 1622 cm
-1
Physical water quality parameters of the water samples. 3230 cm 2850 cm
-1
-1
2925 cm
-1 1415 cm
Parameter WWTP BLG 1729 cm
-1 -1
1555 cm 1065 cm
-1
pH 4.04 3.39
TDS (mg/L) 670 3100
Conductivity (mS) 1000.1 5200.3
4000 3000 2000 1000
Res (Ω) 729 150.7
Redox (mV) 93.5 181.3
Temp (°C) 22.0 22.5
Fig. 1. FTIR spectra of adsorbents.
 S.O. Akpotu, B. Moodley / Journal of Molecular Liquids 261 (2018) 540–549 543

also, the lower O/C ratio of M48G implies that it is less hydrophilic
than M48GO as also observed by [32]. The presence of carbon shown
from the EA results proved that the encapsulation process was success-
80 ful. ASM48 appeared to be very hydrophobic based on the H/C and O/C
indices. This hydrophobicity stems from the presence of the long carbon
chain from the highly hydrophobic surfactant CTAB in the
60 microstructure.
N2 adsorption-desorption isotherms of the adsorbents (Fig. 3)
showed that MCM-48 had the highest amount of N2 adsorbed at low
40 or high pressure. The detailed textural properties are presented in
Table 1. ASM48 had a very low surface area and pore volume when
GO
MCM-48 compared to MCM-48. This is because of the presence of the CTAB sur-
20 ASM48 factant in the template of the ASM48 which caused pore channels/cav-
M48G ities to be blocked, hence, the low surface area. Synthesised GO had a
M48GO surface area and pore volume of 39 m2/g and 0.017 cm3/g, respectively.
0 The surface area and pore volume of M48GO and M48G were signifi-
200 400 600 800 cantly reduced compared to MCM-48, and more so in M48G. This can
be explained as a result of the coating of hydrophobic carbonaceous ma-
terial on MCM-48. M48G had more carbon coating than M48GO, which
accounted for its reduced surface area and pore volume. This confirms
Fig. 2. TGA profiles of adsorbents.
the successful encapsulation and correlates with the results obtained
from the EA. All the adsorbents were of mesoporous nature because
350 °C for ASM48 and MCM-48, respectively. This degradation is attrib- the pore diameter was b50 nm.
uted to the decomposition of residual carbon as was also observed by The XRD profiles of all the adsorbents are presented (Fig. 4). The low
Thuc and Thuc [31]. This explains the drastic weight loss in ASM48, as angle XRD patterns of ASM48, MCM-48, M48GO and M48G are shown
the long chain carbon template of the ionic surfactant CTAB remained in Fig. 4a. Clearly, the Braggs reflection which is characteristic of meso-
as part of the structure. In the third stage between 350 and 650 °C, porous MCM-48 material with unique cubic Ia3d symmetry is promi-
there was minimal weight loss before both adsorbents became ther- nent in all adsorbents. MCM-48 clearly exhibited 5 peaks in the 2θ
mally stable. The loss at this stage was attributed to condensation of ad- range from 2 to 6 with these reflections present at 211, 220, 321, 400
jacent silanol groups. GO is usually characterized by a sharp decline in and 420 with 211 being the principal peak. These reflections were also
weight between 150 °C and 250 °C. This decline is attributed to the re- observed in similar studies by [18,33]. It was observed that reaction con-
moval of oxygen containing functional groups. A similar observation ditions played an important role in the structure of MCM-48. The peaks
was made by Marcano, et al. [21]. From the thermal profiles of M48GO became suppressed when GO was encapsulated and even more so with
and M48G, it was observed that the materials had more stability when RGO. In (Fig. 4b) wide angle XRD scan, GO had a sharp peak at 2(θ) =
compared to GO. This is indicative that the encapsulation of MCM-48 9.74. In contrast, MCM-48, ASM48, M48GO and M48G all had a broad
with GO and G was successful. The weight loss for M48GO and M48G peak at 2(θ) = 24 which is typical of siliceous materials. However,
occurred in 3 phases with the first phase occurring before 150 °C. The M48GO and M48G had marked inflections at 2(θ) = 26.21 and 25.4, re-
second phase occurred between 150 °C and 250 °C and the weight spectively. These peaks are typical of oxygen containing functional
loss recorded was 6.5 and 14.5% for M48GO and M48G, respectively. groups by GO and chemical interaction by the reduced GO (graphene)
This is due to decomposition of carbon. The final stage had a weight on M48GO and M48G as also observed by [34,35]. The presence of this
loss of 1 and 14% for M48G and M48GO, respectively, at a temperature signifies a successful encapsulation process.
above 250 °C. This loss can be attributed to the removal of oxygen func- To further confirm the Ia3d cubic mesostructure of MCM-48, HRTEM
tionality present on the surface of the adsorbents with M48GO having a images (Fig. 5) obtained were in agreement with XRD analysis. Each
higher oxygen content. nanosphere in the FESEM (Fig. 6) images appear as a single crystal-
The elemental compositions of the adsorbents are presented in like structure with ordered mesoporous channels extending through
Table 2. GO, ASM48, MCM-48, M48GO and M48G, had carbon contents the material. The mesoporous nanospheres show discernible cavities
of 37.6%, 35.1%, 0.9%, 16.2% and 22.4%, respectively. It should be noted associated with crystalline MCM-48. GO sheets appeared as large parti-
that elemental analysis did not consider the amount of silica present cles with a transparent outlook. M48GO appeared to be coated with a
and this assumption was necessary to calculate the O/C. The high carbon transparent film of GO on its surface. M48G appeared as stacks of
composition and the highly hydrophobic nature of ASM48 is attributed graphene sheets on MCM-48, thereby losing its transparency due to
to the CTAB present in the template of ASM48. The lower H/C ratio of stacking by RGO sheets due to the in-situ reduction of GO to RGO.
M48G when compared to M48GO implies that it is more aromatic and FESEM image analysis of MCM-48 and ASM48, revealed that it was com-
prised of uniform nano spherical particles with an average diameter of
200 nm with no agglomeration. Interparticle voids were also observed
Table 2 between these spheres. M48GO had a thin layer of GO on its surface
Elemental composition and atomic ratios, and surface areas (SA) of MCM-48, GO, ASM48,
M48GO and M48G.
and GO sheets filling the interparticle voids. In contrast, M48G, had
stacks of RGO sheets layered over MCM-48 surface, thereby losing its
Sample MCM-48 ASM48 GO M48GO M48G surface transparency and forming stacks of RGO sheets on MCM-48.
C/% 0.90 35.1 37.6 16.2 22.4
H/% 2.31 7.07 2.23 2.51 1.84 3.2. Batch adsorption studies
O/% 96.7 55.7 59.5 79.3 72.2
N/% 0.13 2.16 0.58 2.04 3.55
H/C 31.0 2.43 0.71 1.87 0.99 Results for dose study is presented in the SI 6.
O/C 80.9 1.19 0.48 3.69 2.43
SA/m2/g 1028 16 39 119 65.3 3.2.1. Effect of pH
Pore volume/cm3/g 0.67 0.05 0.017 0.20 0.17 There are several ways which pharmaceutical pollutants interact
Average pore size/nm 2.54 12 38.2 6.9 11.8
with adsorbent (i) through electrostatic attraction or repulsion and
544 S.O. Akpotu, B. Moodley / Journal of Molecular Liquids 261 (2018) 540–549

400
(a) 0.20

dV/dlog(D) Pore Volume/cm3/g

350 (b) M48G

Quantity adsorbed/cm3/g
M48GO
300 0.16
MCM-48
250 M48GO
M48G
GO 0.12
200
ASM48
150
0.08
100
0.04
50

0.0 0.2 0.4 0.6 0.8 1.0 0 50 100 150 200 250
Relative pressure/P/Po Pore diameter/nm

6
(C) 0.05

dV/log(D) Pore Volume/cm3/g

5 (d)
dVlogP(D) Pore Volume/cm /g
3

4 0.04 GO
ASM48
3 0.03

2
0.02

1
0.01
0
0.00
0 2 4 6 8 10 0 50 100 150 200 250

Fig. 3. (a) N2 adsorption-desorption of samples and pore size distribution of (b) M48GO and M48G (c) MCM-48 and (d) GO and ASM48.

depending on their pKa values (ii) pH changes causes dissociation of hy- Cb is NaOH concentration, initial volume of suspension is V0, Vb is
drophobic neutral molecules to hydrophilic charged species and (iii) total volume of –OH added at various titration points and Veb1 is NaOH
through π-π interaction [9,36]. Fig. 7 shows the effect of pH on the ad- volume added in the titration at the acidic side. The pHpzc for the fol-
sorption of CAF and PHE on ASM48, M48GO and M48G. Effect of pH lowing materials are 2.32, 2.04, 2.83 and 3.24 for MCM-48, ASM48,
was studied in the range of 2–10. The pHpzc of the adsorbents played M48GO and M48G, respectively. From the pzc values, the CTAB species
an important role in the adsorption behaviour between the adsorbate presence on ASM48 had a pronounced effect on surface properties, thus
and adsorbents. All adsorbents surface contained large amounts of bind- making it more negative which resulted in a lower pH value. The shift
ing sites and the surface density of adsorbents determined from acid- resulting after the addition of CTAB indicates a weak chemisorption or
base titration. TOTH is the total concentration of protons consumed dur- a combination of physical and chemical adsorption.
ing titration and is calculated with the formula: M48G had higher adsorption as compared to ASM48 and
M48GO and this may be attributed to its hydrophobic nature. Ad-
−ðV b −V eb1 ÞC b sorption was strongest at acidic pH (pH b pKa) for caffeine and
TOTH ¼
Vo þ Vb phenacetin on all the adsorbents as compared to basic conditions

(a) (211) GO
MCM-48 (b) MCM48
ASM48
ASM48
M48GO
M48G
M48G
M48GO
Intensity unit/a.u
Intensity/a.u

(220)

(321)

(400) (420)

2 4 6 8 10 10 20 30 40 50

Fig. 4. (a) Low angle scans of adsorbents and (b) Wide angle scans of adsorbents.
 S.O. Akpotu, B. Moodley / Journal of Molecular Liquids 261 (2018) 540–549 545

Fig. 5. HRTEM images of (a) GO (b) ASM48 (c) MCM-48 (d) M48GO (e) M48G.

(pH N pKa). However, the effect of pH on adsorption was not so electrostatic interaction between the π-electrons of the adsorbates
much because the pH range studied was below the pharmaceuti- and π-electrons of the adsorbents surface. Although, other
cals' pKa values. Adsorption may be attributed majorly to non- mechanisms such as H-bonding and hydrophobic-hydrophobic

Fig. 6. FESEM images (magnification = 25,000) of (a) GO (b) ASM48 (c) MCM-48 (d) M48GO (e) M48G.
546 S.O. Akpotu, B. Moodley / Journal of Molecular Liquids 261 (2018) 540–549

(a)
(b)
80
80

Percentage adsorption/%
Percentage adsorbed/% ASM48
60
M48GO 60
M48G ASM48
M48GO
40 M48G
40

20
20

2 4 6 8 10
2 4 6 8 10

Fig. 7. (a) Effect of pH on sorption of CAF and (b) Effect of pH on the sorption of PHE on the adsorbents (Conditions: 15 mL of 12.5 ppm CAF and PHE, equilibration time 180 mins, adsorbent
dose 40 mg, temperature 25 °C).

interaction may have played a part in adsorption for M48GO and heterogeneous surface with several adsorption mechanisms involved.
M48G. Nevertheless, ASM48 adsorption may have been through Freundlich constant KF (adsorption capacity) and N were also calcu-
π-π interaction or π-cation bonding and this is because ASM48 do lated. The KF value for M48G appeared higher in the adsorption of
not possess any aromatic moieties. both CAF and PHE which implied that the adsorption was highest for
M48G. The values of N obtained, indicated that the adsorption was
3.2.2. Adsorption isotherms favourable [39]. The highest qm values gotten from Langmuir isotherm
The isotherm models were calculated from the experimental data in the adsorption of CAF and PHE was obtained with M48G which corre-
obtained from the non-linear regression analysis of Freundlich and lated with the KF value from the Freundlich isotherm.
Langmuir equation isotherm models (Table 3). Fig. 8 shows the experi- M48GO exhibited excellent adsorption capacities for CAF, whilst
mental adsorption isotherm plots of CAF and PHE on M48G, M48GO and M48G and ASM48 gave excellent adsorption for PHE. The results ob-
ASM48, respectively. Based on Giles isotherm classification [37], the ad- tained compared favourably to other studies. M48G performed signifi-
sorption isotherm could be categorised as belonging to type L due to the cantly better for both adsorbate. However, there was no study on
shapes obtained. This is indicative that adsorbed solutes were not verti- phenacetin as at the time this manuscript was written (Table SI 3).
cally oriented or there was no strong competition from the aqueous
phase for the adsorption sites of the adsorbents [38]. From the correla- 3.2.3. Adsorption kinetics
tion coefficient (R2) obtained, values N0.98, showed that the Freundlich Three models were applied to determine the physical or chemical
and Langmuir isotherms was the best fit for M48GO and M48G as com- nature of the adsorbents, namely: pseudo first order (PFO) (Eq. SI 4),
pared to ASM48. Langmuir assumes that the adsorption of CAF and PHE pseudo second order (PSO) (Eq. SI 5) and intraparticle diffusion (IPD)
to the surface of the adsorbents is homogeneous with adsorption occur- (Eq. SI 6) models. Kinetic data were fitted into these models and results
ring on equivalent sites without interference from adjacent sites. The (Table SI 2). PSO model was the best fit for all adsorbents because of the
Freundlich isotherm models assumes that the adsorption occurs on high regression correlation (R2) N0.9999 and there was no significant
difference in qe experimental and qe calculated. This indicated that
Table 3
PSO model was well suited for the description of the kinetics and this
Isotherm parameters for the adsorption of CAF and PHE onto the adsorbents. model is based on the assumption that bimolecular interaction which
involves exchange between adsorbate and adsorbent is the rate limiting
Isotherm Parameters ASM48 M48GO M48G
step responsible for adsorption [19]. Therefore, a route of the adsorption
Caffeine of these pharmaceuticals onto the adsorbents is via electrostatic interac-
Langmuir qm/mg/g 14.9 48.8 153.8
tion with the –OH groups present on the surface of the adsorbents. PFO
b/L/mg 13.9 3.14 0.0065
R2 0.7156 0.9901 0.9839 model was not suited based on the low regression correlation values ob-
Freundlich KF (mg/g(mg/L)1/n) 3.26 10.6 25.4 tained and, there was a significant difference between qe experimental
N 1.03 1.31 1.37 and qe calculated. Also, it was observed that k2 parameter for M48G
R2 0.999 0.998 0.9985 was higher than that of M48GO and ASM48 in the adsorption of both
Temkin B 708 292 131
b/J/mol 3.50 8.50 18.9
CAF and PHE which corresponds to the adsorption capacity of each ad-
A/L/g 9.73 20.5 43.1 sorbent. Therefore, the PSO model best described the adsorption behav-
R2 0.9575 0.9829 0.9778 iour of CAF and PHE on ASM48, M48GO and M48G and this is indicative
Phenacetin
that chemisorption is the most probable form of adsorption mechanism.
Langmuir qm/mg/g 122 75.2 212.7 Adsorption occurred as ion exchange on surface active sites and the in-
b/L/mg 2.17 2.51 0.91 teraction is the rate controlling step, indicative that other factors may
R2 0.9524 0.9967 0.9855 also be responsible for adsorption.
Freundlich KF (mg/g(mg/L)1/n) 10.2 5.24 31.7
For intraparticle diffusion, the adsorption process occurred in 2 steps
N 1.08 1.213 1.385
R2 0.9999 0.9927 0.9984 (i) rapid equilibrium is reached because of the external surface of the
Temkin B 164 232 104 adsorbent adsorbing the adsorbate (ii) slow intraparticle internal diffu-
b/J/mol 15.1 10.7 24.2 sion to diffusion to the internal surface of the adsorbent from the bulk
A/L/g 36.9 25.2 51.0 solution. The solid adsorbent and liquid adsorbate adsorption process
R2 0.9745 0.9753 0.9777
may either be because of mass transfer or intraparticle diffusion or a
 S.O. Akpotu, B. Moodley / Journal of Molecular Liquids 261 (2018) 540–549 547

6 a b c
16
30
5

qe (mg/g)
qe (mg/g)
qe (mg/g)

12
4
20
3
8

2
10
4
1

20 40 60 80 20 40 60 80 100 20 40 60
Ce (mg/g) Ce (mg/g) Ce (mg/L)

30
50

25 d 20
e f
40
20
15

qe (mg/g)
30

qe (mg/g)
15
e

10
10 20

5
5 10

0
20 40 60 20 40 60 80 10 20 30 40 50 60
C (mg/g) Ce (mg/L) Ce (mg/L)

Fig. 8. Experimental isotherm plot of (a) ASM48 on CAF (b) M48GO on CAF (c) M48G on CAF (d) ASM48 on PHE (e) M48GO on PHE (f) M48G on PHE.

combination of both sorption processes. The pharmaceutical adsorption was carried out over a temperature range of 298–318 K (Fig. SI 3). The
mechanism from the solution comprises the following steps [40,41]: results (Fig. SI 3) showed that for all adsorbents, an increase in temper-
ature caused a decrease in adsorption capacity. This may be attributed
a. The movement of CAF and PHE molecules from the bulk solution to to a decrease in surface activity between the adsorbate and adsorbents
the surface of the adsorbents implying that the process was exothermic. An increase in temperature
b. The diffusion of CAF and PHE via a boundary layer to the surface of caused a reduction in the attractive forces between the adsorbents
the adsorbents and adsorbate, thereby causing the adsorbate molecules to escape
c. The adsorption of CAF and PHE at the surface-active site on the ad- from the solid phase (adsorbent) into the liquid phase (pharmaceutical
sorbents solution) [40], which resulted in a decrease in adsorption. The changes
d. Slow intraparticle diffusion of CAF and PHE after the initial rapid ad- in entropy (ΔS°), enthalpy (ΔH°) and Gibbs free energy (ΔG°) were de-
sorption to the internal surface of the adsorbents termined using the Van't Hoff's equation. A physisorption process has
free energy in the range of −20–0 kJ/mol and is −80–400 kJ/mol for
chemisorption. The negative ΔG° values obtained indicated that adsorp-
The slowest step is usually the controlling step for the rate of sorp- tion was a spontaneous physical process. The ΔS° values obtained were
tion and this step may be a combination of external mechanisms and negative and this indicated a decrease in randomness at the solid/liquid
intraparticle diffusion. An IPD plot was obtained by plotting qt (mg/g) interface as well as indicated that the adsorbents did not undergo any
which is weight of sorbent vs. square root of time (t0.5). A larger inter- significant changes [43]. The increase in negative ΔG° values suggested
cept correlates with thickness of boundary and this impacts greatly that the adsorption was thermodynamically favourable and spontane-
the boundary effect. Boundary thickness was found to increase in ous. The negative ΔH° values obtained implied that adsorption was exo-
M48G (encapsulated with more layers of RGO) as compared to thermic [44].
M48GO (encapsulated with GO) (Table SI 3). This implies that adsorp-
tion of both pharmaceuticals onto the adsorbents was influenced by 3.2.5. Adsorption mechanism
boundary-layer of the adsorbents. Therefore, a multi-step process There are several ways which ionisable micropollutants interact
could be involved in the adsorption of CAF and PHE to the adsorbents with adsorbents and this depends on pKa values of the pollutants.
[42]. This is due to higher concentration of CAF and PHE in solution, From the plots on pH (Fig. 7), the removal of both compounds was
thereby resulting in an increased resistance of mass transfer from the slightly improved in acidic pH although consistent over the pH range
adsorbent surroundings. The plots appeared as two separate linear re- (pH b pKa) with both compounds having pKa values over 14. Caffeine
gions (external mass transfer of CAF and PHE to the surface of the adsor- is a basic pharmaceutical with a pKa N 14 and phenacetin is a neutral
bents and intraparticle diffusion to the pores of the adsorbents) and this pharmaceutical with a pKa N 14. This means under acidic conditions,
validated that the adsorption process was due to several processes. both CAF and PHE have both have neutral ion species, hence, have
None of the plots had origins of zero and this deviation could be attrib- higher affinity for the negatively charged adsorbent surface (at pH
uted to a variation in the rate of mass transfer in the preliminary and the N pHpzc) than in basic conditions. As a result, the removal under acidic
last adsorption stages. This is indicative that IPD (pore diffusion) was pH can be attributed to π-π interaction, EDA interaction and hydropho-
not the only rate controlling step but it is the rate-limiting step [40]. bic effects. The slightly reduced adsorption at basic pH can be attributed
The result confirms that adsorption of pharmaceuticals on to the adsor- to electric repulsion between the basic NaOH and the negatively
bents was heterogeneous in nature. charged adsorbents surface as the volume of base is increased in the so-
lution. PHE and CAF becomes ionised (deprotonated) thus leading to an
3.2.4. Effect of temperature electric repulsion between the negative adsorbent surface and the ad-
The effect of temperature helps in understanding the physical or sorbate molecules. At pH N 7, any sorption taking place was due to hy-
chemical nature of the adsorption process. The effect of temperature drophobic effect.
548 S.O. Akpotu, B. Moodley / Journal of Molecular Liquids 261 (2018) 540–549

However, there are other factors, which assisted in the adsorption of for ASM48, were obtained for PHE and CAF, respectively, after 4 cycles
the pharmaceuticals by M48G, because it had better sorption despite its (Fig. SI 5). These results are indicative that all adsorbents can be re-
low surface area and negative surface. The increased sorption by M48G used and are essentially recyclable and stable. Environmental applica-
may be attributed to the presence of aromatic rings in the pharmaceuti- tion studies are presented in SI 10.
cals and GO reduction to RGO by hydrazine. It was observed that phar-
maceutical molecules with aromatic rings are faster/better adsorbed by 4. Conclusion
more hydrophobic carbon based adsorbent [45]. This was consistent
with the higher %C from the EA (Table 1) and also the higher reaction In this study novel adsorbents, GO/reduced GO MCM-48 were fabri-
rate parameter (K1) (Table 3). Therefore, the increased hydrophobicity cated and characterization results showed successful encapsulation
of M48G, which can be attributed to different factors, resulted in a mul- RGO/GO on MCM-48, and the lateral presence of CTAB on ASM48. The
tifaceted sorption of both pharmaceuticals, which is a feature of RGO/ increased hydrophobicity of all the adsorbents led to increased adsorp-
graphene-based adsorbents. These factors are surface defects, high en- tion of CAF and PHE. Additionally, results indicated fast equilibrium of
ergy surface and groove regions [46]. Adsorbed molecules have a ten- (b40 mins) with experimental data best fitted the pseudo second
dency to fill in these voids with strong affinity which also explained order model and the Langmuir isotherm model best described the ad-
the increased sorption [32]. The reduction of GO to RGO on M48G had sorption. Adsorption mechanism was by EDA, hydrophobic and π-π in-
a tremendous effect on adsorption and also conformation because the teraction. Maximum adsorption capacity of CAF was 153.8 and PHE was
encapsulated RGO sheet changed from planar to wrinkle (Figs. 5 and 212.7 mg/g for M48G at a temperature of 25 °C and adsorption behav-
6). The conformational change is as result of the conversion of chemical iour explained by isotherm and kinetic models. Thermodynamic studies
energy to mechanical energy [47]. The wrinkled sites of the M48G re- revealed the adsorption process as spontaneous and exothermic in na-
sults in non-uniform charge distribution, concentrated charges and ture. Regeneration studies showed that adsorbents were reusable for
high chemical activity [32,48], thereby resulting in higher sorption ca- other applications. Adsorbents performed well on real water samples
pacity. Also, increased sorption by M48G may be due to residual oxide and as such are good adsorbents for the removal of pharmaceuticals
groups functionalised by other active sites that are either π- from aqueous solutions.
conjugated structures or vacancies which favours numerous kinds of in-
teraction [49]. Surface defects on M48G grooves promote sorption of or-
ganic molecules independent of charge through hydrophobic Acknowledgements
interaction and π-π conjugation. Thus, this establishes a relationship be-
tween adsorption sites, adsorption capacity and structural morphology We gratefully acknowledge the School of Chemistry and Physics,
transformation of M48G and the pharmaceuticals. University of KwaZulu-Natal, South Africa, for providing the laboratory
Adsorption of the pharmaceuticals by ASM48 was either through π- facilities and instrumentation used in this research and Association of
π interaction, hydrophobic effect or π-cation bonding. Considering CTAB African Universities (AAU) small grant for financial assistance.
and the pharmaceuticals structure, the most likely interaction that pro- Note.
motes the adsorption of CAF and PHE onto ASM48 surface is thought to The authors declare no conflict of interest.
involve head group electrostatic and tail group lateral hydrophobic in-
teraction. At low concentration, the CTAB is bound mainly by electro- Appendix A. Supplementary data
static mechanism [40]. Head group adsorption greatly affects the
extent of hydrophobic tail interaction, therefore an increase in pH Supplementary data to this article can be found online at https://doi.
causes a decrease in the adsorbate electrostatic attraction with nega- org/10.1016/j.molliq.2018.04.046.
tively charged surfaces reducing the chance for intermolecular hydro-
phobic stabilization at the solid/liquid interface, which caused reduced References
sorption when pH exceeds PZC of the ASM48. The negative charge on
[1] S.K. Khetan, T.J. Collins, Human pharmaceuticals in the aquatic environment: a chal-
ASM48 (at pH N pHpzc) increased with increasing pH and this was re-
lenge to green chemistry, Chem. Rev. 107 (2007) 2319–2364.
ported by [50]. ASM48 had high sorption capacity for PHE which could [2] O.A. Jones, J.N. Lester, N. Voulvoulis, Pharmaceuticals: a threat to drinking water?
be attributed to the longer carbon chain length of PHE which encour- Trends Biotechnol. 23 (2005) 163–167.
[3] T.X. Bui, V.H. Pham, S.T. Le, H. Choi, Adsorption of pharmaceuticals onto
ages adsorption laterally along the surfactant molecules, and the conju-
trimethylsilylated mesoporous SBA-15, J. Hazard. Mater. 254 (2013) 345–353.
gated structure resulting in a π-π interaction between the surfactant [4] D.J. Larsson, C. de Pedro, N. Paxeus, Effluent from drug manufactures contains ex-
and PHE molecules in solution as was also observed by [51]. Improved tremely high levels of pharmaceuticals, J. Hazard. Mater. 148 (2007) 751–755.
adsorption by ASM48 observed for both pharmaceuticals compared to [5] J.L. Martínez, Antibiotics and antibiotic resistance genes in natural environments,
Science 321 (2008) 365–367.
MCM-48 was due to increased hydrophobicity by retaining the long [6] S. Castiglioni, R. Bagnati, R. Fanelli, F. Pomati, D. Calamari, E. Zuccato, Removal of
chain cationic SDA and reduced silanol density. Reduced silanol density pharmaceuticals in sewage treatment plants in Italy, Environ. Sci. Technol. 40
causes increased adsorption of hydrophobic pollutants [52]. Therefore, (2006) 357–363.
[7] S.A. Snyder, S. Adham, A.M. Redding, F.S. Cannon, J. DeCarolis, J. Oppenheimer, E.C.
the dominant mechanism of interaction between ASM48 and PHE mol- Wert, Y. Yoon, Role of membranes and activated carbon in the removal of endocrine
ecules in solution may be attributed electrostatic interaction due to the disruptors and pharmaceuticals, Desalination 202 (2007) 156–181.
lateral hydrophobic chain and head group interaction as was also ob- [8] A. Martucci, L. Pasti, N. Marchetti, A. Cavazzini, F. Dondi, A. Alberti, Adsorption of
pharmaceuticals from aqueous solutions on synthetic zeolites, Microporous Meso-
served by [40], π-π bonding and hydrophobic effects. porous Mater. 148 (2012) 174–183.
[9] S.-W. Nam, C. Jung, H. Li, M. Yu, J.R. Flora, L.K. Boateng, N. Her, K.-D. Zoh, Y. Yoon, Ad-
3.3. Desorption studies sorption characteristics of diclofenac and sulfamethoxazole to graphene oxide in
aqueous solution, Chemosphere 136 (2015) 20–26.
[10] L.A. Al-Khateeb, S. Almotiry, M.A. Salam, Adsorption of pharmaceutical pollutants
Recycling and regeneration of adsorbents plays a vital role in the onto graphene nanoplatelets, Chem. Eng. J. 248 (2014) 191–199.
practical application of adsorbents, as it aids in the disposal of used ad- [11] T.X. Bui, H. Choi, Adsorptive removal of selected pharmaceuticals by mesoporous sil-
ica SBA-15, J. Hazard. Mater. 168 (2009) 602–608.
sorbent and a reduction in the amount of secondary pollutants. Model
[12] J. Beck, J. Vartuli, W. Roth, M. Leonowicz, C. Kresge, K. Schmitt, C. Chu, D.H. Olson, E.
adsorbents are adsorbents with outstanding adsorption capacity and Sheppard, A new family of mesoporous molecular sieves prepared with liquid crys-
model desorption ability. These properties would ensure that secondary tal templates, J. Am. Chem. Soc. 114 (1992) 10834–10843.
pollutants are not released into the environment and save cost. Desorp- [13] L. Gibson, Mesosilica materials and organic pollutant adsorption: part B removal
from aqueous solution, Chem. Soc. Rev. 43 (2014) 5173–5182.
tion studies were carried out using acidic ethanol. High desorption effi- [14] P.A. Mangrulkar, S.P. Kamble, J. Meshram, S.S. Rayalu, Adsorption of phenol and o-
ciencies of 83 and 76% for M48G, 75 and 72% for M48GO, and 45 and 30% chlorophenol by mesoporous MCM-41, J. Hazard. Mater. 160 (2008) 414–421.
 S.O. Akpotu, B. Moodley / Journal of Molecular Liquids 261 (2018) 540–549 549

[15] K. Yang, B. Chen, L. Zhu, Graphene-coated materials using silica particles as a frame- [35] Y.-B. Luo, G.-T. Zhu, X.-S. Li, B.-F. Yuan, Y.-Q. Feng, Facile fabrication of reduced
work for highly efficient removal of aromatic pollutants in water, Sci. Rep. 5 (2015) graphene oxide-encapsulated silica: a sorbent for solid-phase extraction, J.
11641–11653. Chromatogr. A 1299 (2013) 10–17.
[16] X. Liu, H. Zhang, Y. Ma, X. Wu, L. Meng, Y. Guo, G. Yu, Y. Liu, Graphene-coated silica [36] F.-f. Liu, J. Zhao, S. Wang, P. Du, B. Xing, Effects of solution chemistry on adsorption
as a highly efficient sorbent for residual organophosphorus pesticides in water, of selected pharmaceuticals and personal care products (PPCPs) by Graphenes and
J. Mater. Chem. A 1 (2013) 1875–1884. carbon nanotubes, Environ. Sci. Technol. 48 (2014) 13197–13206.
[17] U.C. Dubach, B. Rosner, T. Stürmer, An epidemiologic study of abuse of analgesic [37] C. Giles, T. MacEwan, S. Nakhwa, D. Smith, 786. Studies in adsorption. Part XI. A sys-
drugs: effects of phenacetin and salicylate on mortality and cardiovascular morbid- tem of classification of solution adsorption isotherms, and its use in diagnosis of ad-
ity (1968 to 1987), N. Engl. J. Med. 324 (1991) 155–160. sorption mechanisms and in measurement of specific surface areas of solids, J.
[18] R. Peng, D. Zhao, N.M. Dimitrijevic, T. Rajh, R.T. Koodali, Room temperature synthesis Chem. Soc. (1960) 3973–3993.
of Ti–MCM-48 and Ti–MCM-41 mesoporous materials and their performance on [38] C.P. Okoli, G.O. Adewuyi, Q. Zhang, P.N. Diagboya, Q. Guo, Mechanism of dialkyl
photocatalytic splitting of water, J. Phys. Chem. C 116 (2011) 1605–1613. phthalates removal from aqueous solution using γ-cyclodextrin and starch based
[19] S.O. Akpotu, B. Moodley, Synthesis and characterization of citric acid grafted MCM- polyurethane polymer adsorbents, Carbohydr. Polym. 114 (2014) 440–449.
41 and its adsorption of cationic dyes, J. Environ. Chem. Eng. 4 (2016) 4503–4513. [39] Y. Shao, X. Wang, Y. Kang, Y. Shu, Q. Sun, L. Li, Application of Mn/MCM-41 as an ad-
[20] S.O. Akpotu, B. Moodley, Application of as-synthesised MCM-41 and MCM-41 sorbent to remove methyl blue from aqueous solution, J. Colloid Interface Sci. 429
wrapped with reduced graphene oxide/graphene oxide in the remediation of acet- (2014) 25–33.
aminophen and aspirin from aqueous system, J. Environ. Manag. 209 (2018) [40] T.K. Sen, M.T. Thi, S. Afroze, C. Phan, M. Ang, Removal of anionic surfactant sodium
205–215. dodecyl sulphate from aqueous solution by adsorption onto pine cone biomass of
[21] D.C. Marcano, D.V. Kosynkin, J.M. Berlin, A. Sinitskii, Z. Sun, A. Slesarev, L.B. Alemany, Pinus radiate: equilibrium, thermodynamic, kinetics, mechanism and process de-
W. Lu, J.M. Tour, Improved synthesis of graphene oxide, ACS Nano 4 (2010) sign, Desalin. Water Treat. 45 (2012) 263–275.
4806–4814. [41] J. Ma, F. Yu, L. Zhou, L. Jin, M. Yang, J. Luan, Y. Tang, H. Fan, Z. Yuan, J. Chen, Enhanced
[22] S. Akpotu, B. Moodley, Encapsulation of Silica Nanotubes From Elephant Grass adsorptive removal of methyl orange and methylene blue from aqueous solution by
With Graphene Oxide/Reduced Graphene Oxide and its Application in Remedi- alkali-activated multiwalled carbon nanotubes, ACS Appl. Mater. Interfaces 4 (2012)
ation of Sulfamethoxazole From Aqueous Media, 2018https://doi.org/10.1021/ 5749–5760.
acssuschemeng.7b02861. [42] O.A. Oyetade, V.O. Nyamori, B.S. Martincigh, S.B. Jonnalagadda, Nitrogen-
[23] J. Li, S. Zhang, C. Chen, G. Zhao, X. Yang, J. Li, X. Wang, Removal of Cu (II) and fulvic functionalised carbon nanotubes as a novel adsorbent for the removal of Cu (ii)
acid by graphene oxide nanosheets decorated with Fe3O4 nanoparticles, ACS Appl. from aqueous solution, RSC Adv. 6 (2016) 2731–2745.
Mater. Interfaces 4 (2012) 4991–5000. [43] V. Vimonses, S. Lei, B. Jin, C.W. Chow, C. Saint, Kinetic study and equilibrium iso-
[24] I. Langmuir, The adsorption of gases on plane surfaces of glass, mica and platinum, J. therm analysis of Congo red adsorption by clay materials, Chem. Eng. J. 148
Am. Chem. Soc. 40 (1918) 1361–1403. (2009) 354–364.
[25] H. Freundlich, Over the adsorption in solution, J. Phys. Chem. 57 (1906) 385–470. [44] F. Arias, T.K. Sen, Removal of zinc metal ion (Zn2+) from its aqueous solution by ka-
[26] A.K. Kushwaha, N. Gupta, M. Chattopadhyaya, Enhanced adsorption of methylene olin clay mineral: a kinetic and equilibrium study, Colloids Surf. A Physicochem. Eng.
blue on modified silica gel: equilibrium, kinetic, and thermodynamic studies, Asp. 348 (2009) 100–108.
Desalin. Water Treat. 52 (2014) 4527–4537. [45] B. Peng, L. Chen, C. Que, K. Yang, F. Deng, X. Deng, G. Shi, G. Xu, M. Wu, Adsorption of
[27] Y.-S. Ho, G. McKay, Sorption of dye from aqueous solution by peat, Chem. Eng. J. 70 antibiotics on graphene and biochar in aqueous solutions induced by π-π interac-
(1998) 115–124. tions, Sci. Rep. 6 (2016) 31920.
[28] K.Y. Ho, G. McKay, K.L. Yeung, Selective adsorbents from ordered mesoporous silica, [46] P. Lazar, F.E. Karlický, P. Jurečka, M.S. Kocman, E. Otyepková, K.R. Šafářová, M.
Langmuir 19 (2003) 3019–3024. Otyepka, Adsorption of small organic molecules on graphene, J. Am. Chem. Soc.
[29] W.J. Weber, J.C. Morris, Kinetics of adsorption on carbon from solution, J. Sanit. Eng. 135 (2013) 6372–6377.
Div. 89 (1963) 31–60. [47] Z. Xu, M.J. Buehler, Geometry controls conformation of graphene sheets: mem-
[30] H. Meng, Z. Li, F. Ma, X. Wang, W. Zhou, L. Zhang, Synthesis and characterization of branes, ribbons, and scrolls, ACS Nano 4 (2010) 3869–3876.
surface ion-imprinted polymer based on SiO 2-coated graphene oxide for selective [48] B. Pan, D. Lin, H. Mashayekhi, B. Xing, Adsorption and hysteresis of bisphenol a and
adsorption of uranium (vi), RSC Adv. 5 (2015) 67662–67668. 17α-ethinyl estradiol on carbon nanomaterials, Environ. Sci. Technol. 42 (2008)
[31] C.N.H. Thuc, H.H. Thuc, Synthesis of silica nanoparticles from Vietnamese rice husk 5480–5485.
by sol–gel method, Nanoscale Res. Lett. 8 (2013) 1–10. [49] J. Xiao, W. Lv, Z. Xie, Y. Tan, Y. Song, Q. Zheng, Environmentally friendly reduced
[32] J. Wang, Z. Chen, B. Chen, Adsorption of polycyclic aromatic hydrocarbons by graphene oxide as a broad-spectrum adsorbent for anionic and cationic dyes via
graphene and graphene oxide Nanosheets, Environ. Sci. Technol. 48 (2014) π–π interactions, J. Mater. Chem. A 4 (2016) 12126–12135.
4817–4825. [50] H. Ucun, Y.K. Bayhana, Y. Kaya, A. Cakici, O.F. Algur, Biosorption of lead (II) from
[33] P.-K. Chen, N.-C. Lai, C.-H. Ho, Y.-W. Hu, J.-F. Lee, C.-M. Yang, New synthesis of MCM- aqueous solution by cone biomass of Pinus sylvestris, Desalination 154 (2003)
48 nanospheres and facile replication to mesoporous platinum nanospheres as 233–238.
highly active electrocatalysts for the oxygen reduction reaction, Chem. Mater. 25 [51] Y. Zhao, M. Ding, D. Chen, Adsorption properties of mesoporous silicas for organic
(2013) 4269–4277. pollutants in water, Anal. Chim. Acta 542 (2005) 193–198.
[34] X. Li, Z. Wang, Q. Li, J. Ma, M. Zhu, Preparation, characterization, and application of [52] T.X. Bui, S.-Y. Kang, S.-H. Lee, H. Choi, Organically functionalized mesoporous SBA-15
mesoporous silica-grafted graphene oxide for highly selective lead adsorption, as sorbents for removal of selected pharmaceuticals from water, J. Hazard. Mater.
Chem. Eng. J. 273 (2015) 630–637. 193 (2011) 156–163.