Adsorptive denitrogenation of fuel over molecularly imprinted poly-2-(1: H -imidazol-2-yl)-4-phenol microspheres

Article abstract of DOI:10.1039/c8nj02818d

Molecularly imprinted poly-2-(1H-imidazol-2-yl)-4-phenol microspheres (MIPs) were prepared by suspension polymerization of 2-(1H-imidazol-2-yl)-4-vinylphenol in the presence of nitrogen containing compounds (N-compounds). Nitrogen containing compounds suc

Full text of DOI:10.1039/c8nj02818d

NJC
View Article Online
View Journal
PAPER
Adsorptive denitrogenation of fuel over molecularly
imprinted poly-2-(1H-imidazol-2-yl)-4-phenol
microspheres†
Cite this:DOI: 10.1039/c8nj02818d
M. S. Abdul-quadir,a R. van der Westhuizen,b W. Welthagen,b E. E. Ferg,a
a
Z. R. Tshentua and A. S. Ogunlaja
*
Molecularly imprinted poly-2-(1H-imidazol-2-yl)-4-phenol microspheres (MIPs) were prepared by suspension
polymerization of 2-(1H-imidazol-2-yl)-4-vinylphenol in the presence of nitrogen containing
compounds (N-compounds). Nitrogen containing compounds such as quinoline, pyrimidine and
carbazole, largely found in fuel oils, were employed as templates in MIP synthesis. Isothermal titration
calorimetric (ITC) studies revealed that quinoline/2-(1H-imidazol-2-yl)-4-phenol (PIMH) and pyrimidine/
2-(1H-imidazol-2-yl)-4-phenol (PIMH) interactions were exothermic (ꢀ18.0 ꢁ 4.8 and ꢀ28.1 ꢁ 3.6 kJ molꢀ1
,
respectively), while the carbazole/2-(1H-imidazol-2-yl)-4-phenol (PIMH) interaction was endothermic
(3.3 ꢁ 0.7 kJ molꢀ1). Imprinted microspheres show selectivity for the various target nitrogen-containing
compounds in fuel oil with adsorption capacities of 10.56 mg gꢀ1, 11.71 mg gꢀ1 and 10.84 mg gꢀ1 for
pyrimidine, carbazole and quinoline, respectively. The removal of other stubborn nitrogen containing
compounds from fuel oil demonstrates their potential application in the field of fuel denitrogenation.
Received 6th June 2018,
Accepted 30th June 2018
DOI: 10.1039/c8nj02818d
rsc.li/njc
that causes many environmental hazards like acid rain and
photochemical smog. The pollution effect of this NOx also
1. Introduction
Fuel oil is one of the major sources of energy for industries and includes formation of small particles which penetrate the lungs
vehicles in many countries of the world, and it is obtained and cause respiratory diseases such as bronchitis and asthma.
mainly from fossil sources. These fossil fuels contain many Due to the harm posed by these gases, the US EPA aimed at
contaminants known as heteroatoms, e.g. nitrogen containing improving air quality mandated the reduction of nitrogen
compounds, sulfur containing compounds and metals.1–3 The content in diesel fuel to ultra-low levels of o1 mg Lꢀ1 16,17
.
presence of nitrogen containing compounds such as basic
Several methods have been reported on the denitrogenation
nitrogen compounds (e.g. pyridine and its benzologues) and of fuel and these include catalytic hydrodenitrogenation
non-basic nitrogen compounds (e.g. pyrrole and its benzologues (HDN),3 which takes place under extreme temperature and
such as indoles and carbazoles) in fuel constitutes a major pressure,18–20 and it is known to be kinetically slow.4 HDN is
environmental and health concern even though their percentage effective in the removal of the less bulky non-refractory nitro-
in fuel is very small (less than 0.5%).4–9
gen containing compounds. The technique is ineffective in the
These organonitrogen compounds also give rise to fuel elimination of refractory organonitrogen compounds such as
instability during storage due to the formation of gums, colours highly alkylated carbazole and quinoline,21 thus, leading to a
and sediments. Second, compounds such as quinoline and fall out in the mandated nitrogen specification of o1 mg Lꢀ1
.
related molecules are capable of binding strongly on catalyst Extractive denitrogenation is another method that has been
surfaces, hence, deactivating refining catalyst.10–15 The most applied for the elimination of nitrogen containing compounds
harmful effect of nitrogen molecules in fuel is the emission of (NCCs) in fuel.22 This method suffers from the fact that other
NOx into the environment which is a major source of pollution compounds can also be extracted along with NCCs. Adsorptive
denitrogenation (ADN) is the most recent method that is being
applied to remove NCCs from fuel oil. So far, many adsorbents
a Department of Chemistry, Nelson Mandela University, P.O. Box 77000,
such as activated carbon, zeolites, silica, and ion-exchange resins
have been used in ADN.23–26 Nevertheless, most of the adsorbents
Port Elizabeth 6031, South Africa. E-mail: adeniyi.ogunlaja@mandela.ac.za;
Tel: +27 46 504 3061
suffer from limitations such as that they may complex with other
compounds in the fuel. Recently, porous materials, including
b Analytical Technology, Sasol Technology (Pty) Limited, P.O. Box 1,
Sasolburg 1947, South Africa
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c8nj02818d mesoporous materials like metal organic frameworks (MOFs),
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2018
New J. Chem.

View Article Online
Paper
NJC
have been used in the denitrogenation of fuel (offering adsorp- 77 K using a TriStar II 3020 3.02, Surface Area and Porosity
tion capacities of above 150 mg gꢀ1).1 Ahmed et al. (2014) also Analyzer by Micromeritics Instrument Corporation. Prior to
demonstrated that an adsorption capacity of 357 mg gꢀ1 for each measurement, microspheres were degassed at 90 1C
quinoline using MIL-100(Fe) can be achieved at an equilibrium for 24 h. The BET surface area, total pore volume and pore
time of 120 min and an adsorption capacity of 417 mg gꢀ1 can size distribution were calculated from these isotherms. The
be achieved when AlCl3(2.5%)/MIL-100(Fe) was employed.27 solid reflectance spectra of the metal complexes were recorded
A 17% improvement in the adsorption of quinoline was attributed on a Shimadzu UV-VIS-NIR Spectrophotometer UV-3100 with a
to the Lewis acid–base interaction between AlCl3 and quinoline.27 MPCF-3100 sample compartment with samples mounted
These MOFs displayed fast adsorption kinetics/rates, with high between two quartz discs which fit into a sample holder coated
adsorption capacities; nevertheless, MOFs are not equally selective with barium sulfate. The spectra were recorded over the wave-
to NCCs.28–33 Hence, there is a need to develop more selective length range of 2000–250 nm, and the scans were conducted
sorbents that can remove the NCCs from fuel.34 Cao et al. (2014) at a medium speed using a 20 nm slit width. LECO Pegasus
recently developed indole-MIPs which gave an adsorption capacity GC ꢂ GC-HRT was employed to monitor the adsorption of
of 31.80 mg gꢀ1 at an equilibrium time of 90 min. It was further organonitrogen compounds in diesel. Injection: split injection
reported that the synthesised indole-MIPs also selectively removed (100 : 1) at 250 1C; primary column: Stabilwax (Restek), 30 m ꢂ
stubborn non-nitrogen indole compounds from the fuel oil.35 250 mm (0.25 mm); secondary column: Rxi-5 (Restek), 1.5 m ꢂ
Yang et al. (2013) designed SiO2@MIPs for the selective removal of 100 mm (0.1 mm); carrier gas: helium, at a constant flow rate of
indole and carbazole type compounds. In their study, an adsorp- 1.2 mL per minute; primary oven program: 40 1C (0.1 minute) to
tion capacity of 60 mg gꢀ1 was attained in 150 min and a 260 1C (78.4 min) at 3 1C per minute; secondary oven program:
corresponding adsorption amount of 23.54 mg gꢀ1 was attained 45 1C (0.1 minute) to 265 1C (56 minute) at 3 C per minute;
in simulated oil (100 mg Lꢀ1).36
modulator offset: 15 1C; modulation frequency: 8 seconds;
To the best of our knowledge, imprinted 1H-imidazol-2-yl-4- hot time: 2 seconds; MS: LECO Pegasus 4D GC ꢂ GC-TOFMS;
phenol based microspheres have not been prepared as denitro- ionization: electron ionization at 70 eV; source temperature:
genation adsorbents. Hence, for the first time in this study, we 250 1C; stored mass range: 30 to 500 u; acquisition rate:
describe the selective adsorption of nitrogen containing com- 100 spectra per second.
pounds (such as quinoline, carbazole and pyrimidine) over
molecularly imprinted poly-2-(1H-imidazol-2-yl)-4-phenol micro-
2.3 Synthesis of imprinted microspheres
spheres. The formation of intense green coloured pyrimidine Prior to the synthesis of imprinted microspheres, the monomer
imprinted poly-2-(1H-imidazol-2-yl)-4-phenol microspheres was 2-(1H-imidazol-2-yl)-4-vinyl phenol was synthesized by a method
also observed for the first time and characterised spectroscopi- reported by Sellergren and co-workers,37 with a few modifications.
cally for its absorption band. A combined experimental and
2.3.1 Synthesis of the precursor 4-bromo-2-(1H-imidazol-2-yl)
spectroscopic study was adopted to gain a fundamental under- phenol. 4-Bromo-2-hydroxybenzaldehyde (10 g, 0.05 mol) and
standing of the possible interactions responsible for adsorption. glyoxal trimer dehydrate (10.5 g, 0.05 mol) were added into
200 mL of glacial acetic acid and brought to reflux under
nitrogen until they dissolved and then ammonium acetate
(36 g, 0.05 mol) was added followed by continuous reflux for
4 h. After cooling to room temperature, the solution was poured
2. Experimental
2.1 Materials
into 2 L of water and filtered on celite. The dark brown filtrate
4-Bromo-2-hydroxybenzaldehyde, glyoxal trimer dihydrate, ethylene
was made basic by the addition of aqueous ammonia and the
glycol dimethacrylate, quinoline, carbazole and pyrimidine
yellow precipitate formed was filtered and dried. The product
were purchased from Sigma Aldrich, South Africa.
was dissolved in acetone and refluxed with activated carbon for
an hour. This was filtered on celite and concentrated on a
rotavapor to give a purple product weighing 1.5 g. This was
2.2 Instrumentation
FT-IR spectra (4000–400 cmꢀ1) were run on a Bruker, Tensor further purified on a silica column with ethyl acetate and ethanol
27 platinum ATR-FTIR spectrometer. The 1H NMR spectra of as eluents (ratio 1 : 1). This was then concentrated using a rotavap
ligands and monomer were recorded on a Bruker 400 MHz and the solid product was weighed (Scheme 1). Yield = 22%.
spectrometer in DMSO-d6. Thermogravimetric analyses of Melting point = 240–242 1C. FT-IR (cmꢀ1, neat): 3368, 2414, 1586,
imprinted and non-imprinted microspheres were performed 1530, 1494, 1463, 1280, 1252. 1H NMR (400 MHz, DMSO-d6): 12.44
using a Perkin-Elmer TGA 7 thermogravimetric analyser (TGA). (s, 1H), 8.09 (d, 1H), 7.38 (d, 2H), 7.07 (br-s, 2H), 6.92–6.90 (d, 1H).
Typically, the samples were heated at a rate of 10 1C minꢀ1 13C NMR (400 MHz, DMSO-d6): 159.86, 155.49, 131.98, 126.62,
under a constant stream of nitrogen gas. Microspheres were 123.98, 119.90, 115.34, 110.62. Anal. calcd (found) for C9H7BrN2O
imaged using a TESCAN Vega TS 5136LM scanning electron (%): C, 45.22 (44.95); H, 2.95 (3.203); N, 11.72 (10.71).
microscope (SEM). Before images were taken, the microspheres
2.3.2 Synthesis of 2-(1-benzoyl-1H-imidazol-2-yl)-4-bromo-
were coated to prevent surface charging and to protect the phenyl benzoate. 4-Bromo-2-(1H-imidazol-2-yl) phenol (0.8 g,
surface material from thermal damage by the electron beam. 0.003 mol) was dissolved in 30 mL of THF and triethylamine
Nitrogen adsorption/desorption isotherms were measured at (0.61 g, 0.006 mol) was added followed by benzoyl chloride
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2018

View Article Online
NJC
Paper
Scheme 1 The synthesis scheme for 2-(1H-imidazol-2-yl)-4-vinyl phenol.
(0.84 g, 0.006 mol). The solution was stirred overnight and then 133.94, 130.87, 129.54, 128.78, 128.46, 122.95, 121.65, 115.28.
filtered. The filtrate was concentrated to give a purple solid Anal. calcd (found) (%) for C25H18N2O3: C, 76.13 (76.05); H, 4.60
which was digested in diethyl ether and washed in the same (4.84); N, 7.10 (6.28).
solvent. A yield of 40% was observed (Scheme 1). Yield =
2.3.4 Synthesis of 2-(1H-imidazol-2-yl)-4-vinylphenol. 0.45 g
15.13%. Melting point = 131–133 1C. FT-IR (cmꢀ1 neat): 1738, of 2-(1-benzoyl-1H-imidazol-2-yl)-4-vinylphenyl benzoate was dis-
1
1664, 1600, 1577, 1474, 1444, 1396, 1260. H NMR (400 MHz, solved in 10 mL of ethanol. Sodium ethoxide was prepared
DMSO-d6): 7.83 (m, 2H), 7.74 (M, 1H), 7.52 (m, 5H), 7.56 (m s., separately by adding 50 mg of sodium in dry ethanol (50 mg).
1H), 7.42 (M, 4H).13C NMR (400 MHz, DMSO-d6): 166.64, This was then added to the product solution drop wise till the
163.57, 147.12, 134.11, 134.03, 133.07, 133.37, 130.75, 134.11, mixture becomes yellow when it was concentrated and dissolved
130.34, 129.58, 129.17, 128.51, 127.88, 125.04, 121.97, 118.01. in 50 mL of 0.20 mol NaOH. The pH was adjusted to 10 by
Anal. calcd (found) for C23H15BrN2O3 (%): C, 61.76 (60.08); adding 0.1 mol HCl until the solution becomes cloudy; the
H, 3.38 (3.55); N, 6.26 (6.42).
precipitate was filtered and centrifuged and then washed with
2.3.3 Synthesis of 2-(1-benzoyl-1H-imidazol-2-yl)-4-vinyl- ice cold methanol followed by cold diethyl ether to give a
phenyl benzoate. 2-(1-Benzoyl-1H-imidazol-2-yl)-4-bromophenyl pure cream solid (Scheme 1). Yield: 37.7%. Melting point =
benzoate (1 g, 2.24 mol) was dissolved in warm toluene. Palladium 133–134 1C. FT-IR (cmꢀ1, neat): 3369, 2918, 1609, 1532, 1494,
tetrakis(triphenylphosphine) (0.052 g, 0.045 mol) was added 1464, 1413, 1279, 1253. 1H NMR (400 MHz, DMSO-d6): 14.28
followed by the addition of tributyl (vinyl) tin (0.65 mL, (s, 1H), 8.05 (s, 1H), 7.33 (d, 1H), 6.90 (d, 1H), 6.70–6.62 (m,
2.24 mol) and a trace of 3,5-diterbutylcatechol. The mixture 1H vinyl-H), 5.90–5.86 (d, 1H vinyl-H), 5.23–5.20 (d, 1H vinyl-H).
was heated to reflux under nitrogen for 5 h. After cooling, it was 13C NMR (400 MHz, DMSO-d6): 155.50, 144.86, 131.98, 127.81,
filtered and the filtrate was concentrated to give a yellow sticky 126.63, 118.97, 115.35, 109.87. Anal. calcd (found) for C11H10N2O
1
liquid. This was dissolved in 40 mL of acetonitrile and 20 mL of (%): C, 70.97 (71.00); H, 5.40 (5.41); N, 15.02 (15.01). H-NMR
hexane was added for washing the resulting solution. The lower spectra of (A) 4-bromo-2-(1H-imidazol-2-yl) phenol, (B) 2-(1-benzoyl-
layer of the two immiscible liquids was run into a beaker; it is 1H-imidazol-2-yl)-4-bromophenyl benzoate, (C) 2-(1-benzoyl-1H-
yellow and contained the product. This was concentrated and imidazol-2-yl)-4-vinylphenyl benzoate and (D) 2-(1H-imidazol-2-
gave a yellow solid to which diethyl ether was added which yl)-4-vinyl phenol are all presented in Fig. S1, ESI.†
coagulated it into a solid which was filtered. Yield: 85% (Scheme 1).
2.3.5 Synthesis of imprinted poly-2-(1H-imidazol-2-yl)-4-
Melting point = 122–124 1C. FT-IR (cmꢀ1 neat): 1737, 1710, 1600, phenol microspheres. 2-(1H-Imidazol-2-yl)-4-vinylphenol (1.86 g)
1
1500, 1450, 1259. H NMR (400 MHz, DMSO-d6): 7.83 (m, 2H), was dissolved in a solution containing 2 mL of ethylene glycol
7.76 (s, 1H), 7.61–7.48 (m, 5H), 7.34–7.22 (m, 5H), 7.20 (M, 2H), dimethacrylate and 1 mL of toluene in the presence of nitrogen
6.80 (m, 1H vinyl-H), 5.90 (d, 1H vinyl-H), 5.36 (d, 1H vinyl- H); containing compound(s) (carbazole (0.176 g, 0.001 mol), quino-
13C NMR (400 MHz, DMSO-d6): 166.78, 163.79, 147.26, 144.08, line (0.118 mL, 0.001 mol) and pyrimidine (0.08 g, 0.001 mol)
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2018
New J. Chem.

View Article Online
Paper
NJC
were added separately to the organic phase) (Scheme 2). After of nitrogen compounds (120 mg Lꢀ1) alongside other molecules.
homogeneity was achieved, 0.02 g of benzoyl peroxide was The screw-capped vials containing nitrogen compounds were sub-
added to the organic mixture. The resulting organic solution jected to mechanical agitation at 100 rpm for 12 h. The progress of
was transferred in drops into a 100 mL solution containing adsorption of the various nitrogen compounds was followed by
polyvinyl alcohol (PVA) (0.025 g) and NaCl (0.07 g) at 80 1C and withdrawing aliquots for measurement after every hour. Adsorption
under stirring at 400 rpm. The reaction was allowed to proceed capacity, qe (mg gꢀ1), was calculated from eqn (1).
for 7 h under an argon atmosphere, after which the resulting
VðC0 ꢀ CeÞ
microspheres were collected by filtration.
qe ¼
(1)
W
Imprinted template molecules (nitrogen containing com-
pounds) were removed by washing the microspheres via soxhlet
extraction with a solution mixture of warm-to-hot methanol
and acetonitrile (1 : 1) until no templates was detected on the
GC-FID. The synthetic routes for the formation of nitrogen-
imprinted microspheres are similar to the steps described for
the non-imprinted microspheres (Section A of ESI†).
where C0 is the initial concentration (mg Lꢀ1), Ce the equili-
brium concentration (mg Lꢀ1), W the weight of microspheres
(g) and V the volume (L). Molecularly imprinted microspheres
employed for adsorption were regenerated for re-usability by
washing microspheres with a warm solvent mixture of acetonitrile
and methanol (1 : 1). The adsorption kinetics and isotherm pro-
tocols are presented in the ESI.†
2.4 Adsorption and desorption studies of model nitrogen
compounds
2.4.2 Adsorption selectivity studies of imprinted microspheres.
Selectivity studies were carried out to evaluate the loading capacity
2.4.1 Adsorption studies of model nitrogen compounds. and selectivity of molecularly imprinted poly-2-(1H-imidazol-2-yl)-4-
Adsorption studies of nitrogen compounds were performed by phenol microspheres. 50 mg of imprinted microspheres were
(i) batch process for better understanding of the adsorption added to vials and mixed with a 3 mL solution mixture of nitrogen
kinetics, selectivity and isotherms, and (ii) solid phase extrac- containing compounds (quinoline, carbazole and pyrimidine),
tion (SPE) process in the reusability studies. To perform the dibenzothiophene and naphthalene (120 mg Lꢀ1 each); the adsorp-
batch adsorption studies, imprinted polymer microspheres tion equilibrium was reached after 12 h. In order to quantitatively
(50 mg) was weighed into screw-capped vials containing 3 mL estimate the adsorption selectivity of various adsorbents for
Scheme 2 Scheme for the synthesis of quinoline imprinted poly-2-(1H-imidazol-2-yl)-4-phenol microspheres.
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2018

View Article Online
NJC
Paper
each compound a selectivity factor was used, which is expressed
as follows:
ai–r = (qi/qr) ꢀ (Ce,i/Ce,r
)
where qi and qr are the adsorption capacities (mg gꢀ1) of
compound i and the reference compound r at equilibrium,
respectively. Ce,i and Ce,r (mg Lꢀ1) are the equilibrium concentra-
tions of compound i and the reference compound r, respectively.
Naphthalene was selected as a reference compound in this study.
2.5 Thermodynamic studies
2.5.1 Isothermal titration calorimetry (ITC) and Van’t Hoff
experiments. ITC was employed for the determination of the
degree of molecule–molecule interactions in solution.38–47 For
this experiment, N-compounds–PIMH titration was performed
on a modular titration nanocalorimeter with an injection
volume of 10 mL, a time spacing of 10 min between injections
and a stirring speed of 40 rpm and at 25 1C. The reference cell
was left empty.
Van’t Hoff experiments for all model organonitrogen com-
pounds were carried out at 30, 35, 40 and 45 1C by using a batch
adsorption process. Aliquots of analyte solutions were withdrawn
at every 20 min intervals and analysed by using a GC-FID.
Fig. 1 Diffuse reflectance electronic spectra of non-imprinted microspheres,
pyrimidine-imprinted microspheres (PYMMIPs), quinoline-imprinted micro-
spheres (QUNMIPs) and carbazole-imprinted microspheres (CARMIPs).
hence PYMMIPs appear with a light blue-green colour.48 Images
of the various microspheres are provided in Fig. S3 in ESI.†
3.2 Thermogravimetric analysis (TGA) of non-imprinted and
imprinted microspheres
Polymer microspheres (imprinted and non-imprinted) dis-
played distinct weight losses at ranges of around 90–140 1C,
250–360 1C and 400–600 1C under a nitrogen atmosphere.
3.2.1 Non-imprinted microspheres (NIPs). NIPs presented
a decomposition pattern with three distinct weight losses at
around 280 1C, 450 1C and 500 1C (Fig. 2). An observed 4% weight
loss around 310 1C was assigned to loosely bound organic molecules
within the microspheres. The decomposition of non-imprinted
microspheres’ backbone began to occur only at temperatures of
around 320 1C to about 450 1C, hence resulting in a total polymer
weight loss of 68%. A further microsphere collapse (about 12%)
occurred between 500 and 600 1C.
3. Results and discussion
3.1 FT-IR and UV-Vis spectroscopy
FT-IR analyses of imprinted and non-imprinted microspheres
are as follows: non-imprinted microspheres: FT-IR (cmꢀ1):
3371 n(O–H), 1726 n(CQO), 1630 n(4NH), 1289 n(C–N), 1148
n(C–O–C). Carbazole-imprinted microspheres: FT-IR (cmꢀ1):
3393 n(O–H), 1722 n(CQO), 1650 n(4NH), 1233 n(C–N), 1111
n(C–O–C); pyrimidine-imprinted microspheres: FT-IR (cmꢀ1):
3563 n(O–H), 1720 n(CQO), 1650 n(4NH), 1231 n(C–N), 1140
n(C–O–C); quinoline-imprinted microspheres: FT-IR (cmꢀ1):
3408 n(O–H), 1721 n(CQO), 1630 n(4NH), 1260 n(C–N), 1156
n(C–O–C). FT-IR spectra of (i) non-imprinted microspheres,
(ii) carbazole-imprinted microspheres, (iii) pyrimidine-imprinted
microspheres and (iv) quinoline imprinted microspheres are all
presented in Fig. S2 of ESI.† The diffuse reflectance spectra of
non-imprinted microspheres, quinoline-imprinted microspheres
(QUNMIPs), and carbazole-imprinted microspheres (CARMIPs)
exhibited a similar absorption pattern, except for pyrimidine-
imprinted microspheres (PYMMIPs) (Fig. 1). The UV spectra of
the various polymer microspheres show one sharp absorption
maximum at around 346–356 nm which is mostly assigned to
n - p* transition. Transitions (n - p*) for non-imprinted
microspheres, pyrimidine-imprinted microspheres (PYMMIPs),
quinoline-imprinted microspheres (QUNMIPs) and carbazole-
imprinted microspheres (CARMIPs) are observed at 354 nm,
346 nm, 353 nm and 356 nm, respectively. Pyrimidine-imprinted
microspheres (PYMMIPs) also displayed another band appearing
at 733 nm (Fig. 1), thus confirming the interaction between
2-(1H-imidazol-2-yl)-4-phenol and PYM in crosslinked micro-
spheres resulting in a possible intermolecular transition, and
Fig. 2 TGA profile of carbazole-imprinted microspheres (CARMIPs),
quinoline-imprinted microspheres (PYMMIPs), pyrimidine-imprinted
microspheres (PYMMIPs) and non-imprinted microspheres under a nitro-
gen atmosphere.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2018
New J. Chem.

View Article Online
Paper
3.2.2 Quinoline-imprinted microspheres (QUNMIPs). The
NJC
quinoline-imprinted microspheres gave similar decomposition
patterns as NIPs (280 1C, 450 1C and 500 1C). 5% weight loss
attributed to strongly held (intramolecular) solvent molecules
within the microspheres was observed at around 280 1C, followed
by a rapid decomposition of the quinoline-imprinted polymer
backbone occurring from 300 1C to a temperature of about 450 1C
(Fig. 2). A total weight loss of 70% was reported. A further collapse,
410%, was reported to occur between 500 and 570 1C.
3.2.3 Pyrimidine-imprinted microspheres (PYMMIPs). The
pyrimidine-imprinted microspheres (PYMMIPs) presented
three distinct weight losses at around 250–320 1C, 400 1C and
470–500 1C. Pyrimidine-imprinted microspheres displayed a
9% weight loss at 250 1C and this is attributed to bonded solvent
molecules. A further two-step polymer backbone decomposition
was observed in the range of 280–400 1C (weight loss, 51%) and
400–420 1C (weight loss, 20%). The final decomposition (weight
loss, 8%) occurred between 420 and 520 1C (Fig. 2).
3.2.4 Carbazole-imprinted microspheres (CARMIPs). The
carbazole-imprinted microspheres (CARMIPs) presented two distinct
weight losses at around 320 1C and 450 1C. A gradual 10% weight
loss attributed to bonded (or intramolecular) solvents was observed
at a temperature of up to 320 1C, followed by a rapid one-step
polymer backbone disintegration in the temperature range of 320–
450 1C (weight loss, 68%). The final decomposition (weight loss,
20%) was noticed at temperatures between 450 and 580 1C (Fig. 2).
The observed thermograms (Fig. 2) indicated that non-
imprinted microspheres displayed a compact structure with high
thermal stability. By contrast, imprinted microspheres have a
loose structure due to the presence of pores (cavities created)
within the material, and this was evident from the observed
degradation patterns. Generally, porogen (solvent for polymer-
ization) on the template influences the morphology, temperature
stability and swelling capacity of the resultant polymers.49,50
Fig. 3 Scanning electron micrograph (SEM) images of (A) non-imprinted
microspheres, (B) quinoline-imprinted microspheres (QUNMIPs), (C) pyrimidine-
imprinted microspheres (PYMMIPs) and (D) carbazole-imprinted microspheres
(CARMIPs).
of microspheres after template removal was explored. The EDS
study confirmed that the chemical integrity of the micro-
spheres was preserved even after washing the templates off
the microspheres.
3.4 BET surface area
Barrett–Emmett–Teller (BET) and Barrett–Joyner–Halenda (BJH)
models was employed for the calculation of the surface area and the
average pore size of (A) non-imprinted microspheres, (B) quinoline-
imprinted microspheres (QUNMIPs), (C) pyrimidine-imprinted
microspheres (PYMMIPs) and (D) carbazole-imprinted microspheres
3.3 Scanning electron micrography (SEM) and energy
dispersive spectroscopy (EDS) of non-imprinted and imprinted
microspheres
The SEM micrographs of non-imprinted and imprinted micro- (CARMIPs). The nitrogen adsorption–desorption isotherms of
spheres are presented in Fig. 3. Varying diameter ranges non-imprinted microspheres and imprinted microspheres are
were observed with all microspheres, non-imprinted microspheres shown in Fig. S5 (ESI†). Specific pore sizes and surface areas
(290–367 mm), quinoline-imprinted microspheres (106–204 mm), of non-imprinted microspheres and imprinted microspheres
pyrimidine-imprinted microspheres (210–255 mm) and carbazole- are shown in Table 1. The cavities created in non-imprinted
imprinted microspheres (173–280 mm). The images revealed the microspheres, pyrimidine-imprinted microspheres (PYMMIPs),
loose structure of the imprinted microspheres and the presence quinoline-imprinted microspheres (QUNMIPs), and carbazole-
of numerous cavities and roughness on the microsphere imprinted microspheres (CARMIPs) showed well defined pore
surface. By contrast, the structure of non-imprinted micro- openings with pore diameters of 27.1 ꢁ 5.4 Å or 2.7 ꢁ 0.5 nm;
spheres was more compact and presented a smooth surface 51.6 ꢁ 7.9 Å or 5.2 ꢁ 0.8 nm; 60.9 ꢁ 4.1 Å or 6.1 ꢁ 0.4 nm and
(Fig. 3). SEM images confirmed that some of the quinoline and 75.7 ꢁ 3.2 Å or 7.6 ꢁ 0.3 nm, respectively, and fall within the
pyrimidine imprinted polymers are irregular in shape (Fig. 3), mesopore region (2 nm o pore diameter o 50 nm).52
while carbazole imprinted microspheres displayed more pores
Imprinted microspheres exhibited a much smaller surface
on their surface. The observed difference in polymer micro- area than the non-imprinted microspheres. The sharp drop in
sphere size and morphology was probably due to the different surface area was attributed to the presence of hysteresis loop in
template molecules employed in the polymerization step.51
the adsorption–desorption isotherms of imprinted microspheres
EDS images of the various non-imprinted and imprinted micro- (Fig. S5, ESI†). Hysteresis effects similar to type III were
spheres are presented in Fig. S4 (ESI†). Chemical characterization observed (Fig. S5, ESI†). Second, the imprinting effect of
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2018

View Article Online
NJC
Paper
Table 1 Surface area and pore volumes of microspheres (non-imprinted
and imprinted)
and carbazole/2-(1H-imidazol-2-yl)-4-phenol) (Table 2). Quinoline–
2-(1H-imidazol-2-yl)-4-phenol and pyrimidine–2-(1H-imidazol-2-yl)-
4-phenol interactions proceeded spontaneously compared to the
carbazole–2-(1H-imidazol-2-yl)-4-phenol interaction. During the
adsorption process, the randomness of the interacting molecules
decreases, hence giving rise to negative entropies (ꢀVE DSb). The
pyrimidine–2-(1H-imidazol-2-yl)-4-phenol interaction presented
a lower degree of randomness as compared to quinoline–2-
(1H-imidazol-2-yl)-4-phenol and carbazole–2-(1H-imidazol-2-yl)-
4-phenol interactions. The positive value of DS obtained for
carbazole–2-(1H-imidazol-2-yl)-4-phenol suggests increased
Adsorbents (microspheres)
Surface area (m2 gꢀ1
)
Pore size (Å)
NIPs
222.5 ꢁ 52.9
40.1 ꢁ 10.0
76.3 ꢁ 15.1
50.2 ꢁ 17.1
27.1 ꢁ 5.4
51.6 ꢁ 7.9
60.9 ꢁ 4.1
75.7 ꢁ 3.2
PYMMIPs
QUNMIPs
CARMIPs
NIPs, non-imprinted microspheres; QUNMIPs, quinoline-imprinted
microspheres; PYMMIPs, pyrimidine-imprinted microspheres; CAR-
MIPs, carbazole-imprinted microspheres.
templates may have influenced the observed pore diameter and randomness at the solid/solution interface with some structural
surface area of imprinted microspheres.49,50,53,54 Third, the changes in the adsorbate and adsorbent.
thermodynamic properties of template solvents in the presence
Based on the obtained enthalpy (DHb) values, quinoline–2-
of porogen and cross-linker (EGDMA) have been reported to (1H-imidazol-2-yl)-4-phenol and pyrimidine–2-(1H-imidazol-2-
influence the surface area of imprinted microspheres.55,56
yl)-4-phenol offer better interactions (exothermic in nature) as
compared carbazole–2-(1H-imidazol-2-yl)-4-phenol (endothermic
process) (Table 2). The stoichiometry of interaction (n), binding
3.5 Isothermal titration calorimetry (ITC) studies
The ITC interaction titrations between 2-(1H-imidazol-2-yl)-4- constant (K), free energy (DGb), enthalpy (DHb) and entropy (DSb)
phenol and quinoline, pyrimidine and carbazole are presented are presented in Table 2.
in Fig. 4. The amount of heat released with the progress in the
3.6 Van’t Hoff study
interaction between nitrogen containing compounds and
2-(1H-imidazol-2-yl)-4-phenol with time is presented in Fig. 4.
The Van’t Hoff plot was used to determine the thermodynamic
From the obtained thermodynamic parameters, negative free parameters such as enthalpy, entropy and Gibbs free energy of
energies were observed for all interactions (quinoline/2-(1H- the adsorption process.57 Thus, DH1 can be determined from
imidazol-2-yl)-4-phenol, pyrimidine/2-(1H-imidazol-2-yl)-4-phenol the slope of the linear Van’t Hoff plot of Kad versus (1/T).58
Fig. 4 ITC titration involving 2-(1H-imidazol-2-yl)-4-phenol (PIMH) with (A) quinoline, (B) pyrimidine and (C) carbazole. 100 mM (quinoline/pyrimidine/
carbazole) was titrated into 20 mM 2-(1H-imidazol-2-yl)-4-phenol.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2018
New J. Chem.

View Article Online
Paper
NJC
Table 2 Isothermal titration calorimetry (ITC) and Van’t Hoff data
Isothermal titration calorimetry (ITC) data
Free energy (DGb)
Stoichiometry of
the interaction (n)
Enthalpy (DHb)
(kJ molꢀ1
Entropy (DSb)
(kJ molꢀ1
)
)
(kJ Kꢀ1
)
k
Temperature (K)
PIMH–quinoline
PIMH–pyrimidine
PIMH–carbazole
ꢀ28.8
ꢀ30.3
ꢀ4.3
0.987 ꢁ 0.046
0.914 ꢁ 0.050
1.13 ꢁ 0.01
ꢀ18.0 ꢁ 4.8
ꢀ28.1 ꢁ 3.6
3.3 ꢁ 0.7
ꢀ10.3
ꢀ35.3
10.9
1.1 ꢂ 102
2.0 ꢂ 102
1.8 ꢂ 101
298
298
298
Van’t Hoff plot data
Free energy (DGb) (kJ molꢀ1
)
Enthalpy (DHb) (kJ molꢀ1) ꢂ 10ꢀ2
Entropy (DSb) (kJ Kꢀ1
)
Temperature (K)
PIMH–quinoline
PIMH–pyrimidine
PIMH–carbazole
ꢀ16.16
ꢀ37.44
ꢀ12.29
ꢀ6.74
ꢀ48.65
ꢀ7.15
0.094
0.124
0.041
298
298
298
Fig. S6 (ESI†) shows the Van’t Hoff’s plot for carbazole, quino-
Low adsorption capacities were obtained whilst using mole-
line and pyrimidine adsorptions and the calculated thermo- cularly imprinted microspheres as compared to adsorption
dynamic values (DH1, DS1 and DGꢃad) for model organonitrogen
capacities reported in our previous study where molecularly
imprinted poly-2-(1H-imidazol-2-yl)-4-phenol nanofibers were
employed. Molecularly imprinted poly-2-(1H-imidazol-2-yl)-4-
compounds are shown in Table 2. Negative DGꢃad observed for
all interactions indicated the feasibility and spontaneity of the
adsorption process.
phenol nanofibers produced adsorption capacities of 11.7 ꢁ
0.9 mg gꢀ1, 11.9 ꢁ 0.8 mg gꢀ1 and 11.3 ꢁ 1.1 mg gꢀ1 for
The negative DH1 values observed for all organonitrogen
quinoline, pyrimidine and carbazole, respectively, in our recently
compounds confirm the exothermic nature of the overall sorption
published article due to the high surface area and available
process. The positive value of DS1 observed for all model organo-
nitrogen compounds suggests increased randomness at the solid/
functionalities of the adsorbent.51 However, the data obtained
gave a higher adsorption capacity for quinoline as compared to
solution interface with some structural changes in the adsorbate
the reported value for poly-1,10-binaphthyl-2,20-diol nanofibers
and adsorbent and an affinity of molecularly imprinted poly-2-
and polybenzimidazole.40
(1H-imidazol-2-yl)-4-phenol microspheres towards the various
nitrogen containing compounds. The obtained data for carbazole
interaction were in contradiction with the results generated from
Generally, the application of imprinted poly-2-(1H-imidazol-
2-yl)-4-phenol microspheres as adsorbents for adsorptive denitro-
genation of hydrotreated fuel is relatively new and a previous study
the ITC experiment. Van’t Hoff experiments were conducted in
was carried out using nanofibers which had a higher capacity but
two phase systems [solid (microspheres)–liquid (organonitrogen
poor selectivity.60 The crosslinked nature of the microspheres
compounds) interaction], while the ITC experiment was conducted
results in specific voids for the imprinted molecules and hence
gives better selectivity. In another previous study, where quinoline-
in the same phase [liquid (2-(1H-imidazol-2-yl)-4-phenol)–liquid
(organonitrogen compounds) interaction], hence resulting in the
absence of an interface.
N-oxide imprinted polybenzimidazole (PBI) nanofibers were
employed,61 the adsorption capacity was lower (4.8 mg gꢀ1) than
what is experienced in the adsorption of neat quinoline in this
study (6.8 mg gꢀ1). When considering the rate at which equilibrium
3.7 Adsorption studies
3.7.1 Adsorption selectivity studies of imprinted micro- is reached, our microspheres presented a slower kinetics as com-
spheres. Non-imprinted microspheres gave a maximum adsorp- pared to the equilibrium time of 90 min reported by Cao and
tion of 0.31 mg gꢀ1, 0.35 mg gꢀ1 and 0.24 mg gꢀ1, respectively, co-workers.35 The observed faster kinetics was probably due to the
for quinoline, carbazole and pyrimidine. Relatively moderate improved material surface area, particle size and the functional
adsorption capacities with no selectivity were observed when monomer employed (4-vinylpyridine).35 In contrast, our sorbents
NIPs were employed. However, high adsorption capacities were presented slower kinetics (with equilibrium reached at 300 min)
observed when imprinted microspheres were employed to target and low materials’ surface areas. The observed data may have been
specific N-compounds: (i) quinoline-imprinted microspheres influenced by the functional monomer, template and the size of
(QUNMIPs) presented 6.8 ꢁ 0.2 mg gꢀ1 quinoline (Fig. S7, ESI†), the resulting polymer. Similarly, the high adsorption capacity of
(ii) pyrimidine-imprinted microspheres (PYMMIPs) presented 60 mg gꢀ1 reported for indole and carbazole type compounds was
6.3 ꢁ 0.3 mg gꢀ1 pyrimidine (Fig. S8, ESI†) and (iii) carbazole- attained in 150 min,36 in which case again the adsorbent surface
imprinted microspheres (CARMIPs) presented 5.8 ꢁ 0.3 mg gꢀ1 area was above 500 m2 gꢀ1. It is obvious that the surface area and
carbazole (Fig. S9, ESI†). The non-specific binding nature of non- the thermodynamic property of the template influence the adsorp-
imprinted microspheres gave rise to the non-selective adsorption tion kinetics and capacities of MIPs. MOFs and functionalized
reported as compared to the imprinted microspheres whose silica and zeolite are also known to possess large surface areas and
binding sites alongside cavities created via imprinting allow pores which enable the adsorption of high concentrations of
for selective adsorption.59
organonitrogen compounds.1,28–33
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2018

View Article Online
NJC
Adsorption selectivity increased in the order of quinoline
Paper
Table 3 Kinetic data of pseudo-first-order and pseudo-second-order
(ai–r = 136.9) 4 pyrimidine (ai–r = 126.2) 4 carbazole (ai–r = 86.3).
Adsorption selectivity for quinoline was greater than that for pyrimi-
dine and carbazole (Fig. S10, ESI†). This implied that the functional
groups and cavities on the surface of the MIP may be quite different,
leading to the quite different adsorption selectivity. It also indicated
that the surface of imprinted microspheres is likely to have multiple
adsorption sites (surface heterogeneity), which contribute to adsor-
bent–adsorbate interaction. A reduction of o10 mg Lꢀ1 observed for
dibenzothiophene and naphthalene (having a similar structural
orientation as carbazole and quinoline) was attributed to the
imprinting properties of the polymer microspheres. The reduction
in dibenzothiophene and naphthalene concentrations could have
resulted from p–p stacking between absorbent and adsorbates
(dibenzothiophene and naphthalene).
Pseudo-first-order kinetics Pseudo-second-order kinetics
k (hꢀ1
)
R2
k2 (g mgꢀ1 hꢀ1
)
R2
Quinoline 0.448
Carbazole 0.496
Pyrimidine 0.547
0.9992
0.9724
0.9864
0.089
0.033
0.005
0.9078
0.8634
0.1865
3.7.2 Imprinting factor (k). The imprinting constant (k) is
defined as the ratio of the adsorption capacity of imprinted
microspheres (QMIP) to the adsorption capacity of non-imprinted
microspheres (QNIP). The higher the value of k, the better is the
imprinting effect.
QMIPðAdsorption capacity of the imprinted microspheresÞ
k ¼
QNIPðAdsorption capacity of the non-imprinted microspheresÞ
where k is the imprinting factor, QNIP (mg gꢀ1) is the adsorption
capacity of non-imprinted microspheres and QMIP (mg gꢀ1) is
the adsorption capacity of imprinted microspheres.
Fig. 5 Pseudo-first-order plot of the nitrogen compounds: pyrimidine,
carbazole and quinoline.
The imprinting factor (k) value for quinoline-imprinted micro-
spheres is 21.9, for pyrimidine-imprinted microspheres 26.3,
and for carbazole-imprinted microspheres 16.6 (Table S1, ESI†).
3.7.3 Adsorption kinetics and isotherm. Adsorption kinetic
studies using imprinted microspheres showed that nitrogen
molecule adsorption was initially fast due to the availability of
surface adsorption and thereafter the adsorption rate slowed
down as surface saturation was reached, thus, limiting further
nitrogen molecule penetration (Fig. S11, ESI†). The adsorption
equilibrium was reached after 5 h. An approximate adsorption
rate per hour of 1.32 mg gꢀ1, 1.26 mg gꢀ1, and 1.16 mg gꢀ1 was
observed for quinoline, pyrimidine and carbazole, respectively.
The pseudo-first-order and pseudo-second-order parameters
(coefficients) are presented in Table 3. Based on the obtained
correlation coefficients (R2), carbazole, quinoline and pyrimidine
fitted the pseudo-first-order model. The pseudo-first-order model
is presented in Fig. 5.
The adsorption behaviour of nitrogen compounds on imprinted
nanofibers fitted the Freundlich isotherm (better correlation coeffi-
cient R2, Fig. 6 and Table 4)58 (Fig. S12, ESI†). The Freundlich
isotherm indicated multilayer adsorption on the adsorbent
surface, implying nitrogen compound–nitrogen compound
interactions on the adsorbent.
Fig. 6 Freundlich plot for carbazole, quinoline and pyrimidine.
carbazole (Fig. S13, ESI†), quinoline (Fig. S14, ESI†) and pyr-
imidine (Fig. S15, ESI†). This reduction in adsorption capacities
upon employing the microspheres for several cycles indicated
that the imprinting integrity may have been compromised.58
Table 5 presents the various concentrations of nitrogen con-
taining compounds taken up after each cycle.
3.8 Adsorbent reusability studies
Reusability studies on the imprinted microspheres were carried
out using the solid phase extraction (SPE) technique.59,62 The
rebinding adsorption capacities of imprinted microspheres
3.9 Adsorption of nitrogen compounds in diesel fuel
decreased significantly for all nitrogen compounds as we Prior to denitrogenation, some alkylated organonitrogen com-
moved from the 1st adsorption cycle to the 3rd cycle for pounds were identified in the hydrotreated diesel by employing
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2018
New J. Chem.

View Article Online
Paper
NJC
Table 4 Parameters of the Langmuir adsorption model and the Freun-
dlich adsorption model
complexities were observed as the adsorbent also wiped out
some non-imprinted compounds, such as cumidine, 1-phenyl-
1-propanamine and 4-pentyloxyaniline. This was attributed to
the ability of the absorbent to easily interact with these com-
pounds via hydrogen bonding. Imprinted microspheres were
unable to eliminate 3-(N,N-dimethylamino)-9-methylcarbazole
completely due to the highly alkylated nature of the compound,
which inhibited its interactions with the active sites of the
adsorbent. 9H-Carbazol-3-amine, 9-ethyl- (Table 6), on the other
hand, was eliminated as it could easily interact with the polymer
microspheres via hydrogen bond formation and through pi–pi
stacking (a weak interaction). The chemical nature of these
unwanted N-compounds in fuel influences the adsorption prop-
erties of MIPs in targeting these compounds (Fig. 7).
Langmuir parameters
Freundlich parameters
Qm
R2
n
R2
Quinoline
Carbazole
Pyrimidine
65.79
12.56
0.665
0.0670
0.7460
0.8610
1.06
1.02
1.19
0.9845
0.9942
0.9831
Table 5 Various amounts of nitrogen containing compounds absorbed
after each adsorption cycle
First adsorption Second adsorption Third adsorption
cycle (mg Lꢀ1
)
cycle (mg Lꢀ1
)
cycle (mg Lꢀ1
)
Carbazole
Quinoline
Pyrimidine
104.5 ꢁ 4.1
99.6 ꢁ 6.0
98.8 ꢁ 4.2
93.6 ꢁ 5.6
87.2 ꢁ 4.2
89.2 ꢁ 4.1
84.9 ꢁ 3.2
78.4 ꢁ 3.1
81.7 ꢁ 3.6
To confirm the selective ability of imprinted polymers
further, another test was performed. A mixture composed of
pyrimidine, quinoline and carbazole was prepared firstly and
transferred to real diesel (Fig. S16, ESI†). Then 150 mg of MIP
NB: initial concentration = 120 mg Lꢀ1
.
LECO Pegasus GC ꢂ GC-HRT (Fig. 7). A reduction in the mixture (CARMIPs, PYMMIPs, and PYMMIPs) was added to the
concentration of alkylated organonitrogen compounds in the mixture. When the absorption equilibrium was achieved (5 h),
hydrotreated fuel was observed after adsorption (Fig. 7). MS the adsorbed amounts of pyrimidine, quinoline and carbazole
data revealed that the peak area of some identified alkylated were determined using GC (Fig. S16, ESI†). A reduction in
organonitrogen compounds reduced after adsorption (Table 6). the concentrations (peak areas) of pyrimidine, quinoline and
Imprinted adsorbents show promise owing to their ideal physical carbazole was observed in the hydrotreated fuel after adsorp-
and chemical characteristics, thus enabling the adsorption tion (Fig. S16, ESI†). The plot in Fig. 8 shows the percentage
of refractory organonitrogen compounds. However, some reduction of target nitrogen-containing compounds in diesel oil.
Fig. 7 GC ꢂ GC-high-resolution TOF-MS surface plot showing (A) surface contour plot of hydrotreated diesel (XIC) and (B) surface contour plot (XIC)
after adsorption of hydrotreated diesel fuel.
Table 6 Some alkylated organonitrogen compounds found before and after the adsorption of hydrotreated fuel
Name
Formula
R.T. (s)
Base mass
Area (before adsorption)
Area (after adsorption)
Cumidine
1-Phenyl-1-propanamine
4-Pentyloxyaniline
1-Methyl-2,5-dipropyldecahydroquinoline
3-(N,N-Dimethylamino)-9-methylcarbazole
9H-Carbazol-3-amine, 9-ethyl
2-Butyl-1-pyrroline
C9H13
C9H13
C11H17NO
C16H31
C15H16N2
C14H14N2
N
N
1723.39, 3.35143
1915.33, 3.32572
2930.96, 5.05715
4002.58, 2.34857
4002.58, 2.48982
3954.6, 2.50286
3478.83, 2.91429
1 200 932
1 060 777
1 091 012
1 941 088
2 241 195
2 101 401
830 855
4676
95 933
47 285
7818
42 756
30 379
25 596
ND
ND
ND
ND
11 525
ND
ND
N
C8H15N
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2018

View Article Online
NJC
Paper
References
1 J. Ahmed, N. A. Khan, Z. Hasan and J. H. Jhung, J. Hazard.
Mater., 2013, 250–251, 37.
2 V. C. Srivastava, RSC Adv., 2012, 2, 759.
3 Q. Shi, D. Hou, K. H. Chung, C. Xu, S. Zhao and Y. Zhang,
Energy Fuels, 2010, 24, 2545.
4 Q. Wei, S. Wen, X. Tao, T. Zhang, Y. Zhou, K. Chung and
C. Xu, Fuel Process. Technol., 2015, 129, 76.
´
5 G. C. Laredo, P. M. Vega-Merino, F. Trejo-Zarraga and
J. Castillo, Fuel Process. Technol., 2013, 106, 21.
6 M. Z. M. Salleh, M. K. Hadj-Kalic, H. F. Hizaddin and
M. A. Hashima, Sep. Sci. Technol., 2018, 61, 196.
7 G. C. Laredo, N. V. Likhanova, I. V. Lijanova, B. Rodriguez-
Heredia, J. J. Castillo and P. Perez-Romo, Fuel Process.
Technol., 2015, 130, 38.
8 J. W. Bauserman, K. M. Nguyen and G. W. Mushrush, Pet.
Sci. Technol., 2004, 22, 1491.
9 L. Xe., F.-R. Alain, P.-R. Stephane, X. X. Wang, X. Fu,
J. Estager, M. Vrinat and M. Lemaire, Ind. Eng. Chem. Res.,
2008, 47, 8801.
Fig. 8 Percentage distribution of target nitrogen-containing compounds
in diesel oil.
A total of 88.4%, 97.6% and 90.3% was selectively adsorbed for
pyrimidine, carbazole and quinoline, respectively (Fig. 8). The
corresponding adsorption amounts for pyrimidine, carbazole and
quinoline were 10.56 mg gꢀ1, 11.71 mg gꢀ1 and 10.84 mg gꢀ1
,
respectively, in the real diesel oil solution (200 mg Lꢀ1).
10 C. A. Hughey and S. A. Glassow, Fuel, 2007, 86(56), 758.
11 H. Zhang, G. Li, Y. Jia and H. Liu, J. Chem. Eng. Data, 2010,
55, 173.
12 I. Ahmed and S. H. Jhung, J. Hazard. Mater., 2016, 301, 259.
13 B. Pawelec, R. M. Navarro, J. M. Campos-Martin and
J. L. G. Fierro, Catal. Sci. Technol., 2011, 1, 23.
14 J. W. Bauserman, G. W. Mushrush and D. R. Hardy, Ind.
Eng. Chem. Res., 2008, 47, 2867.
15 P. W. Seo, I. Ahmed and S. H. Jhung, Chem. Eng. J., 2016,
299, 236.
16 A. Jayaraman, F. H. Yang and R. T. Yang, Energy Fuels, 2006,
20, 909.
17 V. C. Srivastava, RSC Adv., 2012, 2, 759.
18 X. Fan, G.-F. Liu, Z.-M. Zong, X.-Y. Zhao, J. P. Cao, B.-M. Li,
W. Zhao and X.-Y. Wei, Fuel Process. Technol., 2013, 106, 661.
19 M. Almari, X. Ma and C. Song, Ind. Eng. Chem. Res., 2009,
48, 951.
20 J. H. Maldonado and R. T. Yang, Angew. Chem., Int. Ed.,
2004, 43, 1004.
21 G. C. Laredo, S. Leyva, R. Alvarez, M. T. Mares, J. Castillo
and J. L. Cano, Fuel, 2002, 81, 1341.
4. Conclusions
Molecularly imprinted 2-(1H-imidazol-2-yl)-4-phenol showed
enhanced adsorption capacities and selectivity for individual
nitrogen compounds due to their specific binding nature. A better
regression R2 presented by the Freundlich isotherm confirmed
multilayer adsorption attributed to the interactions between
imprinted microspheres and nitrogen compounds, and possibly
between nitrogen molecules. The adsorbent materials which exhibit
hydrophobic interactions (van der Waals or dispersion forces) and
hydrophilic interactions (hydrogen bonding, pi–pi interactions,
dipole–dipole interactions) based on SPE presented a more efficient
and quantitative extraction of N-compounds in fuels. Isothermal
titration calorimetry (ITC) revealed that quinoline–PIMH and
pyrimidine–PIMH interactions are exothermic in nature, while the
carbazole–PIMH interaction is endothermic in nature. A significant
reduction in the quantity of nitrogen compounds in hydrotreated
fuel was observed (peak area reduction) when imprinted 2-(1H-
imidazol-2-yl)-4-phenol microspheres were employed; however, the
complex nature of N-compounds in fuel influences the imprinting
properties of MIPs for targeting these unwanted compounds. The
spectroscopic analysis of all synthesised microspheres allows for a
clear identification of the possibility of forming a chromophore
when pyrimidine is employed as a template.
22 Y. Sano, K.-H. Choi, Y. Korai and I. Mochida, Energy Fuels,
2004, 18, 644.
23 J. When, X. Han, H. Lin, Y. Zheng and W. Chu, Chem. Eng. J.,
2010, 164, 29.
24 A. L. Nuzhdin, K. A. Kovalenko, D. N. Dybtsev and
G. A. Bukhtiyarova, Mendeleev Commun., 2010, 20, 57.
25 D. Liu, J. Gui and Z. Sun, J. Mol. Catal. A: Chem., 2008, 291, 17.
26 L. L. Xie, A. Farre-Reguillion, X.-X. Wang, X. Fu and
M. Lemaire, J. Chem. Eng. Data, 2010, 55, 4849.
27 I. Ahmed, J. W. Jun, B. K. Jung and S. H. Jhung, Chem.
Eng. J., 2014, 255, 623.
Conflicts of interest
There are no conflicts to declare.
Acknowledgements
´
28 G. Ferey, Chem. Soc. Rev., 2008, 37, 191.
We highly acknowledge financial support to the project in the
form of research grant from the South African National
Research Foundation and Sasol (Pty) Ltd.
29 S. Kitagawa, R. Kitaura and S. I. Noro, Angew. Chem., Int. Ed.,
2004, 43, 2334.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2018
New J. Chem.

View Article Online
Paper
NJC
30 O. M. Yaghi, M. O’Keeffe, N. W. Ockwig, H. K. Chae, 47 J. M. D. Trani, S. De Cesco, R. O’Leary, J. Plescia, C. J. do
M. Eddaoudi and J. Kim, Nature, 2003, 423, 705.
31 N. A. Khan, Z. Hasan and S. H. Jhung, J. Hazard. Mater.,
2012, 244–245, 444.
32 S. H. Jhung, N. A. Khan and Z. Hasan, CrystEngComm, 2012,
14, 7099.
Nascimento, N. Moitessier and A. K. Mittermaier, Nat.
Commun., 2018, 9, 893.
48 W. F. Su, in Principles of Polymer Design and Synthesis,
Lecture Notes in Chemistry, Springer, Berlin, Heidelberg,
2013, vol. 82, p. 89.
33 J. Kim, X. Ma, A. Zhou and C. Song, Catal. Today, 2006, 111, 74. 49 L. A. Tom, N. A. Schneck and C. Walter, J. Chromatogr. B:
34 A. S. Ogunlaja, C. du Sautoy, N. Torto and Z. R. Tshentu,
Anal. Technol. Biomed. Life Sci., 2012, 909, 61.
Talanta, 2014, 126, 61.
50 K. Golker and I. A. Nicholls, Eur. Polym. J., 2016, 75, 423.
35 Y. Cao, L. Liu, W. Xu, X. Wu and W. Huang, J. Appl. Polym. 51 Y. J. Yang, J. Y. Li, Y. R. Liu., J. Y. Zhang., B. Li and X. P. Cai,
Sci., 2014, 131, 40473. Anal. Bioanal. Chem., 2011, 400, 3665.
36 W. Yang, L. Liu, Z. Zhou, C. Qiu, P. Ma, H. Liu and W. Xu, 52 S. A. Nabavi, G. T. Vladisavljevic, E. M. Eguagie, B. Li,
New J. Chem., 2013, 37, 2758. S. Georgiadou and V. Manovic, Chem. Eng. J., 2016, 306, 214.
37 B. Sellergren, R. N. Karmalkar and K. J. Shea, J. Org. Chem., 53 P. Li, C. Mu, T. Wang, F. Lei, H. Li, Q. Huang and J. Zhou, S.
2000, 65, 4009. Afr. J. Chem., 2014, 67, 67.
38 G. Krainer, J. Broecker, C. Vargas, J. Fanghanel and S. Keller, 54 H. Y. Yuan, X. H. Ma and Z. L. Xu, Sci. China: Chem., 2011,
Anal. Chem., 2012, 84, 10715. 54, 257.
39 O. Callies and A. Hernandez Daranas, Nat. Prod. Rep., 2016, 55 M. J. Whitcombe, N. Kirsch and I. A. Nicholls, J. Mol.
33, 881. Recognit., 2014, 27(6), 297.
40 A. S. Ogunlaja, E. Hosten, R. Betz and Z. R. Tshentu, RSC 56 P. A. G. Cormack and A. Z. Elorza, J. Chromatogr. B: Anal.
Adv., 2016, 6, 39024. Technol. Biomed. Life Sci., 2004, 804, 173.
41 C. Schonbeck, R. Holm and P. Westh, Anal. Chem., 2012, 57 F. Gritti and G. Guiochon, Anal. Chem., 2006, 78, 4642.
¨
´
¨
84, 2305.
58 A. S. Ogunlaja, C. J. Mathew, N. Torto and Z. R. Tshentu,
React. Funct. Polym., 2014, 81, 61.
¨
42 K. Rajarathnam and J. Rosgen, Biochim. Biophys. Acta,
Biomembr., 2014, 1838, 69.
59 X. Liu, Y. P. Guan, Q. Wang, X. F. Ren and M. Z. Yang,
J. Appl. Polym. Sci., 2012, 126, 1956.
60 M. S. Abdul-quadir, E. E. Ferg, Z. R. Tshentu and
A. S. Ogunlaja, RSC Adv., 2018, 8, 8039.
61 A. S. Ogunlaja, M. S. Abdul-quadir, P. E. Kleyi, E. E. Ferg,
P. Watts and Z. R. Tshentu, Arabian J. Chem., 2017, DOI:
10.1016/j.arabjc.2017.05.010.
43 L. N. Wafer, W. W. Streicher, S. A. McCullum and
G. I. Makhatadze, Biochemistry, 2012, 51, 7189.
44 O. Khakshoor, S. E. Wheeler, K. N. Houk and E. T. Kool,
J. Am. Chem. Soc., 2012, 134, 3154.
¨
45 X. Wang, E. Matei, A. M. Gronenborn, O. Ramstrom and
M. Yan, Anal. Chem., 2012, 84, 4248.
46 S. Keller, H. Heerklotz and A. Blume, J. Am. Chem. Soc., 62 A. Zwir-Ferenc and M. Bizuik, Pol. J. Environ. Stud., 2006,
2006, 128, 1279.
15(5), 677.
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2018