The adsorptive extraction of oxidized sulfur-containing compounds from fuels by using molecularly imprinted chitosan materials

Article abstract of DOI:10.1016/j.reactfunctpolym.2014.04.006

An innovative approach for desulfurisation of fuels is proposed. It relies on the formation of recognition sites, complementary to oxidized sulfur-containing compounds, on cross-linked chitosan microspheres and electrospun chitosan nanofibers using the molecularly imprinted polymer technique. Benzothiophene sulfone (BTO₂), dibenzothiophene sulfone (DBTO₂) and 4,6-dimethyldibenzothiophene sulfone (4,6-DMDBTO ₂) were used as templates for the preparation of molecularly imprinted polymers (MIPs). The possible molecular interactions between imprinted chitosan adsorbent and oxidized sulfur-containing compounds were investigated by molecular modeling using density functional theory (DFT) and results indicated that interactions took place via hydrogen bonding. The molecularly imprinted polymer adsorbents (cross-linked microspheres and electrospun nanofibers) gave better selectivity for the target sulfonated compounds and the adsorption isothermal studies followed the Freundlich model. Maximum adsorption capacities of 8.5 ± 0.6 mg/g, 7.0 ± 0.5 mg/g and 6.6 ± 0.7 mg/g were observed for model BTO₂, DBTO₂ and 4,6-DMDBTO₂ respectively at 1 mL/h when imprinted nanofibers were employed, and the imprinted microspheres gave maximum adsorption capacity of 4.9 ± 0.5 mg/g, 4.2 ± 0.7 mg/g and 3.9 ± 0.6 mg/g for BTO ₂, DBTO₂ and 4,6-DMDBTO₂ respectively. Application of the nanofibers to oxidized hydro-treated fuel under continuous flow adsorption system at 1 mL/h indicated that 84% of sulfur was adsorbed, with adsorption capacity of 2.2 ± 0.2 mg/g.

Full text of DOI:10.1016/j.reactfunctpolym.2014.04.006

Accepted Manuscript
The adsorptive extraction of oxidised sulfur-containing compounds from fuels
by using molecularly imprinted chitosan materials
Adeniyi S. Ogunlaja, Matthew J. Coombes, Nelson Torto, Zenixole R. Tshentu
PII:
S1381-5148(14)00096-0
DOI:
http://dx.doi.org/10.1016/j.reactfunctpolym.2014.04.006
REACT 3370
Reference:
To appear in:
Reactive & Functional Polymers
Received Date:
Revised Date:
Accepted Date:
11 November 2013
28 January 2014
29 April 2014
Please cite this article as: A.S. Ogunlaja, M.J. Coombes, N. Torto, Z.R. Tshentu, The adsorptive extraction of
oxidised sulfur-containing compounds from fuels by using molecularly imprinted chitosan materials, Reactive &
Functional Polymers (2014), doi: http://dx.doi.org/10.1016/j.reactfunctpolym.2014.04.006
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

The adsorptive extraction of oxidised sulfur-containing compounds from fuels by using
molecularly imprinted chitosan materials
1
Adeniyi S. Ogunlaja , Matthew J. Coombes , Nelson Torto , Zenixole R. Tshentu
2
1
2,∗
1
Department of Chemistry, Rhodes University, P.O. Box 94, Grahamstown 6139, South Africa
Department of Chemistry, Nelson Mandela Metropolitan University, P.O. Box 77000, Port-Elizabeth
2
6
031, South Africa
Abstract
An innovative approach for desulfurisation of fuels is proposed. It relies on the formation of
recognition sites, complementary to oxidized sulfur-containing compounds, on cross-linked chitosan
microspheres and electrospun chitosan nanofibers using the molecularly imprinted polymer technique.
Benzothiophene
sulfone
(BTO
2
),
dibenzothiophene
sulfone
(DBTO
2
)
and
4,6-
dimethyldibenzothiophene sulfone (4,6-DMDBTO
2
) were used as templates for the preparation of
molecularly imprinted polymers (MIPs). The possible molecular interactions between imprinted
chitosan adsorbent and oxidized sulfur-containing compounds were investigated by molecular
modelling using density functional theory (DFT) and results indicated that interactions took place via
hydrogen bonding. The molecularly imprinted polymer adsorbents (cross-linked microspheres and
electrospun nanofibers) gave better selectivity for the target sulfonated compounds and the adsorption
isothermal studies followed the Freundlich model. Maximum adsorption capacities of 8.5 ± 0.6 mg/g,
7
.0 ± 0.5 mg/g and 6.6 ± 0.7 mg/g were observed for model BTO
respectively at 1 mL/h when imprinted nanofibers were employed, and the imprinted microspheres
gave maximum adsorption capacity of 4.9 ± 0.5 mg/g, 4.2 ± 0.7 mg/g and 3.9 ± 0.6 mg/g for BTO
DBTO and 4,6-DMDBTO respectively. Application of the nanofibers to oxidized hydro-treated fuel
2 2 2
, DBTO and 4,6-DMDBTO
2
,
2
2
under continuous flow adsorption system at 1 mL/h indicated that 84% of sulfur was adsorbed, with
adsorption capacity of 2.2 ± 0.2 mg/g.
Keywords: chitosan; organosulfur compounds; nanofibers; cross-linked microspheres, molecularly
imprinted polymers
∗
Corresponding author Tel.: +27 41 504 2074; Fax: +27 41 504 4236.
E-mail address: zenixole.tshentu@nmmu.ac.za (Z.R. Tshentu).

1
.
Introduction
The removal of refractory organosulfur compounds existing in hydro-treated oils is necessary for the
production of cleaner fuels [1]. Sulfur oxides (SO ) are emitted when fuels containing sulfur are
x
combusted and these oxides affect the environment adversely. Sulfur compounds also poison the
catalytic converters of vehicles. The United States environmental protection agency (EPA), European
Committee for standardization and South African regulating agencies over the years mandated the
reduction of sulfur in fuels to ultra-low levels of 10-15 ppm by 2015 (Fig. 1), in order to minimise
sulfur oxides emission [1, 2].
(
Insert Fig.1 here)
The hydrodesulfurisation (HDS) process that is currently being employed for the elimination of sulfur
in fuels is effective in the removal of the less bulky non-refractory organosulfur compounds such as
aliphatic sulfur compounds and thiophene and its derivatives. This technique (HDS) is ineffective in the
elimination of refractory organosulfur compounds such as benzothiophene and dibenzothiophene and
its derivatives which comprises of about 70% of the whole sulfur compounds available in fuels,
leading to fall out in the mandated sulfur specification of 10-15 ppm [2]. When HDS processing
conditions are further pushed to produce these ultra-low sulfur levels, the aromatics and polynuclear
aromatics content falls outside the requisite specification. Hence, the oxidative desulfurization (ODS)
process, a proposed complementary step to HDS, is suggested to be a promising technique to
achieving the anticipated sulfur limit in fuels. The ODS process involves the oxidation of refractory
organosulfur compounds to produce sulfones [2-5] which can be removed by employing either
solvents or sorbents possessing a permanent dipole affinity for polar sulfone compounds. Campos et
al. [6] reported that solvents such as acetonitrile and dimethylsulfoxides (DMSO), when employed for
the extraction of oxidized sulfur compounds (sulfones) in fuels, remove other important compounds
along with the compounds of interest, consequently resulting in fuel properties falling outside the
required specifications. Moreover, adsorption solvents recyclability and recovery becomes challenging
due to the similar properties (such as boiling point) that they share with the extracted compounds [6].
Hence, a need to develop materials (adsorbents) that can selectively extract/remove sulfone
compounds and be easily recycled is increasingly becoming important. Sorbents materials such as

molecularly imprinted polymers (MIPs) fabricated through imprinting of polymers with templates
suffice as potential adsorbents for the adsorption of these compounds [7].
Molecular imprinting of polymers is originally defined as a technique employed for the introduction of
recognition sites into polymeric matrices via the formation of bonds between the imprinting molecule
(
template) and functional groups within the polymer network [7-15].
Chitosan, an amino polysaccharide usually obtained by the deacetylation of chitin, has been
investigated for the development of sorption systems due to the presence of free amine and hydroxyl
groups which allow for chemical modifications through the formation of hydrogen bonding
interactions [16-18]. Another favourable property of chitosan is its ability to swell when in contact
with a thermodynamically compatible solvent [19, 20]. Template imprinting on chitosan gel polymers
improves its adsorptive properties and also provides selectivity due to polymer flexibility which makes
accessing of the recognition site and removal of absorbed template easier. An additional approach of
improving adsorption properties of polymers is to make nanofibers. Electrospinning is a workable
method for fabrication of fibers with diameters in tens to hundreds of nanometers. It provides polymer
fibers with controlled pore size as well as high surface-to-volume ratio, and it decreases adsorption
time and mass transfer constraints [21-23]. Cross-linking of chitosan improves its mechanical
properties, thus enhancing industrial application due to matrix strength. Glutaraldehyde and
epichlorohydrin have been reported as cross-linking agents for chitosan, with the former preserving
hydroxyl groups and later preserving the cationic amino groups [24, 25].
In this article, we describe the fabrication of imprinted chitosan polymer fibers and cross-linked
2 2
microspheres using benzothiophene sulfone (BTO ), dibenzothiophene sulfone (DBTO ), and 4,6-
dimethyldibenzothiophene sulfone (4,6-DMDBTO ) as templates and glutaraldehyde as a cross-linker.
2
Absorption studies focusing on the selectivity and adsorption capacities of the various sulfone
compounds were performed; and application on oxidized hydro-treated diesel was also investigated.
2
.
Experimental
2
.1. Materials
Chitosan with 90% deacetylation degree was purchased from Sigma Aldrich, Germany {Elem. Anal.,
%): C, 41.81; H, 7.14; N, 7.61}. Sulfur in diesel standards was purchased from Matheson Tri-Gas,
(
Texas, USA. Glutaraldehyde, triton X-114 (surfactant agent) and trifloroacetic acid (TFA) were
procured from Sigma Aldrich, South Africa. Dichloromethane {(DCM), analytical grade}, acetonitrile

(
(
HPLC grade) and methanol (HPLC grade) were purchased from Merck, South Africa. Ammonia
25%) solution was purchased from Spellbound South Africa.
2
.2
Sulfone compounds synthesis
Sulfone compounds were synthesised from the catalysed oxidation of model organosulfur compounds
such as benzothiophene, dibenzothiophene and 4,6-dimethyldibenzothiophene using poly[VO(allylSB-
co-EGDMA)] and poly[VO(sal-AHBPD)] as oxidation catalysts [2, 26]. The synthetic scheme for the
formation of benzothiophene sulfone (BTO
2
2
), dibenzothiophene sulfone (DBTO ) and 4,6-
dimethyldibenzothiophene sulfone (4,6-DMDBTO
2
) is presented in the supplementary data (Scheme
S1).
2
.2.1 Benzothiophene sulfone synthesis
.15 g (0.0011 mol) benzothiophene (BT) was placed in a 25 mL round bottom flask containing 10 mL
0
mixture of toluene/hexane (1:4) and 0.015 g (0.0135 mmol) of oxidation catalyst. The mixture was
heated to 40°C with continuous stirring at 500 rpm in an oil bath. An oxidant-to-substrate ratio of 6.8
was then added to the mixture. The oxidation reaction was allowed to proceed for 6 h under continuous
stirring after the addition of tert-butylhydroperoxide. A white precipitate of benzothiophene sulfone
(
BTO
2
) observed was collected through filtration and washed with hexane to remove unreacted
1
benzothiophene. Yield = 81%. H NMR (δ, ppm in DMSO) δ 7.83 (d, J = 7.2, 1H), 7.69 (t, J = 7.4,
1
8 6 2
H), 7.62 (t, J = 9.1, 3H), 7.34 (d, J = 6.8, 1H). Anal. Calcd(found) for C H O S (%): C, 57.81(57.49);
H, 3.64(3.89); S 19.29(19.02).
2
.2.2 Dibenzothiophene sulfone synthesis
.15 g, (0.00081 mol) dibenzothiophene (DBT) was dissolved in a 25 mL round bottom flask
0
containing 10 mL mixture of toluene/hexane (1:4) and 0.015 g (0.0135 mmol) of oxidation catalyst.
After which an oxidant-to-substrate ratio of 7.6 was added to the mixture and the oxidation reaction
was allowed to proceed for 6 h under continuous stirring at 500 rpm at 40°C. A white precipitate of
dibenzothiophene sulfone (DBTO
2
) was collected through filtration and washed with hexane to
1
remove unreacted dibenzothiophene. Yield = 98%. H NMR (δ, ppm in DMSO): 8.22 (d, J = 7.7, 2 H),
8
.00 (d, J = 7.6, 2 H), 7.82 (t, J = 7.6, 2 H), 7.67 (t, J = 7.6, 2 H). Anal. Calcd(found) for C12
10 2
H O S
(
%): C, 66.65(66.02); H, 3.73(4.04); S 14.83(14.83).

2
.2.3 4,6-Dimethyldibenzothiophene sulfone synthesis
0
.15 g (0.00071 mol) 4,6-dimethyldibenzothiophene (4,6-DMDBT) was dissolved in a 25 mL round
bottom flask containing 10 mL mixture of toluene/hexane (1:4) and 0.015 g (0.0135 mmol) of
oxidation catalyst. After which an oxidant-to-substrate ratio of 7.6 was added to the mixture and the
reaction was allowed to proceed for 6 h under continuous stirring at 500 rpm at 40°C. A white
precipitate of 4,6-dimethyldibenzothiophene sulfone (4,6-DMDBTO
2
) was collected through filtration
1
and washed with hexane to remove unreacted dibenzothiophene. Yield = 87%. H NMR (δ, ppm in
DMSO): 7.98 (d, J = 7.6, 2 H), 7.66 (t, 2 H), 7.44 (d, J = 7.5, 2 H), 2.48 (s, 6 H). Anal. Calcd(found)
12 2
for C14H O S (%): C 68.38(68.38); H, 4.95(5.25); S, 13.12(12.83).
2
.3. Instrumentation and methods
-
1
FT-IR spectra (4000-650 cm ) of the adsorbents were obtained on the Perkin Elmer 400 ATR-FTIR
spectrometer. Microanalysis was carried out using Elementar Analysen Systeme Vario® MICRO VI
6
.2 GmbH. Adsorbent images and chemical characterization were carried out by using TESCAN Vega
TS 5136LM scanning electron microscope (SEM) and energy-dispersive spectroscopy (EDS)
respectively. Before images were taken for SEM analysis, adsorbents were coated with a thin film of
gold to prevent surface charging and also to protect the material surface from thermal damage by the
electron beam. No surface coating was needed for EDS analysis. The BET surface area measurements
of adsorbents (microspheres and fibers) were performed by using a Micromeritics ASAP 2020 surface
area analyzer. Prior to analysis, the adsorbents were degassed for 10 days at 50°C. Carbon dioxide
(
2
CO ) was employed as adsorption gas for the surface area measurements. High voltage power supply
from Glassman High Voltage Inc, USA was employed for the electrospinning of polymer solutions.
Adsorption studies on the model compound were monitored by employing an Agilent 7890A gas
chromatograph fitted with flame ionization detector (GC-FID). The GC conditions for the adsorption
analyses was started with an oven temperature of 50°C ramping to 80°C for 2 min, after which it was
-1
increased to 300°C at a rate of 20°C min , and finally held for 1 min. A gas chromatograph fitted with
an atomic emission detector (GC-AED) was used to monitor the adsorption of the sulfur in diesel. The
analyses on the Agilent G2350A, GC-AED was started at an oven temperature of 45°C ramped to
-
1
1
50°C at 10°C min and then held for 5 min, subsequently increased to a final temperature of 320°C at

-1
a rate of 20°C min , and was further maintained at 320°C for 2 min. AED detector for sulfur analysis
was set at a wavelength of 181 nm (and a wavelength of 179 nm was set for carbon analysis) with a
transfer line temperature of 350°C and block cavity temperature of 320°C. In both GCs, helium was
used as carrier gas.
2
2
.3. Adsorption and desorption studies of model sulfone compounds
.3.1. Adsorption procedure of some model sulfone compounds
Batch adsorption studies were carried-out by weighing 100 mg of adsorbent (imprinted and non-
imprinted microspheres and nanofibers) into a screw-capped vials containing 2 mL of 5 mM sulfone
compounds {benzothiophene sulfone (BTO
2
2
), dibenzothiophene sulfone (DBTO ) and 4,6-
dimethyldibenzothiophene sulfone (4,6-DMDBTO
2
)}. The screw-capped vials containing sulfone
compounds were left under mechanical agitation (150 rpm) for 24 h, after which the respective
adsorbent was removed by filtration and the resultant solution analysed by using GC-FID. The
adsorption capacity, q
e
(mg/g), for each sulfone compound was calculated using Eqn. 1.
VꢂC −ꢁC ꢄ
ꢃ
ꢀ
qꢀ ꢁ= ꢁ
ꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢂ1ꢄꢁꢁꢁꢁꢁꢁ
W
o e
where C , C , W and V are the initial concentration (mg/L), equilibrium concentration (mg/L), dry
weight of adsorbent (g) and solution volume (L) respectively.
Conditioning of adsorbents was carried-out by pre-wetting adsorbents with solvents employed in
dissolving the sulfone compounds.
2
.3.2. Desorption procedure for the absorbed sulfone compounds
Adsorbed sulfonated compounds on chitosan nanofibers and microspheres were desorbed through
Soxhlet extraction process using a 50 mL solvent mixture of methanol and acetonitrile (1:1). The
extraction process was carried-out for 12 h after which the absorbents are air-dried. The desorption
procedure prepares MIPs for reuse.
2
2
.4. General procedure for the preparation of non-imprinted and imprinted adsorbent
.4.1. Fabrication of non-imprinted and imprinted chitosan nanofibers by electrospinning
A polymer solution for electrospinning was prepared by dissolving 0.5 g of chitosan in 10 mL of
TFA:DCM (7:3). The mixture was gently stirred overnight at room temperature to obtain a
homogeneous solution which was electrospun and classified as non-imprinted nanofibers. For

imprinted fibers, 1 mL (10 mM in acetonitrile) of the respective sulfone compounds containing 40 µL
triton X-114 (surfactant agent) was added drop-wise to the dissolved chitosan solution. The mixture was
then stirred for 2 h to achieve the desirable solution homogeneity for electrospinning.
Polymer solutions were respectively poured into a 25 mL syringe attached to a needle connected to the
positive electrode of a high voltage power supply. A syringe pump (Model NE-1010, New Era Pump
Systems Inc. USA) was used to supply a constant flow of polymer solution from the syringe during the
electrospinning process. A voltage of 30 kV was applied to the polymer solution which was pumped at a
flow-rate of 1.2 mL/h, with a distance between the needle tip and aluminium collector plate placed at
1
7 cm. The repulsive electrical forces between charged nanofibers enable them to spread smoothly and
the solvent evaporates resulting in solidification while traveling toward the grounded collector. A
schematic diagram of the experimental setup used for electrospinning is shown in Fig. 2.
(
Insert Fig. 2 here)
The collected nanofibers were kept overnight in the fume hood to allow for complete residual solvent
evaporation. Deprotonation of the nanofibers was carried out by soaking 0.5 g of the nanofibers
overnight into 20 mL acetonitrile containing 1 mL of the 25% stock ammonia solution. The template
molecules (sulfone compounds) imprinted within the nanofibers matrix were removed via Soxhlet
extraction using a mixture of methanol and acetonitrile (1:1) until no templates was detected on the
GC-FID. The nanofibers were then dried overnight at 40°C. Elem. Anal., (%): C, 32.33; H, 5.79; N,
5
.64.
2
.4.2. Preparation of cross-linked non-imprinted and imprinted chitosan microspheres
0
.5 g of chitosan was dissolved in 20 mL of water containing 2% (v/v) acetic acid followed by the
drop-wise addition of 1 mL of sulfone compounds (10 mM) containing 40 µL of triton X-114
surfactant agent), and the mixture was stirred until a homogenous solution was obtained. Microspheres
(
were then formed by making droplets of the chitosan solution into a beaker containing 2 M NaOH
solution. The droplets were then allowed to stand for 24 h in order to allow for proper microspheres
formation (Fig. 3).
(
Insert Fig. 3 here)

Cross-linking of the formed chitosan microspheres was necessary in order to preserve the desired
spherical shape and also to strengthen the microspheres. Glutaraldehyde was employed as a suitable
cross-linker, and the mechanism of cross-linking is presented in scheme 1.
To 5 g of the microspheres (wet imprinted and non-imprinted) in a 100 mL flask, 10 mL of 67 mM
NaOH was added, and cross-linking of the microspheres proceeded by the addition of 0.5 mL
glutaraldehyde. The reaction was allowed to proceed for 5 h at 45°C. The cross-linked microspheres
were collected by filtration and were washed extensively with distilled-deionised water to remove any
unreacted glutaraldehyde. Template removal from the glutaraldehyde-free microspheres was carried-
out through Soxhlet extraction. The microspheres were then slowly dried at 40°C for 7 h. Elem. Anal.,
(
%): C, 33.72; H, 4.74; N, 4.85. A glutaraldehyde-to-chitosan ratio of 0.7 was observed indicating that
more carbon element has been incorporated into the polymer microspheres matrix.
(
Insert Scheme 1 here)
3
3
3
.
Results and discussion
.1. Absorbent characterization
.1.1. Energy Dispersive Spectroscopy (EDS)
Chemical characterization of adsorbents after deprotonation and template removal was explored. A
reduction in flourine (F) abundance confirming deprotonation was observed when non-deprotonated
non-imprinted polymer (NIP) nanofibers were compared to the deprotonated counterpart (Table 1).
Disappearance/reduction in sulfur (S) abundance after Soxhlet extraction was also confirmed, EDS
spectra of non-imprinted and imprinted polymer nanofibers are presented in Fig. 4a-d. The use of EDS
on imprinted and non-imprinted microspheres produced no meaningful results; however sulfur
compounds on the microspheres were confirmed by using the GC.
Trifloroacetic acid (TFA) was employed as the main solvent of choice for dissolution and
electrospinning of chitosan due to the following reasons: Firstly, TFA forms salts with amino groups
of chitosan, thus destroying the rigid interaction between chitosan molecules; and secondly, the high
volatility of TFA is advantageous for the rapid solidification of the electrified jet of the chitosan-TFA
solution [27, 28].
(
Insert Table 1 here)

(
Insert Fig. 4 here)
3
.1.2. Scanning electron micrograph (SEM) images of adsorbents
Representative scanning electron micrograph (SEM) images of cross-linked chitosan microspheres and
nanofibers (non-imprinted and imprinted) are presented in Fig. 5. A diameter range of 501-667 µm
was observed for dried cross-linked microspheres, while nanofibers gave a diameter range of 310-540
nm. SEM images for the various imprinted microspheres and nanofibers are presented in the
supplementary file (Fig. S1-S2), while chitosan electrospinning optimization is provided in Fig. S3.
(
Insert Fig. 5 here)
3
.1.3. Surface area measurements of adsorbents
The surface area of adsorbents (microspheres and nanofibers) decreased as template molecular sizes
increased (Table 2). The sulfone compound size increased in the order of benzothiophene >
dibenzothiophene > 4,6-dimethyldibenzothiophene. Nanofibers generally present higher surface area
as compared to microspheres.
(
Insert Table 2 here)
3
.1.4. FT-IR spectra of adsorbents
Representative FT-IR spectra of chitosan (pristine), chitosan nanofibers and cross-linked imprinted
microspheres are shown in Fig. 6. A strong characteristic band attributed to C-O-C stretching in
-
1
polysaccharides was observed between 1000 and 1200 cm in all spectra confirming that the chitosan
-1
structure was preserved (Fig. 6). NH stretch observed around 3300 cm (Fig. S4) represent the level of
chitosan deacetylation (90%) and the band is visible on the pristine chitosan, microspheres and
nanofibers spectra. The representative chitosan further showed two bands, the observed band
-
1
transitions at about 1542 cm was assigned to amino groups -NH
2
(bending) while the other band at
644 cm was assigned to the carbonyl group of acetylamino groups (C=O). These bands are visible
on the pristine chitosan and nanofibers spectra. Comparing the pristine chitosan spectrum to cross-
-
1
1
-
1
-1
linked chitosan microspheres spectrum, two new bands were observed at 1646 cm and 1570 cm , and
the bands were assigned to imine bond (C=N) and ethylenic bond (C=C) respectively [29, 30]. The

carbon-to-nitrogen (C:N) relationship on cross-linking increased to 6.95 from its initial non-cross-
linked ratio of 5.48, further confirming the crosslinking of microspheres. No evidence of the
-
1
characteristic band related to free glutaraldehyde, (aldehydic group, –CHO) around 1720 cm was
observed, indicating that cross-linked imprinted chitosan microspheres were free of the unreacted
-
1
cross-linker. FT-IR full spectra (4000-650 cm ) are provided in Fig. S4.
(
Insert Fig. 6 here)
3
.1.5. Thermo-stability of adsorbents
The TG curves of the pristine chitosan, chitosan nanofibers and cross-linked microspheres can be seen
in Fig. 7. The nanofibers loses solvents within its framework at a temperature of up to 150°C,
thereafter stabilizes up to 200°C, after which structural collapse of the framework occurs. The higher
stability observed on the chitosan nanofibers as compared to pristine chitosan was attributed to the
partial salt formation with the amino group of chitosan on dissolution with TFA. Cross-linked
microspheres lose all the solvents within its framework at a temperature of 160°C, after which
structural collapse began at 280°C. The better stability of the microspheres at higher temperatures can
be attributed to microspheres cross-linking which impacted stability on the chitosan. From the
thermogravimetric curve, glutaraldehyde-cross-linked chitosan microspheres tend to be more stable
when compared to chitosan and chitosan nanofibers as a result of crosslinking (Fig. 7).
(
Insert Fig. 7 here)
3
.2. Molecular modelling of chitosan-sulfone interactions
Chitosan interaction with sulfone compounds through the formation of complexes was evaluated by
employing density functional theory (DFT) molecular modelling. Geometry optimisations and
vibrational analyses of chitosan-sulfone complexes were performed using the Gaussian03 software.
B3LYP functional was employed with a 6-31G(d) basis set [31]. The enthalpies of formation
(
∆∆Hcomplex), Gibb’s free energies (∆∆Gcomplex) and entropies of formation (∆∆Scomplex) for each
complex formed were calculated by using Eqn. 2 and 3 [32].
ΔΔꢅꢆꢃꢇꢈꢉꢀꢊ = Δꢅꢆꢃꢇꢈꢉꢀꢊ − ꢂꢋΔꢅꢆꢌꢍꢎꢃꢏꢐꢑ + ꢒΔꢅꢏꢓꢉꢔꢃꢑꢀꢏꢄꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢂ2ꢄ

where m and t are the stoichiometric amounts of functional polymers and templates involved in
complex formation.
ꢕꢕꢖꢗꢘꢙꢚꢛꢜꢝ = ꢕꢕꢅꢗꢘꢙꢚꢛꢜꢝꢁꢞꢁT∆Δꢟꢗꢘꢙꢚꢛꢜꢝꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢂ3ꢄ
∆∆G, T and ∆∆S are the Gibbs free energy for the complex formation, temperature (273 K) and
entropy for complex formation at standard conditions (i.e. 1 molar concentration for solvents and 1
atm pressure) respectively.
The Gibbs free energy (∆G°) gives information about the feasibility of the adsorption process; entropy
(
∆S°) describes the spontaneous nature of adsorption whereas the sign of ∆H° reflects the endo- or
exothermic nature of the process. For chitosan-sulfone complex formation to be feasible, ∆G° must be
negative, and since ∆G° = ∆H°–T∆S°, a negative enthalpy change is therefore the most important
factor governing spontaneity of the interaction at low temperature. Sites with positive ∆G (Table 3)
indicated the sulfone-chitosan interactions are not favourable.
Four possible interaction sites for complex formations with chitosan were modelled for each
sulfonated compound. From the modelled thermodynamic parameters (preferably the Gibbs free
2
energy) presented in Table 3, the complexes formed between -OH, -NH of the same unit of chitosan
and complexes formed on -NH only of chitosan with sulfone compounds gave promising interactions.
2
The complex formation occurred via hydrogen bond formation through the interactions between
sulfone oxygen groups with -OH and -NH groups of the chitosan (Fig. 8). Other model structure
complex interactions are presented in the supplementary data Fig S4. From the thermodynamic
parameters obtained (Table 3), the order of complex formation is given as BTO
2 2
>DBTO > 4,6-
DMDBTO due to steric effect.
2
(
(
Insert Fig. 8 here)
Insert Table 3 here)
3
.3. Swelling ratio of adsorbents (chitosan microspheres and nanofibers)
Equilibrium swelling experiments of nanofibers and cross-linked microspheres in acetonitrile and
hexanewas carried out at room temperature to confirm the degree of cross-linking of adsorbents and
also to understand their swelling patterns. The equilibrium swelling ratio (ESR) was calculated as
o o o
follows: ESR= (W–W )/W , where W is the weight (mg) of sample before immersion into solvent and

W is the final weight (mg) of adsorbents after immersion into solvent (acetonitrile and hexane) with
time [33].
To perform the swelling experiment, 30 mg of adsorbents (microspheres and nanofibers) was added to
5
ml of each selected solvent in 15 mL vials, the study was carried-out and monitored at room
temperature. Adsorbents were removed by filtration through a Büchner funnel and weighed at specific
time intervals. From the results obtained, the adsorbents (nanofibers and microspheres) swelled more
in acetonitrile (Fig. 9). Swelling in hexane was minimal due to its poor affinity to the hydrophilic
adsorbent.
(
Insert Fig. 9 here)
The low degree of swelling in microspheres as compared to the nanofibers confirmed that the
microspheres were cross-linked (Fig. 9). On drying the swollen nanofiber strands, non-smooth
nanofibers surface morphology was observed (Fig. 10). However this change in the nanofibers
morphology did not affect the re-usability of the adsorbent, as the available functionalities, -OH and -
2
NH determines its reusability.
(
Insert Fig. 10 here)
3
.4. Sulfone compounds adsorption selectivity of imprinted microspheres and nanofibers
Adsorption assays were carried out to evaluate the loading capacity and selectivity of imprinted
chitosan adsorbents. 100 mg of the imprinted adsorbents were added to vials and mixed with 2 mL
2 2 2
solution mixture of organosulfur compounds, BTO , DBTO and 4,6-DMDBTO (5 mM). The
corresponding adsorption assays were also carried out using non-imprinted adsorbents. The suspension
was left under mechanical agitation at 150 rpm for 24 h as described in the adsorption studies.
Maximum adsorption observed for BTO
mg/g, 7.0 ± 0.5 mg/g and 6.6 ± 0.7 mg/g when imprinted chitosan nanofibers were employed, and 4.9
0.5 mg/g, 4.2 ± 0.7 mg/g and 3.9 ± 0.6 mg/g on molecularly imprinted chitosan microspheres (Fig.
1). However, high adsorption capacities with no selectivity were observed for non-imprinted
2 2 2
, DBTO and 4,6-DMDBTO respectively was 8.5 ± 0.6
±
1
nanofibers and microspheres (11.2 ± 1.1 mg/g, 10.4 ± 0.9 mg/g and 8.9 ± 1.2 mg/g when non-
imprinted chitosan nanofibers were employed, and 5.7 ± 0.7 mg/g, 5.6 ± 0.4 mg/g and 5.1 ± 0.6 mg/g

2 2 2
on non-imprinted chitosan microspheres for BTO , DBTO and 4,6-DMDBTO respectively) (Fig. 11).
The non-specific binding nature on non-imprinted hydrogel adsorbents enhanced the high adsorption
as compared to imprinted adsorbents which binding sites are restricted [21, 34]. Desorption of sulfone
compound was relatively high in both the imprinted and non-imprinted polymers.
Imprinted polymer adsorbents were reasonably selective for their target sulfone-containing
compounds. BTO
BTO , while adsorption capacities for DBTO
mg/g respectively on this material. DBTO -imprinted nanofibers’ adsorption capacity for DBTO
and 4,6-DMDBTO adsorption capacities were 2.8 ± 0.5 mg/g and 1.5 ±
.1 mg/g respectively. 4,6-DMDBTO -imprinted nanofibers gave an adsorption capacity of 6.6 ± 0.7
, while BTO and DBTO adsorption capacities were 2.6 ± 0.5 mg/g and 1.3 ±
.1 mg/g respectively. Similar selectivity trend was also observed with imprinted microspheres as
BTO -imprinted microspheres adsorption capacity for BTO was 4.9 ± 0.5 mg/g, while adsorption for
DBTO and 4,6-DMDBTO were 1.3 ± 0.5 mg/g and 1.1 ± 0.3 mg/g respectively. DBTO -imprinted
microspheres’ adsorption capacity for DBTO was 4.2 ± 0.7 mg/g, while BTO and 4,6-DMDBTO
were 1.5 ± 0.3 mg/g and 1.2 ± 0.2 mg/g respectively. 4,6-DMDBTO -imprinted microspheres
adsorption capacity for 4,6-DMDBTO was 3.9 ± 0.6 mg/g, while BTO and DBTO adsorption
capacities were 1.3 ± 0.7 mg/g and 1.8 ± 0.4 mg/g respectively.
2
-imprinted nanofibers showed a high adsorption capacity of 8.5 ± 0.6 mg/g for
2
2
and 4,6-DMDBTO
2
were 2.3 ± 0.2 mg/g and 2.1 ± 0.3
was
2
2
7
0
.0 ± 0.5 mg/g, while BTO
2
2
2
mg/g for 4,6-DMDBTO
2
2
2
0
2
2
2
2
2
2
2
2
2
2
2
2
(
Insert Fig. 11 here)
3
.4.1 Sulfone binding selectivity coefficient
The selectivity coefficient (k) defined for imprinted adsorbents are given by Eqns 4. k is inversely
related to a competitive affinity binding of molecule of interest competing with interfering molecule on
the same binding site of the polymer. The higher the value of k, the better is the imprinting effect.
K ꢂtemplateꢁmoleculeꢄ
ꢠꢁ
kꢁ =ꢁ
ꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢂ4ꢄ
K ꢂinterferingꢁmoleculeꢄ
ꢠꢁ
where K ꢁ=ꢁꢁQ /C ꢁꢁis the distribution coefficient of molecule between the microspheres/nanofibers
ꢠ
ꢀ
ꢀ
and the solution, k is the selectivity coefficient. Two interfering molecules are been considered for
each imprinted adsorbent.

2
The binding selectivity coefficient (k) values for BTO -imprinted microspheres and nanofibers are 3.8
and 4.9, while DBTO -imprinted microspheres and nanofibers are 1.6 and 1.8, and 4,6-DMDBTO -
2
2
imprinted microspheres and nanofibers are 1.1 and 1.4 respectively. From the obtained values, better
template imprinting which enhances adsorption selectivity occurred as the template molecule decreases
in size.
The order of binding selectivity is benzothiophene > dibenzothiophene > 4,6-
dimethyldibenzothiophene.
3
.5. Sulfone compounds adsorption kinetics
Adsorption experimental procedure described in section 2.3.1 was adopted for the adsorption kinetic
studies, in which aliquots are withdrawn at specific time intervals. Adsorption kinetics showed that
sulfone molecule adsorption was initially fast due to the availability of surface adsorption and
thereafter adsorption rate slowed down as a result of surface saturation, thus, limiting further sulfone
molecules penetration. Figure 12 represents the adsorption kinetics curve for the various sulfone
compounds using imprinted adsorbents with adsorption equilibrium achieved after 10 h. Nanofibers
were reported to have a higher adsorption capacity before equilibrium as compared to the
microspheres. The high adsorption capacity presented by nanofibers was attributed to their larger
surface area. The quantity adsorbed on both adsorbents increases in the order of BTO
DMDBTO , the observed adsorption trend was as a result of steric hindrance poised by the bulkiness
of each sulfone compound.
2 2
> DBTO > 4,6-
2
(
Insert Fig. 12 here)
The kinetic mechanism that controls the adsorption process under batch study was monitored by using
pseudo-first-order model (Eqn. 5) and pseudo-second-order model (Eqn. 6) [22]. Kinetic studies were
carried out by using an adsorbent (nanofibers and microspheres) mass of 100 mg contained in 2 mL of
5
mM sulfone compounds solution. Batch study was employed to monitor adsorption capacity with
time.
ꢣꢤ
.303
logꢂꢡ − ꢡ ꢄ = logꢡ −
ꢒꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢂ5ꢄꢁ
ꢜ
ꢢ
ꢜ
2
ꢒ
1
1
ꢡꢜ
=
+
ꢒꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢂ6ꢄꢁ
ꢥ
ꢡꢜ ꢡ ꢣ
ꢜ
ꢥꢁ

where q
equilibrium and time t, respectively. K
calculated by plotting the log (q -q ) versus t (Fig. 13). K
pseudo-second-order and was calculated from the slope and intercept of the plots t/q
The pseudo-first-order and pseudo-second-order parameters (coefficients) are presented in Table 4.
e t
and q (mg/g) are the amounts of sulfone compounds adsorbed on the adsorbent at
−
1
1
(h ) is the pseudo-first-order adsorption rate constant and was
e
t
2
(g/mg/h) is the adsorption rate constant of
t
versus t (Fig. S5).
(
Insert Fig. 13 here)
2
Based on the obtained correlation coefficients (R ) from both graphs, pseudo-first-order model showed
a fitted plot for the adsorption studies.
3
.6. Sulfone compounds adsorption Isotherms
The adsorption behaviour of sulfone compounds on imprinted chitosan adsorbents is an important
factor to investigate. Langmuir and Freundlich isothermal equations (Eqns. 7 and 8) were employed
using the adsorption experimental procedure described in section 2.3.1.
Cꢜ
ꢦꢜ
Kꢨ
=
+
ꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢂ7ꢄ
ꢡꢜ
ꢧꢙ ꢧꢙ
1
ꢩꢪꢫq = ꢩꢪꢫꢦ + ꢩꢪꢫꢣꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢂ8ꢄ
ꢜ
ꢜ
ꢬ
where q
concentration, (Q
the Langmuir constant (related to the affinity of adsorption sites) [35].
Freundlich constants k and n indicating adsorption capacity and intensity respectively were determined
from the linear plot of log q against log C . Freundlich isotherm indicated a multiple layered
adsorptions on adsorbents, which may possibly infer sulfone-sulfone interactions. While Langmuir
equation obtained from a plot of C /q against C is probably proof of chemical adsorption, which may
usually mean monolayer adsorption on the surface of adsorbents [36-38]. The Langmuir constants K
and Q are obtained from the slope and intercept values of the plot C /q against C (Fig. S6). From
e
(mg/g) and C
e
(mg/g) are the amount adsorbed at equilibrium and the equilibrium
m
), is the theoretical maximum adsorption capacity at monolayer (mg/g), and K
d
is
e
e
e
e
e
d
m
e
e
e
plots of the two isotherms, Freundlich adsorption equation fitted better (larger correlation coefficient
2
R , Fig. 14 and Table 5) with equilibrium data as compared to the Langmuir parameters and we also
observed that Freundlich constant ‘n’ value lies within the range 1 to 10, thus indicating that

adsorption on the adsorbent (microspheres and nanofibers) is favourable [39-41]. However, the fact
that the fitting of the Langmuir (Fig S6, Table 5) also shows some correlation infers some form of
chemical adsorption on this system in agreement with the thermodynamic parameters obtained from
modelling.
(
(
Insert Fig. 14 here)
Insert Tables 4 and 5 here)
3
.7. Adsorbents adsorption reusability studies
Reusability studies on the adsorbent were carried out using the similar adsorption procedure described
in the batch adsorption process. The rebinding adsorption capacities of non-imprinted adsorbent
decreased drastically for all sulfone compounds, from 11.2 ± 1.1 mg/g, 10.4 ± 0.9 mg/g and 8.9 ± 1.2
mg/g to 4.2 ± 0.2 mg/g, 3.9 ± 0.3 mg/g and 3.5 ± 0.1 for BTO
respectively when non-imprinted nanofibers was employed (Fig. 15). A reduction from 5.7 ± 0.7
mg/g, 5.6 ± 0.4 mg/g and 5.1 ± 0.6 mg/g to 3.2 ± 0.2 mg/g, 2.9 ± 0.3 mg/g and 3.8 ± 0.5 for BTO
DBTO and 4,6-DMDBTO respectively were observed when non-imprinted microspheres were
employed. From the second adsorption cycle, adsorption capacities of 7.5 ± 0.5 mg/g, 6.2 ± 0.3 mg/g
and 5.8 ± 0.6 mg/g were observed for BTO , DBTO and 4,6-DMDBTO respectively when imprinted
nanofibers were employed, and the imprinted microspheres gave maximum adsorption capacity of 4.5
0.4 mg/g, 3.9 ± 0.6 mg/g and 3.4 ± 0.2 mg/g for BTO , DBTO and 4,6-DMDBTO respectively.
2 2 2
, DBTO and 4,6-DMDBTO
2
,
2
2
2
2
2
±
2
2
2
The imprinted adsorbents presented a slight decrease in adsorption capacity upon the second cycle of
usage as can be seen in figures 11 (first cycle) and 15 (second cycle), and this confirms the integrity of
the imprinting effect which was also reported by Chang et al [38].
Soxhlet extraction using methanol/acetonitrile (1:1) was employed for desorbing the sulfone
compounds from the adsorbents (nanofibers and microspheres). Desorption of sulfone compound was
relatively high in both the imprinted and non-imprinted polymers (Fig. 15) confirming effective
removal of adsorbed sulfone compounds from the polymer matrix.
(
Insert Fig. 15 here)

3
3
.8. Continuous flow adsorption studies
.8.1. Adsorption of model sulfone compounds using imprinted chitosan nanofibers
Continuous flow adsorption technique was employed for the adsorption of sulfone compounds.
Breakthrough volume were evaluated, as it represents the evolution of the concentration of a solution
as a function of adsorption parameters like contact time between liquid and solid phase (adsorbent),
solvent concentration and temperature.
1
00 mg of imprinted nanofibers was packed into a cylindrical tube attached to the tip of a syringe
containing 5 mM solution of 5 mL sulfone compounds as shown in Fig. 16. Nanofibers were employed
for this study because they gave better adsorption capacity in the batch process. The imprinted
nanofibers were easily contained in the tube without leaving much space and the packing was
tightened by conditioning the material with solvent at 1 mL/h. This ensured that the packing density
remained the same for all the imprinted nanofibers. Adsorption progressed as respective sulfone
compounds were pumped through the conditioned adsorbent at a flow-rate of 1 mL/h. From the
adsorption curve, the maximum amount of sulfone compounds was retained only after 1 mL (1000 µL)
of the solution have been dispensed (Fig. 17), and at this volume the nanofibers is said to be saturated.
s
A number of theoretical plates (N), the linear capacity of the column (n ), the capacity factor of the
solute (k) and percentage recovery (r) are calculated from equations 9-12 [42, 43] and presented in
Table 6.
ꢮꢯ
ꢭ =
ꢂꢮ −ꢁꢰ ꢄꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢂ9ꢄ
ꢯ ꢱ
ꢥ
ꢰ
ꢱ
where ꢁꢁꢁꢁꢁ2ꢰ = ꢮ −ꢁꢮ
ꢱ
ꢯ
ꢲ
ꢮꢯ
ꢮꢳ
K =
− 1ꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢂ10ꢄ
ꢬ = ꢮ ꢣꢦ ꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢂ11ꢄ
ꢴ
ꢳ
ꢘ
ꢬꢴ
ꢵ
=
× 100%ꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢁꢂ12ꢄ
ꢦ ꢮ
ꢘ ꢘ
where V
o
B R
is the initial volume of the analyte, breakthrough volume (V ), retention volume (V ) and
hold-up volume (V ) of the analyte (sulfone solution). From Table 6, presents the values of the
M

percentage sulfone recovery of the nanofibers, which is the amount of substrates that can be recovered
in the continuous flow process.
(
(
Insert Figs. 16 and 17 here)
Insert Table 6 here)
3
.8.2. Adsorption of mixture of sulfone compounds (oxidized model fuel)
Adsorption studies on the mixture of model sulfone compound solution under batch process and
continuous flow process were carried out by employing 300 mg of conditioned imprinted nanofibers
(
2 2 2
comprising of 100 mg BTO -imprinted, 100 mg DBTO -imprinted and 100 mg 4,6-DMDBTO -
imprinted). The same procedures as described in sections 2.3.1 and 3.8.1 were employed for the batch
and continuous flow experiments.
The adsorbent was packed into a tube attached to a syringe containing 2 mL of a mixture of 830 ppm
2 2 2
BTO (160 ppm S), 1080 ppm DBTO (163 ppm S) and 1230 ppm 4,6-DMDBTO (167 ppm S), a
similar mixture containing 300 mg of adsorbents was also contained in a screw-capped vial for batch
adsorption process. A reduction in concentration of the sulfone compounds was observed after passing
through the imprinted nanofibers, BTO
DBTO concentration was reduced to 102 ppm (15.1 ± 1.4 ppm S) and 4,6-DMDBTO
was reduced to 120 ppm (15.6 ± 2.6 ppm S) was recorded in the continuous flow adsorption process.
In batch adsorption process, BTO concentration was reduced to 393 ppm (75.7 ± 4.5 ppm S), DBTO
concentration was reduced to 365 ppm (54.1 ± 7.1 ppm S) and 4,6-DMDBTO reduced to 232 ppm
2
concentration was reduced to 148 ppm (28.5 ± 3.4 ppm S),
2
2
concentration
2
2
2
(
30.2 ± 3.1 ppm S). The chromatograms are presented in Fig. 18.
(
Insert Fig. 18 here)
3
.8.3. Adsorption of oxidized sulfur compounds in diesel fuel
Oxidized hydro-treated diesel [44] containing 394 ppm sulfur in the form of sulfonated compounds
were desulfurised by using chitosan-imprinted nanofibers under continuous flow process as described
in model fuel adsorption system. 300 mg of imprinted nanofibers (100 mg BTO
DBTO -imprinted and 100 mg 4,6-DMDBTO -imprinted) was applied as adsorbent for adsorption.
Adsorption was allowed to proceed by employing the conditioned imprinted nanofibers at a flow-rate
2
-imprinted, 100 mg
2
2

of 1 mL/h. Sulfur analysis after adsorption process indicated that a total of 62 ± 3.2 ppm S was left in
the diesel (Fig. 19), the saturated sulfur adsorption capacity of the molecularly imprinted adsorbents
was 2.2 ± 0.2 mg/g. The high adsorption of sulfur through the use of chitosan is in agreement with the
work of Jorge et al. [21] and Chang et al [38]. Desorption of adsorbed sulfone compounds on chitosan
nanofibers were carried out by using a Soxhlet extraction process, and it was observed that 78% of the
absorbed sulfones were desorbed from the nanofibers (Fig. 20).
(
Insert Figs. 19 and 20 here)
4
.
Conclusions
Molecularly imprinted chitosan nanofibers and microspheres were successfully applied to the
adsorption of sulfone compounds found in oxidized hydro-treated diesel. The imprinted adsorbents
exhibited conformational memory for recognizing sulfones. Surface area measurements showed that
non-imprinted microspheres and nanofibers have higher surface area compared to imprinted
adsorbents (nanofibers and microspheres), this phenomenon is due to surface morphological
modifications on imprinting with template of different molecular sizes. The Freundlich isotherm
confirmed multilayer adsorption attributed to the interactions between imprinted adsorbents and
sulfone compounds, and possibly between sulfone molecules. A high adsorption capacity observed by
non-imprinted adsorbents was due to the non-specific binding nature of the adsorbents compared to
imprinted adsorbents where binding environments are restricted through the formation of high specific
sites. Continuous flow adsorption process showed that adsorbent saturation begins only after about 1
mL of the respective sulfone solutions have been passed through the imprinted chitosan nanofibers.
High adsorption of sulfones in oxidized hydro-treated diesel under continuous flow process indicated
its potential use for industrial adsorption purposes. However, the stability of the material under
industrial application process remains uncertain.

References
[
[
[
[
1]
2]
3]
4]
The transition of ultralow-sulfur diesel fuel: effects on prices and supply; Energy Information
Administration (EIA), Washington, DC, 2001.
A.S. Ogunlaja, S. Khene, E. Antunes, T. Nyokong, N. Torto, Z.R. Tshentu, Appl. Catal., A.
4
62-463 (2013) 157-167.
M.R. Maurya, M. Bisht, A. Kumar, M.L. Kuznetsov, F. Avecilla, J.C. Pessoa, Dalton Trans. 40
2011) 6968-6983.
(
R.S. Walmsley, A.S. Ogunlaja, M.J. Coombes, W. Chidawanyika, C. Litwinski, N. Torto, T.
Nyokong, Z.R. Tshentu, J. Mater. Chem. 22 (2012) 5792-5800.
[
[
5]
6]
J.A.L. Da Silva, J.J.R.F. Da Silva, A.J.L. Pombeiro, Coord. Chem. Rev. 255 (2011) 2232-2248.
J.M. Campos-Martin, M.C. Capel-Sanchez, P. Perez-Presas, J.L.G. Fierro, J. Chem. Technol.
Biotechnol. 85 (2010) 879-890.
[
[
[
7]
8]
9]
A. Jorge, S. Le Borgne, Macromolecules 37 (2004) 2938-2943.
J. Matsui, M. Okada, M. Tsuruoka, T. Takeuchi, Anal. Commun. 34 (1997) 85-87.
M.T. Muldoon, L.H. Stanker, Proceedings of the ACS, Symposium Series, Vol. 657, 1997, pp.
3
14-330.
[
[
[
[
[
[
[
[
[
10] L.I. Andersson, A. Paprica, T. Arvidsson, Chromatographia. 46 (1997) 57-62.
11] G. Wulff, Angew. Chem. Int. Ed. Engl. 34 (1995) 1812-1832.
12] A.G. Mayes, K. Mosbach, Trends Anal. Chem. 16 (1997) 321-332.
13] E.N. Vulfson, C. Alexander, M.J. Whitcombe, Chem. Br. 33 (1997) 23-26.
14] M. Kempe, K.J. Mosbach, J. Chromatogr. A 694 (1995) 3-13.
15] I.A. Nicholls, L.I. Andersson, K. Mosbach, B. Ekberg, Trends Biotechnol. 13 (1995) 47-51.
16] M. Ruiz, A.M. Sastre, E. Guibal, React. Funct. Polym. 45 (2000) 155-173.
17] K. Ryu, J. Heo, I.K. Yoo, Biotechnol. Lett. 24 (2002) 143-146.
18] B. Castro, M.J. Whitcombe, E.N. Vulfson, R. Vazquez-Duhalt, E. Barzana, Anal. Chim. Acta.
4
35 (2001) 83-90.
[
[
[
[
[
19] G. Fariba, S. Vasheghani-Farahani, E. Vasheghani-Farahani, Iran Polym. J. 19 (2010) 375-398.
20] O.A.C Monteiro, C. Airoldi, C. Int. J. Biol. Macromol. 26 (1999) 119-128.
21] A. Jorge, A. Mendez-Orozco, S. Le-Borgne, Chem. Eng. Process. 43 (2004) 1587-1595.
22] C. Branger, W. Meouche, A. Margaillan, React. Funct. Polym.73 (2013) 859-875.
23] A. Greiner, J.H. Wendorff, Angew. Chem. Int. Ed. 46 (2007) 5670-5703.

[
[
[
[
24] S.M. Hudson, Y.C. Wei, S.M. Hudson, J.M. Mayer, D.L. Kaplan, J. Polym. Sci. 30 (1992)
187-2193.
25] A.T. Marco, S.V. Rodrigo, M.B. Marisa, J.A. Eduardo, C.S. Cesar, Mater. Res. 10 (2007) 347-
52.
2
3
26] A.S. Ogunlaja, W. Chidawanyika, E. Antunes, M.A. Fernandes, T. Nyokong, N. Torto, Z.R.
Tshentu, Dalton Trans. 41 (2012) 13908-13918.
27] O. Kousaku, C. Dongil, K. Hakyong, N. Ayako, Y. Hiroyuki, Macromol. Rapid Commun. 25
(
2004) 1600-1605.
[
[
[
[
28] M. Hasegawa, A. Isogai, F. Onabe, M.J. Usuda, J. Appl. Polym. Sci. 45 (1992) 1857-1863.
29] A.C. Oyrton, J. Monteiro, A. Claudio, Int. J. Biol. Macromol. 26 (1999) 119-128.
30] E.R. Blout, M. Fields, J.R. Karplus, Am. Chem. Soc. 70 (1948) 194-198.
31] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman, J.A.
Montgomery Jr, T. Vreven, K.N. Kudin, J.C. Burant, J.M. Millam, S.S. Iyengar, J. Tomasi, V.
Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G.A. Petersson, H. Nakatsuji, M. Hada,
M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H.
Nakai, M. Klene, X. Li, J.E. Knox, H.P. Hratchian, J.B. Cross, V. Bakken, C. Adamo, J.
Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A.J. Austin, R. Cammi, C. Pomelli, J.W.
Ochterski, P.Y. Ayala, K. Morokuma, G.A. Voth, P. Salvador, J.J. Dannenberg, V.G.
Zakrzewski, S. Dapprich, A.D. Daniels, M.C. Strain, O. Farkas, D.K. Malick, A.D. Rabuck, K.
Raghavachari, J.B. Foresman, J.V. Ortiz, Q. Cui, A.G. Baboul, S. Cliord, J. Cioslowski, B.B.
Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R.L. Martin, D.J. Fox, T. Keith, M.A.
Al-Laham, C.Y. Peng, A. Nanayakkara, M. Challacombe, P.M.W. Gill, B. Johnson, W. Chen,
M.W. Wong, C. Gonzalez, J.A. Pople, Gaussian 03, Revision E.01. Gaussian, Inc., Wallingford
CT, 2004.
[
32] W.J. Hehre, L.D. Burke, A.J. Shusterman, W.J. Pietro, Experiments in computational organic
chemistry, wave function Inc., CA, USA, 1993.
[
[
33] Z. Gao, J.S. Nahrup, J.E. Mark, A. Sakr, J. Appl. Polym. Sci. 90 (2003) 658-666.
34] T. Enoki, K. Tanaka, T. Watanabe, T. Oya, T. Sakiyama, Y. Takeoka, K. Ito, G. Wang, M.
Annaka, K. Hara, R. Du, J. Chuang, K. Wasserman, A.Y. Grosberg, S. Masamune, T. Tanaka,
Phys. Rev. Lett. 85 (2000) 5000-5003.

[
35] R.I. Masel, in: Principles of adsorption and reaction on solid surfaces, Wiley, 1996, pp. 239-
45, 299-301.
2
[
[
[
36] R.J. Umpleby, C.S. Baxter, M. Bode, M. Anal. Chim. Acta 435 (2001) 35-42.
37] Y.C. Wong, Y.S. Szeto, W.H. Cheung, Langmuir 19 (2003) 7888-7894.
38] C. Yonghui, Z. Lei, Y.Z.L. Hanjie, L. Hao, O. Pingkai, Appl. Biochem. Biotechnol. 160 (2010)
5
93-603.
[
39] E.R. Treybal, Mass Transfer Operations, 3rd edition, McGraw Hill International editions,
Chemical Engineering Series, Singapore, 1981.
[
[
[
[
[
40] M. Rao, A.G. Bhole, Journal of Indian Water Works Association, XXXIII 1 (2001) 97-100.
41] C. Raji, K.P. Shubha, T.S. Anirudhan, Indian J. Environ. Hlth. 39 (1997) 230-238.
42] A. Gelencser, G. Kiss, Z. Krivacsy, Z. Varga-Puchony, J. Chromatogr. A 693 (1995) 217-225.
43] I. Liska, J. Krupcik, P.A. Leclercq, J. High Res. Chromatogr. 12 (1989) 577-590.
44] A.S. Ogunlaja, R.S. Walmsley, C.W. du Sautoy, N. Torto, Z.R. Tshentu, Energy Fuels 27
(
2013) 7714-7723.

List of tables
Table 1
EDS chemical composition of the nanofibers.
Elements
X-ray
keV)
Non-deprotonated
Fiber surface after deprotonation and washing
(
fiber
(non- (imprinted nanofibers)
imprinted
nanofibers)
ND-NIF
56.39
BTO
2
DBTO
61.64
9.60
25.54
3.22
-
2
4,6-DMDBTO
2
Carbon (C)
Nitrogen (N)
Oxygen (O)
Flourine (F)
Sulfur (S)
0.277
0.392
0.525
0.677
2.307
64.15
8.48
67.68
9.04
21.05
2.23
-
8.69
17.04
18.55
8.77
17.88
-
0.05
Table 2
Surface area of microspheres (non-imprinted and imprinted) and nanofibers (non-imprinted and
imprinted).
2
Surface area (m /g)
Adsorbent types
Microspheres
Nanofibers
Non- imprinted (control)
62
59
57
56
101
85
BTO
DBTO
,6-DMDBTO
2
imprinted
2
imprinted
imprinted
87
4
2
73

Table 3
DFT molecular modelling thermodynamic data on the formation of complexes between sulfone compounds {benzothiophene sulfone
(
BTO
2
), dibenzothiophene sulfone (DBTO
2
) and 4,6-dimethyldibenzothiophene sulfone (4,6-DMDBTO
2
)} and chitosan (∆∆H, ∆∆G
and ∆∆S complex).
-
1
-1
-1
∆
∆H (kcal.mol )
∆∆G (kcal.mol )
∆∆S (cal.mol )
BTO
2
DBTO
2
4,6-
DMDBTO
-11.3
BTO
2
DBTO
2
4,6-
DMDBTO
-0.4
BTO
2
DBTO
2
4,6-
DMDBTO
-36.7
2
2
2
Complex OH-NH
Complex OH-NH
2
2
same unit (site A)
diff unit (site B)
-12.5
-6.3
-11.4
-5.6
-1.0
3.8
-0.2
4.8
-38.4
-33.8
-38.3
-36.7
-37.4
-34.9
-38.6
-34.3
-4.7
5.9
-35.7
Complex NH
2
only (site C)
-12.5
-9.4
-12.8
-8.3
-11.6
-1.0
1.5
-1.3
2.0
0.2
-39.4
Complex OH opposite NH
2
only (site
-3.9
5.9
-33.3
D)

Table 4
Kinetic data of pseudo-first-order and pseudo-second-order of the nanofibers and microspheres.
Adsorbents
Pseudo-first-order kinetics
Pseudo-second-order kinetics
BTO
2
DBTO
2
4,6-DMDBTO
2
BTO
2
DBTO
2
4,6-DMDBTO
2
-
1
k (h )
2
-1
k (h )
2
-1
k (h )
2
−1 −1
(g mg h )
2
−1 −1
(g mg h )
2
−1 −1
(g mg h )
2
R
R
R
k
2
R
k
2
R
k
2
R
Microspheres 1.284 0.9908 0.261 0.9793 1.214 0.9844
0.054
0.7649
0.8962
0.05
0.7649
0.8962
0.019
0.006
0.1806
0.034
Nanofibers
0.151 0.9468 0.154 0.9754 0.255 0.9052
-0.11
-0.12
Table 5
Parameters of Langmuir adsorption model and Frendlinch adsorption model for nanofibers and microspheres.
Adsorbents Langmuir parameters Freundlich parameters
BTO
2
DBTO
2
4,6-DMDBTO
2
BTO
2
DBTO
2
4,6-DMDBTO
2
2
2
2
2
2
2
Q
m
R
K
1.7
2.4
d
Q
m
R
K
d
Q
m
R
K
d
n
R
n
R
n
R
Microspheres 5.66 0.8560
Nanofibers 11.17 0.8397
5.52
0.8846 0.6
5.00 0.8065 0.6
8.98 0.7476 0.4
1.92 0.9870 2.03 0.9774 1.95 0.9549
1.34 0.9802 1.83 0.9945 2.01 0.9758
10.70 0.8114
0

Table 6
Influence of sulfone compounds upon adsorption on chitosan nanofibers.
-
2
Sulfone
V
B
(mL)
V
R
(mL)
V
M
(mL)
N
k
n × 10
r %
compounds
(mmol/g)
1.344
BTO
DBTO
,6-DMDBTO
2
0.909
0.895
0.898
2.908
3.237
2.464
0.210
0.230
0.200
5.555
4.893
6.756
12.800
13.100
11.300
53.760
60.260
45.200
2
1.506
4
2
1.130

Legend of figures
5
4
3
2
1
000
000
000
000
000
0
United States
European Union
South Africa
Decreasing sulfur contents in fuels over
the years
1
990
1995
2000
2005
2010
2015
Years
Fig. 1. Trends in diesel sulfur fuel specifications over the years.
Fig. 2. A schematic diagram of experimental setup used for electrospinning.

Fig. 3. Micrograph images of wet chitosan microspheres
Fig. 4. EDS spectra of (A) non-imprinted nanofibers after washing, (B) BTO
nanofibers after deprotonation and washing, (C) DBTO -imprinted nanofibers after
-imprinted nanofibers after deprotonation
2
-imprinted
2
deprotonation and washing and (D) 4,6-DMDBTO
2
and washing.

Surface morphology
A
B
1000 µM
500 µM
200 µM
C
D
1
0 µM
10 µM
Fig. 5. Scanning electron microscopy (SEM) of (A) dried cross-linked chitosan microspheres
after template removal, (B) cross-linked chitosan microspheres showing surface morphology,
(
C) representative electrospun chitosan nanofibers without template (inset: image) and (D)
representative electrospun chitosan nanofibers with template (inset: image).