Fuel 189 (2017) 186–194

Contents lists available at ScienceDirect

Fuel
journal homepage: www.elsevier.com/locate/fuel

Full Length Article

Adsorptive desulfurization of diesel oil using nickel nanoparticle-doped

activated carbon beads with/without carbon nanofibers: Effects of
adsorbate size and adsorbent texture
Yogendra Nath Prajapati a, Nishith Verma a,b,⇑
a
Department of Chemical Engineering, Indian Institute of Technology Kanpur, Kanpur 208016, India
b
Center for Environmental Science and Engineering, Indian Institute of Technology Kanpur, Kanpur 208016, India

h i g h l i g h t s g r a p h i c a l a b s t r a c t

Efficient Ni-doped CNF-containing

adsorbents were prepared for ADS of
diesel oil.

The adsorbate size and adsorbent
textures influenced the adsorption
capacity.

Ni-CNF/ACBs showed higher Thiophene
adsorption capacity than Ni/ACBs in a
commercial oil.

Ni-doped CNFs-containing porous carbon bead V<1 nm (cc/g)

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

Article history: Nickel (Ni)-doped activated carbon beads (ACBs) with/without carbon nanofibers (CNFs) were synthe-
Received 10 April 2016 sized for the efficient desulfurization of diesel oil by adsorption. The materials were characterized for
Accepted 6 October 2016 their physico-chemical properties, using the BET, SEM, XRD, and TPD analysis. The in-situ dispersed Ni
Available online 26 October 2016
nanoparticles in the ACBs served the dual role of catalyzing the growth of the CNFs (average thickness
c.a. 70 nm) across the ACB surface and the removal of S-containing compounds from the liquids. The
Keywords: adsorption capacity of Ni/ACBs (without CNF-containing adsorbents) was approximately 3.5 times higher
Adsorptive desulfurization
than that of Ni-CNF/ACBs for the large sized dibenzothiophene molecules, because of the higher BET sur-
Activated carbon bead
Carbon nanofiber
face area and pore volume of the former adsorbent. On the contrary, the adsorption capacity of Ni-CNF/
Nickel nanoparticle ACBs was higher than that of Ni/ACBs for the small sized thiophene molecules, despite the BET surface
Diesel oil area and pore volume being smaller in the CNF-containing adsorbent, indicating that the p-
complexion interactions were dominant over physisorption in the graphitic CNFs. The comparative
adsorption data indicate that the adsorption capacities of the prepared Ni-containing adsorbents with/
without CNFs in this study were greater than most of the adsorbents discussed in the literature. This
study assumes significance from the perspective of selecting suitable adsorbents for the different
sized-S-compounds in diesel oil.
Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction

Diesel is one of the high energy density (42.5 MJ/kg) fuels

commonly used in transportation vehicles, power generators, and
⇑ Corresponding author at: Department of Chemical Engineering, Indian Institute industrial furnaces. During combustion of diesel, several air
of Technology Kanpur, Kanpur 208016, India. pollutants such as H2S and SO2 are emitted because of the
E-mail address: vermanishith@gmail.com (N. Verma).

http://dx.doi.org/10.1016/j.fuel.2016.10.044
0016-2361/Ó 2016 Elsevier Ltd. All rights reserved.
 Y.N. Prajapati, N. Verma / Fuel 189 (2017) 186–194
decomposition of sulfur-containing aromatic compounds such as
thiophene (TH), benzothiophene (BT), dibenzothiophene (DBT)
and its alkyl-substituted derivatives. These compounds are inher-
ently present in the fuel. The sulfur-containing compounds can
also poison the catalyst of the catalytic converter in automobiles.
Many countries have set strict regulations for the maximum
amounts of sulfur in diesel. European standards organization has
restricted the sulfur concentration below 10 ppm by weight
(ppmw) in all fuels since 2009 [1]. In India, regulations have been
set at 10 ppmw-sulfur in diesel oil from 2017 onwards [2].
Hydrodesulfurization (HDS) is commonly used for removing
sulfur or sulfur-containing compounds from diesel by the catalytic
hydrogenation. The method using Ni/Co/Mo-based catalysts
requires relatively severe operating conditions (350–400 °C and
3–10 MPa hydrogen-pressure), and is considered to be effective
for treating diesel containing sulfur over the 500–2000 ppmw-
concentration range [1,3]. However, further reduction of sulfur in
diesel by HDS may be ineffective, because the hydrogenation of
aromatic hydrocarbons in the fuel is more reactive than that of
the aromatic sulfur compounds [4]. In this context, adsorptive
desulfurization (ADS) has gained much attention because of the
required less severe operating conditions (<200 °C and 0.1–
0.8 MPa) compared to conventional HDS processes. Furthermore,
such processes do not alter the sulfur-free composition and phys-
ical properties of the fuel.
The ADS processes use different transition metals such as Cu, Ni,
Ag, and Zn, which are dispersed in various inorganic supports, viz.,
molecular sieves, silica and activated alumina (AA). Transition
metals have the capability to enhance the adsorption of sulfur
compounds through p-complexation and/or metal-sulfur interac-
tion [5]. Among the transition metals, Ni has attracted much atten-
tion for the ADS of diesel oils, with the adsorption capacity of the
Ni-based adsorbents determined to be 0.2–16 mg-S/g over the sul-
fur concentration range of 14–700 ppmw and temperature range of
25–220 °C [6–8]. In particular, Liao et al. [5] evaluated the desulfu-
rization performance of different transition metals supported in AA
at room temperature (30 °C) and found the adsorption capacities
for TH in the following order: Ag > Ni > Cu > Zn. Similar compara-
tive studies also showed the relatively superior performance of
Ni-based adsorbents [7,9].
Activated carbon (AC) is widely recognized as a versatile mate-
rial having potential to act as an adsorbent, catalyst, and support to
metal catalysts, attributed to its relatively large BET area,
amenability for surface functionalization, and stability in acidic
medium [10]. In the present context of desulfurization also, ACs
have recently gained considerable attention [11–16]. Among the
salient studies performed in this area, Xiao et al. [11] investigated
the comparative effects of AC and Na+/Y zeolite (without metal) for
the removal of BT and DBT from n-octane at 30 °C. The ACs showed
relatively higher adsorption capacity for both sulfur compounds,
attributed to the relatively larger pore size of the adsorbent vis a
vis the molecular size of the solutes, and a strong interaction
between carbon and the BT and DBT molecules. Muzic et al. [12]
investigated the performance of AC and zeolite for the ADS of com-
mercial diesel at different temperatures ranging from 30 to 70 °C.
The adsorption capacity of AC was determined to be approximately
three times higher than that of zeolite. Hernandez and co-worker
[13] compared the performances of the Ni-dispersed AA-SiO2 com-
posite with that of AC with and without Ni for the ADS of different
model oils containing BT, DBT and 4,6-DMDBT. Interestingly, the
data showed that the adsorption capacity of AC without Ni was
higher than that of all AA-SiO2-supported Ni materials synthesized
in the study. The inclusion of Ni into AC showed even higher
adsorption capacity. Similar results have been reported for the
ADS of commercial as well as synthetic diesel containing DBT
and DMDBT [14–16]. The adsorption capacity of AC was found to
187
be higher than that of the metal-dispersed SiO2/AA/zeolite/molec-
ular sieves.
From the overview of the aforesaid studies, it is evident that Ni
dispersed-carbons are potentially efficient materials for the ADS of
diesel. Extending along the theme of those studies, this paper
describes the development of the Ni-containing spherical activated
carbon beads (0.8 mm) having different textural properties for
the ADS of diesel. The transition metal-dispersed porous carbon
beads have been found to exhibit significantly large adsorption
capacity for a number of pollutants in water, such as arsenic, fluo-
ride, vitamin B-12, and Cr(VI) [17,18], and are the focus of the pre-
sent study on desulfurization. Such materials (beads) can also be
packed in a tubular adsorber and efficiently used under flow con-
ditions without maldistribution or channeling of the liquid to be
treated.
The present study investigates the ADS performance of the Ni/
NiO-doped porous carbon beads decorated with/without carbon
nanofibers (CNFs). The effects of textural properties (BET area, pore
volume, pore size distribution (PSD), metal and graphitic contents)
of the prepared materials on desulfurization are investigated using
two different organosulfur compounds, namely, TH and DBT,
which are different in molecular size. Several salient studies have
been recently geared in this direction and are revisited in details
later in the context of discussing the adsorption mechanism for
desulfurization [19–24]. This study shows that suitable textural
modifications to the adsorbent materials are required for the effi-
cient desulfurization of diesel oil, depending upon the type of tar-
get compound to be removed.
2. Materials and methods
2.1. Chemicals
Phenol (>99.5%), formaldehyde (37–38%), hexamethylenete-
tramine (HMTA, 99%), triethylamine (TEA, 99%), polyvinyl alcohol
(PVA) (MW = 95,000), nickel nitrate hexahydrate (>98%), and n-
octane (>99%) were purchased from Merck Germany. The TH
(99%) and DBT (98%) sulfur-containing compounds were purchased
from Alfa Aesar and Spectrochem Pvt. Ltd. Mumbai, respectively.
Commercial diesel (115 ppmw-S) was purchased from a local
market. All gases used in this study were zero grade and purchased
from Sigma Gases India.
2.2. Synthesis and heat treatment of Ni-doped polymeric beads
The Ni-doped phenolic precursor-based polymeric beads were
synthesized according to the procedure described in the previous
study [17]. The in-situ Ni-doped polymeric beads were pre-
oxidized by air (100 standard cubic centimeter per min (sccm)-
flow rate) at 350 °C for 1 h in the horizontal stainless steel (SS)
tubular reactor mounted in a tubular furnace. The temperature of
the reactor was increased at a rate of 5 °C/min to 950 °C. The oxi-
dized beads were carbonized under N2 flow (100 sccm) for 1 h
and then activated at the same temperature for another 1 h, using
a mixture of N2 (100 sccm) and steam (0.3–0.5 g per min). After
activation, the furnace was allowed to cool to room temperature
under N2 flow. Next, the Ni-containing porous carbon beads were
calcined at 450 °C for 4 h under 100 cc/min of N2 flow to convert
Ni-nitrate into NiO. Such samples were termed as NiO/ACB for
the reference purposes in this study. The reduction of NiO into
Ni0 was carried out at 600 °C for 2 h, using H2 as a reducing agent
at 100 sccm-flowrate. These samples were termed as Ni/ACB. Next,
CNFs were grown over Ni/ACBs, using CVD with acetylene gas
(30 sccm) as the carbon source. The dispersed Ni0 in the beads cat-
alyzed the growth of the CNFs. The CVD was carried out at 600 °C
188 Y.N. Prajapati, N. Verma / Fuel 189 (2017) 186–194
for 0.5 h. The heating was stopped and the reactor was cooled to
room temperature. The produced CNFs-decorated beads were sub-
jected to ultrasonication using 0.05 M HNO3 for 2 min to open up
the tips of the CNFs by dislodging the Ni nanoparticles which were
located at the tips. After ultrasonication, the beads were dried at
room temperature and used for adsorption study. These samples
were termed as Ni-CNF/ACB. Some samples were not subjected
to ultrasonication, for the comparison purposes. The carbonized
and activated samples prepared without Ni were termed as ACB.
2.3. Adsorbent characterization
The N2 adsorption-desorption isotherm, BET surface area, pore
volume and pore size distribution (PSD) of the prepared materials
were determined at 77 K, using the Autosorb-1C instrument
(Quantachrome, USA). Prior to the analysis, the samples were
degassed at 150 °C for 12 h. Total pore volume was calculated at
a relative pressure near unity (P/Po = 0.9994). Micropore volume
was calculated from t-plot. Meso-macropore volume was calcu-
lated by subtracting the micropore volume from total pore volume.
Temperature programmed desorption (TPD) was performed for the
qualitative analysis of oxygen-containing functional groups in the
samples, using the Autosorb-1C instrument. Approximately
100 mg of the sample, contained in the U-shaped quartz reactor,
was first outgassed at 100 °C for 20 min under helium stream
(10 sccm) to remove moisture. The temperature was then
increased 10 °C/min from 100 to 800 °C. The surface morphology
of the samples was observed using a high resolution field emission
scanning electron microscope (FE-SEM - MIRA3 series, Tescan).
X-rays diffraction (XRD) patterns of the samples were recorded
using the powder XRD instrument (X0 Pert3, PANalytical, USA) and
Cu Ka radiation (k = 1.5418 Å). The analysis was performed over
the 2h-range of 20–80° at the scanning rate of 3° per min. The aver-
age crystalline size was calculated using Scherrer’s formula.
2.4. ADS experiments
A high pressure SS vessel (I.D. = 45 mm, O.D. = 80 mm,
height = 60 mm), equipped with an electric heater was used for
Fig. 1. Experimental setup for the adsorptive desulfurization study.
performing ADS tests (Fig. 1). The temperature of the vessel was
controlled using a PID controller. A stirrer cum sample holder, fab-
ricated from SS was connected to an electric motor for stirring the
test solution at a constant speed in the vessel. The sample holder
contained perforations of 1 mm-diameter at side and bottom walls.
A SS mesh was fixed inside the sample holder to prevent beads
from escaping the sample holder, while allowing free flow of the
solution into and out from the holder during stirring.
Adsorption tests were performed over different temperatures
(30, 100 and 150 °C). Approximately 0.1 g of the prepared beads
(ACB, NiO/ACB, Ni/ACB, and Ni-CNF/ACB) were mixed into
20 mL of test oil (n-octane containing TH/DBT or a commercial
diesel blended with TH/DBT) contained in the reactor. The reactor
was heated 5 °C/min to the required temperature. After the tem-
perature of the test mixture reached the set temperature, the pres-
sure in the vessel was increased to 0.7 MPa, using N2 gas. The
mixture was stirred for 4 h. The solid beads were separated from
the liquid, using a filter paper (Whatman ashless, grade no. 42).
Total concentration of sulfur in diesel was determined using the
X-ray fluorescence spectrometer (XRF-Rigaku) with an X-ray gen-
erator (4 kW, 60 kV, 150 mA). The concentration of DBT or TH in
n-octane was determined using a gas chromatography equipped
with the flame ionization detector (Nucon-5700, India). The
adsorption capacity, qe (mg-S/g) of the prepared adsorbents was
determined using the S-balance equation:
ðCo  C e ÞW oil
qe ¼ ð1Þ
1000  W ads
where Co and Ce are the initial and final (equilibrium) sulfur concen-
trations (mg/kg or ppmw) in oil, respectively, and Woil and Wads are
the weight (g) of oil and adsorbent, respectively.
3. Results and discussion
3.1. SEM analysis
Fig. 2 shows the SEM images of the synthesized materials. The
ACBs contained hierarchal pores including micropores (pore
size < 2 nm), mesopores (2 nm < pore size < 50 nm) and macrop-
ores (pore size > 50 nm), and these were visible in the image
(Fig. 2a). The NiO/ACB and Ni/ACB samples exhibited similar sur-
face morphologies, with the shiny Ni particles clearly visible on
the external surface and within pores of the materials (Fig. 2b
and c). The average particle size was found to be between 30 and
50 nm. A few relatively larger-sized NiO and Ni particles were also
formed because of the agglomeration of small particles during the
heat treatment. The growth of CNFs was dense and uniform across
the entire surface of the Ni-CNF/ACBs (Fig. 2d). The large
magnification-images showed that the CNFs were hollow, with
the average diameter of the CNFs estimated to be approximately
70 nm (Fig. 2e and f). The ultrasonication caused the removal of
the Ni nanoparticles from the tips of the CNFs, whereas the
nanoparticles dispersed on the surface remained largely intact
(Fig. 2g). The EDX analysis showed a relatively less amount of Ni
nanoparticles in the sonicated sample (Fig. 2h and i). The EDX spec-
tra confirmed the presence of Ni elements (Fig. 2j). The amount (w/
w %) of Ni decreased from 2% in the pre-sonicated samples to
0.45% in the sonicated samples. The Raman spectroscopy analysis
of the CNF-containing materials revealed the high graphitic con-
tent (indicated via low ID/IG ratio) [25].
3.2. BET surface area and PSD
Fig. 3a shows the adsorption-desorption isotherms of the syn-
thesized adsorbents. The N2 uptake in ACBs rapidly increased up
 Y.N. Prajapati, N. Verma / Fuel 189 (2017) 186–194 189

Fig. 2. SEM and EDX images of the prepared adsorbents: (a) ACBs, (b) NiO/ACBs, (c) Ni/ACBs, (d–f) Ni-CNF/ACBs before sonication, (g) Ni-CNF/ACBs after sonication, EDX
mapping of (h) sonicated Ni-CNF/ACBs, (i) pre-sonicated Ni-CNF/ACBs, and (j) EDX spectra and elemental compositions.

to the relative pressure of approximately 0.1. Beyond this pressure
the uptake was almost constant, indicating that the adsorption iso-
therm was type-I. Microporous samples predominantly exhibited
type-I isotherm. The NiO/ACB, Ni/ACB and Ni-CNF/ACB samples
exhibited a mixed type of isotherm, i.e., type–I at low relative pres-
sures and type–IV at medium and high relative pressures [26]. At
low relative pressures, N2 uptake was rapid and there was no pla-
teau at medium relative pressures. An increased N2 volume uptake
was observed at P/Po > 0.85. These materials also exhibited a hys-
teresis loop. The hysteresis was observed to be significant in the
Ni-CNF/ACB samples. Such hysteresis were of type H4, which are
the characteristics of the slit-shaped pores, and indicated that
these materials contained significant amount of mesopores [26].
Fig. 3b shows the t-plots of the different materials prepared in
this study. The dashed black lines provide the micropore volume
in the samples. Fig. 3c shows the textural data of the samples.
The ACBs showed the largest BET surface area (1159 m2/g) and
micropore volume (0.48 cc/g), attributed to the large micropore
content. The inclusion of NiO in the ACBs caused decrease in the
BET surface area. However, mesopore content increased, attributed
to the enlargement of the pores during the heat treatment. The
reduction of NiO into Ni0 resulted in slight increase in the BET sur-
face area and micropore volume. These quantities, however,
decreased in Ni-CNF/ACBs, because micropores were blocked by
the CNFs which covered the pores of the material. The hierarchical
material contained approximately the same amounts of micro- and
190 Y.N. Prajapati, N. Verma / Fuel 189 (2017) 186–194

Fig. 3. Textural data of the adsorbents: (a) N2 adsorption-desorption isotherms, (b) t-plot of the adsorbents. Dashed black line yields actual micropore volume; approximately
40% underestimates for micropore volume yield by using conventional approach shown by red and blue lines [28], (c) Variation of the textural properties in the adsorbents
and (d) BJH-PSD of the adsorbents (inset shows DFT-PSD). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this
article.)

meso-macropores (Fig. 3c). It is mentioned that the measurement
of ultramicropores (pore size 6 0.7 nm), which are available in
the CNFs, can underestimate the BET surface area measurements
by 40%, if using N2 as the probe molecules, because of the diffusion
limitation of the large sized N2 molecules. An improved measure-
ment of ultramicropores has been suggested by using CO2 as probe
molecule [27]. Fig. 3d shows the BJH-PSD of the materials. All sam-
ples showed peak at 39 Å. The peak height was larger for Ni-CNF/
ACBs than for the other materials, indicating that Ni-CNF/ACBs
contained relatively larger numbers of 39 Å-average sized pores.
All adsorbents except ACBs showed a relatively broader PSD in
the range of 100–1000 Å with maxima centered at approximately
500 Å.
The inset of Fig. 3d shows the DFT-PSD. The ACBs showed bimo-
dal pores in micropore range with maxima centered at 14.5 and
18.5 Å. The other materials showed a narrow PSD in micropore
range with maxima centered at 14.5 Å. The relevance of the
described textures (BET and PSD) herein is discussed later.
3.3. X-ray diffraction and TPD analysis
The X-ray diffraction patterns of the synthesized adsorbents are
shown in Fig. 4a. All adsorbents show the existence of parallel gra-
phene layers, indicated by C(0 0 2) peak (2h  23.9°). The peak at
2h  43 ° in ACB samples is attributed to the diffraction of (1 0 0)
plane. The XRD pattern of NiO/ACBs shows the presence of both
Ni and NiO in the sample. The presence of Ni is attributed to the
decomposition of a few NiO crystals into Ni during calcination at
450 °C. The XRD pattern of Ni/ACBs shows the presence of Ni peaks
only, corroborating that all NiO molecules were converted into Ni
during the H2-reduction at 600 °C. The XRD pattern of Ni-CNF/
ACBs confirms the presence of Ni in the material. Small intensity-
peaks for NiO also appear in the Ni-CNF/ACB samples, indicating
that a few Ni NPs were converted into NiO during the CVD at
600 °C. The average crystalline sizes of Ni and NiO were calculated
to be 18.2 and 27.1 nm, respectively.
The TPD analysis of the samples are shown in Fig. 4b. The peaks
at 245 °C observed in all samples are attributed to the decompo-
sition of the oxygen containing strong acidic function groups such
as carboxyl and lactone groups in the form of CO2 [29]. Relatively
higher peak appears in the ACB samples at 245 °C, which indicates
that the material contained relatively larger amounts of the strong
acidic oxygen containing functional groups than the other samples.
The peaks at higher temperatures in the samples are related to the
decomposition of the weak oxygen containing functional groups
such as phenol, carbonyl, and quinone, in the form of CO. The area
under the TPD patterns from their baseline represents the total
amount of oxygen containing functional sites. The areas under
the TPD patterns are calculated to be 7203, 3438, 2076, and 746
(a.u.) for ACBs, NiO/ACBs, Ni/ACBs, and Ni-CNF/ACBs, respectively.
Based on the area under the patterns, the amount of oxygen
containing functional groups in the samples was found in the
 Y.N. Prajapati, N. Verma / Fuel 189 (2017) 186–194 191

Fig. 4. (a) XRD and (b) TPD patterns of the synthesized adsorbents.

following order: ACBs > NiO/ACBs > Ni/ACBs > Ni-CNF/ACBs. We
revisit this aspect of the finding later in the text.
3.4. Adsorption tests
3.4.1. DBT/TH in model oil (n-octane)
Fig. 5a and b shows the adsorption capacity of the adsorbents at
different temperatures for DBT and TH (350 ppmw-S), respectively.
The adsorption capacity for DBT slightly decreased in ACBs with
increase in the temperature (Fig. 5a). On the contrary, the adsorp-
tion capacity of the metal-containing materials, namely, NiO/ACBs,
Ni/ACBs, and Ni-CNF/ACBs, increased with increasing temperatures
and was the highest at 150 °C, indicating the adsorption of the
solute to be endothermic on such materials. Nejad et al. [30] also
reported the endothermic adsorption of DBT over the carbon-
supported Ni.
The adsorption capacity of Ni-CNF/ACBs was significantly smal-
ler than the other adsorbents at all temperatures. The relatively
larger molecular size of DBT (4.93 Å  9.87 Å) [31] and the small
BET area and pore volume of Ni-CNF/ACBs (Fig. 3c) were responsi-
ble for the small adsorption capacity of the CNF-containing mate-
rial. The adsorption capacities of NiO/ACBs and Ni/ACBs were
higher than that of ACBs (i.e., the adsorbent without Ni), signifying
the positive role of the metal towards desulfurization via the direct
Ni-S interaction [5]. The adsorption capacity of Ni/ACBs was higher
than that of NiO/ACBs, attributed to the stronger interaction of the
soft base-DBT with the soft acid-Ni0 than the borderline acid-Ni2+
[32].
The studies have shown that the adsorption capacity of the
carbon-based adsorbents for S-compounds increase with increas-
ing oxygen containing functional groups, attributed to the p-p
and/or p-H interactions between p electrons in the thiophenic aro-
matic ring structure and the carbon surface [19,22,23]. However,
these adsorbents did not contain metal(s). The TPD data, earlier
discussed, showed that the ACB substrate (without Ni) contained
lager amount of oxygen functional groups than in the metal-
included adsorbents. Therefore, it is clear that the direct Ni-S
interaction, rather than p-complexation or p-H interactions was

Fig. 5. Effects of temperature on the adsorption capacity (a) DBT, (b) TH, and (c) effect of pore volume of the adsorbents on adsorption capacity (initial
concentration = 350 ppmw-S in n-octane).
192 Y.N. Prajapati, N. Verma / Fuel 189 (2017) 186–194
dominant for the adsorption of DBT over the metal-containing
materials prepared in this study, viz., NiO/ACBs, Ni/ACBs, and Ni-
CNF/ACBs.
All adsorbents exhibited approximately the same adsorption
capacity for TH, and the effects of temperature on the adsorption
capacity were insignificant (Fig. 5b). Interestingly, Ni-CNF/ACBs
also exhibited the similar adsorption capacity, despite having small
BET surface area and pore volume (Fig. 3c). The molecular size
(4.7 Å  4.7 Å) of TH is relatively smaller [31]. Therefore, the
adsorption of TH was predominantly dependent on the pore filling
[20,24]. The relatively higher adsorption capacity of the CNF-
containing material was attributed to both, the small molecular
size of TH as well as strong Ni-S interactions. The p-p and/or
hydrophobic interactions between the TH molecule and the gra-
phitic planes of CNFs also enhanced the adsorption capacity of
the Ni-CNF/ACBs [33].
The significance of micropore volume of the adsorbents vis a vis
the adsorbate size is further elucidated by analyzing variations in
the adsorption capacity with the pore volume (V<1 nm) of all pores
having less than 1 nm-size in the prepared adsorbents (Fig. 5c).
Whereas the adsorption capacity of the materials for the large
sized molecule (DBT) increased with increasing pore volumes,
the same for the small molecule (TH) was approximately constant,
indicating the effects of the relative size of pore in the adsorbent
and that of adsorbate molecule on adsorption.
3.4.2. Effects of initial solute concentrations on adsorption
Fig. 6 shows the effects of initial concentrations
(350–1200 ppmw-S) of DBT and TH on the adsorption capacity of
Ni/ACBs and Ni-CNF/ACBs at 150 °C. The adsorption capacities of
both adsorbents for DBT are approximately independent of the ini-
Fig. 6. Effects of initial concentrations of (a) DBT and (b) TH in n-octane; diesel blended with (c) DBT and (d) TH on the adsorption capacity of Ni/ACBs and Ni-CNF/ACBs.
tial concentrations (Fig. 6a). However, the average adsorption
capacity of Ni/ACBs is approximately 3.5 times greater than that
of Ni-CNF/ACBs, attributed to the greater amounts of the large
sized DBT molecules adsorbed in the former material, which has
larger BET surface area and pore volume than in the latter material.
The adsorption capacities of both adsorbents, Ni/ACBs and Ni-
CNF/ACBs, increased with increasing initial TH concentrations
(Fig. 6b). Whereas the adsorption capacity of Ni/ACBs leveled off
at or greater than 800 ppmw-S concentration, the adsorption
capacity of Ni-CNF/ACBs increased with increasing concentrations
over the entire concentration range. The higher adsorption capac-
ity of Ni-CNF/ACBs than Ni/ACBs at large concentration
(1200 ppmw-S) is attributed to the adsorption of the small sized
TH on the graphitic planes of the Ni-CNF/ACBs via the p-p and/or
p-H interactions. The dominant role of such chemical interactions
in the Ni-CNF/ACB adsorbents was also earlier discussed.
The adsorption performance of Ni/ACBs and Ni-CNF/ACBs was
tested on a commercial diesel blended with DBT or TH. The com-
mercially available diesel oil contained low S-contents. Therefore,
it was blended with DBT or TH in requisite quantities to increase
the S-concentration in the oil for the adsorption study. Fig. 6c
and d shows variations in the adsorption capacity of Ni/ACBs and
Ni-CNF/ACBs with increasing initial sulfur concentrations (300,
500, 800, and 1200 ppmw-S) in diesel oil. There are two salient
observations. (1) The adsorption capacities of both adsorbents
(Ni/ACBs and Ni-CNF/ACBs) decreased in diesel oil in comparison
to the model oil. The decrease is attributed to the commercial die-
sel oil containing several types of non-S compounds including
polycyclic hydrocarbons [23]. In such case, the competitive adsorp-
tion between S-compounds and hydrocarbons results in the
reduced adsorption capacity of the materials for S-compounds.
 Y.N. Prajapati, N. Verma / Fuel 189 (2017) 186–194 193

Table 1
Comparative adsorption capacity of various adsorbents for organosulfur compounds in model and commercial diesel oils.

S. N. Solvent (ppmw-S)a Adsorbent Adsorbent amount (g/L) Adsorption capacity (mg-S/g) Ref.
1 DBT in diesel (800) Ni/ACBs 5 7.1 Present study
DBT in n-octane (800) 27.3
TH in diesel (800) 41.1
TH in n-octane (800) 85
2 DBT in diesel (800) Ni-CNF/ACBs 5 8.1 Present study
DBT in n-octane (800) 8.2
TH in diesel (800) 62.3
TH in n-octane (800) 88.2
3 Commercial diesel (27) Activated carbon 80 0.41 [12]
4 DBT in commercial diesel (870) GAC 40 5.1 [21]
5 DBT in n-octane (17.4) Activated carbon 5 2.5 [34]
6 BT in n-octane (12–239) ACs – 16.7 [35]
7 DBT in n-octane (71–657) PdCl2/poly-methylbenzene 4 26 [36]
8 DBT in n-hexane (100–1000) Activated Al2O3 20 22.6 [37]
9 TH in isooctane (550) AgCl 4 15.7 [38]
10 DBT in n-hexadecane (44–696) Carbon aerogels 16 15.1 [39]
11 Mixture of TH and DBT in n-octane (700) Ag-Y Zeolite 20 23.8 [40]
12 DBT in n-hexadecane (250) Ni–Cu/c-Al2O3 50–125 3.9 [41]
a
Values given in parentheses are the total amount of sulfur in oil.

The effects are significant for the small TH molecules which could
penetrate both types of adsorbents without hindrance. (2) The
adsorption capacity of Ni-CNF/ACBs is greater than that of Ni/ACBs
for both DBT and TH in the blended diesel oils, re-corroborating the
significance of the graphitic carbons in the enhanced adsorption of
S-compounds via the formation of p-complexation, and therefore,
the suitability of such (CNF-containing) adsorbents for the ADS of
commercial diesel oils.
A comparison of the adsorption capacity of the prepared
adsorbents (Ni/ACB and Ni-CNF/ACB) in this study with several
adsorbents, including carbon-based materials, discussed in the lit-
erature for ADS was also made where possible (Table 1). Although
a direct comparison of the adsorption capacity may not be appro-
priate because of its dependency on initial sulfur concentrations,
the comparative data indicate that the adsorption capacities of
the prepared Ni containing adsorbents with and without CNFs in
this study are greater than most of the adsorbents discussed in
the literature.
4. Conclusions
The Ni-dispersed porous carbon beads with/without CNFs were
synthesized as an adsorbent for the efficient removal of sulfur from
diesel oil. The significance of BET surface area, pore volume, PSD,
and graphitic content were experimentally demonstrated using
two different sized S-compounds, namely, TH and DBT, in n-
octane and a commercial diesel oil. The adsorption capacity of
Ni/ACBs was expectedly higher for the large sized DBT molecules
in n-octane, whereas the Ni-CNF/ACBs exhibited higher adsorption
capacity for the small sized TH molecules, despite having lesser
BET surface area and pore volume. The latter effect was attributed
to the chemical interactions between S-compounds and graphitic
carbons of the Ni-CNF/ACBs via p-complexation. The adsorption
capacity of both adsorbents (with/without CNFs) decreased in the
commercial diesel because of the competitive adsorption between
S-compounds and polycyclic hydrocarbons present in the commer-
cial solvent. This study has shown that the textural properties of
the adsorbent and the size of the adsorbate molecules must be
considered while selecting a suitable material for the efficient
desulfurization of diesel oils by adsorption.
References
[1] Ai S, Sui H, Li H. Coordinative adsorption of thiophene with metallic silver/
adsorbent cotton prepared via aqueous in situ reduction for desulfurization.
Ind Eng Chem Res 2012;51:12337–43.
[2] Bansal G, Bandivadekar A. Overview of India’s vehicle emissions control
program past successes and future prospects. <http://tinyurl.com/h2b2d8w>;
April 2016.
[3] Chen X et al. Brønsted-Lewis acidic ionic liquids and application in oxidative
desulfurization of diesel fuel. Energy Fuels 2015;29:2998–3003.
[4] Li Y-X, Jiang W-J, Tan P, Liu X-Q, Zhang D-Y, Sun L-B. What matters to the
adsorptive desulfurization performance of metal-organic frameworks? J Phys
Chem C 2015;119:21969–77.
[5] Liao J, Wang Y, Chang L, Bao W. Preparation of M/c-Al2O3 sorbents and their
desulfurization performance in hydrocarbons. RSC Adv 2015;5:62763–71.
[6] Sentorun-Shalaby C, Saha SK, Ma X, Song C. Mesoporous-molecular-sieve-
supported nickel sorbents for adsorptive desulfurization of commercial ultra-
low-sulfur diesel fuel. Appl Catal B: Environ 2011;101:718–26.
[7] Ma X, Velu S, Kim JH, Song C. Deep desulfurization of gasoline by selective
adsorption over solid adsorbents and impact of analytical methods on ppm-
level sulfur quantification for fuel cell applications. Appl Catal B: Environ
2005;56:137–47.
[8] Velu S, Ma XL, Song CS, Namazian M, Sethuraman S, Venkataraman G.
Desulfurization of JP-8 jet fuel by selective adsorption over a Ni-based
adsorbent for micro solid oxide fuel cells. Energy Fuels 2005;19:1116–25.
[9] Bhandari VM, Ko CH, Park JG, Han SS, Cho SH, Kim JN. Desulfurization of diesel
using ion-exchanged zeolites. Chem Eng Sci 2006;61:2599–608.
[10] Bandosz TJ. Activated carbon surfaces in environmental
remediation. Amsterdam; London: Elsevier; 2006.
[11] Xiao J, Li Z, Liu B, Xia QB, Yu MX. Adsorption of benzothiophene and
dibenzothiophene on ion-impregnated activated carbons and ion-exchanged y
zeolites. Energy Fuels 2008;22:3858–63.
[12] Muzic M, Sertic-Bionda K, Gomzi Z, Podolski S, Telen S. Study of diesel fuel
desulfurization by adsorption. Chem Eng Res Des 2010;88:487–95.
[13] Hernandez SP, Fino D, Russo N. High performance sorbents for diesel oil
desulfurization. Chem Eng Sci 2010;65:603–9.
[14] Selvavathi V, Chidambaram V, Meenakshisundaram A, Sairam B, Sivasankar B.
Adsorptive desulfurization of diesel on activated carbon and nickel supported
systems. Catal Today 2009;141:99–102.
[15] Li W et al. Deep desulfurization of diesel by integrating adsorption and
microbial method. Biochem Eng J 2009;44:297–301.
[16] Kim JH, Ma X, Zhou A, Song C. Ultra-deep desulfurization and denitrogenation
of diesel fuel by selective adsorption over three different adsorbents: a study
on adsorptive selectivity and mechanism. Catal Today 2006;111:74–83.
[17] Saraswat R, Talreja N, Deva D, Sankararamakrishnan N, Sharma A, Verma N.
Development of novel in situ nickel-doped, phenolic resin-based micro–nano-
activated carbon adsorbents for the removal of vitamin B-12. Chem Eng J
2012;197:250–60.
[18] Sharma A, Verma N, Sharma A, Deva D, Sankararamakrishnan N. Iron doped
phenolic resin based activated carbon micro and nanoparticles by milling:
synthesis, characterization and application in arsenic removal. Chem Eng Sci
2010;65:3591–601.
[19] Triantafyllidis KS, Deliyanni EA. Desulfurization of diesel fuels: Adsorption of
4,6-DMDBT on different origin and surface chemistry nanoporous activated
carbons. Chem Eng J 2014;236:406–14.
[20] Seredych M, Wu CT, Brender P, Ania CO, Vix-Guterl C, Bandosz TJ. Role of
phosphorus in carbon matrix in desulfurization of diesel fuel using adsorption
process. Fuel 2012;92:318–26.
[21] Alhamed YA, Bamufleh HS. Sulfur removal from model diesel fuel using
granular activated carbon from dates’ stones activated by ZnCl2. Fuel
2009;88:87–94.
[22] Shi YW, Zhang XW, Liu GZ. Adsorptive desulfurization performances of
ordered mesoporous carbons with tailored textural and surface properties.
Fuel 2015;158:565–71.
194 Y.N. Prajapati, N. Verma / Fuel 189 (2017) 186–194
[23] Xiao J, Song CS, Ma XL, Lit Z. Effects of aromatics, diesel additives, nitrogen
compounds, and moisture on adsorptive desulfurization of diesel fuel over
activated carbon. Ind Eng Chem Res 2012;51:3436–43.
[24] Zhou A, Ma X, Song C. Effects of oxidative modification of carbon surface on the
adsorption of sulfur compounds in diesel fuel. Appl Catal B: Environ
2009;87:190–9.
[25] Singh S, Verma N. Fabrication of Ni nanoparticles-dispersed carbon micro-
nanofibers as the electrodes of a microbial fuel cell for bio-energy production.
Int J Hydrogen Energy 2015;40:1145–53.
[26] Passe-Coutrin N, Altenor S, Cossement D, Jean-Marius C, Gaspard S.
Comparison of parameters calculated from the BET and Freundlich
isotherms obtained by nitrogen adsorption on activated carbons: a new
method for calculating the specific surface area. Microporous Mesoporous
Mater 2008;111:517–22.
[27] Lozano-Castello D, Cazorla-Amoros D, Linares-Solano A. Usefulness of CO2
adsorption at 273 K for the characterization of porous carbons. Carbon
2004;42:1233–42.
[28] Galarneau A, Villemot F, Rodriguez J, Fajula F, Coasne B. Validity of the t-plot
method to assess microporosity in hierarchical micro/mesoporous materials.
Langmuir 2014;30:13266–74.
[29] Shafeeyan MS, Daud WMAW, Houshmand A, Shamiri A. A review on surface
modification of activated carbon for carbon dioxide adsorption. J Anal Appl
Pyrol 2010;89:143–51.
[30] Nejad NF, Shams E, Amini MK, Bennett JC. Synthesis of magnetic mesoporous
carbon and its application for adsorption of dibenzothiophene. Fuel Process
Technol 2013;106:376–84.
[31] Nuntang S, Prasassarakich P, Ngamcharussrivichai C. Comparative study on
adsorptive removal of thiophenic sulfurs over Y and USY zeolites. Ind Eng
Chem Res 2008;47:7405–13.
[32] Yu MX, Li Z, Xia QB, Xi HX, Wang SW. Desorption activation energy of
dibenzothiophene on the activated carbons modified by different metal salt
solutions. Chem Eng J 2007;132:233–9.
[33] Gupta VK, Kumar R, Nayak A, Saleh TA, Barakat MA. Adsorptive removal of
dyes from aqueous solution onto carbon nanotubes: a review. Adv Colloid
Interface Sci 2013;193–194:24–34.
[34] Nunthaprechachan T, Pengpanich S, Hunsom M. Adsorptive desulfurization of
dibenzothiophene by sewage sludge-derived activated carbon. Chem Eng J
2013;228:263–71.
[35] Jung BK, Jhung SH. Adsorptive removal of benzothiophene from model fuel,
using modified activated carbons, in presence of diethylether. Fuel
2015;145:249–55.
[36] Xia YT et al. Adsorption desulfurization by hierarchical porous organic
polymer of poly-methylbenzene with metal impregnation. Fuel
2016;170:100–6.
[37] Srivastav A, Srivastava VC. Adsorptive desulfurization by activated alumina. J
Hazard Mater 2009;170:1133–40.
[38] Li Y-H, Tan P, Liu X-Q, Zu D-D, Huang C-L, Sun L-B. Facile fabrication of AgCl
nanoparticles and their application in adsorptive desulfurization. J Nanosci
Nanotechnol 2015;15:4373–9.
[39] Haji S, Erkey C. Removal of dibenzothiophene from model diesel by adsorption
on carbon aerogels for fuel cell applications. Ind Eng Chem Res
2003;42:6933–7.
[40] Zhang ZY, Shi TB, Jia CZ, Ji WJ, Chen Y, He MY. Adsorptive removal of aromatic
organosulfur compounds over the modified Na-Y zeolites. Appl Catal B:
Environ 2008;82:1–10.
[41] Mansouri A, Khodadadi AA, Mortazavi Y. Ultra-deep adsorptive desulfurization
of a model diesel fuel on regenerable Ni–Cu/c-Al2O3 at low temperatures in
absence of hydrogen. J Hazard Mater 2014;271:120–30.