Separation and Purification Technology

Adsorptive desulfurization and denitrogenation of model liquid fuel using Mo-AAC:

Central composite design
--Manuscript Draft--

Manuscript Number:

Article Type: Full Length Article

Section/Category: Adsorbent synthesis (carbons, zeolites, metal organic frameworks) for gas or liquid
applications

Keywords: Adsorption; Desulfurization; Denitrogenation; Response surface methodology;

Central composite design

Corresponding Author: Verraboina Subbaramaiah, Ph.D.

Malaviya National Institute of Technology Jaipur
Jaipur, Rajasthan INDIA

First Author: Prerana Sikarwar, Ph.D.

Order of Authors: Prerana Sikarwar, Ph.D.

Vijayalakshmi Gosu, Ph.D.

Verraboina Subbaramaiah, Ph.D.

Tian C. Zhang

Rao Y. Surampalli, Ph.D.

Abstract: Removal of refractory sulfur and nitrogen compounds from model gasoline was
investigated through the batch adsorption method using molybdenum-modified acid-
treated granular activated carbon (Mo-AAC). The synthesized adsorbent was
characterized by using various characterization techniques. Statistical optimization of
parameters (concentration of dibenzothiophene (DBT), indole, adsorbent dose,
temperature) was carried out by employing response surface methodology with a
central composite design. Analysis of variance results are shown for the removal of
both contaminants, and the effect of the influence of parameters and their interaction
on the removal of dibenzothiophene and indole was discussed. The adsorbent dose
was found to be the most significant factor in the removal of both dibenzothiophene
and indole. Interaction between the concentration of dibenzothiophene and indole was
found to be the only interaction that influences the removal of dibenzothiophene.
Temperature and adsorbent dose have a synergistic effect on the removal of both DBT
and indole, whereas concentrations of DBT and indole have an antagonistic effect. The
optimum initial contaminant concentration (DBT and indole), temperature, and
adsorbent dose were found to be 186 mg/L, 50 mg/L, 25.5 °C, and 32 g/L, respectively.
Results demonstrated that Mo-AAC can be effectively used to remove refractory sulfur
and nitrogen compounds from fuel.

Suggested Reviewers: Tawfik A. Saleh, Ph.D.

Professor, King Fahd University of Petroleum & Minerals
tawfik@kfupm.edu.sa
Prof. Saleh Expert in Desulfurization and Material Science.

Dilip H. Lataye, Ph.D.

Professor, SVNIT: Sardar Vallabhbhai National Institute of Technology
dhlataye@civ.vnit.ac.in
Prof. Lataye expert in adsorption and pollution abatement.

Bassim Hamid, Ph.D.

Professor, Qatar University
b.hammadi@qu.edu.qa
Prof.Hamid Expert in Catalysis

Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation
Cover Letter

Dated: Aug 22, 2021

Editor,
Separation and Purification Technology

Dear Editor,

I am hereby submitting the manuscript entitled "Adsorptive desulfurization and

denitrogenation of model liquid fuel using Mo-AAC: Central composite design" for

consideration of publication in Separation and Purification Technology.

We confirm that this manuscript is based on original work carried out by the (co)authors
and it has not been submitted anywhere else for possible publication.

Manuscript Details:
Total Number of Words: 6655
Total Number of Figures: 5
Total Number of Tables: 3
Supporting information: Supporting table 1, and figures 4.

Thank you for your consideration of this manuscript.

Sincerely,
Dr. V Subbaramaiah
Assistant Professor
Department of Chemical Engineering,
MNIT Jaipur,
Jaipur-302017 India
Highlights

Highlights

 Active metal incorporated acid-activated carbon is effective in the adsorption of

heterocyclic sulfur and nitrogen compounds.

 Surface area and pore volume significantly improved in acid-activated carbon.

 Competitive adsorption of heterocyclic DBT and indole possible mechanisms were

summarized.

 Temperature and adsorbent dose have a synergistic effect on the removal of both DBT

and indole.
Graphical Abstract

Graphical Abstract
Manuscript File Click here to view linked References

1 Adsorptive desulfurization and denitrogenation of model liquid fuel using Mo-AAC:

2 Central composite design

3 Prerana Sikarwar1, Vijayalakshmi Gosu1,*, Verraboina Subbaramaiah1,*, Tian C. Zhang2 and Rao

4 Y. Surampalli3

1
5 Department of Chemical Engineering, Malaviya National Institute of Technology Jaipur, 302017,

6 India.

2
7 Department of Civil & Environmental Engineering, University of Nebraska-Lincoln, Omaha, NE

8 68182, USA.

3
9 President, CEO and Chief Technology Officer, Global Institute for Energy, Environment, and

10 Sustainaility (GIEES), P.O. Box 14354, Lenexa, Kansas 66285, USA.

11 Corresponding author E-mail: vlakshmi.chem@mnit.ac.in (VG), vsr.chem@mnit.ac.in (VS).

12

1
 1 ABSTRACT

2 Removal of refractory sulfur and nitrogen compounds from model gasoline was investigated

3 through the batch adsorption method using molybdenum-modified acid-treated granular activated

4 carbon (Mo-AAC). The synthesized adsorbent was characterized by using various characterization

5 techniques. Statistical optimization of parameters (concentration of dibenzothiophene (DBT),

6 indole, adsorbent dose, temperature) was carried out by employing response surface methodology

7 with a central composite design. Analysis of variance results are shown for the removal of both

8 contaminants, and the effect of the influence of parameters and their interaction on the removal of

9 dibenzothiophene and indole was discussed. The adsorbent dose was found to be the most

10 significant factor in the removal of both dibenzothiophene and indole. Interaction between the

11 concentration of dibenzothiophene and indole was found to be the only interaction that influences

12 the removal of dibenzothiophene. Temperature and adsorbent dose have a synergistic effect on the

13 removal of both DBT and indole, whereas concentrations of DBT and indole have an antagonistic

14 effect. The optimum initial contaminant concentration (DBT and indole), temperature, and

15 adsorbent dose were found to be 186 mg/L, 50 mg/L, 25.5 °C, and 32 g/L, respectively. Results

16 demonstrated that Mo-AAC can be effectively used to remove refractory sulfur and nitrogen

17 compounds from fuel.

18

19 Keywords: Adsorption; Desulfurization; Denitrogenation; Response surface methodology;

20 Central composite design.

21

2
 1 1. Introduction

2 Fossil fuels are the primary source of energy production; nearly half of the energy production is

3 dependent on petroleum. Petroleum contained obnoxious impurities, namely, sulfur and

4 nitrogenous compounds. Because of the stringent environmental rules and regulations, removal of

5 these impurities before the use of fuel has become a challenging task. During combustion,

6 heterocyclic sulfur and nitrogen compounds present in transportation fuel can produce SOx and

7 NOx emissions, and these emissions are accountable for haze, acid rain, and global warming.

8 Moreover, nitrogen compounds present in oil tend to make oil more viscous, which creates a

9 hindrance in oil transportation [1]. Sulfur compounds are liable for the deactivation of the catalyst

10 present in the catalytic converter and thus reduce the efficiency of the engine [2–4].

11 Hydrodesulfurization (HDS) and hydrodenitrogenation (HDN) are employed industrially to

12 remove sulfur and nitrogen compounds, respectively. Conventional HDS and HDN utilize the

13 heterogeneous catalyst and hydrogen at higher temperatures and pressures. Therefore, the process

14 demands severe operating conditions as well as pronounced active catalysts, to make the HDS and

15 HDN process an expensive technique for the desulfurization and denitrogenation of liquid fuels.

16 In sequence, this process has several limitations including, side reactions, low cetane number fuel

17 production, and rapid catalyst deactivation.

18 Therefore, there is an urgent need to develop alternative processes to remove heterocyclic nitrogen

19 and sulfur compounds effectively from fuel oil. The alternative to HDS/HDN is the adsorptive

20 desulfurization (ADS) and adsorptive denitrogenation (ADN) process. The ADS and ADN process

21 have been received significant attention nowadays because this process operates an

22 environmentally friendly approach with a moderate temperature (below 50 ℃) and atmospheric

23 pressure without any hydrogen. It is an alternative route for the desulfurization and denitrogenation

3
 1 of liquid fuels under mild operating conditions. The considerable benefits of this process (ADS

2 and ADN) are that the fuel quality doesn’t change during the process and sulfur, nitrogen content

3 can be reduced to a very low level [5]. Further, quality may be changed during the HDS and HDN

4 due to its severe operating conditions [6]. Various adsorbents have been explored for the

5 adsorption of heterocyclic sulfur and nitrogenous compounds from liquid fuels using activated

6 carbon [7], mesoporous material [5], supported reduced metals [8], metal-organic frameworks

7 [9,10], and zeolites [11].

8 Among various adsorbents, activated carbon (AC) is considered a sustainable adsorbent because

9 AC is easy to prepare and easy to alter the functional group, surface area, etc., according to the

10 required application [12,13]. However, mere AC may not have good adsorption capacity and

11 regeneration capacity that limits its industrial applicability. Several authors have been suggested

12 the activity of the catalyst enhanced by tailoring AC with thermal activation or chemical activation,

13 or active metal incorporation into the framework. Sikarwar et al. [13] explored the chemical

14 activation technique for AC to improve the pore volume and surface area, which significantly

15 enhanced the surface area and pore volume from 257 m2/g to 393 m2/g, and 0.1163 cm3/g to 0.2030

16 cm3/g, respectively. The phenomenon may be because of the removal of inert materials from the

17 pores of the AC framework. Similarly, Shah et al. [14] examined the adsorptive desulfurization of

18 dibenzothiophene (DBT) from model fuel oil using tin (Sn) incorporated activated carbon, and

19 results were demonstrated that active metal incorporated activated carbon showed potential

20 adsorption and regeneration capacity than mere activated carbon. However, scarce reports on

21 simultaneous adsorption of heterocyclic nitrogen and sulfur compounds from model fuel oil using

22 active metal incorporated acid-treated AC. In addition, slight attention has been paid to understand

23 the adsorption mechanism of heterocyclic sulfur and nitrogen compounds on the adsorbent surface.

4
 1 In the present study, the removal of refractory sulfur and nitrogen compounds from model gasoline

2 was investigated through the batch adsorption method using molybdenum-impregnated acid-

3 treated activated carbon (Mo-AAC). Mo-AAC was selected because limited literature has been

4 reported on the simultaneous removal of sulfur and nitrogen compounds using Mo-AAC.

5 However, molybdenum has been proved as a potential active metal for the desulfurization of liquid

6 fuel through the HDS process [4]. In addition, globally, Co-Mo or Ni-MO active metals supported

7 catalysts are widely used for HDS or HDN although, in the adoption study, ADN/ADS were

8 limited tested their applicability. In this study, statistical optimization of parameters (concentration

9 of dibenzothiophene (DBT), indole, adsorbent dose, temperature) was carried out by employing

10 response surface methodology (RSM) with a central composite design. RSM was used as a cost-

11 effective method to study the effect of various factors and their interaction on the response variable

12 with a minimum number of experimental runs.

13 2. Materials and Methods

14 2.1. Materials and model oil

15 Iso-octane (Fisher, Mumbai, India), dibenzothiophene (DBT) (Himedia Pvt. Ltd., Mumbai, India),

16 indole (HPLC lab reagents, Mumbai, India), glacial acetic acid (Rankem, Ludhiana, India),

17 ammonium molybdate tetrahydrate and activated carbon (AC) (Loba Chemie Pvt. Ltd., Mumbai,

18 India) were procured from various firms. Model oil was prepared by mixing iso-octane with an

19 appropriate amount of DBT and indole.

20 2.2. Synthesis of adsorbent

21 Firstly, AC was washed with hot deionized (DI) water multiple times to remove the contaminants

22 from AC pores. Afterward, AC was kept in a hot air oven at 110 ℃ to remove the moisture.
5
 1 Subsequently, AC was treated with acetic acid with 1:2 volume ratio with continuous stirring for

2 3 h at 50 ℃ and then washed with DI water multiple times until the used water became neutral pH.

3 Acid-activated AC was dried in a hot air oven for 10 h at 110 ℃, denoted as AAC. Molybdenum

4 species were incorporated in the pretreated AC framework (AAC) through the impregnation

5 technique. A stoichiometric amount of AAC was added to the corresponding metal salt precursor

6 and stirred at room temperature to obtain Mo loading of 1% (which was found to be the best in the

7 preliminary study). The resultant samples were dried in a hot air oven then calcined at 300 ℃ for

8 3 h using a muffle furnace. The synthesized adsorbent is noted as Mo-AAC and then stored in an

9 air-tight container for further use.

10 2.3. Characterization of adsorbent

11 The Fourier transform infrared spectroscopy (FTIR) spectrum of Mo-AAC was recorded on an

12 FTIR spectrophotometer (FTIR Spectrum 2, Perkin Elmer) in the fingerprint region of 400–4000

13 cm-1. Samples were pretreated with KBr powder in a specific weight ratio. The morphology of the

14 synthesized materials was analyzed using Nova Nano FE-SEM 450 (FEI). Before FE-SEM

15 analysis, the samples were pretreated with a gold coating (spin coating system, APEX) to enhance

16 the conductivity of materials. The crystallinity of the synthesized materials was examined using

17 PANalytical X Pert Pro. The PANalytical X Pert Pro instrument operated using Cu-Kα radiation

18 (k = 1.54060 Å) at 40 mA, 40 kV with a step size of 0.02° and a step time of 4 s in the 2θ range of

19 10°–90°. Nitrogen physisorption was performed at 77 K with the help of a Micromeritics

20 instrument (ASAP 2020) to calculate Brunauer–Emmett–Teller (BET) surface area, pore-volume,

21 and average pore size. Before the adsorption measurements, all samples were degassed at 200 ℃

22 for 12 h to remove impurities from the pores of the AC framework. X-ray photoelectron

6
 1 spectroscopy (XPS) spectra of samples were obtained on a Scienta Omicron photoelectron

2 spectrometer equipped with a monochromatic Al Kα source.

3 2.4. Experimentation and analytical procedure

4 To carry out the adsorption experiments, stock solutions of DBT and indole containing model oil

5 were prepared by dissolving 1 g of DBT and 1 g of indole in 1 liter of iso-octane individually. The

6 aforementioned stock solution was used to prepare different concentrations of DBT and indole

7 using iso-octane as a medium. For each test run, 10 ml of the model oil was mixed with the required

8 quantity of Mo-AAC in a conical flask and stirred at the desired temperature for 4.5 h. The

9 performance of Mo-AAC for the removal of DBT and indole was estimated by measuring the

10 concentration of heterocyclic sulfur and nitrogen contaminants in model oil before and after the

11 experiments. The concentration of DBT and indole were determined by using gas chromatography

12 (Trace 1310, Thermo Scientific, USA) equipped with a flame ionization detector. Samples were

13 filtered and injected in the splitless model by injecting 1 µL of the sample into the capillary

14 column. Injector and detector temperature was set at 300 ℃ and 320 ℃, respectively. Colum

15 temperature was kept at 70 ℃ for 1 min hold and then raised to 300 ℃ with a ramp rate of 30

16 ℃/min, and then hold 2.5 min at 300 ℃.

17 The percentage removal of DBT and indole was determined with the aid of the following

18 equations:

COD  C ED
19 DBT removal (%) =  100 (1)
COD

COI  C EI
20 Indole removal (%) = 100 (2)
COI

7
 1 where COD and COI represent the initial concentration of DBT and indole in the model oil,

2 respectively, and CED and CEI indicate the final concentration of DBT and indole in the model oil.

3 2.5. Design of experiments using response surface methodology (RSM)

4 CCD has opted for the statistical design of experiments as it assists to optimize the main parameters

5 with a minimum number of experimental runs and aids in determining the interaction among the

6 variables [15]. In this study, CCD was applied by using Minitab software (version 17, Minitab

7 Inc., USA) to investigate the responses of DBT removal and indole removal as a function of four

8 adsorption parameters (concentration of DBT, concentration of indole, temperature, dose), each

9 with five levels. Generally, CCD is characterized by the following experimental runs, namely, 2k

10 factorial runs (k is the number of variables), 2k axial runs, and the number of central runs [16].

11 The present study consists of total 31 experiments, including 16 factorial points, 8 axial points,

12 and 7 replicates at the center. The factors were chosen on the basis of previously carried out

13 experimental studies [13]. The range and levels of factors are shown in Table S1 (Supporting

14 Information). Furthermore, Mo-AAC was used as an adsorbent, and experiments have conducted

15 for 4.5 h (found to be optimum in the preliminary study).

16 The mathematical association between the four independent factors and their responses was

17 modeled by a second-order polynomial equation comprising linear, quadratic, and cross-product

18 terms [17].

k k k k
19 y   0    i X i    ii X i2    ij X i X j   (3)
i 1 i 1 i 1 j 1

20 where y is the predicted response; k is the number of parameters; Xi and Xj are the coded

21 parameters; βo, βi, βii, βij indicates regression coefficient for intercept, linear, and interaction terms,
8
 1 respectively, and ε represents residual error. For statistical calculations, the following relationship

2 was used to code the independent parameters:

𝑥𝑖 −𝑥𝑜
3 Coded value (𝑋𝑖 ) = (4)
𝛥𝑥𝑖

4 where xi, and x0 are the actual value of an independent variable and the actual value at the center

5 point, respectively. Δxi is the step change value [18].

6 The analysis of variance (ANOVA) was employed to justify the adequacy of the developed

7 regression model and the importance of parameters.

8 3. Results and Discussion

9 3.1. Characterization of adsorbent

10 As shown in Fig. S1(a) (Supporting Information), the surface of AC possess porous nature when

11 compared to the surface of Mo-AAC (Fig. S1 (b)). Further, rod shape particles aggregation can be

12 clearly seen on the surface of Mo-AAC, which indicates the successful impregnation of

13 molybdenum species on the AC framework. Obtained results are in good agreement with the

14 previously reported literature [19,20].

15 In Fig. S2(a), broad humps observed around 24o and 44o in the Mo-AAC indicate the presence of

16 amorphous carbon in the sample (Prahas et al. 2008). Peaks appeared at 12.90°, 23.62°, 25.86°,

17 27.54°, 33.9°, and 39.13° represent the presence of MoO3 with orthorhombic phase [21].

18 Fig. S2(b) shows that the Mo-AAC maintained the characteristic peaks of AC. The Mo-AAC

19 possess various functional groups such as carboxyl and carbonyl group with absorption bands

20 corresponds at 1737 cm-1 and 1628 cm-1, respectively [22]. The C-O and O-H stretching vibration

21 were observed with an absorption peak at 1035 cm-1 and 3441 cm-1, respectively [23]. The

9
 1 observed stretching may be due to the presence of oxygen functional group on carbon surface as

2 well as water vapor molecules on the pores of AC framework [13]. The C-H stretching vibration

3 was observed with an absorption peak corresponds to 2919 cm-1 and 2850 cm-1. These all peaks

4 are presented in the framework of AC (Fig. 2(b)). However, the two additional peaks were obtained

5 in the spectra of Mo-AAC. The peak at 900 cm-1 represented the asymmetric stretching of Mo-O-

6 Mo, whereas the band at 776 cm-1 showed the stretching of Mo=O [19].

7 It can be clearly seen from Fig. 1(a) that AC is highly microporous in nature due to the high uptake

8 of nitrogen at a relative pressure of less than 0.25 [24]. Pore size distribution (Fig. 1(b)) also

9 indicates the high microporous volume. Hysteresis loop can be observed in Mo-AAC, indicating

10 that incorporation of molybdenum species in the AC framework induces mesoporosity in the

11 sample. Further, the surface area significantly enhanced from 257 m2/g to 316 m2/g when bare AC

12 was modified with an acid treatment and active metal incorporation. The increase in surface area

13 may be by the removal of inert sand, slit materials from the pores of the AC framework with the

14 help of pretreatment (Hot DI water washing and acid activation).

15 Fig. 2(a) indicates that the adsorbent sample mainly consists of carbon, oxygen, and molybdenum.

16 Fig. 2(b) shows C 1s XPS spectra can be deconvoluted into four peaks, namely, C-C (284.2 eV),

17 C-O (285.3 eV), C=O (287.4 eV), and O-C=O (289.4 ev). O 1s XPS spectra, as shown in Fig. 2(c),

18 are deconvoluted into five characteristics peaks, namely, MoO3 (530.2 eV), C=O (531.1 eV), C-O

19 (532 eV), O-H (533.1 eV). Fig. 2(d) displays the Mo 3d XPS spectra of Mo-AAC. Binding energies

20 of Mo 3d5/2 and Mo 3d3/2 peak for Mo6+ were found to be 232.6 and 235.9 eV [25].

10
 (a)
(b)

Fig. 1. (a) Nitrogen adsorption-desorption isotherm of AC and Mo-AAC, (b) Pore size
distribution of Mo-AAC.
1

(a) (b)

(c) (d)

2 Fig. 2. XPS spectra (a) full survey scan, (b) deconvoluted C 1s spectra of Mo-AAC, (c)
3 deconvoluted O 1s spectra of Mo-AAC, (d) deconvoluted Mo 3d spectra of Mo-AAC.

11
 1 3.2. Statistical analysis

2 Table 1 shows the details of 31 test conditions as per CCD design and the experimental results.

3 The empirical model equations in terms of coded parameters for percentage removal of DBT and

4 indole are given as below:

5 DBT removal (%) = 69.20 - 6.684A - 6.361B + 3.094C + 7.792D + 0.207 A2 - 0.104B2 + 1.442C2

6 - 0.173D2 + 1 .741A*B + 0.524A*C + 0.318A*D - 0.580B*C + 1.633 B*D - 0.280C*D (5)

7 Indole removal (%) = 67.43 - 6.617A - 6.998B + 4.554C + 7.911D + 0.148A2 - 0.112B2 + 0.558

8 C2 - 0.343D2 + 2.876A*B - 1.954A*C - 0.109A*D + 0.899B*C + 1.922B*D - 1.468C*D

9 (6)

10 For the DBT removal, significant model terms were A, B, D, C*C, A*B, B*D, and insignificant

11 terms were C, A*A, B*B, D*D, A*C, A*D, B*C, and C*D. For the indole removal, significant

12 model terms were A, B, D, A*B, A*C, B*D, and insignificant terms were C, A*A, B*B, C*C,

13 D*D, A*D, B*C, and C*D. The results obtained from the statistical analyses indicated that models

14 were appropriate to predict the removal of both indole and DBT within the range of studied

15 parameters. The positive coefficient in the model equations depicts the positive effect on the

16 response variable, whereas the negative coefficient indicates the adverse effect. Temperature and

17 adsorbent dose have a synergistic effect on the removal of both DBT and indole, whereas

18 concentrations of DBT and indole have an antagonistic effect.

19 ANOVA was used to verify the statistical adequacy of the developed regression model and the

20 importance of parameters. Tables 2 and 3 show the ANOVA for the response surface quadratic

21 model of DBT and indole, respectively. Validation of the model was carried out by using the F

22 value and P-value. When the F value is high and P-value is less than 0.05, it shows that the model

12
 1 and independent parameters are significant [26]. The model F values of 36.13 and 26.78, and the

2 P-value of 0.000 and 0.000 for DBT and indole removal, respectively, indicated that obtained

3 models were significant. The fitness of the developed adsorption models was depicted by their

4 insignificant lack of fit, p = 0.184, and 0.195 for DBT and indole removal.

5 The high values of the coefficient of determination (R2) values for both DBT (0.969) and indole

6 (0.959) adsorption models also indicated that obtained models were suitable to represent the

7 experimental data. The regression (s) standard error indicates how well our experimental

8 observations fit into the developed regression model [23]. It also measures the average distance

9 between the obtained experimental values and the regression line. Here a value of 2.775 and 3.480

10 were obtained for DBT and indole adsorption models, respectively; the low value of S indicates

11 that the model is appropriate to represent the experimental data.

12 Fig. S3 (a) and Fig. S3 (b) compare the experimental and adsorption model predicted values for

13 DBT and indole removal, respectively. To obtain a good correlation between experimental and

14 predicted values, obtained R2 values should be no less than 0.80 [27]. In the present work, the R2

15 value was 0.968 for DBT removal and 0.958 for indole removal, indicating a good fit.

13
1 Table 1. Design matrix for the central composite design and test results.
Run A B C D A A B B
(mg/L) (mg/L) (°C) (g/L) (% removal) (predicted (% removal) (predicted
% removal) % removal)
1 350 350 27.5 25 69.6000 69.2000 69.0000 67.4286

2 200 500 20.0 20 60.3946 57.9848 42.1907 45.6068

3 500 200 20.0 30 73.9844 71.2746 71.0843 70.8313

4 350 350 27.5 25 68.8000 69.2000 65.0000 67.4286

5 350 350 27.5 25 71.0000 69.2000 69.0000 67.4286

6 350 50 27.5 25 81.4211 82.3380 80.2119 80.9750

7 350 350 27.5 25 67.0000 69.2000 64.0000 67.4286

8 350 350 27.5 35 82.1429 84.0901 80.4286 81.8802

9 200 200 35.0 30 91.1606 94.2775 90.0723 94.4110

10 500 500 35.0 20 54.0573 53.0517 51.3989 52.1832

11 200 500 20.0 30 76.3511 76.7579 70.9798 68.4254

12 500 200 35.0 20 66.5898 66.7158 60.6139 62.4730

13 200 500 35.0 20 59.2823 62.5249 63.7969 63.3545

14 350 350 27.5 25 72.0000 69.2000 71.0000 67.4286

15 350 350 42.5 25 82.1265 81.1561 81.5779 78.7696

14
 16 350 650 27.5 25 55.0000 56.8937 52.0000 52.9823

17 500 500 35.0 30 71.8517 71.9758 65.8517 68.6962

18 50 350 27.5 25 83.2450 83.3961 81.2962 81.2549

19 350 350 27.5 25 66.0000 69.2000 65.0000 67.4286

20 200 200 20.0 20 75.8864 76.2952 74.5360 70.9962

21 500 200 35.0 30 80.0432 79.1097 75.7657 71.2995

22 350 350 27.5 25 70.0000 69.2000 69.0000 67.4286

23 350 350 27.5 15 52.0597 52.9231 49.9415 50.2352

24 200 500 35.0 30 83.4630 80.1787 81.6629 80.3026

25 500 500 20.0 20 49.0000 46.4159 47.2841 42.2501

26 200 200 35.0 20 85.4096 83.1537 84.8357 85.1494

27 500 500 20.0 30 67.5467 66.4592 65.9972 64.6335

28 500 200 20.0 20 57.8206 57.7615 55.8240 56.1343

29 350 350 12.5 25 65.0000 68.7810 56.0000 60.5537

30 650 350 27.5 25 54.0000 56.6595 53.0000 54.7867

31 200 200 20 30 90.8759 88.5381 87.9625 86.1282

1

15
1 Table 2. ANOVA for removal of DBT.
Source Sum of squares Degree of freedom F value P value
Model 3896.76 14 36.13 0.000
DBT (A) 1072.27 1 139.20 0.000
Indole (B) 971.12 1 126.07 0.000
Temperature (C) 229.72 1 29.82 0.000
Dose (D) 1457.07 1 189.16 0.000
DBT*DBT(A*A) 1.22 1 0.16 0.695
Indole*Indole (B*B) 0.31 1 0.04 0.844
Temperature*Temperature(C*C) 59.47 1 7.72 0.013
Dose*Dose (D*D) 0.86 1 0.11 0.743
DBT*Indole (A*B) 48.51 1 6.3 0.023
DBT*Temperature (A*C) 4.39 1 0.57 0.461
DBT*Dose (A*D) 1.61 1 0.21 0.653
Indole*Temperature (B*C) 5.38 1 0.7 0.413
Indole*Dose (B*D) 42.64 1 5.54 0.032
Temperature*Dose (C*D) 1.25 1 0.16 0.692
Lack of fit 96.13 10 2.13 0.184
Pure Error 27.12 6
Total 4020.00 30
2

3 Table 3. ANOVA for removal of indole.

Source Sum of squares Degree of freedom F value P value
Model 4540.84 14 26.78 0.000
DBT (A) 1050.85 1 86.76 0.000
Indole (B) 1175.38 1 97.04 0.000
Temperature (C) 497.73 1 41.03 0.000
Dose (D) 1502.11 1 124.01 0.000
DBT*DBT(A*A) 0.63 1 0.05 0.823
Indole*Indole (B*B) 0.36 1 0.03 0.865
Temperature*Temperature 8.91 1 0.74 0.404
(C*C)
Dose*Dose (D*D) 3.36 1 0.28 0.606
DBT*Indole (A*B) 132.37 1 5.57 0.004
DBT*Temperature (A*C) 61.07 1 10.93 0.039
DBT*Dose (A*D) 0.19 1 5.04 0.902
Indole*Temperature (B*C) 12.92 1 0.02 0.317
Indole*Dose (B*D) 59.08 1 1.07 0.042
Temperature*Dose (C*D) 34.46 1 4.88 0.111
Lack of fit 150.09 10 2.06 0.195
Pure Error 43.71 6
Total 4734.64 30

16
 1 3.3. Effect of main parameters on the response variables

2 Fig. S4 (a) and (b) gives information on the mean effect of each variable on the response variable.

3 As shown in Fig. S4(a), on the increase in the concentration of DBT and indole, the removal of

4 DBT decreased significantly, which could be attributed to the decrease in the number of adsorption

5 sites due to the increase in the number of species of DBT and indole (Olgun and Atar, 2011). In

6 addition, the decrease in DBT removal may be due to the competitive adsorption between DBT

7 and indole molecules to get adsorbed on active sites of the adsorbent. Similarly, the removal of

8 indole was reduced significantly on the increase in the concentration of DBT and indole in model

9 oil (Fig. S4(b)). However, temperature and adsorbent dose had a synergistic effect on both DBT

10 and indole removal. Higher amounts of dose provide more vacant active sites for the adsorption of

11 DBT and indole, while higher temperature gives better removal as adsorption of DBT and indole

12 on Mo-AAC is an endothermic process.

13 3.4. Effect of interactions between factors

14 To explore the interaction between the process parameters and response variables, 3D surface and

15 contour plots were constructed. These plots provide the benefit of giving summarized information

16 in one single figure. The plots were constructed by holding two processing variables at central

17 levels and varying the other two variables. For the DBT removal, only one interaction, i.e.,

18 interaction between the concentration of DBT and indole, was significant (Table 2). Fig. 3(a) and

19 (b) show that DBT removal reduced significantly on the increase in the concentration of both DBT

20 and indole. The highest removal of DBT was achieved when the concentration of DBT and indole

21 was lowest in the model oil.

22 Fig. 3(a) and (c) depicted the DBT removal and Indole removal with respect to the initial

23 concentration of heterocyclic sulfur (DBT) and nitrogen (indole) compounds. The DBT and indole

17
 1 removal was significantly declined when an increase in initial concentration. This may be due to

2 an increase in the number of molecules of DBT or indole. The observed trend may not be due to

3 competitive adsorption between DBT molecules and indole molecules. The adsorption of DBT

4 molecules on Mo-AAC surface may be due to acid-base interaction, π complexation, and van der

5 Waals interaction [28]. However, the adsorption of indole on the Mo-AAC surface may be due to

6 hydrogen bond interaction and π complex interaction [29]. The indole adsorption on the Mo-AAC

7 surface differs from the DBT adsorption because the indole molecule is a neutral molecule and the

8 DBT molecule is basic molecules. Fig. 3(b) and (d) show the contour plot of the initial

9 concentration of DBT and indole, it was observed that slightly elliptical structure attributed to the

10 slight interaction between DBT molecule and indole molecules on the removal of DBT removal/

11 indole [30].

12 Fig. 4(a) illustrated the interaction of indole molecule with temperature, it was observed that %

13 removal of indole or DBT (not shown in the figure) increased significantly when an increase in

14 adsorption temperature. However, the removal rate was declined with respect to the initial

15 concentration of indole. The observed trend may be due to an increase in the kinetic energy of

16 molecules when an increase in temperature leads to high pore diffusion and a high adsorption rate

17 at higher temperatures [31]. Fig. 4(b) showed the contour plot of indole concentration vs

18 temperature on indole removal, it observed a slightly elliptical structure attributed to the slight

19 interaction between indole molecules and temperature.

20 Fig. 4(c) shows the interaction of indole molecules with adsorbent dose, it was observed that %

21 removal of indole increased when catalyst dose was increased, maximum DBT removal/indole

22 removal was observed at higher adsorbent dose. As predicted, the % removal of indole decreased

23 with an increase in indole initial concentration. When the increase in the adsorbent dose, the

18
1 available active sites have been enhanced which serves the higher adsorption rate [12]. Fig. 4(d)

2 shows the contour plot of indole concentration vs adsorbent dose on indole removal, it observed

3 an elliptical structure attributed to the strong interaction between indole molecules and adsorbent

4 active site. For indole removal, three interactions (A*B, A*C, and B*D) were found to be

5 important (Table 3).

19
1

(a) Surface plots of DBT removal (b) Contour plots of DBT

(c) Surface plots of indole removal (d) Contour plots of indole

2

3 Fig. 3. Interaction of DBT vs indole on % removal of DBT, % removal of indole

20
 (a) Surface plots of DBT removal (b) Contour plots of DBT

(c) Surface plots of indole removal (d) Contour plots of indole

1
2 Fig. 4. Interaction of indole concentration vs temperature and indole concentration vs dose on %
3 removal of indole
4

5 3.5. Analysis of residual

6 Assessment of residual plots was carried out to study the reliability of the obtained CCD models.

7 Residuals graphs are used to describe the variation within the obtained model with the help of

8 normal distribution. As can be seen from the normal probability plot of both DBT and indole

21
 1 removal (Fig. 5), most data points lie on the normality line, which indicates that the data is

2 normally distributed. From the residual versus fitted values of both responses, it is clear that

3 residuals are arbitrarily scattered around zero, showing that distribution can be considered as

4 normal. The histograms of residual vs frequency for both DBT and indole removal display that the

5 residuals fluctuate randomly and roughly symmetrical around the mean with most of the residual

6 values. The residuals versus observation order graph illustrate that the residuals are scattered

7 randomly and close to the center value. The obtained results indicated that the developed CCD

8 models are reliable to represent the experimental data.

9 (a)
(a)
10

(b)

Fig. 5. Residual plots for (a) DBT and (b) indole removal.

22
 1 3.6. Optimization analysis

2 The optimization of DBT and indole adsorption onto Mo-AAC was conducted by using the

3 response optimizer function available in Minitab software. In the optimization analysis, target

4 values were set at 100%, lower bounds were based on the lowest values of removal obtained for

5 both compounds in CCD design, and upper values were set at 100% as it has to be more than the

6 target value. The optimum adsorption conditions obtained were the adsorbent dosage of 32 g/L,

7 the temperature of 25.5 °C, DBT concentration of 186 mg/L, and indole concentration of 50

8 mg/L in model oil, respectively. The respective predicted and experimental values found at

9 optimum conditions were 98.36% and 97.16% for DBT as well as 99.08% and 98.41% for indole.

10 3.7. Plausible adsorption mechanism

11 The following mechanisms can govern the adsorption of indole onto Mo-AAC: π complexation,

12 hydrogen bonding, and other weak interactions van der Waals interactions. Acid-base interaction,

13 van der Waals interaction, and π complexation might be responsible for the adsorption of DBT.

14 Surface area and porosity play a significant role in the adsorption of indole and DBT via van der

15 Waals interactions [29]. Acid-base interaction is not applicable in the adsorption of indole as it is

16 a neutral compound. The indole could interact with oxygen-containing compounds available on

17 the surface of adsorbent through hydrogen bonding as indole could behave as a hydrogen bond

18 donor through the hydrogen atom attached with the nitrogen atom [28]. Further, the better

19 adsorption capacity of Mo-AAC can be ascribed to the formation of π complexes between

20 molybdenum species and indole. The adsorption of DBT on Mo-AAC is probably an outcome of

21 specific interactions between the acidic centers generated by molybdenum species and the basic

22 structure of DBT molecules [32]. The enhancement in adsorption capacity due to the incorporation

23
 1 of molybdenum species into the framework of AC can also be attributed to the formation of π-

2 complex between molybdenum species and DBT molecule.

3 3.8. Implications

4 Industrially, the HDS process widely used technique to remove heterocyclic sulfur and nitrogen

5 compounds from petroleum products. However, to meet new stringent norms, i.e., BS-VI, the

6 sulfur in transport fuel (gasoline and diesel) needs to be reduced below 10 ppm. To achieve ultra-

7 low sulfur (below 10 ppm), the conventional HDS process demands three times more catalyst

8 volume, 30 to 60% higher hydrogen gas, 10-30 MPa higher hydrogen pressure when compared

9 with the production of BSIV fuel (500 ppm sulfur-containing fuel) [33]. In addition, the HDS

10 catalyst deactivation rate is much rapid under ultra-deep desulfurization conditions because of the

11 sintering of active metal. Furthermore, the HDS process required high capital and operating cost.

12 On the contrary, the ADS/AND process is an environmentally friendly and energy-efficient

13 process [1,13]. It operates at moderate temperature (below 50 ℃) and atmospheric pressure without

14 any hydrogen requirement. However, some challenges still have to be addressed before the

15 implementation of this process industrially. The main hurdle in this process is the

16 selection/development of efficient adsorbents that can be synthesized easily, have high porosity,

17 have good regeneration capacity, and are environmentally friendly process. Various adsorbents

18 were explored for adsorptive ADN or ADS. However, granular activated carbon received

19 significant attention and it had been applied numerous adsorptive applications to separate/remove

20 various unwanted compounds. The main attention on AC may be due to high surface area and easy

21 tailoring of on the surface of AC made them one of the unique adsorbents for wide applications.

22 Furthermore, numerous authors investigated either ADS or ADN using various adsorbents. Very

23 few authors examined the competitive adsorption process (ADS and ADN) since it is one of the

24
 1 significant studies to understand the hindrance between heterocyclic sulfur, nitrogen, and organic

2 compounds [10,34]. In addition, exploration of innovative easy synthesis and regeneration

3 methods are significantly needed to produce the potential adsorbents. Srivastava [35] suggested

4 integrated treatment technology, i.e., tradition HDS followed by ADS process to achieve ultra-low

5 sulfur and nitrogen in the fuel oil and observed a significant reduction of processes cost.

6 4. Conclusions

7 In this study, Mo-AAC was synthesized, characterized, and evaluated as a possible adsorbent for

8 the removal of DBT and indole in model gasoline using the batch sorption technique. The central

9 composite design of response surface methodology was applied to study the influence of operating

10 variables (Concentration of DBT, indole, adsorbent dose, temperature). The responses were then

11 modeled using ANOVA statistical tables, and results showed a high R2 value of the regression

12 model equation, confirming the accuracy of obtained quadratic model. Compared with the other

13 process parameters, the adsorbent dosage was found to be the most significant factor to influence

14 the simultaneous removal of sulfur and nitrogenous compounds in the model oil. Further,

15 temperature and adsorbent dose have a synergistic effect on the removal of both DBT and indole,

16 whereas concentrations of DBT and indole have an antagonistic effect. Residual analysis showed

17 that the developed CCD models are reliable to represent the experimental data. The optimum

18 operating parameters to obtain maximum DBT and indole removal were obtained using the

19 desirability function. The respective predicted and experimental values found at optimum

20 conditions were 98.36% and 97.16% for DBT as well as 99.08% and 98.41% for indole. A

21 plausible adsorption mechanism for the removal of DBT and indole is also proposed. It can be

22 concluded from the present study that molybdenum-modified AAC is an efficient adsorbent with

23 a good capability for the simultaneous removal of sulfur and nitrogen compound.

25
 1

2 References

3 [1] S.K. Thaligari, V.C. Srivastava, B. Prasad, Adsorptive desulfurization by zinc-impregnated

4 activated carbon: characterization, kinetics, isotherms, and thermodynamic modeling, Clean

5 Technol. Environ. Policy. 18 (2016) 1021–1030. https://doi.org/10.1007/s10098-015-1090-

6 y.

7 [2] J.M. Campos-Martin, M.C. Capel-Sanchez, P. Perez-Presas, J.L.G. Fierro, Oxidative

8 processes of desulfurization of liquid fuels, J. Chem. Technol. Biotechnol. 85 (2010) 879–

9 890. https://doi.org/10.1002/jctb.2371.

10 [3] M.A. Safa, X. Ma, Oxidation kinetics of dibenzothiophenes using cumene hydroperoxide

11 as an oxidant over MoO3/Al2O3 catalyst, Fuel. 171 (2016) 238–246.

12 https://doi.org/10.1016/J.FUEL.2015.12.050.

13 [4] P. Sikarwar, V. Gosu, V. Subbaramaiah, An overview of conventional and alternative

14 technologies for the production of ultra-low-sulfur fuels, Rev. Chem. Eng. 35 (2019) 669–

15 705. https://doi.org/10.1515/revce-2017-0082.

16 [5] M. Mohammadian, M. Ahmadi, & Mohammad, R. Khosravi-Nikou, M.R. Khosravi-Nikou,

17 Adsorptive desulfurization and denitrogenation of model fuel by mesoporous adsorbents

18 (MSU-S and CoO-MSU-S), Pet. Sci. Technol. 35 (2017) 608–614.

19 https://doi.org/10.1080/10916466.2016.1265562.

20 [6] Y.N. Prajapati, N. Verma, Fixed bed adsorptive desulfurization of thiophene over Cu/Ni-

21 dispersed carbon nanofiber, Fuel. 216 (2018) 381–389.

22 https://doi.org/10.1016/J.FUEL.2017.11.132.

23 [7] S.K. Thaligari, V.C. Srivastava, B. Prasad, Simultaneous adsorptive desulfurization and
26
 1 denitrogenation by zinc loaded activated carbon: Optimization of parameters, Pet. Sci.

2 Technol. 33 (2015) 1667–1675. https://doi.org/10.1080/10916466.2015.1089283.

3 [8] J.H. Kim, X. Ma, A. Zhou, C. Song, Ultra-deep desulfurization and denitrogenation of

4 diesel fuel by selective adsorption over three different adsorbents: A study on adsorptive

5 selectivity and mechanism, Catal. Today. 111 (2006) 74–83.

6 https://doi.org/10.1016/J.CATTOD.2005.10.017.

7 [9] I. Ahmed, N.A. Khan, Z. Hasan, S.H. Jhung, Adsorptive denitrogenation of model fuels

8 with porous metal-organic framework (MOF) MIL-101 impregnated with phosphotungstic

9 acid: Effect of acid site inclusion, J. Hazard. Mater. 250–251 (2013) 37–44.

10 https://doi.org/10.1016/J.JHAZMAT.2013.01.024.

11 [10] I. Ahmed, S.H. Jhung, Adsorptive desulfurization and denitrogenation using metal-organic

12 frameworks, J. Hazard. Mater. 301 (2016) 259–276.

13 https://doi.org/10.1016/J.JHAZMAT.2015.08.045.

14 [11] D. Liu, J. Gui, Z. Sun, Adsorption structures of heterocyclic nitrogen compounds over

15 Cu(I)Y zeolite: A first principle study on mechanism of the denitrogenation and the effect

16 of nitrogen compounds on adsorptive desulfurization, J. Mol. Catal. A Chem. 291 (2008)

17 17–21. https://doi.org/10.1016/J.MOLCATA.2008.05.014.

18 [12] Priyanka, V. Subbaramaiah, V.C. Srivastava, I.D. Mall, Catalytic oxidation of nitrobenzene

19 by copper loaded activated carbon, Sep. Purif. Technol. 125 (2014).

20 https://doi.org/10.1016/j.seppur.2014.01.045.

21 [13] P. Sikarwar, U. K. Arun Kumar, V. Gosu, V. Subbaramaiah, Synergetic effect of cobalt-

22 incorporated acid-activated GAC for adsorptive desulfurization of DBT under mild

27
 1 conditions, J. Chem. Eng. Data. 63 (2018) 2975–2985.

2 https://doi.org/10.1021/acs.jced.8b00249.

3 [14] S.S. Shah, I. Ahmad, W. Ahmad, Adsorptive desulphurization study of liquid fuels using

4 Tin (Sn) impregnated activated charcoal, J. Hazard. Mater. 304 (2016) 205–213.

5 https://doi.org/10.1016/J.JHAZMAT.2015.10.046.

6 [15] Z.N. Garba, A.A. Rahim, B.Z. Bello, Optimization of preparation conditions for activated

7 carbon from Brachystegia eurycoma seed hulls: A new precursor using central composite

8 design, J. Environ. Chem. Eng. 3 (2015) 2892–2899.

9 https://doi.org/10.1016/J.JECE.2015.10.017.

10 [16] S. Dashamiri, M. Ghaedi, A. Asfaram, F. Zare, S. Wang, Multi-response optimization of

11 ultrasound assisted competitive adsorption of dyes onto Cu (OH)2-nanoparticle loaded

12 activated carbon: Central composite design, Ultrason. Sonochem. 34 (2017) 343–353.

13 https://doi.org/10.1016/J.ULTSONCH.2016.06.007.

14 [17] J.N. Brown, R.C. Brown, Process optimization of an auger pyrolyzer with heat carrier using

15 response surface methodology, Bioresour. Technol. 103 (2012) 405–414.

16 https://doi.org/10.1016/J.BIORTECH.2011.09.117.

17 [18] M. Sarkar, P. Majumdar, Application of response surface methodology for optimization of

18 heavy metal biosorption using surfactant modified chitosan bead, Chem. Eng. J. 175 (2011)

19 376–387. https://doi.org/10.1016/J.CEJ.2011.09.125.

20 [19] L. Qiu, Y. Cheng, C. Yang, G. Zeng, Z. Long, S. Wei, K. Zhao, L. Luo, Oxidative

21 desulfurization of dibenzothiophene using a catalyst of molybdenum supported on modified

22 medicinal stone, RSC Adv. 6 (2016) 17036–17045. https://doi.org/10.1039/c5ra23077b.

28
 1 [20] D. Qu, X. Feng, N. Li, X. Ma, C. Shang, X.D. Chen, Adsorption of heterocyclic sulfur and

2 nitrogen compounds in liquid hydrocarbons on activated carbons modified by oxidation:

3 Capacity, selectivity and mechanism, RSC Adv. 6 (2016) 41982–41990.

4 https://doi.org/10.1039/c6ra06108g.

5 [21] M. Shetty, K. Murugappan, T. Prasomsri, W.H. Green, Y. Román-Leshkov, Reactivity and

6 stability investigation of supported molybdenum oxide catalysts for the

7 hydrodeoxygenation (HDO) of m-cresol, J. Catal. 331 (2015) 86–97.

8 https://doi.org/10.1016/J.JCAT.2015.07.034.

9 [22] T.A. Saleh, K.O. Sulaiman, S.A. AL-Hammadi, H. Dafalla, G.I. Danmaliki, Adsorptive

10 desulfurization of thiophene, benzothiophene and dibenzothiophene over activated carbon

11 manganese oxide nanocomposite: with column system evaluation, J. Clean. Prod. 154

12 (2017) 401–412. https://doi.org/10.1016/J.JCLEPRO.2017.03.169.

13 [23] G.I. Danmaliki, T.A. Saleh, A.A. Shamsuddeen, Response surface methodology

14 optimization of adsorptive desulfurization on nickel/activated carbon, Chem. Eng. J. 313

15 (2017) 993–1003. https://doi.org/10.1016/J.CEJ.2016.10.141.

16 [24] L. Xiong, F.X. Chen, X.M. Yan, P. Mei, The adsorption of dibenzothiophene using

17 activated carbon loaded with cerium, J. Porous Mater. 19 (2012) 713–719.

18 https://doi.org/10.1007/s10934-011-9523-4.

19 [25] W.N. Wan Abdullah, W.A. Wan Abu Bakar, R. Ali, Z. Embong, Oxidative desulfurization

20 of commercial diesel catalyzed by tert-butyl hydroperoxide polymolybdate on alumina:

21 Optimization by Box-Behnken design, Clean Technol. Environ. Policy. 17 (2015) 433–441.

22 https://doi.org/10.1007/s10098-014-0797-5.

29
 1 [26] F. Vafaei, R. Torkaman, M.A. Moosavian, P. Zaheri, Optimization of extraction conditions

2 using central composite design for the removal of Co(II) from chloride solution by

3 supported liquid membrane, Chem. Eng. Res. Des. 133 (2018) 126–136.

4 https://doi.org/10.1016/J.CHERD.2018.03.010.

5 [27] M. Ghanbari, A.M. Hadian, A.A. Nourbakhsh, K.J.D. MacKenzie, Modeling and

6 optimization of compressive strength and bulk density of metakaolin-based geopolymer

7 using central composite design: A numerical and experimental study, Ceram. Int. 43 (2017)

8 324–335. https://doi.org/10.1016/J.CERAMINT.2016.09.159.

9 [28] M. Sarker, J.Y. Song, A.R. Jeong, K.S. Min, S.H. Jhung, Adsorptive removal of indole and

10 quinoline from model fuel using adenine-grafted metal-organic frameworks, J. Hazard.

11 Mater. 344 (2018) 593–601. https://doi.org/10.1016/J.JHAZMAT.2017.10.041.

12 [29] B. Van De Voorde, M. Boulhout, F. Vermoortele, P. Horcajada, D. Cunha, J.S. Lee, J.S.

13 Chang, E. Gibson, M. Daturi, J.C. Lavalley, A. Vimont, I. Beurroies, D.E. De Vos, N/S-

14 heterocyclic contaminant removal from fuels by the mesoporous metal-organic framework

15 MIL-100: The role of the metal ion, J. Am. Chem. Soc. 135 (2013) 9849–9856.

16 https://doi.org/10.1021/ja403571z.

17 [30] V. Gosu, S. Arora, V. Subbaramaiah, Simultaneous degradation of nitrogenous heterocyclic

18 compounds by catalytic wet-peroxidation process using box-behnken design, Environ. Eng.

19 Res. 25 (2020) 488–497. https://doi.org/10.4491/eer.2019.159.

20 [31] V. Subbaramaiah, V.C. Srivastava, I.D. Mall, Catalytic Activity of Cu/SBA-15 for

21 Peroxidation of Pyridine Bearing Wastewater at Atmospheric Condition, AIChE J. 59

22 (2013). https://doi.org/10.1002/aic.14017.

30
 1 [32] A.L. Santos, R.A. Reis, V. Rossa, M.M. Reis, A.L.H. Costa, C.O. Veloso, C.A. Henriques,

2 F.M.Z. Zotin, M.L.L. Paredes, E.B. Silveira, S.S.X. Chiaro, Silica–alumina impregnated

3 with cerium, nickel, and molybdenum oxides for adsorption of sulfur and nitrogen

4 compounds from diesel, Mater. Lett. 83 (2012) 158–160.

5 https://doi.org/10.1016/J.MATLET.2012.06.011.

6 [33] B. Saha, S. Vedachalam, A.K. Dalai, Review on recent advances in adsorptive

7 desulfurization, Fuel Process. Technol. 214 (2021) 106685.

8 https://doi.org/10.1016/J.FUPROC.2020.106685.

9 [34] S. Kumar Thaligari, S. Gupta, V. Chandra Srivastava, B. Prasad, Simultaneous

10 desulfurization and denitrogenation of liquid fuel by nickel-modified granular activated

11 carbon, Energy and Fuels. 30 (2016) 6161–6168.

12 https://doi.org/10.1021/acs.energyfuels.6b00579.

13 [35] V. Chandra Srivastava, An evaluation of desulfurization technologies for sulfur removal

14 from liquid fuels, RSC Adv. 2 (2012) 759–783. https://doi.org/10.1039/c1ra00309g.

15

31
Conflict of Interest

Declaration of Interest Statement

Manuscript Title: Adsorptive desulfurization and denitrogenation of model liquid fuel

using Mo-AAC: Central composite design

The authors whose names are listed immediately below certify that they have NO affiliations

with or involvement in any organization or entity with any financial interest (such as honoraria;

educational grants; participation in speakers’ bureaus; membership, employment, consultancies,

stock ownership, or other equity interest; and expert testimony or patent-licensing arrangements),

or non-financial interest (such as personal or professional relationships, affiliations, knowledge

or beliefs) in the subject matter or materials discussed in this manuscript.

Author names: Prerana Sikarwar1, Vijayalakshmi Gosu1,*, Verraboina Subbaramaiah1,*, Tian C.

Zhang2 and Rao Y. Surampalli3
1
Department of Chemical Engineering, Malaviya National Institute of Technology Jaipur, 302017,
India. 2Department of Civil & Environmental Engineering, University of Nebraska-Lincoln,
Omaha, NE 68182, USA. 3President, CEO and Chief Technology Officer, Global Institute for
Energy, Environment, and Sustainaility (GIEES), P.O. Box 14354, Lenexa, Kansas 66285, USA.

Corresponding author E-mail: vlakshmi.chem@mnit.ac.in (VG), vsr.chem@mnit.ac.in (VS).

Yours Truly

Dr. V. Subbaramaiah
Assistant Professor
Department of Chemical Engineering,
MNIT Jaipur,
Jaipur-302017 India
e-Component

Click here to access/download

e-Component
Supporting Information.docx