Application Of Response Surface Methodology And Artificial Neural Network In Modelling The Adsorption Of Chromium (VI) And Lead (II) Using Corn Cob As Biosorbent
ABSTRACT
This study aimed
at optimizing the adsorption of chromium and
lead with corn cob using Response Surface Methodology (RSM) and Artificial
Neural Network (ANN). The treated and untreated corn cobs were
characterized using Fourier Transform Infra-Red
(FTIR), Scanning Electronic Microscope (SEM) and X-Ray Diffraction (XRD). The experiment was designed with Box-Behnken experimental design of RSM
in Design Expert version 10, with four factors which generated 29 experimental runs in order to optimize the removal of lead (II) and
chromium (VI) ions from waste water. The effect of the process parameters; pH
(2 to 10), contact time (10 to 70 minutes), adsorbent dosage (0.1 to 1.0g) and
initial concentration (20 to 100 mg/L) on the removal efficiency of
the chromium and lead were evaluated using RSM. The 29 experimental data was divided into
subsets, training (70%), validation (15%) and test (15%) for the Artificial
Neural Network modelling in MATLAB 2019 environment and was quantified using
Mean Square Error (MSE) and correlation coefficient (R2). The
experimental result showed maximum lead removal efficiency
of 0.415 at concentration of 20 mg/L, pH of 6, dosage of 0.55 g at 10 minutes
and 0.935 at concentration of 100 mg/L, pH of 6, dosage of 1 g at 10 minutes
were obtained for the treated and untreated corn cob respectively. The maximum chromium removal efficiency of
0.695 at concentration of 60 mg/L, pH of 6, dosage of 0.55 g at 40 minutes and
0.875 at concentration of 100 mg/L, pH of 6, dosage of 1 g at 40 minutes were
obtained for the untreated and treated corn cob respectively. The
multiple regression for the removal of lead (Pb) using untreated and treated
corn cob resulted to 2FI and quadratic model with R-squared values of
0.0.734702 and 0.72372 respectively, while the removal of chromium using
untreated and treated corn cob gave quadratic models with R-squared values of
0.840365 and 0.812564 respectively. The
Cr (VI) and Pb (II) linearity and high values of correlation coefficients (R2)
of 0.9944 and 0.9901 for Langmuir model, 0.9604 and 0.9753 for Freundlich
isotherm and 0.8593 and 0.8034 for Temkin isotherm respectively. The Levenberg–Marquardt algorithm ANN
models for predicting the removal efficiency of lead and chromium using the
treated corn cob generated R2 value of 0. 999999 and 0.999999, compared to
the RSM model with R2 value of 0.734702 and 0.840365 while the
untreated corn cob has ANN R2 value of 0.999999 and 0.999999, compared to
the RSM model with R2 value of 0.72372 and 0.812564. The comparison
of the ANN and RSM models showed ANN to be a better predictor than RSM.
TABLE OF CONTENTS
Cover
page
Title
page i
Declaration
ii
Certification
iii
Dedication
iv
Acknowledgements
v
Table
of contents vi
List
of tables viii
List
of figures ix
Abstract xi
CHAPTER
1
INTRODUCTION
1.1 Background of the study 1
1.2 Statement of problem 5
1.3 Objectives of the study 6
1.4 Justification of the study 6
1.5 Scope of the study and its limitations 7
CHAPTER
2
LITERATURE
REVIEW
2.1 Heavy metals 8
2.1.1 Chromium (VI) 8
2.1.2 Lead 9
2.2 Adsorption 10
2.2.1 Mechanism of Adsorption 12
2.2.2 Types of Adsorption 13
2.2.2.1 Physical adsorption 14
2.2.2.2 Chemical adsorption 14
2.2.3 Factors affecting adsorption 15
2.2.3.1 Nature of adsorbent 15
2.2.3.2 Nature of solute (adsorbate) 16
2.2.3.3 Nature of solvent 17
2.2.3.4 Influence of temperature 17
2.2.3.5 Influence of pH 18
2.2.3.6 Contact time or residence time 18
2.3 Agricultural adsorbents 19
2.3.1 Corn Cob 21
2.4 Response Surface Methodology 23
2.5 Artificial Neural Network 27
2.5.1 Learning Process 28
2.5.2 Generalization 29
2.5.3 Selecting the number of
Hidden Layers 30
2.5.4 Pre-Process and
Post-Process of the Training Patterns 31
2.6 Gap in the literature 33
CHAPTER THREE
MATERIALS AND
METHOD
3.1 Materials/equipment 34
3.2 Preparation of adsorbent 35
3.2.1 Carbonization 35
3.2.2 Impregnation 35
3.3
Preparation of chromium stock
solution/working solution 36
3.4 Batch
adsorption experiment 36
3.4.1 Effect of pH 38
3.4.2 Effect of initial concentration 38
3.4.3 Effect of adsorbent dosage 39
3.5 Design
of experiment 39
3.6 Adsorption isotherm model 40
3.7 Adsorption kinetics 42
3.8 ANN
model 43
CHAPTER
4
RESULT
AND DISCUSSION
4.1 Characterization of the adsorbent 45
4.1.1 Fourier Transform Infra-Red 45
4.1.2 X-ray Diffraction 48
4.1.3 Scanning Electron Microscopy (SEM) 49
4.2 Response surface methodology model for removal efficiency 50
4.2.1 ANOVA for % Lead Removal Using Untreated Corn adsorbent 52
4.2.1.1 Effect of process parameters on the removal of
lead using untreated corncob 53
4.2.2 ANOVA for% Lead Removal Using Treated Corn Cob 56
4.2.2.1 Effect of process parameters on the removal of
lead using treated corncob 57
4.2.3 ANOVA for % Chromium Removal Using Untreated Corn Cob 60
4.2.3.1 Effect of process parameters on the %removal of
chromium using untreated corncob61
4.2.4 ANOVA for % Chromium Removal Using Treated Corn Cob 64
4.2.4.1 Effect of process parameters on the %removal of
chromium using treated corncob65
4.3 Artificial Neural Network (ANN) model 68
4.4 Comparison of Response Surface Methodology and Artificial
Neural Network 72
4.5 Batch adsorption result 73
4.5.1 Effect
of contact time on removal efficiency 73
4.5.2 Effect
of temperature on removal efficiency 74
4.5.3 Effect
of initial concentration on removal efficiency 75
4.5.4 Effect
of pH on removal efficiency 76
4.5.5 Effect
of adsorbent dosage on removal efficiency 77
4.6 Equilibrium adsorption isotherms 78
4.7 Adsorption Kinetics 81
CHAPTER
FIVE
CONCLUSION
AND RECOMMENDATION
5.1 Conclusion 84
5.2 Recommendations 84
5.3 Contribution to knowledge 85
REFERENCES
LIST OF TABLES
Table
Title Page
3.1: List of equipment 34
3.2: Experimental ranges of
the factors for Box Behnken Design 39
3.3 Adsorption Isotherm
models 41
4.1: FTIR result for the
untreated and treated corn cob 46
4.2: Experimental design for removal efficiency 51
4.3: Analysis of Variance for the removal of
lead using untreated corn cob 53
4.4: Analysis of Variance for the removal of
lead using treated corn cob 57
4.5: Analysis of variance for the removal of
chromium using untreated corn cob 61
4.6: Analysis of Variance for the removal of
Chromium using treated corn cob 65
4.7:
ANN comparison of 11 backpropagation
(BP) algorithms for removal efficiency
using untreated
and treated corn cob 69
4.8: Comparison of the correlation coefficients
of RSM and ANN models 72
LIST OF FIGURES
Figure Title Page
2.1: Adsorption mechanism 12
2.2: Pathways of adsorption process 13
2.3:
Training platform of ANN 28
2.4: Result of ANN training platform 30
2.5:
ANN platform showing hidden neurons 31
3.1: Scheme of adsorption of pollutants by the
batch process 38
3.2: Artificial Neural Network for the
adsorption process 44
4.1: FTIR result for untreated corn cob 47
4.2: FTIR of treated corn cob 47
4.3: XRD for untreated corn cob 48
4.4: XRD for treated corn cob 49
4.5:a) SEM for untreated corn cob b) SEM for Treated
corn cob 50
4.6: Effect of
concentration and pH on % removal of lead using untreated corncob 54
4.7: Effect of
concentration and dosage on % removal of lead using untreated corncob 54
4.8: Effect of
concentration and time on % removal of lead using untreated corncob 54
4.9: Effect of dosage
and pH on % removal of lead using untreated corncob 55
4.10: Effect of pH and
time on % removal of lead using untreated corncob 55
4.11: Effect of dosage and
time on % removal of lead using untreated corncob 55
4.12: Effect of
concentration and pH on % removal of lead using treated corncob 58
4.13: Effect of
concentration and dosage on % removal of lead using treated corncob 58
4.14: Effect of
concentration and time on % removal of lead using treated corncob 58
4.15: Effect of dosage and
pH on % removal of lead using treated corncob 59
4.16: Effect of pH and
time on % removal of lead using treated corncob 59
4.17: Effect of dosage and
time on % removal of lead using treated corncob 59
4.18: Effect of
concentration and pH on %removal of chromium using untreated corncob62
4.19: Effect of concentration and dosage on
%removal of chromium using untreated
corncob 62
4.20: Effect of
concentration and time on %removal of chromium using untreated corncob62
4.21: Effect of pH and
dosage on %removal of chromium using untreated corncob 63
4.22: Effect of pH and
time on %removal of chromium using untreated corncob 63
4.23: Effect of time and
dosage on %removal of chromium using untreated corncob 63
4.24: Effect of
concentration and pH on %removal of chromium using treated corncob 66
4.25: Effect of
concentration and dosage on %removal of chromium using treated corncob66
4.26: Effect of
concentration and time on %removal of chromium using treated corncob 66
4.27: Effect of pH and
dosage on %removal of chromium using treated corncob 67
4.28: Effect of pH and
time on %removal of chromium using treated corncob 67
4.29: Effect of dosage and
time on %removal of chromium using treated corncob 67
4.30: Training, validation and test mean squared error using Levenberg–Marquardt
backpropagation
for lead removal using (a) untreated (b) treated corn cob 70
4.31: Training, validation and test mean squared error using Levenberg–Marquardt
backpropagation for chromium removal
using (a) untreated (b) treated corn cob 70
4.32: ANN regression plots for the removal of Lead using (a) untreated
corn cob
(b) treated
corn cob 71
4.33: ANN regression plots for the removal of Chromium using (a)
untreated corn cob
(b) treated
corn cob 71
4.34: Effect of contact time on removal efficiency 73
4.35: Effect of temperature on removal efficiency 74
4.36: Effect of initial concentration on removal
efficiency 75
4.37: Effect of pH on removal efficiency 76
4.38: Effect of adsorbent dosage 77
4.39: Langmuir adsorption isotherm of Pb (II) 79
4.40: Langmuir adsorption isotherm of Cr (VI) 80
4.41: Freundlich adsorption isotherm 80
4.42: Temkins adsorption isotherm 81
4.43: The plots of pseudo-first-order kinetic model for Cr(VI) and
Pb(II)
adsorption
onto corncob 82
4.44: The plots of pseudo-second-order kinetic model for Cr(VI) and
Pb(II)
adsorption
onto corncob 83
0.0
UKANDU, C (2023). Application Of Response Surface Methodology And Artificial Neural Network In Modelling The Adsorption Of Chromium (VI) And Lead (II) Using Corn Cob As Biosorbent. Repository.mouau.edu.ng: Retrieved Oct 31, 2025, from https://repository.mouau.edu.ng/work/view/application-of-response-surface-methodology-and-artificial-neural-network-in-modelling-the-adsorption-of-chromium-vi-and-lead-ii-using-corn-cob-as-biosorbent-7-2
CHRISTIAN, UKANDU. "Application Of Response Surface Methodology And Artificial Neural Network In Modelling The Adsorption Of Chromium (VI) And Lead (II) Using Corn Cob As Biosorbent" Repository.mouau.edu.ng. Repository.mouau.edu.ng, 20 Jul. 2023, https://repository.mouau.edu.ng/work/view/application-of-response-surface-methodology-and-artificial-neural-network-in-modelling-the-adsorption-of-chromium-vi-and-lead-ii-using-corn-cob-as-biosorbent-7-2. Accessed 31 Oct. 2025.
CHRISTIAN, UKANDU. "Application Of Response Surface Methodology And Artificial Neural Network In Modelling The Adsorption Of Chromium (VI) And Lead (II) Using Corn Cob As Biosorbent". Repository.mouau.edu.ng, Repository.mouau.edu.ng, 20 Jul. 2023. Web. 31 Oct. 2025. < https://repository.mouau.edu.ng/work/view/application-of-response-surface-methodology-and-artificial-neural-network-in-modelling-the-adsorption-of-chromium-vi-and-lead-ii-using-corn-cob-as-biosorbent-7-2 >.
CHRISTIAN, UKANDU. "Application Of Response Surface Methodology And Artificial Neural Network In Modelling The Adsorption Of Chromium (VI) And Lead (II) Using Corn Cob As Biosorbent" Repository.mouau.edu.ng (2023). Accessed 31 Oct. 2025. https://repository.mouau.edu.ng/work/view/application-of-response-surface-methodology-and-artificial-neural-network-in-modelling-the-adsorption-of-chromium-vi-and-lead-ii-using-corn-cob-as-biosorbent-7-2