3.1 Characterisation results
3.1.1 Proximate study
Proximate analysis of the precursors as presented in Table 1 shows high moisture values indicating the coagulants’ ability to absorb water, as well as, dissolves colour particles suspended in water
(Obiora-Okafo and Onukwuli 2018)
. The reasonable amount of crude protein contents as recorded indicates the presence of active coagulation components. The values obtained agree with the literature that the protein contents of BE and VS are cationic poly-peptides
(Igwegbe et al. 2021c; Ikegwu et al. 2009)
. Fibre contents present is believed that the precursors were biological polymers having some visible fibrous structures when dispersed in an aqueous medium
(Bolto and Gregory 2007; Onukwuli et al. 2019; Yin 2010)
. The proximate results validate the use of the seed extracts as potential coagulants.
Table 1 Proximate characteristics of the proposed coagulants
S/No.
|
Parameters
|
Values
|
|
|
Brachystegia eurycoma (Black timber)
|
Vigna subterranean (Bambara nut)
|
1.
|
Yield
|
28.31
|
14.6
|
2
|
Bulk density (gm-1 L-1)
|
0.235
|
0.241
|
3.
|
Moisture content (%)
|
7.25
|
10.0
|
4.
|
Ash content (%)
|
3.48
|
2.97
|
5.
|
Protein content (%)
|
19.77
|
18.15
|
6.
|
Fibre content (%)
|
2.20
|
1.64
|
7.
|
Carbohydrate (%)
|
56.76
|
60.94
|
8.
|
Fat content (%)
|
10.53
|
6.30
|
3.1.2 FTIR analysis of the coagulants
The spectra representation of BEC and VSC are shown in Figs. 3a-b, respectively. In Figs. 3a there is a slight absorption peak of 3965.52 - 3780.36 cm-1 attributing to the stretching vibration of –OH, together with vibration of water absorbed
(Igwegbe et al. 2021c)
. Also, the –OH groups with a peak at 3070.58 cm-1 were also evidenced in Fig. 3b. The free hydroxyl groups, confirm the occurrence of carboxylic acids, phenols, and alcohols in the coagulants. This band also links to the O-H vibrations of cellulose, pectin, and lignin. Consequently, there is an agreement between the results of Table 1 and the spectral results indicating the presence of moisture, oil, and carbohydrate. Furthermore, the analysis revealed that the absorption peak for the amines was evidenced in 3348.32 cm-1 for aliphatic primary amine (N-H) and secondary amine of 3070.58 cm-1 for BEC and VSC respectively. Also, the presence of stretching signals, N-H detects the existence of amino compounds, confirming the presence of protein in the powders as demonstrated in Table 1. In addition, a major group in the wider region of 2021.34 cm-1 and 2052.20 cm-1 specifies the existence of a C=O group (carbonyl compound). There was also a strong adsorption peak at 694.36 cm−1 and 632.64 cm−1 for BEC and VSC respectively, showing the distinguishing occurrence for C-H out of plane deformation which is typically comparative to the position and spatial geometry of the double bond
(Coates 2006)
. Finally, the occurrence of moistures, proteins, and esters is confirmed by the FTIR spectral of BEC and VSC, as well as the proximate analysis provided in Table 1, justifying their usage as good sources of coagulants in this research.
3.1.3 Morphological analysis of the coagulant
SEM technology was used to examine the external morphologies of the coagulants in this investigation, as shown in Fig. 4 at 600x magnifications. The 3D reconstructed SEM images revealed well-developed pores of various sizes and shapes. As a result, pore sizes made up of micro-pores, macro-pores, and mesopores, together with their distributions, are confirmed unique features of NOPs. Therefore, a major pore size of 0.41 μm2 was revealed in the histograms, as well as fibre lengths between 1.66 -21.45 μm and 2.11 -17.94 μm for BEC and VSC respectively as shown in Fig. 5. Varying fibre lengths are unique features of NOPs that enhance their multifunctional utilisation as coagulants and adsorbents (Obiora-Okafo et al. 2018). Rough surfaces disclose that the coagulants are rough fibrous solids primarily made of cellulose and lignin, indicating that they are polymeric. The binding of particles to polymer chains via inter-particle bridging or electrostatic interactions improves sweep flocculation. Adsorption as a crucial mechanism in the procedure is also confirmed by small holes and rough surfaces seen on the coagulant morphologies (Igwegbe et al. 2021c; Obiora-Okafo et al. 2018). Furthermore, the structures also retain compact-net arrangements which are more conducive to particle flocculation owing to bridge aggregation. Finally, when compared to the branching structure, the compact-net structure is better for flocculation and particle-bridge creation among flocs (Zhu et al. 2012).
3.2 Colour concentration/removal efficiency dependent on settling time
The flocculation process involves particle interactions and a time-dependent interface of coagulant hydroxide formation, following the hydrolysis reaction (Liang et al. 2016; Obiora-Okafo et al. 2018). The time-dependent influence of colour concentration and its reduction efficiency is presented in Fig. 6. The percentage reduction in concentration as observed in 1000 mgBEC/L and 800 mgVSC/L results to 73.3% and 88.8% respectively. In addition, the sharp time reduction of 30 min specifies a speedy coagulation process that discloses the probable coagulation time (Tag). Moreover, this rapid reduction in concentration may perhaps be attributed to either charge neutralisation or its combination with sweep flocculation mechanism (Beltrán-Heredia et al. 2011a). As a result, after 30 minutes, the amount of particles accessible for flocculation diminishes, showing a gradual drop in colour concentration as the process progresses. This is most likely due to an intricate coagulation-flocculation procedure that may include the development of a net-like structure that does not take a long period. Therefore, the greater flocculation period could be related to the presence of a sorption mechanism that necessitates a longer process time. After 300 min, there was no noticeable change in concentration, indicating that equilibrium has been reached. Consequently, due to the saturation of the active adsorption sites, the aggregate becomes destabilized, preventing further adsorption and, as a result, the settling period is prolonged (Beltrán-Heredia et al. 2011a; Onukwuli and Obiora-Okafo 2019). For these reasons, coagulation-flocculation using NOPs in wastewaters is more efficient at low pH conditions. Analogous to these results, related studies have also been reported by Zhu et al. (2011) and Trinh and Kang (2011).
3.3 Coagulation-flocculation kinetics representing Brownian motion
Analysis was performed on a 95% confidence level to determine the order of coagulation-flocculation response, and the parameters gotten from the data regression analysis for BEC and VSC are provided in Table 2. The intercept and slope of the equation defining the kinetics of agglomeration were used to calculate the coagulation rate constant, K, and the order of reaction (Eq. S3, see Supplementary file). The coagulation proportionality constant that connects the reaction rate to the concentration of the reacting species is called the coagulation rate constant (Schick and Hubbard 2005). This denotes that each minute, 0.000638 mg L-1 and 0.00403 mg L-1 of colour particles were consistently attached to the polymer surfaces creating larger aggregates for BEC and VSC, respectively. From the calculation, the reaction order obtain was in agreement with the conventional principle of coagulation-flocculation being a second-order process (Menkiti et al. 2011; Schick and Hubbard 2005). Hence, the reaction order gotten confirms the optimum order for the process, showing a second-order reaction. Also, the correlation coefficient (R2) demonstrates good agreement that implies that the studied kinetic data is significant. Tag is inversely proportional to the starting concentration of colour particles, proposing that the higher the contaminant concentration, the shorter the coagulation time required for elimination (Obiora-Okafo et al. 2019). Furthermore, the collision efficiency (E) values explain the attainability assumption that particle collision between contaminants and coagulants is 100% efficient throughout the dispersion, implying that particles will stick together after bimolecular collision and that particle distribution or complex formation distribution will occur during the process (Obiora-Okafo et al. 2019).
Table 2 Coagulation Kinetics Parameters from Brownian Theory
Parameters
|
1000 mg BEC/L
|
800 mg VSC/L
|
K (L/ mgmin)
|
6.38 E-04
|
4.03 E-03
|
α
|
1.8
|
1.9
|
R2
|
0.981
|
0.969
|
Rate Equation (-r)
|
6.38 × 10-4 C2
|
4.03 × 10-3 C2
|
Tag (min)
|
31.35
|
26.96
|
K1 (L/ min)
|
3.19 E-04
|
2.02 E-03
|
β (L/ mgmin)
|
0.000638
|
0.00403
|
E (mg-1)
|
1.00
|
1.00
|
3.4 The influence of time on particle behaviour
Particles reduction behaviour as a function of time depicts the pattern at which colour concentrations are reduced. Figure 7 depicts the fluctuations in CT, C1, C2, C3, and C4 for initially monodispersed particles obtained using Eqs. S26 - S29 in the Supplementary file, section S1. With increasing time, both the total colour concentration, Ct, and the concentration of the singlet species, C1, drop monotonically. The concentrations C2 (t), C3 (t), and C4 (t) go through a maximum since they are not present at the initial time and concentration. Due to an increase in the number of particle concentrations to the aggregate formation over time, the number of singlets appears to be decreasing faster than the overall number of particles (Igwegbe et al. 2021a; Taitelbaum and Koza 2000). The resultant effect of the bimolecular reaction results in a drop in the total number of particles. Furthermore, we discovered that the lower the K value, the longer the coagulation time, giving rise to a slow rate and longer coagulation-flocculation process (Menkiti et al. 2009).
3.5 Adsorption models
Some attractions exist between polymer segments and particle surfaces during the flocculation process, which leads to adsorption
(Bolto and Gregory 2007)
. Consequently, some kinetic models such as pseudo-first-order, pseudo-second-order, and Elovich kinetic models (see Supplementary file, section S3) were involved to examine the rate at which particles are adsorbed onto polymer surfaces, as presented in Fig. 8. Thus, the kinetic factors obtained were summarised in Table 3. Consequently, the R2 for the models was quite low when compared to the pseudo-second-order model. Furthermore, the experimental data agree well with the pseudo-second-order kinetic model data, with BEC and VSC having the lowest normalised standard deviation, Δq (%) values of 2.1 % and 1.07 %, respectively evaluated using Eq. S54 (see Supplementary file). Additionally, the coagulation-adsorption process is confirmed as a second-order process owing to an excellent fit of the second-order kinetic with an R2 of 0.999. More importantly, the Elovich model's moderate agreement expanded our knowledge of the adsorption-chemisorption procedure, suggesting selective adsorption without site rivalry, as shown in organic polymers
(Feng et al. 2021; Lanan et al. 2021)
, leading to the position of the Langmuir model in the sorption process
(Obiora-Okafo et al. 2018)
. Thus, chemisorption, which involves valence forces through electron sharing between polymers and pollutants, was found to affect the general rate of the adsorption process
(Ghernaout et al. 2015; Igwegbe et al. 2021b)
.
Table 3 Adsorption factors for colour removal.
Pseudo-first-order kinetics
|
|
qe, exp (mg/g)
|
qe, cal (mg/g)
|
KF1 (min-1)
|
R2
|
Δq (%)
|
1000 mgBEC/L
|
7.3
|
1.95
|
0.01
|
0.911
|
25.911
|
800 mgVSC/L
|
11.1
|
4.899
|
0.009
|
0.851
|
19.75
|
Pseudo-second-order kinetics
|
|
qe, cal (mg/g)
|
K2 (g/mg min)
|
R2
|
h (mg/g min)
|
Δq (%)
|
1000 mgBEC/L
|
7.52
|
0.0836
|
0.997
|
4.73
|
2.1
|
800 mgVSC/L
|
11.76
|
0.0395
|
0.999
|
5.46
|
1.07
|
Elovich kinetics
|
|
a
|
b
|
R2
|
|
|
1000 mgBEC/L
|
3362.83
|
2.028
|
0.925
|
|
|
800 mgVSC/L
|
14.77
|
0.759
|
0.933
|
|
|
3.6 The expectation of particles transfer rate
The mass transfer rate was verified using particle concentration measurements that showed the investigational and projected transfer rates all through the coagulation-flocculation process, as shown in Fig. 9. In consequence, the projected results demonstrate that the rate of concentration reduction, resulting in the rate of mass transfer being rapid at the start of the process, resulting in a tight agreement between the actual and expected results. Due to this, the anticipated equilibrium point is closer to the experimental equilibrium (Oke et al. 2021).
Table 4 displays the results of statistical data comparing the investigational and projected data. The results indicate that the lower the percentage, the better the prediction. The value of M% lesser than 10 specifies a good prediction of investigational data. Also, the correlation coefficient of the predicted results gave positive correlation values of 0.816 and 0.950 for BEC and VSC respectively. Furthermore, the χ2 values greater than 0.05 are more significant than those less than 0.05. During the coagulation-flocculation, the projected contaminant particle decline pattern is likewise similar to Oke et al. (2021)'s earlier study.
Table 4 Modelling verification result
Coagulants
|
M%
|
R2
|
χ2
|
F-test
|
T-test
|
BEC
|
3.263
|
0.816
|
30.30
|
0.711
|
0.0270
|
VSC
|
2.536
|
0.950
|
23.98
|
0.0316
|
0.0316
|