Advertisement

Experimental Design Methodology Applied to the Degradation of a Cytostatic Agent the Imatinib Mesylate Using Fenton Process

  • Fatma Ghrib
  • Taieb Saied
  • Nizar Bellakhal
Original Article
  • 59 Downloads

Abstract

This manuscript investigates the degradation efficiency of a cytostatic agent imatinib mesylate (IM) by Fenton process. The objective of the study is to determine the effect of initial concentrations of hydrogen peroxide, ferrous ion and IM on the degradation process of IM. In this context, the overall research is conducted in two stages. The first stage of experiments involves the evaluation of the experimental parameters impact on the IM removal using the factorial design; whereas the second one includes the optimization conditions for IM degradation using Doehlert design. Different concentrations of H2O2 (10–50 mmol L−1), Fe(II) (0.5–3 mmol L−1) and IM (0.045–1.7 mmol L−1) are adopted both to assess the main effects of factors and their interactions on the anticancer destruction and to establish the optimum process variables that ensure its total removal. The factorial matrix considering different combinations of factors and levels demonstrates that the concentrations of Fe(II) and IM are the most determining parameters on the removal of the target molecule. The Doehlert matrix reveals that the maximum IM abatement is predicted to be 100.59% (± 0.63) when the optimized reaction conditions are set at 30 mmol L−1 for H2O2 and 2.5 mmol L−1 for Fe(II).

Keywords

Cytostatic agent Degradation Doehlert experimental design Factorial matrix Fenton process 

1 Introduction

Cancer is one of the most dangerous diseases around the world. Since its incidence is increasing, the consumption of anticancer drugs has consequently been improved over the last years. These agents have revealed potent cytotoxic, genotoxic, mutagenic, carcinogenic, and/or teratogenic effects in numerous organisms, since they have been designed to disrupt or prevent cellular proliferation, usually by interfering in DNA synthesis [1]. Recent studies have shown that mixtures of anticancer substances in real samples possess a significant toxicological effect comparing with the individual drug [2]. In any case, chemotherapy agents are thus considered as emerging and recalcitrant pollutants, which cause a direct threat to the aquatic life and a potential health risk to humans.

A very interesting field of concern is what to do with these types of wastewater generated especially from pharmaceutical industry that contains structurally complex organic compounds which are either toxic or non-biodegradable. For this reason, conventional treatment methods are often incomplete and inefficient [3, 4] to treat this type of pharmaceutical wastewaters and hence different and more effective treatment options should be explored and evaluated. One of the most promising treatment is the advanced oxidation processes (AOP). Owing to their powerful capability to oxidize several organic compounds into CO2 and H2O, AOPs have been selected for various applications [5, 6]. They are characterized by the production of hydroxyl free radicals (HO·) which have a high oxidizing potential. Among these, Fenton’s oxidation which is a very efficient process used for the elimination of different hazardous organic pollutants from wastewaters as mentioned in a comprehensive review by Neyens and Baeyens [7]. In addition, Fenton process (FP) is based on the highly reactive HO· formed under acidic conditions as a result of the reaction between hydrogen peroxide (H2O2) and ferrous ions (Fe2+) which is called as Fenton’s reagent [6, 8]. This is an attractive oxidative system for water treatment since required reagents are readily available, easy to manage, safe and they do not cause environmental damages [6, 9, 10]. Another advantage of the Fenton reagent is that the activation of H2O2 does not require any energy input. Therefore, this method offers a cost-effective source of hydroxyl radicals. However, the only disadvantage in using this process is the formation of a substantial amount of Fe(OH)3 precipitate which contains large quantities of adsorbed organic compounds. Therefore, works are performed to reduce the Fe2+ content in Fenton’s reagent. The optimization of Fe2+ is necessary as an excess of iron would improve the generation of a lot of ferrous ion, which consume and then scavenge free radicals resulting in reduced degradation efficiency. Also treatment cost will rise by using an excessive quantity of iron salts as the amount of sludge increases introducing enhanced treatment costs [8].

The oxidation is based on Fenton’s reagent, and it takes advantage from the very high reactivity of the free radical formed in acidic solution by the catalytic decomposition of H2O2 [6, 11]. The conventional Fenton Process is primarily described by Haber and Weiss [12]. They have shown that this catalytic decomposition followed a complex radical chain mechanism as described by the Eqs. (17) [6, 12, 13, 14, 15]:

Chain initiation
$${\text{Fe}}^{ 2+ } + {\text{ H}}_{ 2} {\text{O}}_{ 2} \to {\text{Fe}}^{ 3+ } + {\text{ HO}}^{.} + {\text{ OH}}^{ - } .$$
(1)
Chain propagation
$${\text{Organics }} + {\text{ HO}}^{.} \to {\text{H}}_{ 2} {\text{O}} + {\text{ CO}}_{ 2} + {\text{ products,}}$$
(2)
$${\text{HO}}^{.} + {\text{ H}}_{ 2} {\text{O}}_{ 2} \to {\text{H}}_{ 2} {\text{O}} + {\text{ HO}}_{ 2}^{.} ,$$
(3)
$${\text{H}}_{ 2} {\text{O}}_{ 2} + {\text{ Fe}}^{ 3+ } \to {\text{Fe}}^{ 2+ } + {\text{ HO}}_{ 2}^{.} + {\text{ H}}^{ + } ,$$
(4)
$${\text{Fe}}^{ 3+ } + {\text{ HO}}_{ 2}^{.} \to {\text{Fe}}^{ 2+ } + {\text{ O}}_{ 2} + {\text{ H}}^{ + } .$$
(5)
Chain termination
$${\text{Fe}}^{ 2+ } + {\text{ HO}}_{ 2}^{.} \to {\text{Fe}}^{ 3+ } + {\text{ OH}}^{ - } ,$$
(6)
$${\text{HO}}^{.} + {\text{ Fe}}^{ 2+ } \to {\text{OH}}^{ - } + {\text{ Fe}}^{ 3+ } .$$
(7)

Besides, FP depends on various factors: pH, H2O2 concentration, ferrous ions concentration, temperature and reaction time. Generally, for the homogeneous Fenton reaction, the optimum pH should be between 2 and 4 to produce the greatest amount of hydroxyl radicals to oxidize organic compounds [6, 8, 16, 17]. Therefore, pH value should not be too low because at these pHs excessive H+ reacts with H2O2 to produce H3O2+, which is stable and cannot react with Fe(II) to form the HO· species. Additionally, hydroxyl radicals can also be scavenged by excessive H+ [5]. On the other hand, at high pH (> 4), the formation of Fe(II) complexes decreases the production of free radicals and also the regeneration of ferrous ion gets prevented by the precipitation of ferric oxyhydroxides thereby bringing down the pollutant degradation rate. In addition, higher pH enables the oxidation potential of hydroxyl radicals to be reduced which also leads to a decline in the degradation rate of the pollutant [8].

To our knowledge, no study was performed to investigate the degradation of imatinib mesylate (IM) (Fig. 1); also known as STI 571 and which is one of the most frequently administered anticancer drug worldwide; by Fenton process using the experimental design methodology. Therefore, the objective of this paper is to adopt the Fenton process in order to remove this pharmaceutical pollutant of emerging concern. In this present study, the effect of experimental conditions such as ferrous ions catalyst, H2O2 and IM concentrations was investigated while the pH was set at its optimal value 3 based on literature cited before and also on the study achieved by Governo et al. [10]. This latter study had evaluated the Fenton’s Process performance in a real wastewater spiked with a cytostatic agent and the pH was fixed at 3 its optimum value.
Fig. 1

IM chemical structure

Although many researchers have usually only focused on the single-factor-at-a-time approach studying the effect of each experimental parameter on the process performance while keeping all other conditions constant, this type of approach does not take into account cross-effects from the factors considered and leads to a poor optimization result [11]. Hence, two experimental design methodologies (the factorial matrix and the Doehlert experimental design) have been adopted successively to evaluate the effect of factors, their interactions and to determine the optimal operating conditions that ensure the total removal of IM.

2 Experimental

2.1 Chemicals

Imatinib mesylate; which was procured from Dr. Reddy’s Laboratories Limited; is designated chemically as (4-[(4-methyl-1-piperazinyl) methyl]-N-[4-methyl-3-[[4-(3-pyridinyl)-2-pyrimidinyl] amino]-phenyl] benzamide methane sulfonate). It is a protein tyrosine kinase inhibitor that was developed in the 1990s and is used to deal with chronic myeloid leukaemia, gastrointestinal stromal tumours [18] and a number of other malignancies. It kills cancer cells by blocking the action of tyrosine kinases.

Fenton’s reagent was composed of hydrogen peroxide (H2O2, 30% w/w) and heptahydrated ferrous sulfate (FeSO4·7H2O) used as a ferrous ions source. These reagents were supplied by Fluka (Switzerland) and VWR BDH Prolabo (France), respectively. Sulfuric acid (H2SO4) and sodium hydroxide (NaOH) were procured from Sigma Aldrich (Germany). The other chemicals used for chromatographic analysis such as acetonitrile, 1-octane sulphonic acid sodium salt and the ortho phosphoric acid were purchased from Carlo Erba (France), Loba Chemie (India) and VWR Chemicals (France), respectively. High purity water, obtained from a Millipore Milli-Q plus purification system, was used for the preparation of all the solutions.

2.2 Experimental Procedure

2.2.1 Preparation of the Synthetic Solution

Stock solutions were composed of requisite amounts of the cytostatic agent, weighed and dissolved in Milli-Q water reaching concentrations of 0.045 mmol L−1 and 1.7 mmol L−1.

2.2.2 Fenton Degradation Procedure

In accord with the preliminary tests we have conducted and based on previous research revealing that the optimum pH value was about 3, Fenton treatments were carried out by varying H2O2, Fe(II) and IM initial concentrations. Indeed, all these trials were performed, at ambient temperature (20 ± 2 °C), in an open and undivided glass cell containing; in each run; 250 mL of synthetic IM wastewater at a specified initial concentration. The pH of solutions, measured with a Mettler Toledo SE S470-K pH-meter, was first adjusted as we have mentioned to 3 with H2SO4 to prevent the formation of iron oxyhydroxides. Then, in order to initiate the oxidation, appropriate quantity of the Fenton reagent was added to the solution. This mixture was stirred continuously with a magnetic stirrer. Samples were taken out periodically from the cell. Consequently, while rising the pH value to 8 with NaOH, the production of hydroxyl radical is effectively prevented [19], and thus the FP is stopped. Finally, the liquid was immediately filtered and analysed for pollutant concentration.

2.3 Experimental Design

Experimental design of the FP for IM degradation was carried out using successively full factorial design and Doehlert design methodologies. Experimental data and the variance analysis (ANOVA) were performed using NEMRODW software.

Interactions between the process variables were graphically illustrated by plotting the three-dimensional (3D) response surface and the two-dimensional (2D) isoresponse curves.

In our case, a two level full factorial design 2k was firstly employed to evaluate the main and interaction effects of the factors on the removal rate of our cytostatic agent. We have investigated the influence of three main factors; hydrogen peroxide initial concentration [H2O2] (U1), ferrous ion initial concentration [Fe2+] (U2) and imatinib mesylate initial concentration [IM] (U3); on the anticancer agent degradation efficiency. This kind of design has the advantage of calculating the average and the principle effects of each factor and their interaction 2–2, 3–3 until k factors. The experimental response associated to a 23 factorial design is represented by a linear polynomial model with interaction:
$${\text{Y }} = {\text{ b}}_{0} + {\text{ b}}_{ 1} {\text{X}}_{ 1} + {\text{ b}}_{ 2} {\text{X}}_{ 2} + {\text{ b}}_{ 3} {\text{X}}_{ 3} + {\text{ b}}_{ 1 2} {\text{X}}_{ 1} {\text{X}}_{ 2} + {\text{ b}}_{ 1 3} {\text{X}}_{ 1} {\text{X}}_{ 3} + {\text{ b}}_{ 2 3} {\text{X}}_{ 2} {\text{X}}_{ 3} + {\text{ b}}_{ 1 2 3} {\text{X}}_{ 1} {\text{X}}_{ 2} {\text{X}}_{ 3} ,$$
(8)
where Y represents the experimental response which is the IM removal rate, b0 is the constant coefficient of the model, Xi is the coded variable (− 1 or + 1), bi represents the estimation of the principal effect of the factor i for the response Y, whereas bij represents the estimation of the interactions between factor i and j.
Subsequently, the Doehlert experimental design was used to investigate the optimal conditions of the two quantitative factors (U1 and U2); in the all experimental field; for the removal of IM by FP. The experiment region investigated for IM degradation and the coded values; for both factorial and Doehlert design methodologies; were summarized in Table 1.
Table 1

Experimental range and levels of independent process variables_ Factorial and Doehlert design methodologies

Coded variables (Xi)

Factors (Ui)

Unit

Experimental field

Ui (0)

ΔUi(0)

   

(− 1)

(+ 1)

  

X1

U1: H2O2 initial concentration

mmol L−1

10

50

30

20

X2

U2: Fe2+ initial concentration

mmol L−1

0.50

3

2

1.3

X3

U3: IM initial concentration

mmol L−1

0.045

1.7

Where Ui (0) = (Ui,max + Ui,min)/2 represents the value of Ui at the center of the experimental field and ΔUi (0) = (Ui,max − Ui,min)/2 represents the variation step; with Ui,max and Ui,min are respectively the upper and the lower limits of Ui.

The choice of Doehlert design is justified by its multiple advantages such as its spherical experimental domain with an uniformity in space filling, its ability to explore the whole of the domain, and its potential for sequentially where the experiments can be reused when the boundaries have not been well chosen at first [20]. It consists on N experiments with N = k2 + k + 1, where k is the number of variables. For k = 2, the matrix is composed of 7 experiments which were uniformly distributed within the space of the coded variables (Xi). To obtain an estimation of the experimental error three replicates of the center point were added. Indeed, the experimental response is described by a second order model which allows us to predict the response in all experimental regions from the following equation:
$${\text{Y }} = {\text{ b}}_{0} + {\text{ b}}_{ 1} {\text{X}}_{ 1} + {\text{ b}}_{ 2} {\text{X}}_{ 2} + {\text{ b}}_{ 1 1} {\text{X}}_{ 1}^{ 2} + {\text{ b}}_{ 2 2} {\text{X}}_{ 2}^{ 2} + {\text{ b}}_{ 1 2} {\text{X}}_{ 1} {\text{X}}_{ 2} ,$$
(9)
where Y, experimental response; b0, constant of the model; bi, estimation of the principal effect of the factor i; bii, estimation of the second order effects and bij is the estimation of the interactions between factors i and j for the response Y.

All the experiments were carried out in triplicates and each data point represents the average value of triplicates.

2.4 Analytical Determination

2.4.1 High Performance Liquid Chromatography (HPLC)

IM degradation was monitored by a Waters High Performance Liquid Chromatography system (Alliance e2695 model) equipped with a photodiode array detector (2998) under the following conditions: C-18 reverse phase column, two mixtures composed of 1-octane sulphonic acid sodium salt, acetonitrile, ortho phosphoric acid and purified water as two mobile phases, gradient elution mode and 1 mL min−1 flow rate. The injection volume was 10 μL and detection was performed at 234 nm and 267 nm. The chromatographic data were monitored with the EMPOWER 3 manager software.

2.4.2 Total Organic Carbon (TOC)

The mineralization rates of IM under optimal operating conditions were monitored by measuring the TOC by the mean of a TOC analyser (Sievers 900, GE). Indeed, measurements were applied to the solution containing the target molecule, having the highest and the lowest studied concentrations, before and after degradation.

3 Results and Discussions

3.1 Effect of the Experiment Parameters on the IM Degradation Using the Experimental Factorial Design Methodology

The experimental design and the obtained results are represented in Table 2. The coefficients of the polynomial model were calculated by means of NEMRODW Software according to Eq. (8):
Table 2

Factorial experimental design, experimental plan and results for IM degradation by Fenton process

Experiment number

Experimental design

Experimental plan

Response: removal rate of IM

X1

X2

X3

U1 (mmol L−1)

U2 (mmol L−1)

U3 (mmol L−1)

Y (%)

1

− 1

− 1

− 1

10

0.5

0.045

100

2

1

− 1

− 1

50

0.5

0.045

100

3

− 1

1

− 1

10

3

0.045

100

4

1

1

− 1

50

3

0.045

100

5

− 1

− 1

1

10

0.5

1.7

56.71

6

1

− 1

1

50

0.5

1.7

54.34

7

− 1

1

1

10

3

1.7

99.66

8

1

1

1

50

3

1.7

98.19

$${\text{Y }} = { 88}. 6 1 2 { } - \, 0. 4 80{\text{ X}}_{ 1} + { 1}0. 8 50{\text{ X}}_{ 2} - { 11}. 3 8 8 {\text{ X}}_{ 3} + \, 0. 1 1 3 {\text{ X}}_{ 1} {\text{X}}_{ 2} - \, 0. 4 80{\text{ X}}_{ 1} {\text{X}}_{ 3} + { 1}0. 8 50{\text{ X}}_{ 2} {\text{X}}_{ 3} + \, 0. 1 1 3 {\text{ X}}_{ 1} {\text{X}}_{ 2} {\text{X}}_{ 3} .$$
(10)
The effects of the various studied factors are represented in Fig. 2. This latter indicates that:
Fig. 2

Graphical analysis of the effect of H2O2 concentration, Fe(II) concentration, IM concentration and of their interactions on the degradation of IM

  • The degradation efficiency of the cytostatic solution is very influenced by the initial concentration of IM. Its effect is negative (b3 = − 11.39). Consequently, a complete removal is achieved with low IM concentrations.

  • The second most significant factor on the elimination of IM by FP is the ferrous ions concentration which has a positive effect on the studied response (b2 = + 10.85). Thus an augmentation in the ferrous ions concentration results in a meaningful increase in the removal rate of IM.

  • The effect of H2O2 concentration is very negligible (b1 = − 0.48).

  • Among the interaction terms, X2X3 has the most important value (b23 = + 10.85).

Therefore, the effect of Fe(II) concentration (X2) depends on the level of IM concentration (X3) and vice versa. Although this interaction has a positive effect, the other interactions have a low influence on the studied response (removal rate of IM).

The importance of factors and their interactions has been put into evidence using graphical Pareto analysis [21] which gives more detailed and important information to interpret this result. In fact, this analysis calculates the percentage effect of each factor on the response, according to the following relation [22]:
$${\text{P}}_{\text{i}} = {\text{ (b}}_{\text{i}}^{ 2} / \, \sum {\text{ b}}_{\text{i}}^{ 2} ) { } \times { 1}00{\text{ (i }} \ne \, 0 ).$$
(11)
Figure 3 shows the Pareto graphic analysis. It confirms that the IM concentration and the ferrous ions concentration are the most critical and decisive factors on the removal rate of the cytostatic agent. The contributions of their principal effects on the removal rate of IM are 35.471% and 32.198% respectively.
Fig. 3

Graphical Pareto analysis

Moreover, these two factors have an important and meaningful interaction. Therefore, more than 99.9% of the response is generated by these two factors and their interaction. However, the concentration of H2O2 and the other interactions (X1X2, X1X3 and X1X2X3) have a negligible effect. They represent only 0.1% of the response.

The interpretation of the interactions X1X2, X1X3 and X2X3 can be facilitated by inspecting Fig. 4.
Fig. 4

Interactions: a b12 between Fe(II) concentration and H2O2 concentration, b b13 between IM concentration and H2O2 concentration and c b23 between IM concentration and Fe(II) concentration

Each summit of the square represents a combination of the levels of the two factors [23]; for example: Fe(II) concentration and H2O2 concentration.

For instance, in Fig. 4a, the corner at the top (at the left) of this figure corresponds to Fe(II) concentration equal to 3 mmol L−1 and H2O2 concentration of 10 mmol L−1. The maximum removal rate of 99.83% was obtained under these conditions. The ferrous ions concentration (X2) had a notable and significant impact on the response. When the concentration of hydrogen peroxide (X1) was fixed at the lowest level (10 mmol L−1) or at the highest level (50 mmol L−1), the average rate of the cytostatic agent degradation passed from 99.83 to 78.35% and from 99.09 to 77.17%, respectively; while decreasing the concentration of X2.

The same approach can be applied to interpret X1X3 and X2X3 interactions (Fig. 4b, c).

It can be noticed that the concentration of IM (X3) had a significant impact. The maximum removal rate of IM (100%) was reached with the lowest IM concentration independently of the effect of X1 and X2. The concentration of H2O2 (Fig. 4b) had a negligible effect on the response which decreased only from 78.18 to 76.26% at the lowest level of X3 (0.045 mmol L−1) and was the same (100%) at the highest level of IM. Although the ferrous ions concentration had not an impact on the response when the concentration of IM (X3) was fixed at the lowest level (0.045 mmol L−1), it had a remarkable and significant influence at the highest level of X3 (1.7 mmol L−1). The average rate of the cytostatic agent degradation increased from 55.52 to 98.92%.

These conclusions are reliable based on the tendency of the response, which is extremely influenced by the factors having a significant effect [23]. A total removal of the cytostatic agent was achieved while proceeding with the lowest concentration of IM.

This conclusion was confirmed by the measurement of the TOC of this lowest concentration of the target molecule when it was degraded with 0.5 mmol L−1 of Fe(II) and 10 mmol L−1 of H2O2. Indeed, the obtained TOC value was about 0.11 µg L−1 while the initial one was about 14.9 mg L−1 which is indicative of a 100% mineralization rate.

Considering now high concentration of IM, the best result for its degradation was recorded (98.92%) at a high ferrous ions concentration (3 mmol L−1). We can conclude that the Fenton treatment performance is optimal when working with a high concentration of ferrous ions.

Finally, the factorial design was useful to determine the interactions affecting the response and to indicate if the lowest or the highest levels of the factors are favorable or not [24]. Indeed, this kind of model did not seem suitable for the prediction and the determination of the optimal conditions. Therefore, a surface response methodology should be used as a second step to optimize the experimental parameters in order to improve the degradation efficiency of high levels of IM.

3.2 Optimization Conditions for IM Degradation Using Doehlert Experimental Design Methodology

The experiments and experimental results; defined by the Doehlert matrix; are presented in Table 3.
Table 3

Doehlert matrix and experimental results for IM degradation by FP

Experiment number

Experimental design

Experimental plan

Response: removal rate of IM

X1

X2

U1 (mmol L−1)

U2 (mmol L−1)

Y (%)

1

1

0

50

2.0

85.91

2

− 1

0

10

2.0

92.19

3

0.5

0.866

40

3.1

98.38

4

− 0.5

− 0.866

20

0.9

66.81

5

0.5

− 0.866

40

0.9

66.04

6

− 0.5

0.866

20

3.1

99.63

7

0

0

30

2.0

96.27

8

0

0

30

2.0

96.07

9

0

0

30

2.0

95.00

The results obtained from the three replicates of the center point (0, 0, 0) were nearly the same which indicates the reproducibility of the data.

The calculation of coefficients is carried out in the experimental field (Table 3) using the least squares method [24]:
$${\text{B }} = {\text{ (X}}^{\text{T}} {\text{X)}}^{ - 1} {\text{X}}^{\text{T}} {\text{Y,}}$$
(12)
with B is the vector of estimates of the coefficients, XT is the transposed model matrix, X is the model matrix and Y is the vector of the experiment results. The transformation of experimental values (Ui) into coded variables (Xi) was calculated using the following equation [25]:
$${\text{X}}_{\text{i}} = \, ({\text{U}}_{\text{i}} - {\text{U}}_{{{\text{i(}}0 )}} ) \, /\Delta {\text{U}}_{\text{i}} .$$
(13)
According to the obtained results, polynomial model coefficients were determined by the mean of NEMRODW software:
$${\text{Y }} = { 95}. 7 80 \, - { 2}. 4 30{\text{ X}}_{ 1} + { 18}. 8 1 1 {\text{ X}}_{ 2} - { 6}. 7 30{\text{ X}}_{1}^{2} - 1 5. 1 7 8 {\text{ X}}_{2}^{2} - \, 0. 2 7 7 {\text{ X}}_{ 1} {\text{X}}_{ 2} .$$
(14)

The obtained coefficients related to \({\text{X}}_{ 1} ,{\text{ X}}_{1}^{2} ,{\text{ X}}_{2}^{2}\) and X1X2 are negative. They indicate the unfavourable effects on the IM removal rate. Besides, the positive coefficient related to X2 indicates favourable effect on the cytostatic agent degradation yield.

The analysis of variance (ANOVA) and the p value significance levels were adopted to test the statistical significance of the second-order polynomial model. The ANOVA results are shown in Table 4.
Table 4

ANOVA results for the response surface model for IM removal by FP

Source of variation

Sum of squares

Degree of freedom

Mean square

F ratio

p level

R2

\({\text{R}}_{\text{Adj}}^{2}\)

Regression

1372.69

5

274.538

208.1304

p < 0.1***

0.997

0.992

Residual

3.95720

3

1.31907

  

Lack of fit

3.02460

1

3.02460

6.4864

12.6%

Pure error

0.93260

2

0.46630

  

Correlation total

1376.65

8

  

***Significant at 99.9 % confidence level

A test based on the Fisher distribution (F test) indicated that the fitted equation is statistically significant (F = 208.1304 > 9.013) and has a low probability value (p < 0.1%). The best fit of the model was also justified by the lack of fit (F = 6.4864 < 18.51) and by the high value of the determination coefficient (R2 = 0.997). This latter shows that 99.7% of the total variation in the residual activity was explained by the second-order polynomial predicted Eq. (14). In addition, the value of the adjusted coefficient of determination (\({\text{R}}_{\text{Adj}}^{2}\) = 0.992) is also very high to advocate for a high significance of the model [26].

Hence, the closer the values of R2 to 1, the better the model would explicate the variability between the measured and the predicted values [27]. So, we can deduce that the response Y is adequately described by the second-order polynomial model and the Y values could be calculated in the studied field.

The comparison between experimental and model predicted values of IM degradation rate is reported in Fig. 5. These results are in good agreement and in harmony with the statistical significance of the polynomial model presented in Table 4.
Fig. 5

Comparison of measured and predicted rates for IM degradation

The isoresponse curve and its corresponding three-dimensional response surface plot related to IM degradation rate as a function of Fe(II) and H2O2 initial concentrations are represented in Fig. 6. This graphic analysis showed that the IM degradation rate increases with hydrogen peroxide and ferrous ions initial concentrations. The cytostatic agent removal percentage is maximum in the region from 16 to 37 mmol L−1 (for H2O2 concentration) and from 2.3 to 3.2 mmol L−1 (for Fe(II) concentration). This can be justified by the increase of hydroxyl radicals’ production which degrades more the anticancer agent.
Fig. 6

Effect of Fe(II) and H2O2 concentration on IM degradation yield: a two-dimensional (2D) isoresponse curve plot and b three-dimensional (3D) response surface plot

Further, minimum removal efficiency was noted with higher concentration of H2O2 (> 37 mmol L−1). This slow decomposition of IM was caused by the hydroxyl radical scavenging effect of H2O2 (Eqs. 15, 16) and the recombination of hydroxyl radical (Eq. 17) [15, 28]:
$${\text{HO}}^{.} + {\text{ H}}_{ 2} {\text{O}}_{ 2} \to {\text{HO}}_{ 2}^{.} + {\text{ H}}_{ 2} {\text{O,}}$$
(15)
$${\text{HO}}_{ 2}^{.} + {\text{HO}}^{.} \to {\text{H}}_{ 2} {\text{O }} + {\text{ O}}_{ 2} ,$$
(16)
$$2 {\text{ HO}}^{.} \to {\text{H}}_{ 2} {\text{O}}_{ 2} .$$
(17)

According to the literature, applying a large initial dose of H2O2 is less efficient than its stepwise addition [29]. Thus, it is judicious to control the initial concentration of H2O2 to have a sufficient amount for the degradation of IM without having an adverse effect on the removal rate or the cost of the treatment.

As it can be also observed from Fig. 6, the degradation efficiency of IM was improved by increasing the initial concentration of ferrous ions. This can be also explained by the increase of hydroxyl radicals. Indeed, the removal rate had increased from 62 to 100% while adding Fe(II) from 0.7 to 2.3 mmol L−1.

Finally and based on the predicted results, we can conclude by moving along the major and minor axis of the contour that the optimal values for the variables allowing to achieve a total removal of the cytostatic wastewater (Ypred = 100.59% with an error of 0.63) were 30 mmol L−1 of hydrogen peroxide concentration and 2.5 mmol L−1 of ferrous ions concentration. These values were experimentally validated. In fact, the experimental obtained result was 100% which is in accordance with the statistical one. The mineralization of this target anticancer was also evaluated under these optimal operating conditions. A final TOC value, measured after degradation, was equivalent to 366 mg L−1 while the initial one was equal to 532 mg L−1 which implies a mineralization rate of 31.2%. This latest rate, showing the impossibility of overall mineralization, could be explained by the generation of carboxylic acids especially Fe(III)–carboxylic acid complexes derived from the mineralization of the cytostatic agent [6].

4 Conclusions

In this study, the degradation of a hazardous substance the IM was investigated; to our knowledge for the first time; by applying the Fenton process. Hence, two experimental design methodologies have been adopted successively to evaluate the effect of factors, their interactions and to define the optimal operating conditions that ensure the total removal of IM.

The factorial matrix, which is the first experimental design employed in this work, demonstrated that the IM concentration and the ferrous ions concentration are the most determining parameters on the removal of IM. Indeed, more than 99.9% of the response is bringing by these two factors and their interaction. It is evident from this model that a total removal of the cytostatic agent was achieved with a lowest concentration of IM. Nevertheless, while operating with high levels of the pollutant concentration, the best performance was 98.92% at 3 mmol L−1 of Fe(II).

Optimal experimental conditions for the removal of the anticancer wastewater by Fenton Process were defined by the Doehlert matrix. The maximum removal rate was predicted to be 100.59% when the optimum molar ratio of Fenton reagent (H2O2/Fe2+) was set at 12 (30 mmol L−1 H2O2/mmol L−1 Fe(II)). This estimation was deduced by a statistical significant model (R2 = 0.997).

Therefore, it is appropriate to apply the Fenton process in the advanced treatment of industrial effluents containing other cytotoxic substances. It is an efficient method that does not require expensive apparatus, hardly available reagents and which yields nontoxic products (carbon dioxide and water) or simpler biodegradable particles.

Notes

Acknowledgements

This research project is in the framework of a doctoral thesis MOBIDOC belonging to the Project for the Support of the Research and Innovation System in Tunisia (PASRI) which is funded by the European Union and managed by the National Agency for scientific Research Promotion (ANPR). The authors gratefully acknowledge the members of the Oncology Department at Neapolis Pharma-Medis Group especially Mr. Mohamed Amine BOUJBEL, Mr. Aymen JELASSI and Mr. Riadh BANI for their encouragement and support to carry out this work.

Compliance with ethical standards

Conflict of interest

Authors declare no conflict of interest.

References

  1. 1.
    Ferrando-Climent L, Rodriguez-Mozaz S, Barceló D (2014) Incidence of anticancer drugs in an aquatic urban system: from hospital effluents through urban wastewater to natural environment. Environ Pollut 193:216–223CrossRefGoogle Scholar
  2. 2.
    Mater N, Geret F, Castillo L, Faucet-Marquis V, Albasi C, Pfohl-Leszkowicz A (2014) In vitro tests aiding ecological risk assessment of ciprofloxacin, tamoxifen and cyclophosphamide in range of concentrations released in hospital wastewater and surface water. Environ Int 63:191–200CrossRefGoogle Scholar
  3. 3.
    Zhang J, Chang VWC, Giannis A, Wang JY (2013) Removal of cytostatic drugs from aquatic environment: a review. Sci Total Environ 445–446:281–298CrossRefGoogle Scholar
  4. 4.
    Wang XH, Lin AYC (2014) Is the phototransformation of pharmaceuticals a natural purification process that decreases ecological and human health risks? Environ Pollut 186:203–215CrossRefGoogle Scholar
  5. 5.
    Santos MSF, Alves A, Madeira LM (2011) Paraquat removal from water by oxidation with Fenton’s reagent. Chem Eng J 175:279–290CrossRefGoogle Scholar
  6. 6.
    Oturan MA, Aaron JJ (2014) Advanced oxidation processes in water/wastewater treatment: Principles and applications: A review. Crit Rev Environ Sci Technol 44(23):2577–2641.  https://doi.org/10.1080/10643389.2013.829765 CrossRefGoogle Scholar
  7. 7.
    Neyens E, Baeyens J (2003) A review of classic Fenton’s peroxidation as an advanced oxidation technique. J Hazard Mater 98:33–50CrossRefGoogle Scholar
  8. 8.
    Bagal MV, Gogate PR (2014) Wastewater treatment using hybrid treatment schemes based on cavitation and Fenton chemistry: a review. Ultrason Sonochem 21:1–14CrossRefGoogle Scholar
  9. 9.
    Babu Ponnusami A, Muthukumar K (2014) A review on Fenton and improvements to the Fenton process for wastewater treatment. J Environ Chem Eng 2:557–572CrossRefGoogle Scholar
  10. 10.
    Governo M, Santos MSF, Alves A, Madeira LM (2016) Degradation of the cytostatic 5-fluorouracil in water by Fenton and photo-assisted oxidation processes. Environ Sci Pollut Res 24:844–854CrossRefGoogle Scholar
  11. 11.
    Ramirez JH, Costa CA, Madeira LM (2005) Experimental design to optimize the degradation of the synthetic dye orange II using Fenton’s reagent. Catal Today 107–108:68–76CrossRefGoogle Scholar
  12. 12.
    Haber F, Weiss J (1934) The catalytic decomposition of hydrogen peroxide by iron salts. Proc R Soc Lond A 147(861):332–351CrossRefGoogle Scholar
  13. 13.
    Wu Y, Zhou S, Qin F, Zheng K, Ye X (2010) Modeling the oxidation kinetics of Fenton’s process on the degradation of humic acid. J Hazard Mater 179:533–539CrossRefGoogle Scholar
  14. 14.
    Sina MA, Mohsen M (2017) Advances in Fenton and Fenton based oxidation processes for industrial effluent contaminants control—a review. Int J Environ Sci Nat Res 2(4):555–594.  https://doi.org/10.19080/IJESNR.2017.02.555594 Google Scholar
  15. 15.
    Brillas E, Sirés I, Oturan MA (2009) Electro-Fenton process and related electrochemical technologies based on Fenton’s reaction chemistry. Chem Rev 109:6570–6631CrossRefGoogle Scholar
  16. 16.
    Kuo WG (1992) Decolorizing dye wastewater with Fenton’s reagent. Water Res 26:881–886CrossRefGoogle Scholar
  17. 17.
    Kavitha V, Palanivelu K (2005) Destruction of cresols by Fenton oxidation process. Water Res 39:3062–3072CrossRefGoogle Scholar
  18. 18.
    Bianchi F, Caffarri E, Cavalli S, Lagrastac C, Musci M, Quaini F, Savi M (2013) Development and validation of an high performance liquid chromatography-tandem mass spectrometry method for the determination of imatinib in rat tissues. J Pharm Biomed Anal 73:103–107CrossRefGoogle Scholar
  19. 19.
    Muruganandham M, Swaminathan M (2004) Decolourisation of reactive orange 4 by Fenton and photo-Fenton oxidation technology. Dyes Pigment 63:315–321CrossRefGoogle Scholar
  20. 20.
    Bensalah W, Feki M, Wery M, Ayedi HF (2010) Thick and dense anodic oxide layers formed on aluminum in sulphuric acid bath. J Mater Sci Technol 26:113–118CrossRefGoogle Scholar
  21. 21.
    Haaland DP (1989) Experimental design in biotechnology. Marcel Dekker Inc, New YorkGoogle Scholar
  22. 22.
    Hammami S, Oturan N, Bellakhal N, Dachraoui M, Oturan MA (2007) Oxidative degradation of direct orange 61 by electro-Fenton process using a carbon felt electrode: application of the experimental design methodology. J Electroanal Chem 610:75–84CrossRefGoogle Scholar
  23. 23.
    Zaviska F, Drogui P, Blais JF, Mercier G, Lafrance P (2011) Experimental design methodology applied to electrochemical oxidation of the herbicide atrazine using Ti/IrO2 and Ti/SnO2 circular anode electrodes. J Hazard Mater 185:1499–1507CrossRefGoogle Scholar
  24. 24.
    Myers RH, Montgomery DC (2002) Response surface methodology: process and product optimization using designed experiments, 2nd edn. Wiley, New YorkGoogle Scholar
  25. 25.
    Mansour D, Fourcade F, Soutrel I, Hauchard D, Bellakhal N, Amarane A (2015) Mineralization of synthetic and industrial pharmaceutical effluent containing trimethoprim by combining electro-Fenton and activated sludge treatment. J Taiwan Inst Chem Eng 53:58–67.  https://doi.org/10.1016/j.jtice.2015.02.022 CrossRefGoogle Scholar
  26. 26.
    de Sena AR, Valasques Júnior GL, Barretto IKSP, Assis SA (2012) Application of Doehlert experimental design in the optimization of experimental variables for the Pseudozyma sp. (CCMB 306) and Pseudozyma sp. (CCMB 300) cell lysis. Food Sci Technol (Campinas) 32:761–767CrossRefGoogle Scholar
  27. 27.
    Sayyad SA, Panda BP, Javed S, Ali M (2007) Optimization of nutrient parameters for lovastatin production by Monascus purpureus MTCC 369 under submerged fermentation using response surface methodology. Appl Microbiol Biotechnol 73:1054–1058CrossRefGoogle Scholar
  28. 28.
    Nidheesh PV, Gandhimathi R (2012) Trends in electro-Fenton process for water and wastewater treatment: an overview. Desalination 299:1–5CrossRefGoogle Scholar
  29. 29.
    Mansoorian HJ, Bazrafshan E, Yari A, Alizadeh M (2014) Removal of azo dyes from aqueous solution using Fenton and modified Fenton processes. Health Scope 3(2):15507.  https://doi.org/10.17795/jhealthscope-15507 Google Scholar

Copyright information

© The Tunisian Chemical Society and Springer Nature Switzerland AG 2018

Authors and Affiliations

  1. 1.Research Unit of Catalysis, Electrochemistry, Nanomaterials and their Applications and Didactic, National Institute of Applied Sciences and TechnologiesCarthage UniversityTunisTunisia
  2. 2.Neapolis Pharma, MediS GroupNabeulTunisia

Personalised recommendations