**4. Materials and methods**

**Figure 2.** Pseudo-catalyst conversion of cyanide by free hydroxyl functional groups [61]

*vulgaris*

62 Biotechnology

are shown in Table 1.

**3. Biodegradation of cyanide by** *Fusarium oxysporum* **grown on** *Beta*

hours to produce minute quantities of ammonium-nitrogen (5.24 mg NH4

A number of different studies report on the application of cyanide degrading fungi. For instance, white rot fungi, *Trametes versicolor*, have been shown by Cabuk *et al*. [9] to tolerate cyanide concentration up to 130 mg F-CN/L, with complete degradation observed within 42

cyanide degrading fungi were examined by Pereira *et al*. [57] such as *Fusarium sp*. including *Aspergillus sp.* by Santos *et al.* [57, 62], and were found to tolerate cyanide concentration up to 520 mg F-CN/L. A list of other cyanide degrading species including degradation conditions

There has been limited emphasis on the effect of carbon or nitrogen sources used in the biodegradation of cyanide. The viability of the agrowaste depends on the type of bioremedia‐ tion required and the microorganism used. When the cultivating conditions are conducive, the minerals, proteins, carbohydrates and water in the agrowaste become easily accessible to the microorganisms [46]. Monosaccharides such as mannose, glucose and fructose present in the agrowaste can effectively support and/or enhance microbial growth [2]. Other overriding factors which directly influence cyanide degradation include exposure to direct sunlight, temperature and pH. Cyanide compounds are soluble in water, thus dissociate and evaporate easily at low pH (i.e. pH<9) while under high salinity, the solubility decreases. Also at neutral

+


The experiments were carried out in batch cultures. *B. vulgaris* waste was milled to ≤ 100 µm. A broth of 0.5 g of milled waste in 10 mL distilled water was autoclaved at 116° C for 15 min to prevent thermal breakdown of reducing sugars [51]. To the waste broth, wastewater (20 mL) with 1 mL of a spore solution (2.25 x 106 spore/mL) of *Fusarium oxysporum* was added to the *B. vulgaris* broth. The wastewater used had characteristics similar to the goldmine wastewater reported by Acheampong *et al.* [1] having metals such as arsenic, iron, copper, lead and zinc. The mixture was incubated for 48 hours in a rotary shaker at 70 rpm at the desired temperature and pH (- see Table 2). After this, KCN in distilled water, was added to make a final cyanide concentration of 500 mg CN- /L in the mixture. Thereafter, the mixture was incubated for a further 72 hours at 70 rpm at the desired temperature (- see Table 2). All experiments were carried out in duplicate in airtight multiport round bottom Erlenmeyer flasks (n = 28; final volume of 51 mL). Cyanide (CN) (09701) and ammonium-nitrogen (NH4 + -N) (00683) test kits (MERCK®) were used to quantify the residual free cyanide and ammonium-nitrogen concen‐ trations using a NOVA 60 spectroquant. Free cyanide volatilised was accounted for using the mass balance equations below:

$$\mathbf{CN}\_s \cdot \left(\mathbf{CN}\_r^\cdot + \mathbf{CN}\_v^\cdot\right) = \mathbf{CN}\_b^\cdot \tag{1}$$

$$\text{CNN}\_{\text{v}} = \text{CN}^{\cdot}\_{\text{vo}} - \text{CN}^{\cdot}\_{\text{vf}} \tag{2}$$

where *CN<sup>s</sup>* is the initial free cyanide concentration in the culture broth; *CN<sup>r</sup>* is the measured residual free cyanide after incubation; *CN<sup>v</sup>* is the volatilised free cyanide during incubation; *CN<sup>b</sup>* is the bioremediated free cyanide; *CNvo* is the initial free cyanide in control cultures (500 mg F-CN/L); and *CNvf* is the final free cyanide in control cultures. The control was prepared under the same conditions as other cultures without the *Fusarium oxysporum*.


**Table 1.** Cyanide degrading microbial species using different nutritional sources under different temperature and pH conditions

Utilization of *Beta vulgaris* Agrowaste in Biodegradation of Cyanide Contaminated Wastewater http://dx.doi.org/10.5772/59668 65


**Table 2.** Experimental variation of pH and temperature

**Microorganism C-source N-source Temperature**

*Burkholderia cepacia* Fructose, glucose,

*Citrobacter sp., Pseudomonas sp.* Sugarcane molasses,

*Klebsiella oxytoca immobilised cell* Alginate and cellulose

Mixed culture of bacteria Glucose CN-

*Fusarium oxysporum* immobilised

on sodium alginate

64 Biotechnology

*Gloeocercospora sorghi, Stemphylium loti*

Mixed culture of bacteria immobilised on ultrafiltration

*Pseudomonas fluorescens* immobilised on calcium alginate

*Pseudomonas fluorescens* immobilised on zeolite

*pseudoalcaligenes CECT5344*

on sodium alginate

*Pseudomonas putida* immobilised

*Pseudomonas putida* immobilised

membranes

*Pseudomonas*

conditions

glucose

triacetate

on sodium alginate NaCN, sodium alginate NaCN, Cyanates

*Pseudomonas sp.*(CM5, CMN2) Glycerol CN-

*Fusarium oxysporum Beta vulgaris* **KCN 30 11 This study** *Agrobacterium tumefaciens* Starch KCN - 7.2 [58] *Aspergillus awamori Citrus sinesis* extract KCN 40 8.84 [62] *Baccillus pumilus* Glucose KCN 40 8.5-9 [64] *Baccillus stearothermphilus* - NaCN 27±2 7.8 [6]

**(**

mannose KCN <sup>30</sup> <sup>10</sup> [2]

Formamide Cyanides 25-30 8 [11]

Glucose KCN 35, 28 5.3-5.7, 7.0 [48]

Phenol Cyanides 25 - [35]

Glucose Ferrocyanide 25-35 4-7 [20]

CH3COONa NaCN 30 9.5 [38]

NaCN NaCN 25 6.7 [7]

and thiocyanates

**Table 1.** Cyanide degrading microbial species using different nutritional sources under different temperature and pH

nickelate (II)

[Cu(CN)4] 2-,

[Zn(CN)4]

*Cryptococcus humicolus* MCN2 Glucose KCN 25 7.5 [36] *Eschericia coli* Glucose KCN 30 9.2 [26] *Fusarium solani* Glucose K2Ni(CN)4, KCN 25 7.0 [8] *Fusarium solani* Yeast KCN 30 9.2-10.7 [19] *Fusarium oxysporum* Glucose KCN 25 8.0 [57]

*Klebsiella oxytoca* Glucose KCN 30 7 [34]

*Pseudomonas fluorescens* Glucose Ferrocyanide 25 5 [21]

*Pseudomonas putida* BCN3 Glucose [K2[Ni(CN)4]] 30 - [63]

*Pseudomonas stutzeri* AK61 - KCN 30 7.6 [68]

*Stemphilium loti* KCN 25 6.5, 7.5 [27] *Trametes versicolor* Citrate KCN 30 10.5 [9] *Trichoderma sp.* Glucose CN- 25 6.5 [24] *Scenedesmus obliquus* NaCN NaCN - 10.3 [33] *Rhodococcus* UKMP-5M Glucose KCN 30 6.6 [41]

Zeolite Tetra-cyano-

**oC) pH Reference**

2- 35 7.5 [56]

KCN 30 7 [13]

WAD 22 7.0 [70]

30 - [66]

25 7.5 [12]

WAD 30 9.2-11.4 [4]

The response surface methodology was used for the statistical design of the experiments to assess the influence of temperature and pH for optimal degradation of cyanide. A central composite design was used for the determination of optimal operating conditions with a minimum residual ammonium-nitrogen as one of the objectives. Design Expert software® version 6.0.8 (Stat-Ease Inc., USA) was used to generate the experimental runs.


A and B represent coded level of variables.

**Table 3.** Coded experimental design variables and the corresponding response

The results (Table 3) indicated a variation in responses measured. There was appreciable degradation of cyanide in Runs 9, 4, 1, 3, 6, 8, 13, and 14, with the highest cyanide degraded being 263 mg F-CN/L (Run 9) and the lowest (83 mg F-CN/L) being observed for Run 11. However, both cases had a high residual ammonium-nitrogen of 210 mg NH4 + -N/L and 120 mg NH4 + -N/L, respectively. Both Runs 9 and 11 were axial points. Run 9 with an extremely high pH resulted in high residual ammonium-nitrogen while Run 11 with an extremely low temperature was observed to have minimal microbial activity despite the presence of a suitable quantity of *B. vulgaris* used as a carbon source. A similar scenario had earlier been reported by Zilouei *et al.*[72] and Zou *et al*. [73], whereby a low temperature was found to inhibit the growth of microorganisms, thus resulting in low removal of contaminants (ammonium-nitrogen, nitrate and nitrite). On the other hand, Runs 1, 3, 4, 6, 7, 8, 13 and 14 had up to 99% correlation with the predicted values for cyanide degradation which indicated a high accuracy of the model (Equation 4) used for predicting cyanide degradation. However, only Runs 4 and 7, which showed minimal residual ammonium-nitrogen presence, can be used for optimisation for a pilot scale process.

#### **5. Statistical model analysis**

The statistical model summary clarifies the fitness of the mean and quadratic models for the two responses based on the Sequential Model Sum of Squares and Lack of Fit Test. The responses were analysed using ANOVA to assess the significance of the variables in the model. A quadratic model was found to give the best fit for the experimental results.


S = significant; NS = Not significant; CL = Confidence Level; DF = Degree of freedom; "Prob > F" less than 0,05 indicates the model term is significant while values greater than 0.1 indicates the model term is not significant; Std. Dev. = 24.58; R2 = 0.8907; Adj. R2 = 0.8127; Pred. R2 = -0.1858; Adeq. Precision = 10.341

**Table 4.** ANOVA for F-CN Reponse Surface Quadratic Model

The predicted response (Y) for the biodegradation of free cyanide in terms of the coded values was:

$$\text{Y = } 239 + 23.6 \text{A} + 38.09 \text{B} - 49.87 \text{A}^2 - 3.62 \text{B}^2 + 3.25 \text{AB} \tag{3}$$

where *A* and *B* are the coded values of temperature and pH, respectively. When coefficients with significant effects were considered, Eq. (3) became;

$$\text{Y} = \text{239} + \text{23.6A} + \text{38.09B} - \text{49.87A}^2 \tag{4}$$

A model reduction was appropriate since there were many insignificant model terms. Excluding these terms improved the model. The Model F-value of 11.41 for the cyanide biodegradation was significant; therefore, there was only a 0.29% chance that a "Model F-Value" this large could occur due to noise for the quadratic model. Statistically, an adequate ratio greater than 4 is desireable for measuring a signal to noise ratio; therefore, the adequate precision of 10.341 observed in this study indicates a passable signal that can be used to further navigate the design space. Figure 3 further justifies the fitness of the model with normality in the error term.

## **6. Representation of the response surface model**

The results (Table 3) indicated a variation in responses measured. There was appreciable degradation of cyanide in Runs 9, 4, 1, 3, 6, 8, 13, and 14, with the highest cyanide degraded being 263 mg F-CN/L (Run 9) and the lowest (83 mg F-CN/L) being observed for Run 11.

high pH resulted in high residual ammonium-nitrogen while Run 11 with an extremely low temperature was observed to have minimal microbial activity despite the presence of a suitable quantity of *B. vulgaris* used as a carbon source. A similar scenario had earlier been reported by Zilouei *et al.*[72] and Zou *et al*. [73], whereby a low temperature was found to inhibit the growth of microorganisms, thus resulting in low removal of contaminants (ammonium-nitrogen, nitrate and nitrite). On the other hand, Runs 1, 3, 4, 6, 7, 8, 13 and 14 had up to 99% correlation with the predicted values for cyanide degradation which indicated a high accuracy of the model (Equation 4) used for predicting cyanide degradation. However, only Runs 4 and 7, which showed minimal residual ammonium-nitrogen presence, can be used for optimisation

The statistical model summary clarifies the fitness of the mean and quadratic models for the two responses based on the Sequential Model Sum of Squares and Lack of Fit Test. The responses were analysed using ANOVA to assess the significance of the variables in the model.

**95% CL Low 95% CL High F Value Prob > F Significance**

A quadratic model was found to give the best fit for the experimental results.

Intercept 239 1 10.03 215.27 262.73 11.41 0.0029 S A 23.6 1 8.69 3.05 44.50 7.37 0.0300 S B 38.09 1 8.69 17.54 58.63 19.21 0.0032 S A2 -49.87 1 9.05 -71.26 -28.49 30.40 0.0009 S B2 -3.62 1 9.05 -25.01 17.76 0.16 0.7005 NS AB 3.25 1 12.29 -25.81 32.31 0.07 0.7991 NS

S = significant; NS = Not significant; CL = Confidence Level; DF = Degree of freedom; "Prob > F" less than 0,05 indicates the model term is significant while values greater than 0.1 indicates the model term is not significant; Std. Dev. = 24.58;

The predicted response (Y) for the biodegradation of free cyanide in terms of the coded values

2 2 Y = 239 + 23.6A + 38.09B - 49.87A – 3.62B + 3.25AB (3)

= -0.1858; Adeq. Precision = 10.341

**Standard Error**


+


However, both cases had a high residual ammonium-nitrogen of 210 mg NH4

mg NH4 +

66 Biotechnology

for a pilot scale process.

**Factor**

R2

was:

= 0.8907; Adj. R2

**5. Statistical model analysis**

**Coeff. Estimate**

**DF**

= 0.8127; Pred. R2

**Table 4.** ANOVA for F-CN Reponse Surface Quadratic Model

The interaction between independent variables can be studied by plotting three dimensional (3-D) curves of the response against the variables. It allows for the interpretation of experi‐ mental results and determination of optimal conditions. Elliptical contour shows the interac‐ tion between the independent variables is perfect while a circular contour indicates the variables are non-interactive [44, 47].

**Figure 3.** Normal probability plot of the residual F-CN

**Figure 3. Normal probability plot of the residual F-CN** 

**Figure 4.** D plot showing interaction of independent variables on cyanide degradation

**Figure 4. 3-D plot showing interaction of independent variables on cyanide degradation** 

11

12

+


**Figure 5. 3-D plot showing interaction of independent variables on ammonium-nitrogen formation Figure 5.** D plot showing interaction of independent variables on ammonium-nitrogen formation

mg F-CN/L and minimum ammonium-nitrogen formation of 74.285 mg NH4

C and pH of 11.

#### **7. Cyanide biodegradation optimisation 7. Cyanide biodegradation optimisation**

be at temperature of 30<sup>o</sup>

The optimisation was done using the Design-Expert software® numerical optimisation option where input factors were selected to achieve a desired perfomance. The numerical optimisation The optimisation was done using the Design-Expert software® numerical optimisation option where input factors were selected to achieve a desired perfomance. The numerical optimisation

can maximise, minimise or achieve a targeted value: a single response; a single response subjected to upper and/or lower boundaries on other responses; and combinations of two or more responses. The desired goal for each variable and response is selected and the weight is chosen to show the degree of importance of individual goals. In this analysis, temperature and pH were set within range, cyanide degradation response was set at maximum while ammonium-nitrogen formation response was set at a minimum. The software gave three different solutions for this criteria with different desirability. The optimum point with the highest desirability was selected as shown in Fig.6 and 7. The optimal point with the maximum cyanide degradation of 250.436 can maximise, minimise or achieve a targeted value: a single response; a single response subjected to upper and/or lower boundaries on other responses; and combinations of two or more responses. The desired goal for each variable and response is selected and the weight is chosen to show the degree of importance of individual goals. In this analysis, temperature and pH were set within range, cyanide degradation response was set at maximum while ammo‐ nium-nitrogen formation response was set at a minimum. The software gave three different solutions for this criteria with different desirability. The optimum point with the highest desirability was selected as shown in Fig.6 and 7. The optimal point with the maximum cyanide degradation of 250.436 mg F-CN/L and minimum ammonium-nitrogen formation of 74.285 mg NH4 + -N /L was found to be at temperature of 30o C and pH of 11.

**Figure 6.** Desirability ramp for the numerical optimisation of cyanide degradation and ammonium-nitrogen formation

11

12

+


**Figure 3. Normal probability plot of the residual F-CN** 

(mg/L)

68 Biotechnology

(mg/L)

**7. Cyanide biodegradation optimisation** 

**7. Cyanide biodegradation optimisation**

be at temperature of 30<sup>o</sup>

**Figure 4. 3-D plot showing interaction of independent variables on cyanide degradation** 

**Figure 5. 3-D plot showing interaction of independent variables on ammonium-nitrogen formation** 

**Figure 5.** D plot showing interaction of independent variables on ammonium-nitrogen formation

The optimisation was done using the Design-Expert software® numerical optimisation option where input factors were selected to achieve a desired perfomance. The numerical optimisation can maximise, minimise or achieve a targeted value: a single response; a single response subjected to upper and/or lower boundaries on other responses; and combinations of two or more responses. The desired goal for each variable and response is selected and the weight is chosen to show the degree of importance of individual goals. In this analysis, temperature and pH were set within range, cyanide degradation response was set at maximum while ammonium-nitrogen formation response was set at a minimum. The software gave three different solutions for this criteria with different desirability. The optimum point with the highest desirability was selected as shown in Fig.6 and 7. The optimal point with the maximum cyanide degradation of 250.436

The optimisation was done using the Design-Expert software® numerical optimisation option where input factors were selected to achieve a desired perfomance. The numerical optimisation

mg F-CN/L and minimum ammonium-nitrogen formation of 74.285 mg NH4

C and pH of 11.

**Figure 4.** D plot showing interaction of independent variables on cyanide degradation

**Figure 7.** Desirability histogram for numerical optimisation of cyanide degradation and ammonium-nitrogen forma‐ tion
