Degradation and Detoxification of Banana Postharvest Treatment Water Using Advanced Oxidation Techniques ()
The elimination and detoxification of water contaminated with the principal commercial fungicides—thiabendazole (Textar 60 T) and imazalil sulphate (Fruitgard IS 7.5)—used in the postharvest treatment of bananas was successfully achieved through the use of advanced oxidation processes (AOPs). The techniques tested were heterogeneous photocatalysis with TiO2 and TiO2-activated carbon (TiO2-CA) and Fenton and photo-Fenton processes. The evolution of fungicide concentration during degradation was studied together with mineralization and toxicity levels. When treated with any of the methods under study, the water meets the conditions stipulated by the legislation currently in force for its reuse in irrigation. However, this study proposes that the Fenton process is the best option for treatment of this type of water due to the efficiency of its reactions and its low economic cost.
1. Introduction
The water under study represents a typical wastewater obtained after bananas have been washed as part of their postharvest treatment. In this postharvest treatment the fruit are normally washed two or three times in order to remove larger particles and any dust that might have been brought with the fruit from its growing area. A fungicide mixture is also applied to combat possible diseases, particularly crown rot, that might affect the fruit. Among the most commonly used fungicide mixtures are those studied in this work which are thiabendazole-based or imazalil sulphate-based [1].
This water contaminated with the postharvest fungicide is normally washed away into the sewage system. For a banana packing system of between 10 and 60 tons/ day, this represents a total daily volume of some 10000 L of water that needs to be washed away. The various wastewater samples analysed for this study contained concentrations of between 10 and 15 mg·L–1 of imazalil sulphate or between 25 and 30 mg·L–1 of thiabendazole. By law, this fungicide concentration must be removed before it is washed away into the sewer system. Legislation in the Canary Islands (Spain), one of the biggest banana producing regions, requires that concentrations of this type of substance must not exceed 0.05 mg·L–1 [2]. However, because of the growing scarcity of the water resource in the Canary Islands as a result of overexploitation and contamination of the aquifers, the use of treated water for irrigation purposes is of particular interest. For this to be possible, restrictions related to quality indicators such as BOD5, COD, nitrates, etc., must also be complied with in addition to the required fungicide threshold concentrations [3].
Studies have been published on the effectiveness of advanced oxidation techniques, specifically heterogeneous TiO2-based photocatalysis and the Fenton processes for the degradation of pesticides and fungicides in water [4-6]. These processes, although making use of different reacting systems, are all characterised by the same chemical feature: production of ·OH radicals [7].
In heterogeneous photocatalysis radiant energy is absorbed by the heterogeneous photocatalyst, a broadband semiconductor, in this case TiO2. The degradation reactions of the pollutants take place in the interfacial region without the catalyst undergoing chemical changes.
To improve the efficiency of the TiO2, an inert surface like activated carbon is used as a support for the TiO2 to provide a large surface area. There are two main reasons for using adsorbents like this as support for the photoactive catalyst: firstly, to increase the size of the photoactive surface and secondly, to promote contact between the compound to be degraded and the catalyst by preconcentration of the pollutant at the semiconductor surface [8].
Studies have been undertaken which report that activated carbon increases not only the accessible surface area of the TiO2, as it impedes its aggregation, but also its decantability, acid-base properties and absorbance spectrum. This all leads to an improved photocatalytic performance. It has been established that the amount of activated carbon that maximises these benefits lies between 7% and 13% [9].
Fenton’s reagent is an aqueous solution of hydrogen peroxide and ferrous ions which provides an important source of hydroxyl radicals and which, under acidic conditions (pH 2 - 4), is a powerful oxidizing agent of organic compounds. Degradation rates increase substantially when the process is complemented with UV/visible radiation. In these cases it is known as a photo-Fenton process. The process becomes a photocatalytic one, as the Fe2+ oxidized to Fe3+ in the Fenton reaction is reduced again to Fe2+ by the action of radiation. Both reactions generate hydroxyl radicals [10].
The Fenton process has proven to be effective for the degradation of various compounds, including chlorinated aliphatic and aromatic compounds, nitroaromatics, azodyes, chlorobenzene, phenols, formaldehyde, herbicides, soil pollutants, etc. [11]. Fenton’s reagent has also been successfully applied in the reduction of COD in municipal waste and ground waters and in leachate treatment. It is useful for the pre-treatment of non-biodegradable compounds [12]. Among the very few compounds that cannot be attacked by this reagent are acetone, acetic acid, oxalic acid, paraffins and organochlorine compounds [13].
The use of Fenton and photo-Fenton processes in the treatment of pesticide-containing wastewater has also been studied: the degradation of triazine herbicides [14], methyl-parathion [15], fenuron [16], the fungicides pyrimethanil and triadimenol [17], and commercial mixtures of pesticides [6, 18] are known examples.
The objective of this study is the removal of both thiabendazole and imazalil sulphate from the water used in the postharvest washing of bananas, using advanced oxidation techniques to comply with legislation governing wastewater disposal and/or water reuse for agricultural purposes. The effectiveness of four techniques is compared: heterogeneous photocatalysis based on TiO2 used alone and in combination with activated carbon, Fenton and photo-Fenton processes.
2. Experimental
2.1. Characterisation of the Water to be Treated
Of the two fungicide mixtures under study in this work, one is thiabendazole-based (C10H7N3S) and the other imazalil sulphate-based (C14H16Cl2N2O5S), shown in Figure 1. The latter belongs to the imidazole chemical group and was developed to control postharvest diseases in fruit, particularly when there are problems of pathogen tolerance to benzimidazole fungicides like thiabendazole [19].
The fungicide mixtures were provided by the North Gran Canaria Agricultural Cooperative (Cooperativa Agrí- cola del Norte de Gran Canaria), which forms part of the Canary Island Banana Cooperative (COPLACA). The thiabendazole-based fungicide mixture was a 0.075% (v/v) solution of Textar 60 T which contained 60% (w/v) of thiabendazole. The imazalil sulphate based fungicide mixture was a 0.6% (v/v) solution of Fruitgard IS 7.5. The wastewater samples analysed in this study revealed concentrations of 25 - 30 mg·L–1 of thiabendazole and 10 - 15 mg·L–1 of imazalil sulphate. Other chemical characteristics of both waste products are shown in Table 1.
2.2. Materials and Methods
2.2.1. Reagents/Chemicals
The reagents used for the photocatalytic tests were: TiO2 (Degussa P25) and activated carbon supplied by Panreac (surface area: 778.28 m2/g, pore volume: 0.454 cm3/g, ash content: <1%, impurities: Cl– ≈ 0.2%).
35% H2O2 (Scharlau) was used for the Fenton tests. Panreac dehydrated ferrous sulphate (99%) was used as the Fe2+ source.
pH was adjusted with diluted H2SO4 and NaOH.
Catalase was used to remove the H2O2 and avoid interferences with the analytical results (Micrococcus Lysodeikticus obtained from Fluka with 243075 U·mL–1).
2.2.2. Analytical Determinations
Remaining concentrations of thiabendazole at different reaction times were HPLC-measured using a Supelco Discovery C18 (25 cm × 4.6 mm ID, 5 µm particles) and an acetonitrile-water-ammonia 40:60:0.6 solution as mobile phase, using a UV detector (λ = 280 nm). Concentrations of imazalil sulphate were measured by the same method, using in this case an acetonitrile-10mM KH2PO4 solution (35:65) and 200 mg·L–1 of sodium 1-octanedulfonate as mobile phase (adjusted to pH 3 with phosphoric
Figure 1. Thiabendazole molecule (left) and imazalil sulphate molecule (right).
Table 1. Parameters of interest of the water under study. Thiabendazole-based or imazalil-based fungicide mixtures.
acid), with detection conducted at a wavelength of λ = 225 nm. The quantification of both fungicides was performed by least-squares fit [20].
Total organic carbon (TOC) was measured by using a TOC Shimadzu 5000-A.
The concentration of ions in solution was determined on a DIONEX Ionic Chromatograph equipped with a GP50 gradient pump, ED50 electrochemical detector and an IonPac AS11-HC column (4 × 250 mm) using aqueous NaOH (30 mM) as eluent and a flowrate of 1 mL·min–1.
The toxicity of samples was determined by using the Lemna minor toxicity test [21]. Glass Petri dishes containing 13 ± 2 fronds of common duckweed (L. minor) were placed under constant visible radiation (one 18 W fluorescent tube placed approximately 25 cm above the test chambers) for 96 ± 2 h in a chamber with an ambient temperature of 23 ± 1˚C. Four replicates were used for each sample, which included a control without pollutant and samples taken at different reaction times. To 50 mL of solution, 0.5 mL of concentrated nutrient solutions were added. To each dish 15 mL of sample at pH 7.5 - 8 were added. Growth inhibition percentage (I) was calculated with respect to the control without pollutant according to
(1)
where C and T are the frond number mean increments for the control and the sample, respectively.
BOD5 was analysed using special opaque and lightproof containers with an Aqualytic BOD-Sensor. COD was determined using the recovery technique with potassium dichromate (K2Cr2O7) and SS using the filtration technique with 0.45 µm filters [21].
2.2.3. Experimental Conditions
Prior to the degradation tests, both fungicide mixtures were filtered to remove suspended solids.
Degradation tests were conducted in 250 mL capacity Pyrex glass batch reactors.
For the heterogeneous photocatalytic degradation studies 200 mL volumes with 1 g·L–1 of Degussa P-25 TiO2 or Degussa P-25 TiO2 with 7% activated carbon were used. Aeration was maintained with an aquarium pump (EOLO AC3000 model: 2.5W power, 2 L·min–1 output and pressure > 0.02 MPa) and a constant stirring rate of 450 rpm. A 60W Philips Solarium HB175 equipped with four 15 W Philips CLEO fluorescent tubes with emission spectrum from 300 to 400 nm (maximum around 365 nm) was used as UV source in the degradation studies. The tests were conducted at natural pH and pH 5, as various studies have reported that photocatalytic processes are more efficient at slightly acidic pHs [22].
For the Fenton and photo-Fenton tests various concentrations of Fe2+ and H2O2 were used, with reaction volumes of 200 mL and a working pH equal to 3, adjusted with sulphuric acid. A constant stirring of 450 rpm was maintained. For the photo-Fenton reactions the Philips Solarium as described above was used as UV source. Catalase was used to remove excess H2O2 and to avoid interferences in the analytical and toxicity measurements.
After adsorption equilibrium was established, samples were taken to monitor the reaction in time intervals of 30 minutes or less, according to the kinetics established for each reaction. Before being measured the samples were filtered using 0.45 µm syringe filters.
3. Results
3.1. Thiabendazole
3.1.1. Photocatalysis: TiO2 and TiO2 + AC
Figure 2 shows the results of thiabendazole degradation at natural pH (pH = 6.3) using the TiO2 based photocatalytic process. It can be seen that after 120 minutes of reaction time 95% of the thiabendazole present in the sample has been removed and after 390 minutes the amount of thiabendazole falls below the wastewater limit (0.05 mg·L–1) as laid down in Canary Island (Spain) legislation.
Figure 2. Thiabendazole concentration, TOC and toxicity evolution during the photocatalytic treatment of a 25 mg·L–1 thiabendazole aqueous solution at natural pH (6.3) using 1 g·L–1 of TiO2.
TOC kinetics are much lower than for degradation. So, after 120 minutes only 37% of the compound had been mineralized, rising to 76% after 240 minutes.
As can be observed, the initial sample had a toxicity value equivalent to an inhibition level of 77%. This falls to non-toxic values after 120 minutes of reaction time and, though small toxicity values can be detected in the following minutes, the sample becomes completely nontoxic after 210 minutes. Negative values are now given as a result of the nutritive character the sample acquires with the growth of Lemna minor.
Figure 3 shows the results for photocatalysis with 7% AC-TiO2. Figure 3(a) shows the results for photocatalysis with 7% AC-TiO2 and illumination from commencement of the test. It can be seen that the thiabendazole degrades rapidly up to 90 minutes of reaction time, when 61% of the initial concentration has been removed. However, it is not until 240 minutes that the amount of this fungicide has fallen below the limit of 0.05 mg·L–1. The mineralization rate falls more progressively, with a remaining TOC of 39.2% after 90 minutes and 18.1% after 240 minutes.
A fall in toxicity is observed from initial inhibition levels of 77% to 6% at 90 minutes. Toxicity ceases after 120 minutes.
Figure 3(b) underlines the adsorbent effect of activated carbon, as the reactor is left in darkness for the first hour before illumination commences. During the first 30 minutes a rapid decrease can be observed of up to 33% of the initial concentration. Thirty minutes after illumination commences 97.8% of the initial thiabendazole has been degraded, though it is not until 180 minutes of illumination that the concentration falls below the limit of 0.05 mg·L–1. The same values are obtained after one hour of illumination as when illumination is used from the beginning of the reaction. This shows that the period of darkness has had no other effect than to delay the mineralization and detoxification processes. The toxicity