**2. Experimental**

#### **2.1 Materials**

Polyvinyl alcohol (PVA), poly(3-hydroxybutyrate) (PHB, MW = 312,000 g mol-1), poly(3 hydroxybutyrate-co-hydroxyvalerate) (PHBV, MW = 238,000 g mol-1) and ametryn (Pestanal®) were purchased from Sigma Chem. Co. The solvents employed in the chromatographic analyses were acetonitrile, HPLC grade methanol (JT Baker) and Milli-Q water. The solutions were filtered using 0.22 µm nylon membranes (Millipore, Belford, USA).

#### **2.2 Methodology**

4 Herbicides – Properties, Synthesis and Control of Weeds

al., 2002). It is therefore desirable to develop techniques whereby the physico-chemical properties of these chemicals can be altered and their usage made safer. The goal is to enable the use of soil management strategies that can produce foods at the current high levels of

Micro- and nanostructured polymeric materials can be used as transport systems for active chemicals. Advantages of these materials include good physical, chemical and biological stability, simple and reproducible preparation procedures, and applicability to a wide range of chemicals. In use, the active principle is released slowly and continuously, enabling the use of smaller quantities with greater efficiency, which reduces the risk of adverse

Controlled release systems have been extensively used in the food and pharmaceutical industries for active substances including nutrients, drugs and aromas (El Bahri & Taverdet, 2007; Grillo et al., 2008; Mello et al., 2008; Moraes et al., 2010), and there has been a recent increase in their application in medicine (Natarajan et al., 2011; Parajo et al., 2010; Vicente et

Amongst the new controlled-release system technologies under development, the use of polymeric micro- and nanoparticles is of special interest in agribusiness. Several studies have investigated controlled-release systems for bioactive compounds in agricultural applications (Ahmadi & Ahmadi, 2007; Bin Hussein et al., 2010; El Bahri & Taverdet, 2005, 2007; Grillo et al., 2010; Hirech et al., 2003; Li et al., 2010; Lobo et al., 2011; Silva et al., 2010; Singh et al., 2008, 2010). Materials that have been used include silica, bentonite and sepiolite clays, and polymeric substances such as alginate, lignin and synthetic polymers. The latter include the poly(hydroxyalkanoates) (PHAs) (Salehizadeh & Loosdrecht, 2004), of which poly(3-hydroxybutyrate) (PHB) and its hydroxyvalerate copolymer (PHBV) have been most widely used (Amass & Tighe, 1998). The advantages of using polymers such as PHB and PHBV are that they are fully biodegradable, inexpensive and easily prepared by bacterial fermentation (Pouton & Akhtarb 1996; Reis et al., 2008). These polymers are isotactic and highly crystalline (55-80 %), so that their degradation rates are relatively slow compared to

The objective of this work was to develop a novel release system for ametryn, employing microparticles prepared using two different polymers, PHB and PHBV (either individually or as mixtures). It was envisaged that the encapsulation of the herbicide in these microparticles would improve its chemical stability and enable the use of smaller quantities

Polyvinyl alcohol (PVA), poly(3-hydroxybutyrate) (PHB, MW = 312,000 g mol-1), poly(3 hydroxybutyrate-co-hydroxyvalerate) (PHBV, MW = 238,000 g mol-1) and ametryn (Pestanal®) were purchased from Sigma Chem. Co. The solvents employed in the chromatographic analyses were acetonitrile, HPLC grade methanol (JT Baker) and Milli-Q water. The solutions were filtered using 0.22 µm nylon membranes (Millipore, Belford,

those of lactate (PLA) and glycolate (PGA) copolymers (Sudesh et al., 2000).

of the chemical, hence reducing the risk of environmental contamination.

demand, without significant human or environmental risk.

environmental impacts (Sinha et al., 2004; Sopena et al., 2009).

al., 2010).

**2. Experimental** 

**2.1 Materials** 

USA).

#### **2.2.1 Determination of ametryn**

The HPLC analyses were performed using a Varian ProStar instrument fitted with a PS 210 pump, a UV-VIS detector (PS 325), a Metatherm oven and an automatic injector (PS 410). The chromatograms were acquired and processed using Galaxy Workstation software. The eluent used was acetonitrile/water (70:30, v/v), at a flow rate of 1.4 mL min–1, and separation was achieved using a Phenomenex Gemini C18 reversed phase column (5 μm, 110 Å, 150 mm x 4.60 mm i.d.). Ametryn was detected at a wavelength of 260 nm. The injection volume was 100 μL, and all samples were previously filtered through 0.22 µm nylon membranes.

#### **2.2.2 Preparation of the polymeric microparticles containing ametryn**

Microparticles were prepared with the PHB and PHBV polymers, used either individually or as a mixture, by formation of oil in water emulsions using the emulsification-solvent evaporation technique (Coimbra et al., 2008; Conti et al., 1995; Lionzo et al., 2007; Lobo et al., 2011). 200 mg of polymer (PHB, PHBV or a mixture of the two polymers, as described in Table 1) and 10 mg of herbicide were dissolved in 10 mL of chloroform to form the organic phase. The aqueous phase (200 mL) was prepared using 0.5 % (w/v) polyvinyl alcohol, at 50 oC. The organic phase was transferred to the aqueous phase (at 50 oC) with magnetic stirring (1000 rpm for 15 min). The chloroform was then evaporated from the emulsion. The suspension of microparticles formed was stored in an amber flask (to avoid any photodegradation of the herbicide). The final concentration of ametryn was 50 mg L-1.


Table 1. Proportions of polymers used to prepare the different formulations.

#### **2.2.3 Measurements of encapsulation efficiency**

Portions (10 mg) of the different microparticles containing herbicide were dissolved in 50 mL of acetonitrile, and the association rate of the herbicide with the microparticles was determined by the technique described previously, which involves ultrafiltration/ centrifugation and analysis using HPLC (Kilic et al., 2005; Schaffazick et al., 2003). The samples were centrifuged in regenerated cellulose ultrafiltration filters that had a molecular size-exclusion pore size of 30 KDa (Microcon, Millipore), and the filtrate was analyzed using HPLC. The ametryn concentration was obtained from an analytical curve. The association

Preparation and Characterization

Ritger & Peppas, 1987a, 1987b).

**3. Results and discussion** 

% for atrazine in PHBV microparticles.

**2.2.5.1 Mathematical modeling of ametryn release** 

of Polymeric Microparticles Used for Controlled Release of Ametryn Herbicide 7

Mathematical modeling is increasingly used to investigate the release profiles of bioactive compounds in polymeric systems, since it can provide important information concerning the release mechanism. Analysis of the mechanism of release of ametryn from the microparticles employed the zero order, first order, Higuchi and Korsmeyer-Peppas models (Colombo et al., 1995, 2005; Costa & Lobo, 2001; Ferrero et al., 2000; Hariharam et al., 1994;

The encapsulation efficiency values obtained for the different microparticles are listed in Table 2. Formulation A (100 % PHBV) showed the highest encapsulation efficiency (76.5 %). The efficiency decreased as the proportion of PHBV decreased, and formulation E (100 % PHB) provided the lowest encapsulation efficiency (26.2 %). The values obtained for formulations A and B were fairly high, relative to values that have been reported in the literature for other active principles (Bazzo et al., 2009; Grillo et al., 2010; Lobo et al., 2011; Sendil et al., 1999). Grillo and colleagues (2010) showed that the encapsulation efficiency of the herbicide atrazine in PHBV microparticles was in excess of 30 %. Lobo et al. (2011), using an experimental design optimization procedure, obtained an encapsulation efficiency of 24

**Formulation PHBV (%) PHB (%) EE (%)**

**A** 100 0 76.5

**B** 75 25 54.7

**C** 50 50 40.5

**D** 25 75 29.3

**E** 0 100 26.2

The relationship between the percentage of PHBV and the encapsulation efficiency is illustrated in Figure 2. There was a polynomial relationship between the encapsulation efficiency and the PHBV concentration, which was positive for PHBV and negative for PHB. This can probably be explained by the structural differences between the microparticles, due

The morphological characteristics of the microparticles, as well as the influence of the encapsulation of ametryn, were analyzed using the SEM procedure. Electron micrographs of the microparticles containing ametryn are illustrated in Figure 3. All types of microparticle were spherical, although the surface structures were different. Most of the PHB microparticles possessed smooth surfaces with few pores, while most of the PHBV microparticles were rough-surfaced with many cavities and pores, some of which were quite large, as can be clearly seen for formulation A (Figure 3, a1 and a2). Grillo et al. (2010) also found that PHBV microparticles, prepared using the same methodology as that

Table 2. Encapsulation efficiencies (EE, %) of the different microparticles.

to the different polymer ratios used in their preparation (Table 1).

rate of ametryn was calculated from the difference between the concentration measured in the filtrate and the total concentration (100 %) in the microparticle suspension. The total concentration was measured after diluting the suspension with acetonitrile, which dissolved the polymer and ensured complete release of the herbicide. The measurements were performed in triplicate for each formulation. The encapsulation efficiency (EE, %) was expressed as the ratio:

$$\text{EE} \left(\% \right) = \frac{\text{W}\_{\text{S}}}{\text{W}\_{\text{TOTAL}}} \times 100\% \tag{1}$$

Where, *Ws* is the quantity of ametryn in the microparticles and *Wtotal* is the amount of ametryn used in the formulation.

#### **2.2.4 Scanning electron microscopy (SEM)**

A scanning electron microscope (Model JSM-6700F, JEOL, Japan) was used to investigate the size distribution and surface morphology of the microparticles. Suspensions of microparticles containing the herbicide were filtered and the particles were then washed with 150 mL of distilled water. The solid residues were dried overnight over Na2SO4 in a desiccator. The samples were then attached to metallic supports (stubs) with double-sided tape, and metalized by deposition of a gold layer at a current of 25 mA for 150 s. Images (electron micrographs) of the samples were then generated using the microscope. Particle sizes were measured using the ImageJ 1.42 program, and the size distributions of the different microparticles were obtained using OriginPro 7.0. At least 1000 individual particles of each sample were used for these measurements.

#### **2.2.5 Release of ametryn from the microparticles**

The release profiles of ametryn, either free or associated with the microparticles, were investigated using a two-compartment experimental system. A cellulose membrane (Spectrapore, with a molecular exclusion pore size of 1000 Da) separated the donor compartment, containing 4 mL of solution (or suspension) of the herbicide, from the acceptor compartment, which contained 50 mL of deionized water maintained under gentle agitation at ambient temperature (Paavola et al., 1995). The pore size of the membrane only allowed passage of the free herbicide, while the herbicide associated with the microparticles was retained in the donor compartment until the equilibrium was shifted so as to release the ametryn present within the particles. The size of the microparticles prevented their passage through the pores of the membrane. These experiments were conducted under *dilution sink* conditions, whereby the volume of the dissolution medium was sufficiently large that the herbicide concentration never exceeded 10 % of the value of its saturation concentration (Aulton et al., 2002).

Samples were retrieved from the acceptor compartment as a function of time, and analyzed by HPLC at a detector wavelength of 260 nm. During the first hour, samples were collected every 15 min, during the second hour every 30 min, and subsequently at hourly intervals until the peak area stabilized. The peak area values were then converted into the percentage of herbicide released as a function of time (De Araújo et al., 2004).

#### **2.2.5.1 Mathematical modeling of ametryn release**

Mathematical modeling is increasingly used to investigate the release profiles of bioactive compounds in polymeric systems, since it can provide important information concerning the release mechanism. Analysis of the mechanism of release of ametryn from the microparticles employed the zero order, first order, Higuchi and Korsmeyer-Peppas models (Colombo et al., 1995, 2005; Costa & Lobo, 2001; Ferrero et al., 2000; Hariharam et al., 1994; Ritger & Peppas, 1987a, 1987b).
