**1. Introduction**

The fertilizer industry faces a continuing challenge to improve its products and increase the efficiency of their use, particularly of nitrogenous fertilizers, and to minimize any possible adverse environmental impact. This is done either through improvement of fertilizers already in use, or through development of new specific fertilizer types [1]. One specific objective in Europe is the development of new sustainable fertilizers with the aim of reducing the consumption and/or import of conventional mineral fertilizers from third-world countries, and of N fertilizers based on energy-intensive production process. The development of innovative controlled-released fertilizers (CRFs) could limit the consumption of the traditional ones. Indeed, CRFs present the interest to better match the plants' need for nutrients over time.

Several CRFs products are already available in the market. They are manufactured from blends of fertilizers as active principle and organic polymers. There are two major CRFs types [2]. In the first, the fertilizer is encapsulated into the polymer matrix. In the second, the fertilizer is coated by a polymer layer. The fertilizer bioavailability for the plant uptake is controlled by the diffusion rate through the pores of the polymer matrix or through the surrounding polymer shell. Although several commercial CRFs can achieve

**Citation:** Evon, P.; Labonne, L.; Padoan, E.; Vaca-Garcia, C.; Montoneri, E.; Boero, V.; Negre, M. A New Composite Biomaterial Made from Sunflower Proteins, Urea, and Soluble Polymers Obtained from Industrial and Municipal Biowastes to Perform as Slow Release Fertiliser. *Coatings* **2021**, *11*, 43. https://doi.org/10.3390/coatings11010043

Received: 17 December 2020 Accepted: 30 December 2020 Published: 2 January 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

satisfactory fertilizers release rate, the plastic polymer in the blend raises concern for its low biodegradability and the accumulation of plastic impurities in soil. The new EU regulation on fertilizing products addresses the degradability of the organic matter of CRFs by stipulating 90% conversion of the organic carbon into CO2 in a maximum of 24 months. The specific challenge focuses on finding organic polymers that comply with the biodegradability EU requirements, while achieving the controlled release of nutrients in the best possible manner.

Much research work has been carried out to develop new CRFs containing natural polymers, instead of synthetic polymers. The substitution of synthetic polymers derived from fossil sources with natural polymers is a current major environmental issue [3]. Although biodegradable polymers can be also synthesized from monomers obtained from the petrochemical industry, the consumption of chemicals from fossil sources is a major source of GHG emission. In addition, it contributes to the depletion of fossil sources. Natural polymers such as lignin, starch, chitin, cellulose, and other polysaccharides (e.g., *k*-carrageenan) have been investigated to make CRFs. Generally, these cannot be used in natural pure form as effective coating material due to their inherent properties, such as poor mechanical properties, hydrophobicity, high industrial demand, poor solubility, and processability. These imply limitations for their use.

To overcome the above criticalities, researchers have focused their work on chemical and physical modifications of natural polymers. Biopolymers from monomers obtained from renewable source have also been obtained, e.g., polylactic acid (PLA) or poly hydroxyl alkanoates (PHA). Aims of current research on CRFs are to reduce production cost, improve nutrient release rate, optimize organic matter biodegradability rate, and guarantee crushing strength to bear crack resistance under environmental stresses. These three properties need to be compatible with each other. Indeed, a poor mechanical strength or high biodegradability may result in faster nutrient release rate, excess nutrient availability over the plant uptake rate, high nutrient leaching from soil into ground, and low CRFs efficiency. High crushing strength also facilitates the handling and transportation and reduces the water absorption [4].

In the present work, the current CRFs issues have been addressed by fabricating a new material out of three components. These are a sunflower protein concentrate (SPC) obtained from sunflower oil cake (SOC), a biopolymer (BP) obtained from municipal biowaste (MBW), and commercial urea (U). All three components contain N. Three different blends were manufactured, i.e., SPC-10% U, SPC-10% BP, SPC-5% U-5% BP. The underlying rationale of the work was to use SPC as polymer matrix, and BP and U as fillers/active principles. The reason for this approach is based on the following literature data reported for SPC, BP and U.

The SPC is well known biopolymer used in human nutrition for its nutritional value [5], and potential antioxidant properties and benefits for disease prevention and aging retardation [6]. SPC are reported also to have also good processability to manufacture films by solvent casting [7] and extrusion at 160 ◦C [8]. This material therefore seemed a potential eco-friendly matrix to be used for manufacturing new CRFs.

The BP is obtained by chemical hydrolysis of fermented unsorted urban food wastes. The BPs sourcing food wastes contain all major animal and vegetable natural polymers. Upon anaerobic fermentation, the most readily degradable fats, polysaccharides, and proteins are converted to methane and carbon dioxide. The solid anaerobic digestate (AD) contains the recalcitrant lignocellulosic fraction. The BP, obtained from AD chemical hydrolysis, contains a mix of water-soluble molecules with molecular weight from 5 to above 750 kDa [9]. These molecules are water soluble lignocellulosic fragments. They keep the memory of the macromolecularity and functional groups, and of the mechanical properties of the proximates in the parent food wastes. However, BP is water soluble and thermally stable up to 200 ◦C. This allows for fabricating, by solvent casting and melt extrusion, composite articles containing BP in blends with synthetic polymers such as polyethyleneco-vinyl alcohol [10] and polyethylene-co-acrylic acid [11]. The blends have been proven

to have higher mechanical resistance, compared to the article manufactured from the neat synthetic polymer only. The BP is not biodegradable. However, based on its sourcing material, no adverse effect is expected from its accumulation in soil. Indeed, MBW anaerobic digestates and/or compost are used as soil amending agents and/or fertilizers. The BP has also been proven [9] more efficient soil fertilizer and plant growth biostimulant in the cultivation of several food and ornamental plants, in comparison with its sourcing fermented MBW materials and with commercial organo-inorganic peat and leonardite derived products claimed by the vendor as plant biostimulants. This material seemed therefore a potential processable filler, capable of increasing the mechanical resistance of the SPC matrix and to perform at the same time as active fertilizer principle in CRFs.

Urea is a main commercial fertilizer. Its world consumption amounts to approximately 51 Mt/yr [12]. In the present work, urea has been chosen as reference commercial N fertilizer. Urea and lignin materials have also been investigated for use in the CRF field. Urea coated with pine lignin [13] mixed with various types of additives exhibited 59% higher crushing strength than the uncoated fertilizer. Urea coated with four types of lignin, recovered from the effluent liquor of the paper and paper industry [14], gave products with film forming properties, but unsatisfactory release rate. The same occurred for urea granules fabricated by mixing urea with kraft lignin under melting conditions [15]. Kraft lignin was added to a tapioca starch-urea-borate matrix to modulate the starch matrix hydrophilicity properties and reduce the urea release in water [16]. The final film product remained intact after one month of contact time with water. The retardation of the urea release rate by Kraft lignin was confirmed also for other formulations [15]. Based on these results, the BP keeping the memory of the pristine lignin matter present in the sourcing food waste seemed to the authors of the present work to offer intriguing perspectives for use together with urea and the SPC matrix in the manufacture of new CRFs.

The main aims of the present work were to develop a twin-screw extrusion process tailored to the fabrication of the new SPC-BP-U pellets and to test the pellets mechanical behavior and release rate in solution of organic and inorganic nitrogen. The principal conclusions of the work are that the developed twin-screw extrusion process successfully produces the SPC-BP-U pellets and that the pellets behave as CRFs. The results also sugges<sup>t</sup> that BP has positive effects on the mechanical properties of the pellets and on the chemical behavior of the organic and inorganic N species released in solution.

#### **2. Materials and Methods**

#### *2.1. SPC, BP, U, and Other Reagents*

The BP was available from previous work [9]. It was obtained from the anaerobic digestate of the MBW processed in the ACEA waste treatment plant in Pinerolo, Italy. The sampled digestate was hydrolyzed at 60 ◦C in pH 13 water. The liquid hydrolysate was separated from the insoluble residue by centrifugation and then filtered through polysulphone membrane with 5 kDa cut off. The membrane retentate was dried at 60 ◦C to yield the solid BP.

The SPC was obtained from a sunflower oil cake (SOC). To obtain it, SOC was sieved using a Ritec (France) 600 vibrating sieve shaker fitted with a 1 mm grid. The oversize was mainly composed of solid particles from the seed hull, and therefore rich in fibers. Conversely, the undersize (SPC) was enriched with the smaller particles coming from the kernel of the seed, thus having a high protein content. This resulted in a higher content of SPC in proteins, the latter having been evaluated at 50.7% ± 0.1% (in proportion to the SPC dry weight) using the Kjeldahl method [17]. The other chemicals inside SPC were minerals, lipids, cellulose, hemicelluloses, lignins, and water-solubles, with contents of 8.6% ± 0.1%, 1.4% ± 0.1%, 11.0% ± 0.9%, 11.2% ± 1.7%, 0.8% ± 0.2%, and 26.3% ± 0.1%, respectively. The SPC components were determined according to the following literature methods: ISO 749 standard [18] for minerals, ISO 659 standard [19] for lipids, and the ADF-NDF method [20,21] for cellulose, hemicelluloses, and lignin. The content in watersoluble components was estimated by measuring the mass loss of the test sample after 1 h in boiling water. All analyses were carried out in duplicate.

#### *2.2. Fabrication of Extruded Pellets and Injection-Molded Pieces* 2.2.1.ExtrudedPellets

The extruded pellets were obtained by destructuring and then plasticizing the proteins present in SPC. Thermo-mechano-chemical destructuring was carried out [22,23]. Protein plasticization was conducted in the presence of an aqueous solution of sodium sulfite (1 kg sodium sulfite for 10 kg water) or even urea (U) using a Clextral (France) Evolum HT 53 co-rotating and co-penetrating twin-screw extruder (Figure 1). The latter was composed of nine modules, each 4D in length, D corresponding to the screw diameter (i.e., 53 mm). The total barrel length was thus 36D (i.e., 1.908 m).

The screw rotation speed was 300 rpm for all the formulations produced. The temperature profile along the barrel was 20 ◦C for the feeding module (module 1), 80 ◦C for module 2, and 100 ◦C for the seven other modules (i.e., modules 3 to 9).

SPC was introduced at the level of module 1 using a Coperion K-Tron (Coperion K-Tron (Schweiz) GmbH, Niederlenz, Switzerland) SWB 300-N weight feeder at a 64.5 kg/h inlet flow rate. The sodium sulphite solution was injected at the end of the second module at a 23.3 kg/h inlet flow rate using a DKM (Clextral, Firminy, France) Super K CAMP 112 piston pump. The resulting formulation was referred to as SPC. When incorporated in the formulation, BP and U were also introduced at the level of module 1 using a Coperion K-Tron K-ML-KT20 weight feeder (Coperion K-Tron (Schweiz) GmbH, Niederlenz, Switzerland) at a 6.5 kg/h inlet flow rate (i.e., 10% (*w*/*w*) in proportion to SPC). The corresponding formulations have been designated as follows: SPC-BP and SPC-U, respectively. For its part, the SPC-BP-U formulation was obtained by adding simultaneously BP and U in module 1, both introduced at a 3.25 kg/h inlet flow rate.

The intimate mixing of the solids each other, and the impregnation of the liquid into the solid(s) was made possible by the use of two consecutive pairs of bilobe paddles (BL22 type), each 1D in length, positioned at the beginning of module 4. The plasticization of sunflower proteins was obtained through intense mechanical shear using four consecutive pairs of reversed elements (CF2C type), each 0.5D in length, positioned at the end of module 7 and beginning of module 8. A die equipped with six holes, each 3 mm in diameter, was positioned at the end of the barrel, and the extruded pellets were obtained using a Clextral HC 45 granulating system. They were then dried using a Clextral Evolum 600 continuous belt dryer up to a 10% moisture content before their packaging inside sealed plastic bags.

#### 2.2.2. Injection-Molded Pieces

The extruded pellets were molded into standard bending and tensile specimens by thermoplastic injection using a Negri Bossi (Cologno Monzese, Italy) VE 160–720 machine with a clamping force of 160 ton, and a mold with two cavities. All formulations were rehydrated by adding water to the extruded pellets prior to thermoplastic injection. The moisture content of the extruded pellets was 20% at the time of molding.

The conditions used to produce the test specimens were as follows: 70 ◦C/90 ◦C/110 ◦C for the temperature profile along the plasticizing screw, 130 ◦C for the temperature of the nozzle, 150 rpm for the rotation speed of the plasticizing screw up to a 21 mm length for the shot building, 30 bar for the counter pressure, 150 mm/s for the injection speed, 800 bar for the follow-up pressure applied during a 2.5 s duration, 1600 kN for the clamping force, 50 ◦C for the mold temperature, and 20 s for the cooling time before opening the mold and ejecting the test specimens from the cavities.

Once obtained, the test specimens were placed in a climatic chamber at 60% relative humidity (RH) and 25 ◦C for three weeks for equilibration. Once equilibrated, they were then used for characterization.

#### *2.3. Characterization of Mechanical Properties and Water Sensitivity Test for SPC Composites* 2.3.1. Density of Extruded Pellets and Injection-Molded Pieces

The density of the extruded pellets was measured using a 50 mL pycnometer, and cyclohexane as immersion liquid. Cyclohexane was chosen because of its marked hydrophobic character. Indeed, cyclohexane contains only carbon and hydrogen atoms whose bonds are not polarized, thus classifying this solvent in the category of the apolar solvents. On the contrary, the granules are hydrophilic as they are composed of biomolecules and biopolymers with many polarized bonds (e.g., CO and NH bonds for proteins, OH bonds for cellulose, etc.). In consequence, there is no absorption of cyclohexane by the granules and no change in their volume by swelling. All determinations were carried out in duplicate.

The density of the injection-molded pieces was measured from three test specimens with an 80 mm length, a 10 mm width, and a 4 mm thickness. Their thickness and their width were measured at three points, and their length at two points, with a 0.01 mm resolution electronic digital sliding caliper. Thickness (*t*), width (*b*), and length (*l*) mean values were recorded to calculate the specimen volume, and test specimens were all weighed to calculate their density (*d*). Mean apparent density of the injection-molded pieces from the same formulation was the mean value of measurements made on the three test specimens.

#### 2.3.2. Resistance to Mechanical Abrasion

The resistance to mechanical abrasion of both pellets and injection-molded pieces was estimated from an unstandardized test that was specifically developed during this study. For this test, a cylindrical plastic container having a 115 mm diameter and a 130 mm height was used. The bottom of the container was fixed to an axis, inclined at 15◦ to the top with respect to the horizontal direction, and rotating by means of a motor at an 80 rpm rotation speed. About 30 g of pellets or injection-molded pieces having a 10 × 10 × 4 mm<sup>3</sup> volume were positioned inside the container before the beginning of the test. During the test, 10 metal parts for a total mass of 125 g were also placed inside the container, their addition aiming to simulate the mechanical abrasion to which the pellets or the injection-molded pieces will be subjected during their spreading to the field. Two wooden blades having the same length as the height of the container, both 1 cm high and 1 cm wide, were fixed inside the cylinder, 180◦ to each other. Here, their objective was to amplify the effect of mechanical abrasion on the tested sample by the metal parts. The test duration was 1 h. At the end of the test, the plant objects were recovered and then sieved on a 2 mm grid to quantify the fines generated during the mechanical abrasion test. The result was expressed as the ratio of fines to the initial mass of the sample analyzed (%). The higher this ratio, the more sensitive the sample was to mechanical abrasion. All determinations were carried out in duplicate.

#### 2.3.3. Tensile and Bending Properties

The tensile properties of the injection-molded pieces were determined according to ISO 527-4 standard [24]. In particular, the two points chosen at the beginning of the stress-strain curve for the Young's modulus calculation were associated with the following elongations: 0.0005 (i.e., 0.05%) and 0.0025 (i.e., 0.25%).

The flexural properties were determined according to the ISO 178 standard [25], i.e., using the three points bending method. For both tensile and bending properties, an Instron (Instron, Norwood, MA, USA) 33R 4204 universal testing system fitted with a 5 kN load cell was used.

The distance between jaws for the tensile tests was 105 mm, and a testing speed of 3 mm/min was applied. The grip separation for the bending tests was 40 mm, and a testing speed of 2 mm/min was applied. All determinations were carried out through four repetitions.

#### 2.3.4. Shore D Surface Hardness

The Shore D surface hardness of the injection-molded pieces was determined according to the ISO 868 standard [26] with a Bareiss (Bareiss Prüfgerätebau GmbH, Oberdischingen, Germany) durometer. The indentation direction was perpendicular to the upper face of the injection-molded piece. For each formulation, determinations were carried out from four different bending test specimens, and measurements were made 10 times per each test specimen (five times for each of its sides).

#### 2.3.5. Water Sensitivity

The water sensitivity of the injected pieces was determined according to ISO 16983 standard [27]. Measurements were conducted after soaking of the analysed samples in water during 1, 3, 6, and 24 h, respectively. Thickness swelling (TS, %) and water absorption (WA, %) were calculated. All determinations were carried out in triplicate.

#### *2.4. Measurements of N Release from SPC Composites*

The study of the release of urea and other N-containing compounds was performed according to a previously reported procedure [1]. The samples (4 g each) were immersed in 80 mL demineralized water in a 100 mL glass flask. Three replicates were carried out for each material. The flasks were stored at 25 ◦C in the dark. At given sampling times (1, 2, 4, 10, 15, 25 days), the liquid and solid phases were separated by centrifugation (15 min at 3000 rpm). The liquid phase was used for the determination of the concentrations of urea, ammonia, and total nitrogen with the methods described below. At each sampling, the same amount of demineralized water as that of the discarded supernatant was added to the solid phase.

#### 2.4.1. Determination of the Urea Concentration

The concentration of urea in the supernatant was measured following the spectrophotometric method of Chen et al. [28]. One ml aliquot solution was diluted with water (1:1000 (*v*/*v*) after 1 day extraction, 1:500 (*v*/*v*) after 2, 4, and 7 days of extraction, 1:100 (*v*/*v*) after 10 days of extraction, and 1:50 (*v*/*v*) after 15 and 25 days of extraction).

To the diluted samples (18 mL), transferred in a 100 mL amber glass bottles, 0.9 mL di DAM (diacetylmonoxime 0.5 g/10 mL), 0.15 mL TSC (thiosemicarbazide 0.2 g/100 mL), 0.15 mL FeCl3, 6H2O 40.56 mg/100 mL), 12 mL (5%, *v*/*v*) H2SO4 were added in sequence. The reaction mixtures were heated in a water bath at 80 ◦C for one hour.

The absorbance of the complex was measured at 520 nm in a spectrophotometer (Hitachi U-2000), previously calibrated with solutions of urea at given concentrations treated as described above.

#### 2.4.2. Determination of Ammonia Nitrogen

A 10 mL aliquot of the supernatant was transferred to a Kjeldahl tube. After addition of 1 g MgO and 70 mL water, NH3 was distilled in a Kjeldahl instrument and collected in a beaker containing 20 mL boric acid and 2 drops of methyl red and bromocresol green indicator.

#### 2.4.3. Determination of Total Nitrogen

A 1 mL aliquot of the supernatant was suspended in 25 mL sulfuric acid and 0.5 g selenic mixture and heated to boiling point until complete mineralization. The solution was transferred to a 25 mL volumetric flask and brought to volume with water. A 10 mL aliquot was transferred to a Kjeldahl tube and treated as above after addition of 40 mL 40% NaOH instead of MgO.

#### 2.4.4. Determination of Organic Nitrogen

Organic nitrogen was calculated by subtracting ammonia and urea N from total N.

#### *2.5. Statistical Treatment of Data*

The data were evaluated by one-way ANOVA (*p* < 0.05) followed by the Tukey's test for multiple comparison procedures.
